Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1999-06-02
2001-08-14
Le, Vu (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06275535
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to the decoding of compressed images and in particular to the decoding of images which are compressed according to an MPEG standard. The invention also relates to the display of the decoded images. More particularly, the invention relates to the decoding of bidirectional images.
BACKGROUND OF THE INVENTION
According to various image compression standards, especially MPEG (“Motion Pictures Experts Group”), the images are decoded in blocks, or macroblocks, generally of 16×16 pixels. The macroblocks can be of various formats. The most commonly used format is the one termed 4:2:0 according to which each macroblock contains four blocks of 8×8 luminance pixels of eight bits and two blocks of 8×8 chrominance pixels of eight bits.
The images processed are essentially of three types, namely an “intra” type, a “predicted” type and a “bidirectional” type. The person skilled in the art is aware that the macroblocks of an “intra” image do not undergo any motion compensation. In a predicted image, each macroblock can undergo a motion compensation which includes combining the macroblock with another macroblock, the “predictor”, fetched from a previously decoded image. Each macroblock of a bidirectional image can undergo a motion compensation which includes combining the macroblock with two other predictor macroblocks, fetched respectively from two previously decoded images. The positions of the predictor macroblocks are determined by motion vectors.
An MPEG decoding/display system, referred to more simply hereafter as an “MPEG decoder”, must communicate to a dynamic memory to carry out the decoding and the displaying of the decoded images. Such a memory plays an essential role in the decoding and displaying of these images. In certain modes of operation (freeze frame for example), certain images have to be decoded several times. These multiple decodings require repeated access to the area of the image memory storing the compressed image data awaiting processing. In particular, it is especially important to be able to redecode, without the risk of errors originating from the addressing of the memory, the right compressed data actually corresponding to an image already previously decoded.
SUMMARY OF THE INVENTION
It is an object of the invention to provide a solution to this problem, which is simple to implement and to embody. The invention applies especially, but not exclusively, to the “on-the-fly” decoding of bidirectional images, which decoding is effected twice while the two frames of the image are displayed directly, and which does not require storage of the decoded bidirectional image in the image memory, thus reducing the image memory size.
The invention therefore proposes a method for decoding an inbound image, for example a bidirectional image in which a memory area of a dynamic memory is used to store a stream of compressed data comprising successive groups of compressed data. These groups relate respectively to successive inbound images, some of which require two successive decodings. Each data group associated with an image comprises a start-of-image identifier (or picture start code), followed by a header containing a specific identifier of the image (for example the temporal reference of this image), this header being followed by useful data. The stream of data is read in packets of bits located at consecutive addresses of the memory area. For each packet extracted from the memory area, the presence or the absence of a start-of-image identifier is detected, and when a detected start-of-image identifier corresponds to an image requiring two successive decodings, the address of the relevant packet is stored together with the specific identifier of the image. Then a first decoding of the useful data of the image is performed, on completion of which the packet situated at the stored address is extracted again from the memory area. The second decoding of the image is then carried out by decoding the data of the packet after detection of the specific identifier. The data of the packet preceding the specific identifier is ignored for the decoding.
In an MPEG data stream, one difficulty in carrying out two successive decodings of the same image resides in the fact that the intervals between the various start-of-image identifiers of the various images are not constant and depend on the contents of the images. Moreover, the image memory is in practice read in packets of bits situated at successive addresses of the memory.
A first characteristic of the invention therefore includes storing the address of the bit packet in which the start identifier of an image requiring two successive decodings, for example a bidirectional image, has been detected. This being so, if, the first decoding having been performed, a return is made to the memory area to repoint to the bit packet containing the start-of-image identifier, then the position of this start-of-image identifier is not accurately known in the packet. Moreover, it is not certain that the packet might not comprise other start-of-image identifiers corresponding to previous images. Stated otherwise, if a second decoding of the data arising from the packet situated at the address returned to for positioning, were then performed without precaution, a second, erroneous, decoding of the image would be obtained.
The invention solves these additional difficulties by providing, in addition to the above described features, a second characteristic including storing a specific identifier (for example, the temporal reference of the image) making it possible to tag the image in a one-to-one manner in the sequence. Thus, after having returned back and having again extracted the packet containing the start-of-image identifier of the image to be redecoded, the bits of the packet are analyzed sequentially until the presence of the specific identifier of the bidirectional image to be redecoded is detected again. And it is only from this instant onwards that the pipeline circuit conventionally used in an MPEG decoder will be able to take into account the useful data of this image in such a way as to perform the second decoding thereof.
The subject of the invention is also a device for decoding an inbound image, comprising a dynamic memory including a memory area storing a stream of compressed data comprising successive groups of compressed data, which groups relate respectively to successive inbound images, certain of which require two successive decodings. Each data group is associated with an image comprising a start-of-image identifier, followed by a header containing a specific identifier of the image, followed by useful data. The device also includes a first address pointer and a second address pointer, each making it possible to read the memory area in packets of bits, as well as a start-of-image identifier detector able to detect in each bit packet extracted at the address pointed at by the first address pointer, the presence or the absence of a start-of-image identifier.
Furthermore, the device includes a first storage means (for example a register) able to store in the presence of a first control signal representative of the presence of a current image requiring two successive decodings, the specific address of the packet containing the start-of-image identifier of the current image, and a second storage means (for example a second register) able to store in the presence of the first control signal, the specific identifier of the said current image. The device further includes processing means (for example a microprocessor) linked to the start-of-image identifier detector, able to deliver the first control signal, and a decoding circuit (pipeline circuit) able to decode the bits of each packet pointed at by the second address pointer. Also, the device includes pointer management means able, in response to a second control signal transmitted by the decoding circuit and representative of the end of a first decoding of an image, to move the second address pointer to the specific address stor
Bramley Richard
Woodward Patrice
Allen Dyer Doppelt Milbrath & Gilchrist, P.A.
Galanthay Theodore E.
Le Vu
STMicroelectronics S.A.
LandOfFree
Method and device for decoding an image compressed in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and device for decoding an image compressed in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for decoding an image compressed in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2516087