Image collating method and apparatus

Claim 2

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second pixel data into a plurality of verification blocks; extracting representative pixel data from each of the base blocks; extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; and comparing the representative pixel data with the feature amount data and partially collating the base block with the verification block.

Claim 3

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second pixel data into a plurality of verification blocks; means for extracting representative pixel data from each of the base blocks; means for extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; and means for comparing the representative pixel data with the feature amount data and partially collating the base block with the verification block.

Claim 4

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second pixel data into a plurality of verification blocks; means for extracting representative pixel data from each of the base blocks; means for extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; and means for comparing the representative pixel data with the feature amount data and partially collating the base block with the verification block.

Claim 5

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second pixel data into a plurality of verification blocks; extracting representative pixel data from each of the base blocks; extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; comparing the representative pixel data with the feature amount data and generating an evaluated value that represents a matching degree of the base block and the verification block corresponding to the compared result; and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 6

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second pixel data into a plurality of verification blocks; extracting representative pixel data from each of the base blocks; extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; comparing the representative pixel data with the feature amount data and generating an evaluated value that represents a matching degree of the base block and the verification block corresponding to the compared result; and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 7

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second pixel data into a plurality of verification blocks; means for extracting representative pixel data from each of the base blocks; means for extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; means for comparing the representative pixel data with the feature amount data and generating an evaluated value that represents a matching degree of the base block and the verification block corresponding to the compared result; and means for detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 8

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second pixel data into a plurality of verification blocks; means for extracting representative pixel data from each of the base blocks; means for extracting feature amount data from each of the verification blocks that contains pixel data corresponding to the representative pixel data, the feature amount data representing a feature amount of the verification block; means for comparing the representative pixel data with the feature amount data and generating an evaluated value that represents a matching degree of the base block and the verification block corresponding to the compared result; and means for detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 9

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the representative pixel data is pixel data nearly at the center position of the base block.

Claim 10

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the feature amount data is two out of the maximum value, the small value, and the dynamic range of a plurality of pixels of the verification block.

Claim 11

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the feature amount data is the average value and the standard deviation of a plurality of pixels of the verification block.

Claim 12

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the feature amount data is two out of the maximum value, the small value, and the dynamic range of a plurality of pixels of the verification block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as an evaluated value when the value of the representative pixel data is included in the range from the maximum value to the minimum value and for outputting the difference between the representative pixel data and the maximum value and the difference between the representative pixel data and the minimum value as evaluated values when the value of the representative pixel data is not included in the range from the maximum value to the minimum value.

Claim 13

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the feature amount data is the average value and the standard deviation of a plurality of pixels of the verification block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as an evaluated value when the representative pixel data is included in the range between the sum value and the difference value of the average value and the standard deviation and for outputting the difference between the representative pixel data and the sum value and the difference between the representative pixel data and the difference value as evaluated values when the representative pixel data is not included in the range between the sum value and the difference value of the average value and the standard deviation.

Claim 14

The method or the apparatus as set forth in claim 1, 2, 3, or 4, wherein the base block and the verification block are composed of a plurality of small blocks, the representative pixel data and the feature amount being extracted for each of the small blocks, the evaluated value being generated for each of the small blocks, the evaluated value being cumulated, the base block and the verification block being partially collated corresponding to the cumulated value.

Claim 15

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount data of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount data of all pixels of the modified verification block; and comparing the first feature amount data with the second feature amount data and to partially collating the base block with the verification block.

Claim 16

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount data of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount data of all pixels of the modified verification block; and comparing the first feature amount data with the second feature amount data and to partially collating the base block with the verification block.

Claim 17

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount of all pixels of the modified verification block; and means for comparing the first feature amount data with the second feature amount data and to partially collating the base block with the verification block.

Claim 18

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount of all pixels of the modified verification block; and means for comparing the first feature amount data with the second feature amount data and to partially collating the base block with the verification block.

Claim 19

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount of all pixels of the modified verification block; comparing the first feature amount data with the second feature amount data and generating an evaluated value representing a matching degree of the base block and the verification block corresponding to the compared result; and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value, the evaluated value being generated by moving the verification block in a predetermined search range.

Claim 20

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount data representing the feature amount of all pixels of the modified verification block; comparing the first feature amount data with the second feature amount data and generating an evaluated value representing a matching degree of the base block and the verification block corresponding to the compared result; and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value, the evaluated value being generated by moving the verification block in a predetermined search range.

Claim 21

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and a generating modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the feature amount of all pixels of the modified verification block; means for comparing the first feature amount data with the second feature amount data and generating an evaluated value representing a matching degree of the base block and the verification block corresponding to the compared result; and means for detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value, the evaluated value being generated by moving the verification block in a predetermined search range.

Claim 22

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and a generating modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the feature amount of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the feature amount of all pixels of the modified verification block; means for comparing the first feature amount data with the second feature amount data and generating an evaluated value representing a matching degree of the base block and the verification block corresponding to the compared result; and means for detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the evaluated value, the evaluated value being generated by moving the verification block in a predetermined search range.

Claim 23

The method or the apparatus as set forth in claim 11, 12, 13, or 14, wherein the first feature amount data and the second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the base block and the modified verification block.

Claim 24

The method or the apparatus as set forth in claim 11, 12, 13, or 14, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the base block and the modified verification block.

Claim 25

The method or the apparatus as set forth in claim 11, 12, 13, or 14, wherein the first feature amount data and the second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the base block and the modified verification block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as an evaluated value when the maximum value and the minimum value of the first feature amount data are included in the range from the maximum value to the minimum value of the second feature amount data and for outputting the difference between the difference between the maximum values of the first feature amount data and the second feature amount data and the difference between the minimum values thereof as the evaluated values when the maximum value and the minimum value of the first feature amount data are not included in the range from the maximum value to the minimum value of the second feature amount data.

Claim 26

The method or the apparatus as set forth in claim 11, 12, 13, or 14, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the base block and the modified verification block, and wherein said evaluated value generating step or means is adapted for generating the sum and the difference of the average value and the standard deviation, for outputting nearly 0 as an evaluated value when the sum and the difference of the average value and the standard deviation of the first feature amount data are included in the range between the sum and the difference of the average value and the standard deviation of the second feature amount data, and for outputting the difference between the sums of the average values and the standard deviations of the first feature amount data and the second feature amount data and the differences thereof as evaluated values when the sum and the difference of the average value and the standard deviation of the first feature amount data are not included in the range between the sum and the difference of the average value and the standard deviation of the second feature amount data.

Claim 27

The method or the apparatus as set forth in claim 11, 12, 13, or 14, wherein the base block and the verification block are composed of a plurality of small blocks, the first feature amount and the second feature amount being extracted for each of the small blocks, the evaluated value being generated for each of the small blocks, the generated evaluated values being cumulated, the base block and the verification block being partially collated corresponding to the cumulated value.

Claim 28

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks; extracting first representative pixel data from each of the base blocks; extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; extracting second representative pixel data from the verification block; extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 29

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks; extracting first representative pixel data from each of the base blocks; extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; extracting second representative pixel data from the verification block; extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 30

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks; means for extracting first representative pixel data from each of the base blocks; means for extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; means for comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; means for extracting second representative pixel data from the verification block; means for extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; means for comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 31

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks; means for extracting first representative pixel data from each of the base blocks; means for extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; means for comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; means for extracting second representative pixel data from the verification block; means for extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; means for comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 32

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of the verification block necessary for a phase compensation; extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 33

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of the verification block necessary for a phase compensation; extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 34

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of the verification block necessary for a phase compensation; means for extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; means for comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; means for generating a modified base block containing pixels in the vicinity of the base blocks necessary for a phase compensation; means for extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; means for extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; means for comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 35

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of the verification block necessary for a phase compensation; means for extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; means for comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; means for generating a modified base block containing pixels in the vicinity of the base blocks necessary for a phase compensation; means for extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; means for extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; means for comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 36

An image collating method for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising the steps of: generating a first evaluated value corresponding to a base block, the first evaluated value representing a collating degree of the base block and a verification block; generating a second evaluated value corresponding to the verification block, the second evaluated value representing a collating degree of the verification block and the base block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 37

An image collating method for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising the steps of: generating a first evaluated value corresponding to a base block, the first evaluated value representing a collating degree of the base block and a verification block; generating a second evaluated value corresponding to the verification block, the second evaluated value representing a collating degree of the verification block and the base block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 38

An image collating apparatus for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising: means for generating a first evaluated value corresponding to a base block, the first evaluated value representing a collating degree of the base block and a verification block; means for generating a second evaluated value corresponding to the verification block, the second evaluated value representing a collating degree of the verification block and the base block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 39

An image collating apparatus for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising: means for generating a first evaluated value corresponding to a base block, the first evaluated value representing a collating degree of the base block and a verification block; means for generating a second evaluated value corresponding to the verification block, the second evaluated value representing a collating degree of the verification block and the base block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 40

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks; extracting first representative pixel data from each of the base blocks; extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; extracting second representative pixel data from the verification block; extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 41

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks; extracting first representative pixel data from each of the base blocks; extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; extracting second representative pixel data from the verification block; extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 42

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks; means for extracting first representative pixel data from each of the base blocks; means for extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; means for comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; means for extracting second representative pixel data from the verification block; means for extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; means for comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 43

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks; means for extracting first representative pixel data from each of the base blocks; means for extracting first feature amount data from each of the verification blocks containing pixel data corresponding to the first representative data, the first feature amount representing a feature amount of the verification block; means for comparing the first representative pixel data with the first feature amount data so as to generate a first evaluated value corresponding to the base block; means for extracting second representative pixel data from the verification block; means for extracting second feature amount data from the base block containing pixel data corresponding to the second representative data, the second feature amount representing a feature amount of the base block; means for comparing the second representative pixel data with the second feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 44

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value.

Claim 45

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks; segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; extracting first feature amount data from the base block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value.

Claim 46

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; means for comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; means for generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; means for extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; means for extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; means for comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value.

Claim 47

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks; means for segmenting second image data into a plurality of verification blocks and generating a modified verification block containing pixels in the vicinity of each of the verification blocks necessary for a phase compensation; means for extracting first feature amount data from each of the base blocks, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification block, the second feature amount representing the range of all pixels thereof; means for comparing the first feature among data with the second feature amount data so as to generate a first evaluated value corresponding to the base block; means for generating a modified base block containing pixels in the vicinity of the base block necessary for a phase compensation; means for extracting third feature amount data from the verification block, the third feature amount data representing the range of all pixels thereof; means for extracting fourth feature amount data from the modified base block, the fourth feature amount data representing the range of all pixels thereof; means for comparing the third feature amount data with the fourth feature amount data so as to generate a second evaluated value corresponding to the verification block; and means for generating a totally evaluated value corresponding to the first evaluated value and the second evaluated value and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value.

Claim 48

The method or the apparatus as set forth in claim 20, 21, 26, or 27, wherein the first representative pixel data and the second representative pixel data are pixel data nearly at the center position of the block.

Claim 49

The method or the apparatus as set forth in claim 20, 21, 26, or 27, wherein the first feature amount data and the second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block.

Claim 50

The method or the apparatus as set forth in claim 20, 21, 26, or 27, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the block.

Claim 51

The method or the apparatus as set forth in claim 20, 21, 26, or 27, wherein the first feature amount data and the second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as the first evaluated value and the second evaluated value when the values of the first representative pixel data and the second representative pixel data are included in the range from the maximum value to the minimum value and for outputting the difference between the first representative pixel data and the minimum value and the difference between the second representative pixel data and the minimum data as the first evaluation value and the second evaluation data, respectively, when the values of the first representative pixel data and the second representative pixel data are not included in the range from the maximum value to the minimum value.

Claim 52

The method or the apparatus as set forth in claim 20, 21, 26, or 27, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the block, and wherein said first evaluated value generating means or step and said second evaluated value generating means or step are adapted for outputting nearly 0 when the first representative pixel data and the second representative pixel data are included in the range between the sum and the difference of the average value and the standard deviation and for outputting the difference between the sum and the difference of the first representative pixel data and the second representative pixel data as evaluated values when the first representative pixel data and the second representative pixel data are not included in the range between the sum and the difference of the average value and the standard deviation.

Claim 53

The method or the apparatus as set forth in claim 20, 21 26, or 27, wherein the base block and the verification block are composed of a plurality of small blocks, the first representative pixel data, the second representative pixel data, the first feature amount, and the second feature amount being extracted for each of the small blocks, and wherein the first evaluated value and the second evaluated value are generated for each of the small blocks, the first evaluated value and the second evaluated values being cumulated as a first cumulative value and a second cumulative value, respectively, the base block and the verification block being partially collated corresponding to the first cumulative value and the second cumulative value.

Claim 54

The method or the apparatus as set forth in claim 22, 23, 28, or 29, wherein the first feature amount data, the second feature amount data, the third feature amount data, and the fourth feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block.

Claim 55

The method or the apparatus as set forth in claim 22, 23, 28, or 29, wherein the first feature amount data, the second feature amount data, the third feature amount data, and the fourth feature amount data are the average value and the standard deviation of a plurality of pixels of the block.

Claim 56

The method or the apparatus as set forth in claim 22, 23, 28, or 29, wherein the first feature amount data, the second feature amount data, the third feature amount data, and the fourth feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block, wherein said first evaluated value generating step or means is adapted for outputting nearly 0 as the first evaluated value when the maximum value and the minimum value of the first feature amount data are included in the range from the maximum value to the minimum value of the second feature amount data and for outputting the difference between the difference between the maximum values of the first feature amount data and the second feature amount data and the difference between the minimum values thereof as the first evaluated value when the maximum value and the minimum value of the first feature amount data are not included in the range from the maximum value to the minimum value of the second feature amount data, and wherein said second evaluated value generating step or means is adapted for outputting nearly 0 as the second evaluated value when the maximum value and the minimum value of the third feature amount data are included in the range from the maximum value to the minimum value of the fourth feature amount data and for outputting the difference between the difference between the maximum values of the third feature amount data and the fourth feature amount data and the difference between the minimum values thereof as the second evaluated value when the maximum value and the minimum value of the third feature amount data are not included in the range from the maximum value to the minimum value of the fourth feature amount data.

Claim 57

The method or the apparatus as set forth in claim 22, 23, 28, or 29, wherein the first feature amount data, the second feature amount data, the third feature amount data, and the fourth feature amount data are the average value and the standard deviation of a plurality of pixels of the base block and the modified verification block, wherein said first evaluated value generating step or means is adapted for generating the sum and the difference of the average value and the standard deviation, for outputting nearly 0 as the first evaluated value when the sum and the difference of the average value and the standard deviation of the first feature amount data are included in the range between the sum and the difference of the average value and the standard deviation of the second feature amount data, and for outputting the difference between the sums of the average values and the standard deviations of the first feature amount data and the second feature amount data and the differences thereof as the first evaluated value when the sum and the difference of the average value and the standard deviation of the first feature amount data are not included in the range between the sum and the difference of the average value and the standard deviation of the second feature amount data, and wherein said second evaluated value generating step or means is adapted for generating the sum and the difference of the average value and the standard deviation, for outputting nearly 0 as the second evaluated value when the sum and the difference of the average value and the standard deviation of the third feature amount data are included in the range between the sum and the difference of the average value and the standard deviation of the fourth feature amount data, and for outputting the difference between the sums of the average values and the standard deviations of the third feature amount data and the fourth feature amount data and the differences thereof as the second evaluated value when the sum and the difference of the average value and the standard deviation of the third feature amount data are not included in the range between the sum and the difference of the average value and the standard deviation of the fourth feature amount data.

Claim 58

The method or the apparatus as set forth in claim 22, 23, 28, or 29, wherein the base block and the verification block are composed of a plurality of small blocks, the first feature amount, the second feature amount, the third feature amount, and the fourth feature amount being extracted for each of the small blocks, the first evaluated value and the second evaluated value being generated for each of the small blocks, the first evaluated values and the second evaluated values being independently cumulated, the base block and the verification block being partially collated corresponding to the first cumulated value and the second cumulated value, respectively.

Claim 59

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting feature amount data that represents the range of the selected pixels of the verification sub-block; extracting second representative pixel data from the base sub-block; and comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 60

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting feature amount data that represents the range of the selected pixels of the verification sub-block; extracting second representative pixel data from the base sub-block; and comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 61

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting feature amount data that represents the range of the selected pixels of the verification sub-block; means for extracting second representative pixel data from the base sub-block; and means for comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 62

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting feature amount data that represents the range of the selected pixels of the verification sub-block; means for extracting second representative pixel data from the base sub-block; and means for comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 63

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of the verification sub-block necessary for a phase compensation; extracting first representative pixel data from the base sub-block; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 64

An image collating method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of the verification sub-block necessary for a phase compensation; extracting first representative pixel data from the base sub-block; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 65

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of the verification sub-block necessary for a phase compensation; means for extracting first representative pixel data from the base sub-block; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and means for comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 66

An image collating apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of the verification sub-block necessary for a phase compensation; means for extracting first representative pixel data from the base sub-block; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and means for comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 67

An image collating method for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising the steps of: segmenting a base block and a verification block into base sub-blocks and verification sub-blocks, respectively; selecting pixel data used for the collation on the assumption of rigid body; generating an evaluated value for each of the verification sub-blocks and the base sub-blocks, the evaluated value representing a collating degree of selected pixels of the verification sub-block and the base sub-block; and generating a totally evaluated value for the block corresponding to the evaluated value for the sub-block and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 68

An image collating method for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising the steps of: segmenting a base block and a verification block into base sub-blocks and verification sub-blocks, respectively; selecting pixel data used for the collation on the assumption of rigid body; generating an evaluated value for each of the verification sub-blocks and the base sub-blocks, the evaluated value representing a collating degree of selected pixels of the verification sub-block and the base sub-block; and generating a totally evaluated value for the block corresponding to the evaluated value for the sub-block and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 69

An image collating apparatus for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising: means for segmenting a base block and a verification block into base sub-blocks and verification sub-blocks, respectively; means for selecting pixel data used for the collation on the assumption of rigid body; means for generating an evaluated value for the verification sub-block and the base sub-block, the evaluated value representing a collating degree of selected pixels of the verification sub-block and the base sub-block; and means for generating a totally evaluated value for the block corresponding to the evaluated value for the sub-block and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 70

An image collating apparatus for compensating a phase difference that takes place due to simplification in partially collating first image data with second image data, comprising: means for segmenting a base block and a verification block into base sub-blocks and verification sub-blocks, respectively; means for selecting pixel data used for the collation on the assumption of rigid body; means for generating an evaluated value for the verification sub-block and the base sub-block, the evaluated value representing a collating degree of selected pixels of the verification sub-block and the base sub-block; and means for generating a totally evaluated value for the block corresponding to the evaluated value for the sub-block and partially collating the first image data with the second image data corresponding to the totally evaluated value.

Claim 71

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting feature amount data that represents the range of the selected pixels of the verification sub-block; extracting second representative pixel data from the base sub-block; and comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 72

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting feature amount data that represents the range of the selected pixels of the verification sub-block; extracting second representative pixel data from the base sub-block; and comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 73

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of each of the base sub-blocks and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting feature amount data that represents the range of the selected pixels of the verification sub-block; means for extracting second representative pixel data from the base sub-block; and means for comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 74

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks and segmenting each of the verification blocks into a plurality of verification sub-blocks; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of each of the base sub-blocks and second representing data that represents a pixel group at a position corresponding to each of the verification sub-blocks, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting feature amount data that represents the range of the selected pixels of the verification sub-block; means for extracting second representative pixel data from the base sub-block; and means for comparing the second representative pixel data with the feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 75

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of each of the verification sub-blocks necessary for a phase compensation; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 76

A motion vector detecting method, comprising the steps of: segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of each of the verification sub-blocks necessary for a phase compensation; extracting first representative pixel data from each of the base sub-blocks; generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 77

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of each of the verification sub-blocks necessary for a phase compensation; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and means for comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 78

A motion vector detecting apparatus, comprising: means for segmenting first image data into a plurality of base blocks and segmenting each of the base blocks into a plurality of base sub-blocks; means for segmenting second image data into a plurality of verification blocks, segmenting each of the verification blocks into a plurality of verification sub-blocks, and generating a modified verification sub-block that contains pixels in the vicinity of each of the verification sub-blocks necessary for a phase compensation; means for extracting first representative pixel data from each of the base sub-blocks; means for generating first representing data that represents first representative pixel data group of the base sub-block and second representing data that represents a pixel group at a position corresponding to the verification sub-block, comparing the first representing data with the second representing data, and selecting pixel data used for the collation; means for extracting first feature amount data from the base sub-block, the first feature amount data representing the range of all pixels thereof; means for extracting second feature amount data from the modified verification sub-block, the second feature amount data representing the range of all pixels of the modified verification sub-block, the all pixels being selected pixels; and means for comparing the first feature amount data with the second feature amount data, generating an evaluated value for the sub-block, generating a totally evaluated value corresponding to the evaluated value for the sub-block, and detecting a motion vector corresponding to the position of the verification block that most accords with the base block based on the totally evaluated value that is generated by moving the verification block in a predetermined search range.

Claim 79

The method or the apparatus as set forth in claim 41, 42, 47, or 48, wherein the first representative pixel data and the second representative pixel data are pixel data nearly at the center potions of the respective blocks.

Claim 80

The method or the apparatus as set forth in claim 41, 42, 47, or 48, wherein the feature amount data is two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block.

Claim 81

The method or the apparatus as set forth in claim 41, 42,47, or 48, wherein the feature amount data is the average value and the standard deviation of a plurality of pixels of the block.

Claim 82

The method or the apparatus as set forth in claim 41, 42, 47, or 48, wherein the feature amount data is two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as an evaluated value for the sub-block when the value the second representative pixel data is included in the range from the maximum value to the minimum value and for outputting the difference between the second representative pixel data and the maximum value and the difference between the second representative pixel data and the minimum value as evaluated values for the sub-block when the value of the second representative pixel data is not included in the range from the maximum value to the minimum value.

Claim 83

The method or the apparatus as set forth in claim 41, 42, 47, or 48, wherein the feature amount data is the average value and the standard deviation of a plurality of pixels of the block, and wherein said evaluated value generating step or means is adapted for outputting nearly 0 as an evaluated value for the sub-block when the first representative pixel data and the second representative pixel data are included in the range between the sum value and the difference value of the average value and the standard deviation and for outputting the difference between the second representative pixel data and the sum value and the difference between the second representative pixel data and the difference value as evaluated values for the sub-block when the first representative pixel data and the second representative pixel data are not included in the range between the sum value and the difference value of the average value and the standard deviation.

Claim 84

The method or the apparatus as set forth in claim 43, 44, 49, or 50, wherein the first feature amount data and second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block.

Claim 85

The method or the apparatus as set forth in claim 43, 44, 49, or 50, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the block.

Claim 86

The method or the apparatus as set forth in claim 43, 44,49, or 50, wherein the first feature amount data and second feature amount data are two out of the maximum value, the minimum value, and the dynamic range of a plurality of pixels of the block, and wherein said evaluated value generating means is adapted for outputting nearly 0 as an evaluated value for the sub-block when the maximum value and the minimum value of the first feature amount data are included in the range from the maximum value to the minimum value of the second feature amount data and for outputting the difference between the maximum values of the first feature amount data and the second feature amount data and the difference between the minimum values thereof as evaluated values for the sub-block when the maximum value and the minimum value of the first feature amount data are not included in the range from the maximum value to the minimum value of the second feature amount data.

Claim 87

The method or the apparatus as set forth in claim 43, 44, 49, or 50, wherein the first feature amount data and the second feature amount data are the average value and the standard deviation of a plurality of pixels of the base block and the verification block, and wherein said evaluated value generating step or means is adapted for generating the sum value and the differential value of the average value and the standard deviation, for outputting nearly 0 as an evaluated value for each of sub-blocks when the sum value and the difference value of the average value and the standard deviation are included in the range from the sum value to the difference value of the average value and the standard deviation of the second feature amount data, and for outputting the difference between the sum values of the average values and the standard deviations of the first feature amount data and the second feature amount data and the difference between the difference values of the average values and the standard deviations of the first feature amount data and the second feature amount data as evaluated values for the sub-block when the sum value and the difference value of the average value and the standard deviation of the first feature amount data are not included in the range from the sum value to the difference value of the average value and the standard deviation of the second feature amount data.

Claim 88

The method or the apparatus as set forth in claim 41, 42, 43, 44, 47, 48, 49, or 50, wherein the first representing data and the second representing data are average values.