Video encoder using adjacent pixel difference for quantizer control

Claim 3

The method of claim 1, wherein the segment is one field of the digital video signal.

Claim 4

The method of claim 1, wherein the segment is one frame of the digital video signal.

Claim 5

The method of claim 1, wherein the segment is one track of the digital video signal.

Claim 6

The method of claim 1, wherein the segment is less than one track of the digital video signal.

Claim 7

A compressive encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data, comprising: a block converter for converting said segment into rectangular image area blocks of pixel values; a quantization controller coupled to said block converter, for separately calculating, for each block, a sum of absolute values of differences between horizontally and vertically adjacent pixel values, for determining, from each separately calculated sum, a standard quantization level, for adding together each of the separately determined predicted amounts for all blocks in said segment to form a total predicted amount and for selecting a segment quantization level that will produce an amount of encoded data substantially equal to the nominal amount of encoded data based upon the total predicted amount, thereby forming a selected segment quantization level; an orthogonal transform processor coupled to said quantization controller, for executing an orthogonal transform on each of said blocks of pixel values, thereby generating coefficients; a quantizer coupled to said orthogonal transform processor, for quantizing said generated coefficients according to said selected segment quantization level to generate quantized coefficients; a variable-length encoder coupled to said quantizer, for encoding said quantized coefficients to generate encoded data; and a buffer memory coupled to said variable-length encoder, for storing said generated encoded data.

Claim 8

A compressive encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data, comprising: a block converter for converting said segment into rectangular image area blocks of pixel values; a quantization controller coupled to said block converter, for separately calculating, for each block, a sum of absolute values of differences between horizontally and vertically adjacent pixel values, for determining, from each separately calculated sum, a standard quantization level, for adding together each of the separately determined predicted amounts for all blocks in said segment to form a total predicted amount and for selecting a segment quantization level that will produce an amount of encoded data substantially equal to the nominal amount of encoded data based upon the total predicted amount, thereby forming a selected segment quantization level; an orthogonal transform processor coupled to said quantization controller, for executing an orthogonal transform on each of said blocks of pixel values, thereby generating coefficients; a quantizer coupled to said orthogonal transform processor, for quantizing said generated coefficients according to said selected segment quantization level to generate quantized coefficients; a variable-length encoder coupled to said quantizer, for encoding said quantized coefficients to generate encoded data; and a buffer memory coupled to said variable-length encoder, for storing said generated encoded data.

Claim 9

The encoder of claim 6, wherein said segment comprises a luminance component, a first chrominance component, and a second chrominance component, wherein said block converter converts each of said luminance component, said first chrominance component and said second chrominance component separately, thereby generating luminance blocks, first chrominance blocks, and second chrominance blocks, and wherein said quantization controller comprises: a segment memory for storing the blocks of pixel values generated by said block converter and for providing said blocks of pixel values to said orthogonal transform processor; a luminance signal processor for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values in a luminance block; a first chrominance signal processor for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values in a first chrominance block; a second chrominance signal processor for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values in a second chrominance block; a luminance block data predictor coupled to said luminance signal processor for determining, from the sum calculated by said luminance signal processor, a predicted amount of encoded data that would be generated from a luminance block at a standard quantization level; a first chrominance block data predictor coupled to said first chrominance signal processor for determining, from the sum calculated by said first chrominance signal processor, a predicted amount of encoded data that would be generated from a first chrominance block at said standard quantization level; a second chrominance block data predictor coupled to said second chrominance signal processor for determining, from the sum calculated by said second chrominance signal processor, a predicted amount of encoded data that would be generated from a second chrominance block at said standard quantization level; a first accumulator coupled to said luminance block data predictor, for adding the predicted amounts of encoded data for different luminance blocks to determine a total predicted amount of luminance data; a second accumulator coupled to said first chrominance block data predictor, for adding the predicted amounts of encoded data for different first chrominance blocks to determine a total predicted amount of first chrominance data; and a third accumulator coupled to said second chrominance block data predictor, for adding the predicted amounts of encoded data for different second chrominance blocks to determine a total predicted amount of second chrominance data.

Claim 10

The encoder of claim 7, wherein said quantization controller further comprises: a quantization level selector coupled to said first accumulator, said second accumulator, and said third accumulator, said quantization level selector adding said total predicted amount of luminance data, said total predicted amount of first chrominance data, and said total predicted amount of second chrominance data to determine a total predicted amount of encoded data for said segment at said standard quantization level, and selecting said segment quantization level based upon the total predicted amount of encoded data; and a control table coupled to said quantization level selector, for storing quantization levels predicted to yield the nominal amount of encoded data corresponding to a plurality of different total predicted amounts at said standard quantization level.

Claim 11

The encoder of claim 7, wherein said first accumulator, said second accumulator, and said third accumulator add predicted amounts of encoded data for all luminance blocks, all first chrominance blocks, and all second chrominance blocks in said segment, respectfully.

Claim 12

The encoder of claim 7, wherein said first accumulator, said second accumulator, and said third accumulator add predicted amounts of encoded data for a subset of all luminance blocks, a subset of all first chrominance blocks, and a subset of all second chrominance blocks in said segment, respectively.

Claim 13

The encoder of claim 6, further comprising: a motion estimator coupled to said block converter, for generating motion vectors and error values of blocks relative to previous image data; and a switch coupled between said block converter and said quantization controller, for selecting an output of one of said block converter and said motion estimator as an input to said quantization controller.

Claim 14

The encoder of claim 6, wherein the segment is one field of the digital video signal.

Claim 15

The encoder of claim 6, wherein the segment is one frame of the digital video signal.

Claim 16

The encoder of claim 6, wherein the segment is one track of the digital video signal.

Claim 17

The encoder of claim 6, wherein the segment is less than one track of the digital video signal.

Claim 18

An encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data comprising: first means for formatting the segment of the digital video signal into blocks corresponding to rectangular image areas; second means for calculating, separately for each block, a sum of absolute values of differences between horizontally and vertically adjacent pixel values of said segment of the digital video signal; third means for predicting, separately from each sum calculated by the second means, an amount of encoded data that would be generated at a predetermined quantization level; fourth means for adding each of the separate predicted amounts for all blocks in said segment together, to thereby form a total predicted amount; fifth means for selecting one of a plurality of prestored quantization levels that will generate an amount of encoded data substantially equal to the nominal amount of encoded data, based on the total predicted amount of encoded data, and for quantizing the segment of the digital video signal by the quantization level selected by the fifth means; and sixth means for encoding the quantized segment of the digital video signal quantized by the fifth means.

Claim 19

An encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data comprising: first means for formatting the segment of the digital video signal into blocks corresponding to rectangular image areas; second means for calculating, separately for each block, a sum of absolute values of differences between horizontally and vertically adjacent pixel values of said segment of the digital video signal; third means for predicting, separately from each sum calculated by the second means, an amount of encoded data that would be generated at a predetermined quantization level; fourth means for adding each of the separate predicted amounts for all blocks in said segment together, to thereby form a total predicted amount; fifth means for selecting one of a plurality of prestored quantization levels that will generate an amount of encoded data substantially equal to the nominal amount of encoded data, based on the total predicted amount of encoded data, and for quantizing the segment of the digital video signal by the quantization level selected by the fifth means; and sixth means for encoding the quantized segment of the digital video signal quantized by the fifth means.

Claim 20

The encoder of claim 16, wherein the fifth means includes: quantization control means for selecting one of the plurality of prestored quantization levels; and quantizing means for quantizing the segment of the digital video signal by the selected quantization level.

Claim 21

The encoder of claim 17, further comprising: orthogonal transforming means for orthogonally transforming the pixels of each block of the segment prior to input to the quantizing means.

Claim 22

The encoder of claim 18, further comprising: memory means for storing the encoded quantized segment of the digital video signal.

Claim 23

The encoder of claim 16, further comprising: motion estimation means for comparing each block to a reference image and for generating motion vectors and error values for each block; and selection means for selecting an output of one of the first means and the motion estimation means for input into the second means.

Claim 24

The encoder of claim 20, wherein the motion estimation means receives the selected quantization level from the fifth means, for constructing a reference image.

Claim 25

The encoder of claim 20, wherein the selection means selects an output of the first means for intrafield encodation and selects an output of the motion estimation means for interfield encodation.

Claim 26

The encoder of claim 16, wherein the second means comprises: luminance processing means for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values of a luminance component of the segment; first chrominance processing means for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values of a first chrominance component of the segment; and second chrominance processing means for calculating a sum of absolute values of differences between horizontally and vertically adjacent pixel values of a second chrominance component of the segment.

Claim 27

The encoder of claim 23, wherein the first means includes block converting means for converting said segment of an input digital video signal into blocks, each block including a predetermined number of pixels, wherein each of the luminance component, first chrominance component, and second chrominance component are converted separately by the block converting means and wherein a sample rate of the first and second chrominance components is half the sample rate of the luminance component.

Claim 28

The encoder of claim 24, wherein the third means comprises: luminance predicting means for predicting, from the calculated sum of the luminance processing means for each block of pixels, an amount of encoded luminance data; first chrominance predicting means for predicting, from the calculated sum of the first chrominance processing means, for each block of pixels, an amount of encoded first chrominance data; and second chrominance predicting means for predicting, from the calculated sum of the second chrominance processing means for each block of pixels, an amount of encoded second chrominance data.

Claim 29

The encoder of claim 25, further comprising: first accumulating means for accumulating the predicted amounts of encoded luminance data, for each block of pixels in the segment, from the luminance predicting means; second accumulating means for accumulating the predicted amounts of encoded first chrominance data, for each block of pixels in the segment, from the first chrominance predicting means; and third accumulating means for accumulating the predicted amounts of encoded second chrominance data, for each block of pixels in the segment, from the second chrominance predicting means.

Claim 30

The encoder of claim 26, wherein the fourth means adds the first, second and third accumulated predicted amounts from the first, second, and third accumulating means, respectively, and to generate the total predicted encoded amount of data for the segment, divides the total predicted encoded amount of data for the segment by the nominal amount of encoded data to derive a ratio, and the fifth means selects one of a plurality of prestored quantization levels based on the derived ratio.

Claim 31

The encoder of claim 16, wherein the first, second and third means of the encoder operate separately on luminance and chrominance components of the segment of the digital video signal.

Claim 32

The encoder of claim 16, wherein the segment is one field of the digital video signal.

Claim 33

The encoder of claim 16, wherein the segment is one frame of the digital video signal.

Claim 34

The encoder of claim 16, wherein the segment is one track of the digital video signal.

Claim 35

The encoder of claim 16, wherein the segment is less than one track of the digital video signal.

Claim 36

An encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data comprising: converting means for converting the segment of the digital video signal into blocks, each block including a predetermined number of pixels, prior to input to a second means; the second means for calculating, separately for each block, at least one pixel intercomparison value for each block based on intercomparisons of pixel values of said segment of the digital video signal; third means for predicting, separately from each calculated pixel intercomparison value calculated by the second means, an amount of encoded data that would be generated at a predetermined quantization level; fourth means for adding each of the separate predicted amounts for all blocks in said segment together, to thereby form a total predicted amount; fifth means for selecting one of a plurality of prestored quantization levels that will generate an amount of encoded data substantially equal to the nominal amount of encoded data, based on the total predicted amount of encoded data, and for quantizing the segment of the digital video signal by the quantization level selected by the fifth means; and sixth means for encoding the quantized segment of the digital video signal quantized by the fifth means.

Claim 37

An encoder for encoding a segment of a digital video signal to obtain a nominal amount of encoded data comprising: converting means for converting the segment of the digital video signal into blocks, each block including a predetermined number of pixels, prior to input to a second means; the second means for calculating, separately for each block, at least one pixel intercomparison value for each block based on intercomparisons of pixel values of said segment of the digital video signal; third means for predicting, separately from each calculated pixel intercomparison value calculated by the second means, an amount of encoded data that would be generated at a predetermined quantization level; fourth means for adding each of the separate predicted amounts for all blocks in said segment together, to thereby form a total predicted amount; fifth means for selecting one of a plurality of prestored quantization levels that will generate an amount of encoded data substantially equal to the nominal amount of encoded data, based on the total predicted amount of encoded data, and for quantizing the segment of the digital video signal by the quantization level selected by the fifth means; and sixth means for encoding the quantized segment of the digital video signal quantized by the fifth means.

Claim 38

The encoder of claim 33, wherein the fifth means includes: quantization control means for selecting one of the plurality of prestored quantization levels; and quantizing means for quantizing the segment of the digital video signal by the selected quantization level.

Claim 39

The encoder of claim 34, further comprising: orthogonal transforming means for orthogonally transforming the pixels of each block of the segment prior to input to the quantizing means.

Claim 40

The encoder of claim 35, further comprising: memory means for storing the encoding quantized segment of the digital video signal.

Claim 41

The encoder of claim 33, further comprising: motion estimation means for comparing each block to a reference image and for generating motion vectors and error values for each block; and selection means for selecting an output of one of the converting means and the motion estimation means for input into the second means.

Claim 42

The encoder of claim 37, wherein the motion estimation means receives the selected quantization level from the fifth means, for constructing a reference image.

Claim 43

The encoder of claim 37, wherein the selection means selects an output of the converting means for intrafield encodation and selects an output of the motion estimation means for interfield encodation.

Claim 44

The encoder of claim 33, wherein the first means comprises: luminance processing means for calculating a pixel intercomparison value based on intercomparisons of pixel values of a luminance component of the segment; first chrominance processing means for calculating a pixel intercomparison value based on intercomparisons of pixel values of a first chrominance component of the segment; and second chrominance processing means for calculating a pixel intercomparison value based on intercomparisons of pixel values of a second chrominance component of the segment.

Claim 45

The encoder of claim 40 wherein the converting means includes block converting means for converting said segment of an input digital video signal into blocks, each block including a predetermined number of pixels, wherein each of the luminance component, first chrominance component, and second chrominance component are converted separately by the block converting means and wherein a sample rate of the first and second chrominance components is half the sample rate of the luminance component.

Claim 46

The encoder of claim 41, wherein the third means comprises: luminance predicting means for predicting, from the calculated pixel intercomparison value of the luminance processing means for each block of pixels, an amount of encoded luminance data; first chrominance predicting means for predicting, from the calculated pixel intercomparison value of the first chrominance processing means, for each block of pixels, an amount of encoded first chrominance data; and second chrominance predicting means for predicting, from the calculated pixel intercomparison value of the second chrominance processing means for each block of pixels, an amount of encoded second chrominance data.

Claim 47

The encoder of claim 42, further comprising: first accumulating means for accumulating the predicted amounts of encoded luminance data, for each block of pixels in the segment, from the luminance predicting means; second accumulating means for accumulating the predicted amounts of encoded first chrominance data, for each block of pixels in the segment, from the first chrominance predicting means; and third accumulating means for accumulating the predicted amounts of encoded second chrominance data, for each block of pixels in the segment, from the second chrominance predicting means.

Claim 48

The encoder of claim 43, wherein the fourth means adds the first, second and third accumulated predicted amounts from the first, second, and third accumulating means, respectively, to generate the total predicted encoded amount of data for the segment, divides the total predicted encoded amount of data for the segment by the nominal amount of encoded data to derive a ratio, and the fifth means selects one of a plurality of prestored quantization levels based on the derived ratio.

Claim 49

The encoder of claim 33, wherein the first, second and third means of the encoder operate separately on luminance and chrominance components of the segment of the digital video signal.

Claim 50

The encoder of claim 33, wherein the segment is one field of the digital video signal.

Claim 51

The encoder of claim 33, wherein the segment is one frame of the digital video signal.

Claim 52

The encoder of claim 33, wherein the segment is one track of the digital video signal.

Claim 53

The encoder of claim 33, wherein the segment is less than one track of the digital video signal.