Method and device for encoding object contour by using centroid

Claim 2

A centroid-based object contour encoding method comprising the steps of: (a) determining a centroid of an object; (b) determining an incrementing angle for sampling; (c) obtaining distances between the centroid and a plurality of boundary pixels on a contour of the object at angular positions spaced by the incrementing angle; and (d) encoding the distances and a coordinate of the centroid.

Claim 3

A centroid-based object contour encoding method as claimed in claim 1, wherein the centroid (C x ,C y ) determined in step (a) is defined by ##EQU4## where (i,j) is a coordinate of a point in a region in which the object is included, and k.sub.(i,j) equals "1" when (i,j) falls inside the object and equals "1" otherwise.

Claim 4

A centroid-based object contour encoding method as claimed in claim 1, wherein step (a) further comprises: determining a pseudo-centroid, when the centroid does not fall inside the object, the pseudo-centroid comprising a central point among a plurality of points inside the object and positioned along a line, wherein the line passes through (i) the centroid, and (ii) a geometrical center of a minimum rectangle encompassing the object.

Claim 5

A centroid-based object contour encoding method as claimed in claim 1, wherein the incrementing angle is adjusted according to object size and contour complexity.

Claim 6

A centroid-based object contour encoding method as claimed in claim 1, wherein step (c) further comprises choosing a farthest pixel as a boundary pixel, when there are a plurality of pixels on the contour at a single angle.

Claim 7

A centroid-based object-contour encoding method as claimed in claim 1, wherein the distances are obtained sequentially at said angular positions.

Claim 8

A centroid-based object contour encoding method comprising the steps of: (a) determining a centroid of an object; (b) determining an incrementing angle for sampling; (c) obtaining distances between the centroid and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle; (d) repeating steps (a) through (c) for a plurality of sequential frames; (e) encoding the difference between (i) the distances of step (c) in a frame of the plurality of sequential frames, and (ii) the distances of step (c) in a previous frame of the plurality of sequential frames; and (f) encoding the difference between (i) a coordinate of the centroid of the object in the frame and (ii) a coordinate of the centroid of the object in the previous frame.

Claim 9

A centroid-based object contour encoding method comprising the steps of: (a) determining a centroid of an object; (b) determining an incrementing angle for sampling; (c) obtaining distances between the centroid and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle; (d) repeating steps (a) through (c) for a plurality of sequential frames; (e) encoding the difference between (i) the distances of step (c) in a frame of the plurality of sequential frames, and (ii) the distances of step (c) in a previous frame of the plurality of sequential frames; and (f) encoding the difference between (i) a coordinate of the centroid of the object in the frame and (ii) a coordinate of the centroid of the object in the previous frame.

Claim 10

A centroid-based object contour encoding method as claimed in claim 7, wherein the centroid (C x ,C y ) determined in step (a) is defined by ##EQU5## where (i,j) is a coordinate of a point in a region in which the object is included, and k.sub.(i,j) equals "1" when (i,j) falls inside the object and equals "0" otherwise.

Claim 11

A centroid-based object contour encoding method as is claimed in claim 7, wherein step (a) further comprises: determining a pseudo-centroid, when the centroid does not fall inside the object, the pseudo-centroid comprising a central point among a plurality of points inside the object and positioned along a line, wherein the line passes through (i) the centroid, and (ii) a geometrical center of a minimum rectangle encompassing the object.

Claim 12

A centroid-based object contour encoding method as claimed in claim 7, wherein the incrementing angle is adjusted according to object size and contour complexity.

Claim 13

A centroid-based object contour encoding method as claimed in claim 7, wherein step (c) further comprises choosing a farthest pixel as a boundary pixel, when there are a plurality of pixels on the contour at a single angle.

Claim 14

A centroid-based object contour encoding method as claimed in claim 7, wherein the distances are obtained sequentially at said angular positions.

Claim 15

A centroid-based object contour encoding device comprising: centroid determining means for determining a centroid inside an object; incrementing angle determining means for determining an incrementing angle for sampling; calculating means for obtaining distances between the centroid and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle; and encoding means for encoding the distances and a coordinate of the centroid.

Claim 16

A centroid-based object contour encoding device comprising: centroid determining means for determining a centroid inside an object; incrementing angle determining means for determining an incrementing angle for sampling; calculating means for obtaining distances between the centroid and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle; and encoding means for encoding the distances and a coordinate of the centroid.

Claim 17

A centroid-based object contour encoding device as claimed in claim 13, wherein said centroid determining means determines the centroid (C x ,C y ) defined by ##EQU6## where (i,j) is a coordinate of a point in a region in which the object is included, and k.sub.(i,j) equals "1" when (i,j) falls inside the object and equals "0" otherwise.

Claim 18

A centroid-based object contour encoding device as claimed in claim 13, wherein said centroid determining means determines a pseudo-centroid, when the centroid does not fall inside the object, the pseudo-centroid comprising a central point among points which are inside the object and are positioned on a line, wherein the line passes through (i) the centroid, and (ii) a geometrical center of a minimum rectangle encompassing the object.

Claim 19

A centroid-based object contour encoding device as claimed in claim 13, wherein said incrementing angle determining means adjusts the incrementing angle according to object size and contour complexity.

Claim 20

A centroid-based object contour encoding device as claimed in claim 13, wherein said calculating means chooses a distance to a farthest pixel, when there are a plurality of boundary pixels on the contour at a single angle.

Claim 21

A centroid-based object contour encoding device as claimed in claim 13, wherein said calculating means obtains the distances sequentially at said angular positions.

Claim 22

A centroid-based object contour encoding device comprising: distance extracting means for obtaining distances between a centroid of an object and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle, wherein said distance extracting means also outputs a plurality of sequential frames; and means for encoding a first difference between (i) the distances of the centroid and the plurality of boundary pixels on a contour of the object in a frame of the plurality of frames and, (ii) the distances of the centroid and the plurality of boundary pixels on a contour of the object in a previous frame of the plurality of frames, and wherein the means for encoding further encodes a second difference between (i) a coordinate of the centroid of the object in the frame and (ii) a coordinate of the centroid of the object in the previous frame.

Claim 23

A centroid-based object contour encoding device comprising: distance extracting means for obtaining distances between a centroid of an object and a plurality of boundary pixels on a contour of the object at angular positions spaced by said incrementing angle, wherein said distance extracting means also outputs a plurality of sequential frames; and means for encoding a first difference between (i) the distances of the centroid and the plurality of boundary pixels on a contour of the object in a frame of the plurality of frames and, (ii) the distances of the centroid and the plurality of boundary pixels on a contour of the object in a previous frame of the plurality of frames, and wherein the means for encoding further encodes a second difference between (i) a coordinate of the centroid of the object in the frame and (ii) a coordinate of the centroid of the object in the previous frame.

Claim 24

A centroid-based object contour encoding device according to claim 19, wherein said encoding means further comprises: a discrete cosine transform (DCT) circuit for performing a DCT with respect to the first distance and second distance, and for outputting a result of the DCT as DCT data; a quantizer for quantizing the DCT data and for outputting quantized DCT data; a variable-length coder for variable length coding the quantized DCT data.

Claim 25

A centroid-based object contour encoding device as claimed in claim 19, wherein said distance extracting means determines the centroid (C x ,C y ) defined by ##EQU7## where (i,j) is a coordinate of a point in a region in which the object is included, and k.sub.(i,j) equals "1" when (i,j) falls inside the object and equals "0" otherwise.

Claim 26

A centroid-based object contour encoding device as claimed in claim 19, wherein said centroid determining means determines a pseudo-centroid, when the centroid does not fall inside the object, the pseudo-centroid comprising a central point among points which are inside the object and are positioned on a line, wherein the line passes through (i) the centroid, and (ii) a geometrical center of a minimum rectangle encompassing the object.

Claim 27

A centroid-based object contour encoding device as claimed in claim 19, wherein said distance extracting means adjusts the incrementing angle according to object size and contour complexity.

Claim 28

A centroid-based object contour encoding device as claimed in claim 19, wherein said distance extracting means chooses a distance to a farthest pixel, when there are a plurality of boundary pixels on the contour at a single angle.

Claim 29

A centroid-based object contour encoding device as claimed in claim 19, wherein said distance extracting means obtains the distances sequentially at said angular positions.