Error correction and concealment technique

Details Error correction and concealment technique Error correction and concealment technique

1998-05-26

2001-10-16

Decady, Albert (Department: 2133)

Error detection/correction and fault detection/recovery

Pulse or data error handling

Digital data error correction

C714S754000

Reexamination Certificate

active

06304990

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention-relates to an image processing apparatus, and more particularly to an image processing apparatus equipped with image decoding means which has a function of expanding compressed image data, for example.
2. Related Background Art
Conventionally, there is known a technique for transmitting image data after performing a frequency transform (transformation) of image data for each block to make data compression. In particular, by separating frequency components into a high-frequency component and a low-frequency component and encoding those components separately from each other, compression efficiency of images is increased.
However, the above prior art has had a disadvantage that if an error occurs during transmission of encoded data, an image signal including false image data is reproduced and image quality deteriorates remarkably. In the case of using a variable-length code in encoding the above high-frequency component, it may sometimes happen that even the number of data is not reproduced accurately and a gross deterioration of image quality such as missing image data occurs.
Meanwhile, in the field of transmitting images in the digital form, there have also conventionally been proposed a variety of encoding methods for cutting down the quantity of information before transmitting images because the quantity of information is very large.
Among them, there is known a method of carrying out an orthogonal transform, quantizing a coefficient after the transform, and forming a variable-length code.
With the above-mentioned prior art including no error detecting means, however, if an error occurs in a transmission path, such an error cannot be corrected, leading to a gross deterioration of image quality.
Especially, in the event of exceeding a dynamic range capable of decoding and reproducing when encoded data after transmission is decoded, for example, image quality deteriorates.
Moreover, in an image transmission system of the type in which a moving image signal is digitized and image data is transmitted via transmission paths such as optical fibers or communication satellites and recording media such as magnetic tapes, an error correcting code (ECC) for detecting and correcting a transmission error is utilized to correct the transmission error in the reception side (or the reproduction side) based on the error correcting code. That error which cannot be corrected by using the error correcting code is subjected to interpolation processing to form an approximate value from surrounding pixels.
In order to avoid a deterioration of image quality caused by the interpolation processing, it is required that the surrounding pixels utilized for the interpolation are free from errors. When adopting highly efficient encoding (image compression) which has been widely used in recent years, the surrounding pixels capable of being utilized for the interpolation are restricted. In a DPCM (difference pulse code modulation) method wherein reset is made for each line, for example, if a correction impossible error occurs, the original data cannot be reproduced by using the line inclusive of the error. Accordingly, only those pixels which are included in the upper and lower lines can be utilized for interpolation in this case. Further, in an encoding method that uses an orthogonal transform such as a discrete cosine transform (DCT), if a correction impossible error occurs, all the pixels included in a transmission block (e.g., 8 pixels vertical×8 pixels horizontal) cannot be used to reproduce the original signal. As a result, a deterioration of image quality cannot be prevented even by the interpolation using the upper and lower lines.
Thus, even in an attempt of performing the interpolation in the same frame, no effect may be expected depending on the coding methods.
Additionally, in the field of transmitting information such as images and voices in the digital form, there have been proposed various encoding methods for cutting down the amount of data transmitted. One known run-length coding method is to encode data by combining the number of successive 0s and a value other than 0 into a set. On the other hand, a forecast difference encoding method or a difference pulse code modulation (hereinafter referred to as DPCM) encoding method is also known which compresses information by utilizing correlation between sample values close to each other. There is further known an encoding method that the above two methods are combined with each other.
FIG. 14
is a block diagram showing the encoding method in a combination of the two methods. Image data of 8 bits inputted through an input terminal
611
is applied to a DPCM encoder
613
for DPCM encoding into 4 bits. Specifically, the DPCM encoder
613
assigns DPCM codes of 4 bits to respective difference values as shown in Table 1 below.
TABLE 1
Representative
Range of Differ-
Value of Differ-
ence Values
DPCM Code
ence
−255~−94
15
−140
−93~−70
13
−80
−69 ~−50
11
−58
−49~−34
9
−40
−33~−22
7
−27
−21~−13
5
−17
−12~−6
3
−8
−5~−2
1
−3
−1~1
0
0
2~5
2
3
6~11
4
8
12~20
6
15
21~35
8
27
36~53
10
44
54~93
12
70
94~255
14
150
An image generally has correlation between sample values close to each other such that many difference signals occur as 0 and many DPCM codes occur as 0. The DPCM codes encoded into 4 bits are outputted to a run-length encoder
615
. As shown in
FIG. 15A
, the run-length encoder
615
combines the DPCM code other than 0 and the number of successive 0s (hereinafter referred to as a 0 run-length) before the DPCM code into a set for each DPCM code of 4 bits.
At this time, by limiting the 0 run-length to 16 at maximum, the run-length codes can be each expressed in 8 bits. The 8-bit run-length codes thus encoded by the run length encoder
615
are added with a reset synch bit in a transmission format unit
617
as shown in FIG.
15
B and then delivered to an output terminal. The reset synch bit is added, for example, after processing of each line.
FIG. 13
is a block diagram showing a configuration of a decoder device in cooperation with the encoding section of FIG.
14
. The run-length code and the reset synch bit both transmitted through a transmission path are applied to an input terminal
621
, and a run-length decoder
623
decodes the DPCM code by inserting 0s in the number corresponding to the 0 run-length before each DPCM code until arrival of the reset synch bit. The data decoded to the DPCM code is outputted to a DPCM decoder
625
for decoding to image data of 8 bits, which are then delivered to an output terminal
627
.
With the above conventional method, however, if an error occurs in a transmission path, such an error cannot be corrected and image quality deteriorates remarkably because of including no error detecting means.
As mentioned before, in the field of transmitting information such as images and voices in the digital form, there have been proposed various encoding methods for cutting down the amount of data transmitted. One known run-length coding method is to encode data by combining the number of successive 0s and a value other than 0 into a set. On the other hand, a transform encoding method is also known which performs an orthogonal transform of pixels to be transmitted and quantizes transmitted data. There is further known an encoding method that the above two methods are combined with each other.
With the above conventional methods, however, if an error occurs in a transmission path, such an error cannot be corrected and image quality deteriorates remarkably because of including no error detecting means.
Meanwhile, U.S. Pat. No. 5,023,919 has been proposed with an intention to correct or avoid a deterioration on the side decoding compressed data, but has not yet succeeded in sufficiently solving the problem.
SUMMARY OF THE INVENTION
The present invention has been made in view of the situations i

Affiliated with

Also associated with

Rating

Error correction and concealment technique does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Error correction and concealment technique, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Error correction and concealment technique will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2611151