Image analysis – Image compression or coding – Lossless compression
Patent
1995-11-03
1998-08-11
Boudreau, Leo
Image analysis
Image compression or coding
Lossless compression
382239, G06K 900
Patent
active
057938970
DESCRIPTION:
BRIEF SUMMARY
TECHNICAL FIELD
The present invention relates to adaptive variablelength coding and decoding methods for digital image data, and more particularly, to adaptive variable-length coding and decoding methods which improve compression efficiency of transmission data by performing variable-length coding and decoding adaptively, according to statistical characteristics of image data.
BACKGROUND ART
Recently, in an apparatus for transmitting and receiving video and audio signals, a method by which the video and audio signals are coded to be digital signals to then be transmitted or stored in a memory and the digital signals are decoded to then be reproduced, has been widely adopted.
However, in the case of coding a video signal into digital data, the data quantity is large. Thus, in order to decrease the overall data quantity by removing redundant data contained in the digital video signal, discrete cosine transform (DCT) coding, differential pulse code modulation (DPCM), vector quantization, or variablelength coding (VLC) should be performed.
FIG. 1 is a schematic block diagram of a general coding system for image data. The apparatus includes means 11 and 12 for performing a DCT function with respect to an N.times.N block and for quantizing DCT coefficients, means 13 and 14 for variable-length-coding the quantized data and for further compressing data quantity, and means 15, 16, 17, 18, 19, A1, A2, SW1 and SW2 related to the inverse quantization and DCT operations with respect to the quantized data to then perform a motion compensation, which codes image data in an intra mode or inter mode.
FIG. 2 is a schematic block diagram of a general decoding system for image data. The apparatus decodes and reproduces the image data coded by the coding system shown in FIG. 1.
The operation of the coding and decoding systems respectively shown in FIGS. 1 and 2 will be briefly described.
In FIG. 1, the video signal input through an input port 10 becomes a signal of a frequency domain in the units of N.times.N blocks in DCT 11, where although the magnitude of a block is generally N.sub.1 .times.N.sub.2, it is assumed that N.sub.1 =N.sub.2 =N, for the sake of convenience. The energy of transform coefficients is chiefly concentrated in a low frequency domain. Data transforms for each block are performed by a discrete cosine transform, Walsh-Hadamard transform, discrete Fourier transform, or discrete sine transform method. Here, the transform coefficients are obtained by DCT operation.
Quantizer 12 changes the DCT coefficients into representative values of a constant level through a predetermined quantization process.
Variable-length encoder 13 variable-length-codes the representative values using their statistical characteristics, thereby further compressing the data.
Meanwhile, a quantization step size Q.sub.ss, which is varied depending on the state (a fullness) of a buffer 14 wherein the variable-length-coded data is stored, controls quantizer 12 to thereby adjust a transmission bit rate. The quantization step size Q.sub.ss is also transmitted to a receiver side, to be used in a decoding system.
Also, in general, there are many similar portions between consecutive screens. Therefore, in the case of a screen having motion, a motion vector MV is obtained by estimating the motion, and data is compensated using the motion vector MV. Then, a differential signal between adjacently positioned screens becomes very small, thereby allowing transmission data to be more compressed.
In order to perform such motion compensation, an inverse quantizer (Q.sup.-1) 15 shown in FIG. 1 inverse-quantizes the quantized data output from quantizer 12. Thereafter, the inverse-quantized data is inverse-DCT-operated in an inverse DCT means (DCT.sup.-1) 16 to then be a video signal of a spatial domain. The video signal output from inverse DCT means 16 is stored in a frame memory 17 in frame units. Motion estimator 18 searches a block having the most similar pattern to that of an N.times.N block of input port 10 among the frame data stored in frame mem
REFERENCES:
patent: 5329318 (1994-07-01), Keith
patent: 5377051 (1994-12-01), Lane et al.
Jeong Je-Chang
Jo Jae-moon
Boudreau Leo
Samsung Electronics Co,. Ltd.
Tadayon Bijan
LandOfFree
Adaptive variable-length coding and decoding methods for image d does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adaptive variable-length coding and decoding methods for image d, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive variable-length coding and decoding methods for image d will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-398303