High-efficiency coding apparatus for compressing a digital video signal while controlling the coding bit rate of the compressed digital data so as to keep it constant

Claim 1

An apparatus for high-efficiency coding of a digital video image, comprising: a means for dividing a digital video signal into a plurality of blocks each composed of a plurality of pixel values; a means for extracting a minimum value and a maximum value of the pixel values in each of the plurality of blocks; a means for subtracting the minimum value from each of the pixel values in each block to obtain difference values; a means for calculating a dynamic range of each block from the maximum and minimum values; a first memory means for storing the difference values of each block; a second memory means for storing the dynamic range of each block; a third memory means for storing the minimum value of each block; a means for detecting a frequency distribution of dynamic ranges of the plurality of blocks; a means having N kinds of first characteristic tables each having as a parameter a decoded image quality visually observed which are equal in loss of picture quality and showing a relationship between the dynamic range and distortion caused by encoding, and a second characteristic table having a quantization width as a parameter and showing a relationship between the dynamic range and distortion caused by encoding; a means for calculating a total amount of coded data from the detected frequency distribution of dynamic ranges and quantization widths determined by a combination of each of the N kinds of first characteristic tables and the second characteristic table to obtain N kinds of total amount values corresponding to the N kinds of characteristic tables, respectively, and for comparing each of the N kinds of total amount values with a predetermined value, and for selecting one of the N kinds of total amount values which is smaller than the predetermined value and closest to the predetermined value, and outputting a table number indicating one of the N kinds of first characteristic tables which gives the selected total amount value; a means for encoding the difference values of each block stored in the first memory means by using the dynamic range stored in the second memory and the quantization widths determined by a combination of the selected first characteristic table indicated by the table number and the second characteristic table to obtain coded data; and a means for outputting the minimum value stored in the third memory, the dynamic range stored in the second memory and the coded data together with the table number.

Claim 2

An apparatus for high-efficiency coding of a digital video image, comprising: a means for dividing a digital video signal into a plurality of blocks each composed of a plurality of pixel values; a means for extracting a minimum value and a maximum value of the pixel values in each of the plurality of blocks; a means for subtracting the minimum value from each of the pixel values in each block to obtain difference values; a means for calculating a dynamic range of each block from the maximum and minimum values; a first memory means for storing the difference values of each block; a second memory means for storing the dynamic range of each block; a third memory means for storing the minimum value of each block; a means for detecting a frequency distribution of dynamic ranges of the plurality of blocks; a means having N kinds of first characteristic tables each having as a parameter a decoded image quality visually observed which are equal in loss of picture quality and showing a relationship between the dynamic range and distortion caused by encoding, and a second characteristic table having a quantization width as a parameter and showing a relationship between the dynamic range and distortion caused by encoding; a means for calculating a total amount of coded data from the detected frequency distribution of dynamic ranges and quantization widths determined by a combination of each of the N kinds of first characteristic tables and the second characteristic table to obtain N kinds of total amount values corresponding to the N kinds of characteristic tables, respectively, and for comparing each of the N kinds of total amount values with a predetermined value, and for selecting one of the N kinds of total amount values which is smaller than the predetermined value and closest to the predetermined value, and outputting a table number indicating one of the N kinds of first characteristic tables which gives the selected total amount value; a means for encoding the difference values of each block stored in the first memory means by using the dynamic range stored in the second memory and the quantization widths determined by a combination of the selected first characteristic table indicated by the table number and the second characteristic table to obtain coded data; and a means for outputting the minimum value stored in the third memory, the dynamic range stored in the second memory and the coded data together with the table number.

Claim 3

An apparatus for high-efficiency coding of a digital video image, comprising: a means for dividing a digital video signal in each frame into a plurality of blocks each composed of a plurality of pixel values; a means for extracting a minimum value and a maximum value of the pixel values in each of the plurality of blocks; a means for subtracting the minimum value from each of the pixel values in each block to obtain difference values; a means for calculating a dynamic range of each block from the maximum and minimum values; a first memory means for storing the difference values of each block; a second memory means for storing the dynamic range of each block; a third memory means for storing the minimum value of each block; a means for detecting a frequency distribution of dynamic ranges in a frame; means having N kinds of first characteristic tables each having as a parameter a decoded image quality visually observed which are equal in loss of picture quality and showing a relationship between the dynamic range and distortion caused by encoding, and a second characteristic table having a quantization width as a parameter and showing a relationship between the dynamic range and the distortion caused by encoding; a means for calculating a total amount of coded data from the detected frequency distribution of dynamic ranges and quantization widths determined by a combination of each of the N kinds of first characteristic tables and the second characteristic table to obtain N kinds of total amount values corresponding to the N kinds of characteristic tables, respectively, and for comparing each of the N kinds of total amount values with a predetermined value, and for selecting one of the N kinds of total amount values which is smaller than the predetermined value and closest to the predetermined value, and outputting a table number indicating one of the N kinds of first characteristic tables which gives the selected total amount value; a means for encoding the difference values of each block stored in the first memory means by using the dynamic range stored in the second memory and the quantization widths determined by a combination of the selected first characteristic table indicated by the table number and the second characteristic table to obtain coded data; and a means for outputting the minimum value stored in the third memory, the dynamic range stored in the second memory and the coded data together with the table number.

Claim 4

An apparatus for high-efficiency coding of a digital video image, comprising: a means for dividing a digital video signal in each frame into a plurality of blocks each composed of a plurality of pixel values; a means for extracting a minimum value and a maximum value of the pixel values in each of the plurality of blocks; a means for subtracting the minimum value from each of the pixel values in each block to obtain difference values; a means for calculating a dynamic range of each block from the maximum and minimum values; a first memory means for storing the difference values of each block; a second memory means for storing the dynamic range of each block; a third memory means for storing the minimum value of each block; a means for detecting a frequency distribution of dynamic ranges in a frame; means having N kinds of first characteristic tables each having as a parameter a decoded image quality visually observed which are equal in loss of picture quality and showing a relationship between the dynamic range and distortion caused by encoding, and a second characteristic table having a quantization width as a parameter and showing a relationship between the dynamic range and the distortion caused by encoding; a means for calculating a total amount of coded data from the detected frequency distribution of dynamic ranges and quantization widths determined by a combination of each of the N kinds of first characteristic tables and the second characteristic table to obtain N kinds of total amount values corresponding to the N kinds of characteristic tables, respectively, and for comparing each of the N kinds of total amount values with a predetermined value, and for selecting one of the N kinds of total amount values which is smaller than the predetermined value and closest to the predetermined value, and outputting a table number indicating one of the N kinds of first characteristic tables which gives the selected total amount value; a means for encoding the difference values of each block stored in the first memory means by using the dynamic range stored in the second memory and the quantization widths determined by a combination of the selected first characteristic table indicated by the table number and the second characteristic table to obtain coded data; and a means for outputting the minimum value stored in the third memory, the dynamic range stored in the second memory and the coded data together with the table number.