Image analysis – Image compression or coding
Reexamination Certificate
1997-12-19
2001-03-06
Boudreau, Leo H. (Department: 2721)
Image analysis
Image compression or coding
C382S236000, C382S250000, C358S438000, C348S404100, C348S415100, C375S240150
Reexamination Certificate
active
06198849
ABSTRACT:
BACKGROUND INFORMATION
The method according to the invention proceeds from a method according to the species defined in the main claim. A video coding standard, provided for the conducted transmission of video images at a transmission rate of up to 64 kbit/sec, has already been proposed by the International Telecommunications Union. The standard is entitled “Draft Recommendation H.263” and comes from Study Group 15, Working Party 15/1, Document LBC-95, Apr. 7, 1995. This document contains essential information regarding source coding of the video signals, the syntax and semantics for data transmission, and decoding of the data in the receiver.
This standard, however, is not necessarily suitable for the transmission of video images over error-prone mobile radio networks, since the quality of the video images transmitted is insufficient because transmission errors occur. To solve this problem, the method according to the present invention provides various modifications to the conventional methods utilizing a conventional H.263 standard. Expressions which are known to those of ordinary skill in the art from this H.263 standard are referred to in the description below.
SUMMARY OF THE INVENTION
The method according to the present invention has the advantage that the quality of the video images transmitted, in particular when image data are transmitted via data channels in which multiple transmission errors may occur, is improved. Because those image data which have a fixed word width are combined into a particular class and are transmitted coherently and consecutively, not all of the subsequent information of the class being transmitted is lost after a transmission error occurs. The subsequent information words can be analyzed correctly without being influenced. This leads to an improvement in image quality after a transmission error occurs.
Because the coded image data of the macroblocks are inserted into the stream of image data in such a way that in the receiver station, one quantity of coded macroblocks is decodable forward while a second quantity of coded macroblocks is decoded backward, it is possible for a quantity of coded macroblocks to be decoded correctly even after a transmission error has occurred and has been recognized. This occurs because the macroblocks are divided into two groups, and that after transmission of the coded data of the macroblocks into the receiver station, a very wide binary word is present which consists of the data of both groups but is decoded from both ends. If an error occurs during decoding of the one group, only the remaining data of the same group of macroblocks should be discarded. The other group, i.e. the group of macroblocks which is being decoded from the other end, can still be decoded correctly. As a result after a transmission error has occurred, much fewer image regions fail to be updated from the previous video image than in the case where the entire transmitted binary word is uniformly decoded from one end only.
It is further advantageous for the selection of the macroblocks to be decoded forward and the selection of the macroblocks to be decoded backward is made in such a way that adjacent macroblocks are in each case decoded from different ends. This ensures that the respective macroblocks to be decoded differently are distributed as uniformly as possible over the entire image, so that an error occurring in one of the two groups of macroblocks cannot be detected as easily in the image.
It is further advantageous if the macroblocks to be decoded in the same manner are inserted into the stream of image data in a specific sequence, so that in each case those macroblocks which are positioned in the center of the image can be decoded first, and those macroblocks which are positioned in the edge regions of the image can be decoded last. As a result the particularly important image sections in the center of the image are always decoded first. If a transmission error then occurs, then always only the less important image information elements in the edge regions which are lost.
The spiral form according to the present invention is a particularly favorable sequence for transmission of the image data of the macroblocks to be decoded in the same manner.
It is further advantageous that the image data whose word length is defined a priori are equipped with a parity bit, so that error recognition upon transmission of said words becomes easier. Image data of the HEADER and ADMIN+DC-INTRA class are particularly important, namely the quantization words GQUANT and the DC-INTRA coefficients of the discrete cosine transformation.
It is further advantageous that in the transmission of coded image data, such as the MCBPC codes for chrominance information in the image (of the ADMIN+DC-INTRA class), and the MVD codes for the movement vectors (of the MW class), the shortest code used is a two-digit binary number, rather than a single-digit number as in the proposed H.263 standard. This also facilitates error recognition for these frequently used codes.
REFERENCES:
patent: 5317396 (1994-05-01), Fujinami
patent: 5591956 (1997-01-01), Longacre, Jr. et al.
patent: 5598216 (1997-01-01), Lee
patent: 5771102 (1998-06-01), Vogt et al.
patent: 5905815 (1999-05-01), Mack et al.
patent: 5917948 (1999-06-01), Holtz
patent: 0 546 339 A1 (1993-06-01), None
patent: WO 96/12993 (1996-05-01), None
PCT Written Opinion issued in PCT application PCT/US97/17983 dated Jul. 20, 1998.
Contact Person: Karel Rijkse, “Line Transmission Of Non-Telephone Signals—Video Coding For Low Bitrate Communication,” Draft ITU-T Recommendation H.263 (May 2, 1996), International Telecommunication Union.
Schäfer, R. et al., Bildverarbeitung und Bildkompression für die digitale HDTV-Magnethandaufzeichnung und für den digitalen TV/HDTV Rundfunk, pp. 69-87.
Fischer Ralf
Nitsche Gunnar
Boudreau Leo H.
Desire Gregory
Kenyon & Kenyon
Robert & Bosch GmbH
LandOfFree
Process for producing and evaluating a video data stream for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for producing and evaluating a video data stream for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for producing and evaluating a video data stream for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2512778