Method and apparatus for image data compression using combined luminance/chrominance coding

• US 5,434,623 A
• Filed: 12/20/1991
• Issued: 07/18/1995
• Est. Priority Date: 12/20/1991
• Status: Expired due to Term

• Current Assignee: Ampex Corporation (Delta Information Systems, Inc.)
• Original Assignee: Ampex Corporation (Delta Information Systems, Inc.)
• Inventors: Coleman, Charles H., Miller, Sidney D., Smidth, Peter
• Primary Examiner(s): Chin, Tommy P.
• Assistant Examiner(s): LEE, RICHARD J
• Application Number:
  US07/811,486
• Time in Patent Office: 1,306 Days
• Field of Search: 358/13, 358/133, 358/136
• US Class Current: 375/240.04
• CPC Class Codes: H04N 11/044 involving transform coding
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/14 Coding unit complexity, e.g. amount of activity or edge presence estimation H04N19/146 takes precedence
• CPC Class Codes: H04N 19/146 Data rate or code amount at the encoder output
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/60 using transform coding
• CPC Class Codes: H04N 19/88 involving rearrangement of data among different coding units, e.g. shuffling, interleaving, scrambling or permutation of pixel data or permutation of transform coefficient data among different blocks
• CPC Class Codes: H04N 5/9264 using transform coding
• CPC Class Codes: H04N 9/8047 using transform coding

Abstract

A technique is provided for a color imaging system having a luminance component and two color difference (chrominance) components, which allots more space in the compressed data stream to the signal component of greater information content or complexity; that is, which combines the encoding of the components of the data representing a common portion of an image in response to the actual needs of the component, prior to encoding the combined data. To this end, the calculation of the quantizing factors considers the combined information content of both the luminance and chrominance components for the same image location. Thus, low chrominance complexity allows the transmission of more of the luminance information, and vice versa. The system also provides for predetermining the relative complexity of the signal components to thus pre-allot the amount of space for each signal component in the compressed data stream in proportion to its measured complexity.

Claims

1. A method for compressing image data formed of two or more signal components of respective complexities at a common location in an image which together define a combined signal information content corresponding to the complexities, wherein the compressed image data defines a data stream formed of a preselected number of data bits comprising the steps of;

selecting a joint quantizing factor from the combined signal information content corresponding to the complexities of said two or more signal components derived from the common image location, wherein each of the signal components are distinct from each other;

detecting, prior to the step of selecting, a relative complexity index of the signal components taken from the common location in the image;

performing said step of selecting by modifying said joint quantizing factor in response to the detected relative complexity index to pre-assign a proportion of said preselected number of data bits for each of the signal components during compression; and

quantizing said signal components from said common image location in response to said joint quantizing factor to assign said proportion of said preselected number of data bits to each signal component in the dam stream in accordance with the respective complexities and thus the actual needs of the signal components from the common location.

2. Apparatus for compressing image data formed of two or more signal components having different characteristics and derived from a common location in an image, and wherein the compressed image data are allotted a preselected amount of data stream space, comprising:

means for combining processed versions of the signal components taken from the common location into a combined data stream;

means for determining a joint numerical processing parameter for the combined data stream which allots an amount of data stream space to each signal component in proportion to its complexity of the signal component, wherein the processing parameter is applied during compression to the combined data stream;

means for detecting a relative complexity between said two or more signal components derived from the common location;

means for generating a relative complexity index indicative of the relative complexity; and

.

means for modifying the joint numerical processing parameter in response to said relative complexity index.

3. The apparatus of claim 2 wherein:

said detecting means includes image complexity detecting means responsive to said processed versions of the respective signal components, for detecting the relative complexity thereof; and

said generating means includes complexity comparator means in the form of look-up tables representative of values of said relative complexity index.

4. Apparatus for compressing image data formed of two or more signal components derived from a common location in an image and having varying complexities, wherein the compressed image data is allotted a preselected data stream space, comprising:

means for providing separate streams of data each formed of a respective signal component from said common location;

means receiving said separate streams of data for generating respective streams of transform coefficients corresponding to said separate streams of data;

means receiving said respective streams of transform coefficients for combining the transform coefficient streams into a combined stream of transform coefficients formed of distinct coefficients of said two or more signal components;

means for determining a joint quantizing factor for the combined information content corresponding to the complexities of said two or more signal components from the common location, wherein each of the transform coefficients in the stream are distinct from each other;

means for predetermining a relative complexity index between said two or more signal components derived from the common location;

means for modifying said joint quantizing factor in response to the relative complexity index to pre-allot the amount of said data stream space to each signal component prior to applying the joint quantizing factor; and

means for applying said joint quantizing factor to the combined stream of distinct transform coefficients to allot an amount of said data stream space to each signal component in proportion to the complexity of each signal component at the common location while fitting the compressed image data into said preselected data stream space.

5. The apparatus of claim 4 including:

means coupled to said applying means for supplying a weighting value to the joint quantizing factor which varies for each transform coefficient in accordance with the characteristics of the signal component and with a human observer'"'"'s visual sensitivity to errors in the signal component;

said means for supplying including;

a weighting function memory for storing weighting lookup tables indicative of said visual sensitivity to errors; and

a timing generator for supplying, to said weighting function memory, component identification and index signals which identify the signal components.

6. The apparatus of claim 4 wherein said signal components comprise the luminance and chrominance at said common location in the image, and said combined stream is formed of cosine transform coefficients of one signal component distinctly interspersed with the cosine transform coefficients of another signal component.

7. A method for selecting a quantizing factor for two or more signals derived from a common location of an image and having different complexities, wherein application of the quantizing factor during quantization provides a stream of compressed signals confined to a preselected number of data bits, comprising the steps of:

supplying separate streams of data each representative of a respective signal from the common location;

combining the separate streams of data into a combined stream of signal data;

selecting a joint quantizing factor based on the combined stream of signal data, wherein said joint quantizing factor allots to each of the two or more signals a proportion of data bits commensurate with the respective complexity of the two or more signals, while confining the combined stream of signal data into said preselected number of data bits upon said quantization;

determining a relative complexity between said two or more signals at said common location;

generating, in response to the step of determining, an index signal indicative of the relative complexity; and

modifying said joint quantizing factor in response to said index signal during the step of selecting.

8. The method of claim 7 wherein said two or more signals respectively define the luminance and chrominance of the image at said common location.

9. The method of claim 7 further including modifying said joint quantizing factor during quantization with a weighting function indicative of a human observer'"'"'s sensitivity to errors in the two or more signals.