Robust encoded domain pixel decoding

Details Robust encoded domain pixel decoding Robust encoded domain pixel decoding

2000-11-28

2003-12-16

Tran, Phuoc (Department: 2621)

Image analysis

Image compression or coding

Including details of decompression

C382S260000, C382S309000

Reexamination Certificate

active

06665443

ABSTRACT:

FIELD OF INVENTION
The present invention is related to the field of image processing.
BACKGROUND OF THE INVENTION
Conventional image compression techniques, such as JPEG 2000, use subband transform domain coding. Pyramid coding is a similar technique where a small version of the image is used to predict a larger version and the residual errors are tracked. Though good for compression, these methods are usually considered fragile in the presence of errors.
FIGS. 1 and 2
show dependency graphs for typical subband and pyramid decompositions.
FIG. 1
the branches are labeled either L or H to indicated low-pass or high-pass coefficients respectively.
FIG. 2
sorts the decomposition into images and residuals. Each small image is obtained by low-pass filtering and down sampling the larger image. This indicated by the L in FIG.
2
. Residual images are formed by differencing the larger image from its prediction based on the smaller image, as shown in FIG.
2
. Coding systems that use the structures depicted in
FIGS. 1 and 2
transmit only the “leaves” of the graphs. That is, only those coefficients corresponding to nodes that have no outgoing arrows are transmitted.
Subband decoding is generally accomplished by an iterative technique where data from two leaf nodes is combined to reconstruct a parent node. This is illustrated in FIG.
3
. Similarly, pyramid decoders begin with the anchor image, predict the next larger image, and add the residual error. This process is indicated in FIG.
4
.
Subband and pyramid codes are generally thought of as fragile because errors may propagate from smaller to larger reconstruction images. The anchor image is especially vulnerable since this is where the most important reconstruction information is packed. Though subband wavelet coders are at risk, pyramid coders are especially subject to this frailty because they may use highly non-linear predictors.
SUMMARY OF THE INVENTION
A method and apparatus for robust encoded domain pixel decoding is disclosed. In one embodiment, robust encoded domain pixel decoding includes receiving transform domain coefficients, and identifying erroneous coefficients. Then, for each erroneous pixel resulting from erroneous coefficients, a virtual tap is determined. For each virtual tap, transform domain coefficient dependencies are determined. A filter to estimate a value for the erroneous pixel is determined using the corresponding transform domain coefficient dependencies.


REFERENCES:
patent: 5450506 (1995-09-01), Gillard et al.
patent: 6327307 (2001-12-01), Brailean et al.
patent: 6438267 (2002-08-01), Kondo et al.

Affiliated with

Also associated with

Rating

Robust encoded domain pixel decoding does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Robust encoded domain pixel decoding, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Robust encoded domain pixel decoding will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3159938