Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-08-01
2004-02-24
Kelley, Chris (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06697433
ABSTRACT:
This application is the national phase under 35 U.S.C. §371 of PCT International Application No. PCT/JP99/05820 which has an International filing date of Oct. 21, 1999, which designated the United States of America.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image decoding apparatus in which an error partial image region, which receives the influence of a bit error occurring in a coded bit stream, is detected in each of various types of image compressing and expanding methods.
2. Description of Related Art
As a method for effectively coding an image signal, as is adopted in MPEG-4 (Moving Picture Experts Group Phase-4) which is currently in progress toward standardization by ISO/IEC JTC11/SC29/WG11, a method, in which a discrete cosine transformation (DCT) is performed for an image signal for each block and a variable-length coding is performed for quantization-transformation coefficients obtained after a quantization of the image signal, is used.
As is described above, on a decoding side, after a variable-length decoding is performed for a coded bit stream for each block of data, an inverse quantization and an inverse DCT are performed for the coded bit stream, and the coded bit stream is reproduced as the image signal. In this decoding, there is a case that a bit error occurs in a variable-length code-word existing in a block data region. For example, in cases where the variable-length code-word, in which the bit error occurs, has a value defined in a predetermined variable-length decoding table, because the variable-length code-word is decoded without detecting the bit error, an erred image signal is reproduced.
Also, because data of each block has a variable-length, an end of data of the block cannot be specified until the data of the block is decoded. In cases where a variable-length code-word is erroneously decoded, there is a high possibility that the end of the data of the block is also erroneously specified, so that pieces of data of blocks placed after this block are erroneously decoded. In this case, the degradation of an image quality caused by the error is propagated to data of succeeded blocks.
Because a conventional image decoding apparatus has a configuration described above, for example, in cases where a bit error occurs in a block data region, because the decoding of coded data is continued until a value not defined in a variable-length decoding table is detected in the decoding, an image, which is obtained by erroneously decoding pieces of data placed in a range from an error occurring position to an error detected position, is considerably degraded. Also, because the degraded image is used as a predicted image in the decoding for obtaining one or more images of frames succeeding the frame of the degraded image, there is a problem that the decoded images of the succeeded frames are also degraded.
As a prior art relating to the above problem, there is a prior art disclosed in the Published Unexamined Japanese Patent Application No. H7-154783 (1995). In this specification, when pieces of coded data of a particular block having a prescribed size are decoded, a change degree between each peripheral pixel of the particular block and each peripheral pixel of a block adjacent to the particular block is calculated, and an error is detected by comparing the change degree with a prescribed critical value. However, there is a case that it is difficult to reliably detect the error according to only the change degree between each peripheral pixel of the particular block and each peripheral pixel of the adjacent block.
SUMMARY OF THE INVENTION
The present invention is provided to solve the above problem, and an object of the present invention is to obtain an image decoding apparatus in which a block relating to an error occurrence is reliably detected according to a statistical performance to suppress the degradation of an image quality or the propagation of the error caused by the error occurrence to the minimum degree.
An image decoding apparatus of the present invention, in which a coded bit stream produced by compressing and coding an input image signal for each partial image region is decoded, comprises partial image region data decoding means for decoding the coded bit stream for each partial image region, and error partial image region detecting means for calculating a first pixel level change degree between pixels adjacent to each other through a boundary between a decoded partial image region decoded by the partial image region data decoding means and a partial image region adjacent to the decoded partial image region, calculating a second pixel level change degree between pixels of the partial image region adjacent to the decoded partial image region, and detecting according to the first pixel level change degree, the second pixel level change degree and a prescribed threshold value relating to the first pixel level change degree and the second pixel level change degree that the decoded partial image region is an error partial image region in which an influence of a bit error occurring in the coded bit stream is received.
Therefore, the error occurrence is reliably detected in an early stage, so that an error propagation or the degradation of an image quality caused by the error can be suppressed to the minimum degree.
In an image decoding apparatus of the present invention, the error partial image region detecting means calculates a first pixel level difference value between pixels adjacent to each other through the boundary between the decoded partial image region and the partial image region adjacent to the decoded partial image region, calculates a second pixel level difference value between pixels of the partial image region adjacent to the decoded partial image region, calculates a third pixel level difference value between pixels of the partial image region adjacent to the decoded partial image region, and detects the error partial image region according to a difference value between the first pixel level difference value and the second pixel level difference value, a difference value between the second pixel level difference value and the third pixel level difference value and the prescribed threshold value.
Therefore, the error occurrence is reliably detected in an early stage, so that an error propagation or the degradation of an image quality caused by the error can be suppressed to the minimum degree.
In an image decoding apparatus of the present invention, the error partial image region detecting means comprises threshold value control means for controlling the prescribed threshold value used for the detection of the error partial image region according to a quantization step size decoded from the coded bit stream.
Therefore, an erroneous judgment for a macro-block, in which an error occurs, can be avoided.
An image decoding apparatus of the present invention, in which a coded bit stream produced by compressing and coding an input image signal for each partial image region is decoded, comprises partial image region data decoding means for decoding the coded bit stream for each partial image region, and error partial image region detecting means for calculating a representative value of a plurality of first pixel levels in a decoded partial image region decoded by the partial image region data decoding means, calculating a representative value of a plurality of second pixel levels in a partial image region placed in the neighborhood of the decoded partial image region, and detecting according to the representative value of the first pixel levels, the representative value of the second pixel levels and a prescribed threshold value relating to the representative value of the first pixel levels and the representative value of the second pixel levels that the decoded partial image region is an error partial image region in which an influence of a bit error occurring in the coded bit stream is received.
Therefore, the error occurrence is reliably detected in an early stage, so that an error propagation or th
Asai Kohtaro
Isu Yoshimi
Kuroda Shin-ichi
Nishikawa Hirofumi
Ogawa Fuminobu
Kelley Chris
Mitsubishi Denki & Kabushiki Kaisha
Vo Tung
LandOfFree
Image decoding apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image decoding apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image decoding apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3340662