Facsimile and static presentation processing – Facsimile – Specific signal processing circuitry
Patent
1987-06-04
1989-05-02
Groody, James J.
Facsimile and static presentation processing
Facsimile
Specific signal processing circuitry
358135, H04N 712
Patent
active
048273384
DESCRIPTION:
BRIEF SUMMARY
The present invention relates to a hybrid coding process by transformation for the transmission of image or picture signals. The invention relates to the real time processing of sequential information signals, such as video signals, with a view to compressing the data volume necessary for representing said picture, so as to transmit it with a minimum bit rate on a transmission line. The inventive process more particularly applies in television, visioconference or visiophony. It can also be applied to pictures with slow variations, such as in remote monitoring.
Over the last few years, numerous processes have been proposed for compressing a picture signal in order to reduce the rate necessary for the digital transmission of said picture. This has been based on the wish to limit the digital rate, so as to be able to transmit video signals on a low rate line or transmit several of these simultaneously on one line.
In general terms, the video signal is firstly converted into a digital signal, which is then compressed in several stages. A first compression of the video signal is obtained by subsampling of the digitized luminance signal and digitized chrominance signals, e.g. for a 384 kbit/s visioconference application, whereby said first processing can make it possible to obtain a compression rate of approximately 10 to 15. The rate corresponding to a picture according to CCIR recommendation 601 is consequently brought from 166 to 11.7 mbits/s. The second compression stage is a stage involving the transformation of the picture in the frequency range, followed by interframe or intraframe coding of the coefficients of the transformed picture. The third stage is a statistical coding, generally of the HUFFMANN type, of the sequential signal constituted by the transformed and coded coefficients.
The inventive process relates to the second stage and aims at a greater compression of the picture signal, whilst maintaining an appropriate picture quality.
In general terms, in order to perform the second coding stage, the picture is subdivided into a plurality of identical blocks. Each of these blocks is separately coded, either by an interframe code, or by an intraframe code. Interframe coding consists of coding the difference between the coefficients of one block of a frame of a picture and the coefficients of the corresponding block of the frame of the previous picture. Intraframe coding consists of a direct coding of the block coefficients. In the interframe mode, the differences can relate to blocks in the spatial range, the blocks then representing part of the picture, or to blocks in the frequency range, the blocks being obtained by transforming picture blocks and representing the frequency components of said picture.
The transform can be a cosine transform, a FOURIER transform, a HADAMARD transform, a Haar transform or the like. The coefficients of the transformed block are called in general frequency coefficients, or sequential coefficients in the case of the HADAMARD transform. The coefficients of the transformed block do not exactly represent the frequency components of the picture corresponding to the untransformed block. However, by linguistic misuse, the transformed block range will be called the frequency range.
A picture compression system using interframe coding of the blocks is described in European patent application EP-A2-0084270. In this known system, the corresponding blocks of two successive frames are compared and then, if the difference between the two blocks exceeds a given threshold, the block is converted in the frequency range by a discrete cosine transform before being coded by statistical coding.
European patent application EP-A2-0123456 discloses a coding system using interframe and intraframe modes in a combined manner. This system uses variable prediction factors to take account of the fact that the correlation between the corresponding transformed coefficients between individual frames is generally variable and non-linear. This system eliminates the redundant data between the blocks of
REFERENCES:
patent: 4281344 (1981-07-01), Mounts et al.
patent: 4394774 (1983-07-01), Widergren et al.
patent: 4541012 (1985-09-01), Tescher
patent: 4546386 (1985-08-01), Matsumoto et al.
patent: 4581638 (1986-04-01), Chiariglione et al.
patent: 4704628 (1987-11-01), Chen et al.
Transactions on Communications, vol. COM-25, No. 11, Nov. 1977, IEEE New York (U.S.) Ali Habibi: "Survey of Adaptive Image Coding Techniques", pp. 1275-1279, Line 6, p. 1281, Line 20-p. 1282, Line 7: Figure 4.
Groody James J.
Kostak Victor R.
LandOfFree
Hybrid coding process by transformation for the transmission of does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Hybrid coding process by transformation for the transmission of , we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Hybrid coding process by transformation for the transmission of will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-589277