Data compression system having perceptually weighted motion vectors

Claim 2

A data compression method for compressing a digitally encoded video signal organized in a succession of image frames, said method comprising the steps of: receiving said digitally encoded video signal; comparison/matching selected portions of said video signal in a current image frame to portions of said video signal in an image frame preceding said current image frame to produce corresponding error values indicative of the differences therebetween; generating a motion vector for each comparison in said comparison/matching step in association with each corresponding error value; weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and selecting one of said motion vectors by applying a selection criteria to each of said perceptually weighted error values and selecting the motion vector associated with the one of said perceptually weighted error values which best fits said criteria.

Claim 3

The method of claim 1 wherein said comparing step includes the steps of: dividing a current image frame into a plurality of current frame pixel blocks; dividing said preceding image frame into a plurality of previous frame pixel blocks; selecting one of said current frame pixel blocks; and comparing said selected one of said current frame pixel blocks to each of said previous frame pixel blocks within a predetermined range to produce said error values.

Claim 4

The method of claim 2 wherein said step of generating a motion vector includes the steps of: establishing each motion vector as the direction and distance between the position of each of said previous frame pixel blocks within said previous image frame and said selected one of current frame pixel blocks; and associating each generated motion for each previous frame pixel block with the error value corresponding thereto.

Claim 5

The method of claim 2 wherein said weighting step includes the step of multiplying each of said error values by a spatial weighting factor which is inversely related to the spatial separation between said selected current frame pixel block and said previous frame pixel block.

Claim 6

The method of claim 2 further including the steps of: converting said digitally encoded video signal to a frequency domain signal; weighting said frequency domain signal to reduce the higher frequency terms thereof to produce a frequency weighted frequency domain signal; and performing the inverse of said converting step upon said frequency weighted frequency domain signal.

Claim 7

A data compression system for compressing a digitally encoded video signal organized in successive image frames, said system comprising: means for receiving said digitally encoded video signal; means for comparison/matching selected portions of said video signal in a current image frame to portions of said video signal in an image frame preceding said current image frame to produce corresponding error values indicative of the differences therebetween; means for generating a motion vector for each comparison produced by said means for comparison/matching in association with each corresponding error value; means for weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and means for selecting one of said motion vectors by applying a selection criteria to each of said perceptually weighted error values and selecting the motion vector associated with the one of said perceptually weighted error values which best fits said criteria.

Claim 8

A data compression system for compressing a digitally encoded video signal organized in successive image frames, said system comprising: means for receiving said digitally encoded video signal; means for comparison/matching selected portions of said video signal in a current image frame to portions of said video signal in an image frame preceding said current image frame to produce corresponding error values indicative of the differences therebetween; means for generating a motion vector for each comparison produced by said means for comparison/matching in association with each corresponding error value; means for weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and means for selecting one of said motion vectors by applying a selection criteria to each of said perceptually weighted error values and selecting the motion vector associated with the one of said perceptually weighted error values which best fits said criteria.

Claim 9

A data compression system as set forth in claim 6 wherein said means for comparing includes: means for dividing a current image frame into a plurality of current frame pixel blocks; means for dividing an image frame preceding the current image frame into a plurality of previous frame pixel blocks; means for selecting one of said current frame pixel blocks; and means for comparing said selected one of said current frame pixel blocks to each of said previous frame pixel blocks within a predetermined range to produce said error values.

Claim 10

A data compression system as set forth in claim 7 wherein said means for generating a motion vector includes: means for establishing each motion vector as the direction and distance between the position of each of said previous frame pixel blocks within said previous image frame and said selected one of current frame pixel blocks; and means for associating each generated motion for each previous frame pixel block with the error value corresponding thereto.

Claim 11

A data compression system as set forth in claim 7 wherein said means for weighting includes means for multiplying each of said error values by a spatial weighting factor which is inversely related to the spatial separation between said selected current frame pixel block and said previous frame pixel block.

Claim 12

A data compression system as set forth in claim 7 further including: means for converting said digitally encoded video signal to a frequency domain signal; means for weighting said frequency domain signal to reduce the higher frequency terms thereof to produce a frequency weighted frequency domain signal; and means for performing the inverse of said converting means upon said frequency weighted frequency domain signal.

Claim 13

For use in a television transmission system processing a video image signal organized into a plurality of successive image frames, a data compression method comprising the steps of: receiving said video image signal; organizing each image frame into a plurality of pixel blocks; providing a current image frame and a preceding image frame concurrently; comparing each pixel block in the current image frame to each pixel block in at least a portion of the preceding image frame and producing an error value for each comparison indicative of the degree of difference between the pixel block in the current image frame and the pixel block in the preceding image frame; generating motion vectors indicative of the spatial relationship between each pixel block in the current image frame and each pixel block in the preceding image frame; associating each error value produced with the corresponding motion vector generated; weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and selecting the motion vector from said generated motion vectors for each pixel block in the current image frame which has associated therewith the lowest magnitude perceptually weighted error value.

Claim 14

For use in a television transmission system processing a video image signal organized into a plurality of successive image frames, a data compression method comprising the steps of: receiving said video image signal; organizing each image frame into a plurality of pixel blocks; providing a current image frame and a preceding image frame concurrently; comparing each pixel block in the current image frame to each pixel block in at least a portion of the preceding image frame and producing an error value for each comparison indicative of the degree of difference between the pixel block in the current image frame and the pixel block in the preceding image frame; generating motion vectors indicative of the spatial relationship between each pixel block in the current image frame and each pixel block in the preceding image frame; associating each error value produced with the corresponding motion vector generated; weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and selecting the motion vector from said generated motion vectors for each pixel block in the current image frame which has associated therewith the lowest magnitude perceptually weighted error value.

Claim 15

The method of claim 11 wherein said perceptual factor includes the spatial relationship between the pixel block in the current image frame and the pixel block in the preceding image frame.

Claim 16

The method of claim 11 further including the step of deemphasizing the higher frequency components of said video signal to frequency weight said motion vector selecting step.

Claim 17

For use in a television transmission system processing a video image signal organized into a plurality of successive image frames, a data compression system comprising: means for receiving said video image signal; means for organizing each image frame into a plurality of pixel blocks; means for providing a current image frame and a preceding image frame concurrently; means for comparing each pixel block in the current image frame to each pixel block in at least a portion of the preceding image frame and producing an error value for each comparison indicative of the degree of difference between the pixel block in the current image frame and the pixel block in the preceding image frame; means for generating motion vectors indicative of the spatial relationship between each pixel block in the current image frame and each pixel block in the preceding image frame; means for associating each error value produced with the corresponding motion vector generated; means for weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and means for selecting the motion vector from said generated motion vectors for each pixel block in the current image frame which has associated therewith the lowest magnitude perceptually weighted error value.

Claim 18

For use in a television transmission system processing a video image signal organized into a plurality of successive image frames, a data compression system comprising: means for receiving said video image signal; means for organizing each image frame into a plurality of pixel blocks; means for providing a current image frame and a preceding image frame concurrently; means for comparing each pixel block in the current image frame to each pixel block in at least a portion of the preceding image frame and producing an error value for each comparison indicative of the degree of difference between the pixel block in the current image frame and the pixel block in the preceding image frame; means for generating motion vectors indicative of the spatial relationship between each pixel block in the current image frame and each pixel block in the preceding image frame; means for associating each error value produced with the corresponding motion vector generated; means for weighting each of said error values in accordance with a perceptual factor to produce perceptually weighted error values; and means for selecting the motion vector from said generated motion vectors for each pixel block in the current image frame which has associated therewith the lowest magnitude perceptually weighted error value.

Claim 19

A data compression system as set forth in claim 14 wherein said perceptual factor includes the spatial relationship between the pixel block in the current image frame and the pixel block in the preceding image frame.

Claim 20

A data compression system as set forth in claim 14 further including the step of deemphasizing the higher frequency components of said video signal to frequency weight said means for motion vector selecting.

Claim 21

For use in a television transmission system processing a video signal organized into a plurality of successive image frames, a data compression method comprising the steps of: organizing each current image frame into a plurality of current frame pixel blocks and each preceding image frame into a corresponding plurality of previous frame pixel blocks; for each current frame pixel block, and each previous frame pixel block within a predetermined range in the preceding image frame, producing a perceptually weighted error value related to the perceptually weighted difference therebetween and a motion vector related to the spatial displacement therebetween; and selecting a motion vector for each current frame pixel block in response to a comparison of said perceptually weighted error values.

Claim 22

For use in a television transmission system processing a video signal organized into a plurality of successive image frames, a data compression method comprising the steps of: organizing each current image frame into a plurality of current frame pixel blocks and each preceding image frame into a corresponding plurality of previous frame pixel blocks; for each current frame pixel block, and each previous frame pixel block within a predetermined range in the preceding image frame, producing a perceptually weighted error value related to the perceptually weighted difference therebetween and a motion vector related to the spatial displacement therebetween; and selecting a motion vector for each current frame pixel block in response to a comparison of said perceptually weighted error values.

Claim 23

The method of claim 17 wherein said step of producing includes the steps of: comparing the pixels within said current frame pixel block to those of each previous frame pixel block within a predetermined range to provide an error value magnitude; and multiplying said error value magnitude by a perceptual weighting factor to produce a perceptually weighted error value.

Claim 24

The method of claim 18 wherein said perceptual weighting factor includes spatial displacement between said current frame pixel block and each previous frame pixel block.

Claim 25

For use in a television transmission system processing a video signal organized into a plurality of successive image frames, a data compression system comprising: means for organizing each current image frame into a plurality of current frame pixel blocks and each preceding image frame into a corresponding plurality of previous frame pixel blocks; producing means operative for each current frame pixel block, and each previous frame pixel block within a predetermined range in a preceding image frame, for producing a perceptually weighted error value related to the perceptually weighted difference therebetween and a motion vector related to the spatial displacement therebetween; and means for selecting a motion vector for each current frame pixel block in response to a comparison of said perceptually weighted error values.

Claim 26

For use in a television transmission system processing a video signal organized into a plurality of successive image frames, a data compression system comprising: means for organizing each current image frame into a plurality of current frame pixel blocks and each preceding image frame into a corresponding plurality of previous frame pixel blocks; producing means operative for each current frame pixel block, and each previous frame pixel block within a predetermined range in a preceding image frame, for producing a perceptually weighted error value related to the perceptually weighted difference therebetween and a motion vector related to the spatial displacement therebetween; and means for selecting a motion vector for each current frame pixel block in response to a comparison of said perceptually weighted error values.

Claim 27

A data compression system as set forth in claim 20 wherein said producing means includes: means for comparing the pixels within said current frame pixel block to those of each previous frame pixel block to provide an error value magnitude; and means for multiplying said error value magnitude by a perceptual weighting factor to produce a perceptually weighted error value.

Claim 28

A data compression system as set forth in claim 21 wherein said perceptual weighting factor includes spatial displacement between said current frame pixel block and each previous frame pixel block.

Claim 31

For use in a television signal transmission system processing a video signal organized into a plurality of successive image frames, a data compression method comprising the steps of: organizing each image frame into a plurality of pixel blocks; and for each current frame pixel block, and each pixel block within at least a portion of a preceding image frame, producing a motion vector representing the smallest perceptually weighted error therebetween.

Claim 32

For use in a television signal transmission system processing a video signal organized into a plurality of successive image frames, a data compression method comprising the steps of: organizing each image frame into a plurality of pixel blocks; and for each current frame pixel block, and each pixel block within at least a portion of a preceding image frame, producing a motion vector representing the smallest perceptually weighted error therebetween.

Claim 33

The method of claim 24 wherein said producing step comprises producing a motion vector representing the smallest spatially weighted error therebetween.

Claim 34

The method of claim 24 wherein said producing step comprises producing a motion vector representing the smallest frequency weighted error therebetween.