Image compression coding system

Claim 2

An image coding apparatus comprising: image dividing means for dividing an image signal of a predetermined unit into a plurality of blocks; amount of-data calculating means, coupled to said image dividing means, for computing an amount of data representing fineness of a picture and included in each of said blocks; bit-assignment determining means, coupled to said amount-of-data calculating means, for determining a number of coding bits to be used to assign a predetermined total number of coding bits allowable for encoding said image signal of said predetermined unit to each of said blocks in accordance with said amount of data of said each blocks; quantizing means for quantizing said image signal for each of said blocks and outputting a quantization block signal; and encoding means, coupled to said quantizing means, for encoding said quantization block signal in accordance with said determined number of coding bits.

Claim 3

An apparatus according to claim 1, wherein said image signal is an image signal of one of one field and one frame constituted by a number of pixel signals, and said amount-of-data calculating means calculates variance and standard deviation of pixel values in each of said blocks of said image signal for each of said blocks to thereby compute said amount of data.

Claim 4

An apparatus according to claim 1, wherein said image signal is an image signal of one of one field and one frame having a number of pixel values, and, for each of said blocks, said amount-of-data calculating means calculates a difference between maximum and minimum values of the pixel values in said each block of said image signal to thereby compute said amount of data.

Claim 5

An apparatus according to claim 3, wherein said amount-of-data calculating means comprises: comparing means for sequentially comparing values of pixels contained in each of said blocks with each other to arrange said pixel values in an ascending or descending order; selecting means for selecting a maximum pixel value and a minimum pixel value when all pixels in each of said blocks are arranged; and means for calculating a difference between said maximum and minimum pixel values to thereby determine said amount of data from said difference.

Claim 6

An apparatus according to claim 1, wherein said bit-assignment determining means determines the number of coding bits to be assigned to each of said blocks from a ratio between a total amount of data attained by accumulation of an amount of data of each of said blocks for one of one field and one frame and said amount of data of each of said blocks and said total number of coding bits.

Claim 7

An apparatus according to claim 1, wherein said encoding means subjects said quantization block signal to Huffman coding.

Claim 8

An apparatus according to claim 1, wherein based on said number of coding bits determined for each of said blocks by said bit-assignment determining means, said encoding means subjects said quantization block signal to Huffman coding within said number of bits and executes such a coding process for all of said blocks of one of one field and one frame, thereby ensuring execution of a coding process of the entire one of one field and one frame within said total number of coding bits for one of one field and one frame.

Claim 9

An apparatus according to claim 1, wherein said encoding means includes means for computing a difference between said number of coding bits for each of said blocks and a number of coding bits actually needed to encode said each block, adding remaining bits to a number of coding bits for an ext block to be encoded, and encoding said next block to be encoded using a resultant number as a new number of coding bits.

Claim 10

An apparatus according to claim 1, wherein said amount-of-data calculating means subjects an image signal of each of said blocks to discrete cosine transformation, and computes an amount of each of said blocks from one of an entire transformation coefficient signal resulting from said discrete cosine transformation and part thereof.

Claim 11

An apparatus according to claim 1, wherein said amount-of-data calculating means comprises: a predictor for outputting a prediction signal; means for computing a difference between said prediction signal from said predictor and an image signal of each of said blocks from said image dividing means; means for selecting a specific block to be actually encoded from a relationship between said difference and a predetermined threshold value; and means of computing an amount of data only from said specific block.

Claim 12

An apparatus according to claim 1, wherein said quantizing means comprises vector quantizing means for selecting one of several kinds of predetermined code books and using said selected code book for vector quantization of said image signal for each of said blocks in accordance with said number of bits determined by said bit-assignment determining means.

Claim 13

An apparatus according to claim 1, wherein said amount-of-data calculating means comprises: means for computing amounts of data of all of said blocks in one of one field and one frame and detecting amounts of data indicating maximum and minimum values among said computed amounts of data; means for separating a range between said maximum amount of data and said minimum amount of data, into a predetermined number of classes; classifying means for classifying said blocks into associated ones of said predetermined number of classes by checking said amount of data for each of said blocks; and means for outputting a predetermined, fixed amount of data corresponding to each of said classes to said bit-assignment determining means as a new amount of data for said blocks classified into said each class.

Claim 14

An image coding apparatus comprising: image dividing means for dividing an image signal of a predetermined unit into a plurality of blocks; amount-of-data calculating means for computing an amount of data representing fineness of a picture and included in each of said blocks; bit-assignment determining means for determining a number of coding bits to be used to assign a predetermined total number of coding bits allowable for encoding said image signal of said predetermined unit to each of said blocks in accordance with said amount of data of said each blocks; quantizing means for quantizing said image signal for each of said blocks and outputting a quantization block signal; and encoding means for encoding said quantization block signal within said determined number of coding bits for each of said blocks, thereby ensuring execution of a coding process within said total number of coding bits.

Claim 15

An image coding apparatus comprising: image dividing means for dividing an image signal of a predetermined unit into a plurality of blocks; amount-of-data calculating means for computing an amount of data representing fineness of a picture and included in each of said blocks; bit-assignment determining means for determining a number of coding bits to be used to assign a predetermined total number of coding bits allowable for encoding said image signal of said predetermined unit to each of said blocks in accordance with said amount of data of said each blocks; quantizing means for quantizing said image signal for each of said blocks and outputting a quantization block signal; and encoding means for encoding said quantization block signal within said determined number of coding bits for each of said blocks, thereby ensuring execution of a coding process within said total number of coding bits.

Claim 16

An image coding apparatus comprising: image dividing means for diving an image signal of one of one frame and one field into a plurality of blocks, said image signal including DC and AC components; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of data contained in each of said blocks; first quantizing means, coupled to said image dividing means, for quantizing said DC components of said image signal for each of said blocks and outputting a DC component quantization signal; coding bit adding means for accumulating numbers of coding bits required to encode said quantization signal for all of said blocks within one of one field and one frame; bit-assignment determining means, coupled to said amount-of-data calculating means, for computing an umber of coding bits left by subtracting the number of coding bits required to encode the DC components for all of said blocks in said one of one field and one frame from the total number of coding bits determined in advance to encode said image signal of said one of one field and one frame, and for determining the number of coding bits to be assigned to coding of the AC components of each of said blocks in accordance with said computed number of bits left and an amount of data of said each block; second quantizing means, coupled to said image dividing means, for quantizing said AC components included in said image signal, and outputting an AC component quantization signal; and encoding means, coupled to said bit-assignment determining means and said second quantizing means, for encoding said AC component quantization signal in accordance with said determined number of bits.

Claim 17

An image coding apparatus comprising: image dividing means for diving an image signal of one of one frame and one field into a plurality of blocks, said image signal including DC and AC components; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of data contained in each of said blocks; first quantizing means, coupled to said image dividing means, for quantizing said DC components of said image signal for each of said blocks and outputting a DC component quantization signal; coding bit adding means for accumulating numbers of coding bits required to encode said quantization signal for all of said blocks within one of one field and one frame; bit-assignment determining means, coupled to said amount-of-data calculating means, for computing an umber of coding bits left by subtracting the number of coding bits required to encode the DC components for all of said blocks in said one of one field and one frame from the total number of coding bits determined in advance to encode said image signal of said one of one field and one frame, and for determining the number of coding bits to be assigned to coding of the AC components of each of said blocks in accordance with said computed number of bits left and an amount of data of said each block; second quantizing means, coupled to said image dividing means, for quantizing said AC components included in said image signal, and outputting an AC component quantization signal; and encoding means, coupled to said bit-assignment determining means and said second quantizing means, for encoding said AC component quantization signal in accordance with said determined number of bits.

Claim 18

An apparatus according to claim 14, wherein said encoding means subjects said quantization signals to Huffman coding.

Claim 19

An apparatus according to claim 14, wherein said second quantizing means comprises means for subjecting an image signal of each of said blocks to discrete cosine transformation and quantizing a transformation coefficient signal resulting from said discrete cosine transformation.

Claim 20

An apparatus according to claim 14, wherein said encoding means transfers the AC components of each of said blocks from those AC components which are significant from a view point of value evaluation or subjective evaluation or stops transferring of said AC components from those AC components which are insignificant, and, when a number of coding bits assigned in advance to each of said blocks is reached, said encoding means stops encoding of an associated block, and executes this encoding process for all of said blocks within one frame, thereby ensuring encoding of said image signal within a total number of coding bits determined in advance for one frame.

Claim 21

An image coding apparatus according to claim 14, wherein said bit-assignment determining means subtracts one of first, second, and third numbers of bits, the first number of bits being required for encoding said DC components of each of said blocks, the second number of bits being needed to encode a signal that is required for reproduction at a receiver side, and the third number of bits being needed to encode a signal attained by adding a signal that is required for reproduction at said receiver side and part of a different signal, from the total number of coding bits assigned to one frame, and distributes remaining coding bits as coding bits for said AC components of each of said blocks in accordance with said amount of data of said each block.

Claim 22

An apparatus according to claim 14, wherein said apparatus is designed to execute quantization and encoding of said DC components and a process for computing an amount of data in parallel simultaneously.

Claim 23

A color image coding apparatus comprising: converting means for converting a color image signal into predetermined image signal components; amount-of-data calculating means for computing an amount of data of each of said predetermined image signal components acquired by said converting means; bit-assignment determining means for distributing a total number of coding bits assigned in advance to said color image signal to each of said image signal components as a number of coding bits used to encode said each image signal component in accordance with said amount of data of said each image signal component computed by said amount-of-data calculating means; and means for encoding each of said image signal components using said number of coding bits distributed by said bit-assignment determining means.

Claim 24

A color image coding apparatus comprising: converting means for converting a color image signal into predetermined image signal components; amount-of-data calculating means for computing an amount of data of each of said predetermined image signal components acquired by said converting means; bit-assignment determining means for distributing a total number of coding bits assigned in advance to said color image signal to each of said image signal components as a number of coding bits used to encode said each image signal component in accordance with said amount of data of said each image signal component computed by said amount-of-data calculating means; and means for encoding each of said image signal components using said number of coding bits distributed by said bit-assignment determining means.

Claim 25

An apparatus according to claim 20, wherein said amount-of-data determining means has means for performing a weighting process in consideration of a visual characteristic at a time said amount of data of each of said image signal components is acquired and setting an amount of data after said weighting process as a final amount of data of said each image signal component.

Claim 26

An apparatus according to claim 20, wherein said amount-of-data calculating means divides each of said image signal components into a plurality of blocks, computes an amount of data for each of said blocks, repeats this computation for all of the blocks of one frame of said each image signal component, then averaging resultant amounts of data for said each image signal component to thereby determine an amount of data of said each image signal component.

Claim 27

An apparatus according to claim 20, wherein when said image signal components have Y and color difference signal components, Y is given a highest weight, and when said image signal components have R, G and B signal components, G is given a highest weight.

Claim 28

An image coding apparatus comprising: image dividing means for dividing an image signal of a predetermined unit into a plurality of blocks to output a data signal for each of said blocks; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of power of a middle frequency component included in each of said blocks; bit-assignment determining means, coupled to said amount-of-data calculating means, for determining the number of coding bits to be assigned to each of said blocks in accordance with said amount of power of said each block; quantizing means for quantizing and said data signal for each of said blocks and outputting a quantization block signal; and encoding means, coupled to said quantizing means, for encoding said quantization block signal in accordance with said determined number of bits.

Claim 29

An image coding apparatus comprising: image dividing means for dividing an image signal of a predetermined unit into a plurality of blocks to output a data signal for each of said blocks; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of power of a middle frequency component included in each of said blocks; bit-assignment determining means, coupled to said amount-of-data calculating means, for determining the number of coding bits to be assigned to each of said blocks in accordance with said amount of power of said each block; quantizing means for quantizing and said data signal for each of said blocks and outputting a quantization block signal; and encoding means, coupled to said quantizing means, for encoding said quantization block signal in accordance with said determined number of bits.

Claim 30

An image coding apparatus comprising: image dividing means for dividing an image signal of one of one frame and one field into a plurality of blocks; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of data contained in said each of said blocks; first quantizing means, coupled to said image dividing means, for quantizing, for each of said blocks, a specific signal required for decoding of said each block and outputting a quantization signal; coding bit adding means for accumulating numbers of coding bits required to encode said quantization signal for all of said blocks within one of one field and one frame; bit-assignment determining means, coupled to said amount-of-data calculating means, for computing a number of coding bits left by subtracting the number of coding bits required to encode said specific signal for all of said blocks in said one of one field an one frame from the total number of coding bits determined in advance to encode said image signal of said done of one field and one frame, and for determining the number of coding bits to be assigned to coding of AC components of each of said blocks in accordance with said computed number of bits left and an amount of data of said each block; second quantizing means, coupled to said image dividing means, for quantizing said AC components included in said image signal, and outputting an AC component quantization signal; and encoding means, coupled to said bit-assignment determining means and said second quantizing means, for encoding said AC component quantization signal in accordance with said determined number of bits.

Claim 31

An image coding apparatus comprising: image dividing means for dividing an image signal of one of one frame and one field into a plurality of blocks; amount-of-data calculating means, coupled to said image dividing means, for computing an amount of data contained in said each of said blocks; first quantizing means, coupled to said image dividing means, for quantizing, for each of said blocks, a specific signal required for decoding of said each block and outputting a quantization signal; coding bit adding means for accumulating numbers of coding bits required to encode said quantization signal for all of said blocks within one of one field and one frame; bit-assignment determining means, coupled to said amount-of-data calculating means, for computing a number of coding bits left by subtracting the number of coding bits required to encode said specific signal for all of said blocks in said one of one field an one frame from the total number of coding bits determined in advance to encode said image signal of said done of one field and one frame, and for determining the number of coding bits to be assigned to coding of AC components of each of said blocks in accordance with said computed number of bits left and an amount of data of said each block; second quantizing means, coupled to said image dividing means, for quantizing said AC components included in said image signal, and outputting an AC component quantization signal; and encoding means, coupled to said bit-assignment determining means and said second quantizing means, for encoding said AC component quantization signal in accordance with said determined number of bits.