Method of compressing image signals by vector quantization

Claim 1

A method of compressing image signals by vector quantization, which comprises the steps of: i) dividing two-dimensional image signals into blocks B1 each of which is composed of signals at M×N number of picture elements adjacent to one another, ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks B1 from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and iii) encoding a number representing the selected vector to correspond to each of the blocks B1, wherein the improvement comprises the steps of: a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, b) dividing group signals of said numbers selected for the respective blocks B1 into blocks B2, each of which is composed of P×Q number of samples, by putting together said group signals at the blocks B1 adjacent to one another, c) carrying out orthogonal transformation of said group signals in each of said blocks B2, and d) encoding the transformed signals obtained by said orthogonal transformation respectively with an intrinsic code length.

Claim 2

A method of compressing image signals by vector quantization, which comprises the steps of: i) dividing two-dimensional image signals into blocks B1 each of which is composed of signals at M×N number of picture elements adjacent to one another, ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks B1 from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and iii) encoding a number representing the selected vector to correspond to each of the blocks B1, wherein the improvement comprises the steps of: a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, b) dividing group signals of said numbers selected for the respective blocks B1 into blocks B2, each of which is composed of P×Q number of samples, by putting together said group signals at the blocks B1 adjacent to one another, c) carrying out orthogonal transformation of said group signals in each of said blocks B2, and d) encoding the transformed signals obtained by said orthogonal transformation respectively with an intrinsic code length.

Claim 5

A method of compressing image signals by vector quantization, which comprises the steps of: i) dividing two-dimensional image signals into blocks each of which is composed of signals at M×N number of picture elements adjacent to one another, ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and iii) encoding a number representing the selected vector to correspond to each of the blocks, wherein the improvement comprises the steps of: a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, and b) carrying out prediction encoding of group signals of said numbers selected for the respective blocks.

Claim 6

A method of compressing image signals by vector quantization, which comprises the steps of: i) dividing two-dimensional image signals into blocks each of which is composed of signals at M×N number of picture elements adjacent to one another, ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and iii) encoding a number representing the selected vector to correspond to each of the blocks, wherein the improvement comprises the steps of: a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, and b) carrying out prediction encoding of group signals of said numbers selected for the respective blocks.

Claim 7

A method of compressing image signals by vector quantization, which comprises the steps of: (i) dividing two-dimensional image signals into blocks each of which is composed of signals at M×N number of picture elements adjacent to one another. (ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and (iii) encoding a number representing a selected vector to correspond to each of the blocks, wherein the improvement comprises the steps of: (a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, and (b) carrying out prediction encoding of group signals of said numbers selected for the respective blocks, wherein said prediction encoding comprises previous value prediction plus Huffman encoding processing.

Claim 8

A method of compressing image signals by vector quantization, which comprises the steps of: (i) dividing two-dimensional image signals into blocks each of which is composed of signals at M×N number of picture elements adjacent to one another. (ii) selecting a vector that corresponds with the minimum distortion to said M×N number of image signals in each of the blocks from a code book comprising a plurality of vectors different from one another and prepared in advance by defining M×N number of vector elements, and (iii) encoding a number representing a selected vector to correspond to each of the blocks, wherein the improvement comprises the steps of: (a) defining said number representing said selected vector so that said number is larger or smaller as the distortion of said vector in said code book with respect to a predetermined vector is larger, and (b) carrying out prediction encoding of group signals of said numbers selected for the respective blocks, wherein said prediction encoding comprises previous value prediction plus Huffman encoding processing.