Image analysis – Image compression or coding – Quantization
Reexamination Certificate
2000-01-20
2003-06-10
Wu, Jingge (Department: 2623)
Image analysis
Image compression or coding
Quantization
C374S204000
Reexamination Certificate
active
06577773
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a method and arrangement for quantizing data values with a given quantization step size.
BACKGROUND OF THE INVENTION
A known method of quantizing data values is disclosed in “An MPEG decoder with Embedded Compression for Memory Reduction”, IEEE Transactions on Consumer Electronics, Vol. 44, No. 3, August 1998. Herein, a data compressor including a quantizer is used to efficiently store pictures that are required for decoding subsequent pictures. In such a system, which includes a repetitive encoding-decoding loop, quantization errors can accumulate if the quantization errors are not symmetric about zero. To avoid this, a robust quantization scheme is disclosed in
FIG. 7
(right) of the IEEE publication.
However, it has been found that the robust quantization scheme appears to be robust only for one of the disclosed compression strategies, viz. the block-predictive coding which is performed in the spatial domain. The quantization scheme is not robust if the image is compressed in the frequency domain, i.e. if it is subjected to an orthogonal transform such as the Discrete Cosine Transform (DCT), and quantization is applied to the transform coefficients.
OBJECT AND SUMMARY OF THE INVENTION
It is an object of the invention to provide a quantization scheme that is also robust in the transform domain.
This is achieved by the method as claimed in claim
1
and the corresponding arrangement claim
4
. Advantageous embodiments of the invention are defined in the dependent claims.
The invention is based on the recognition that symmetry with respect to the level zero is not sufficient. A quantizer is robust in the DCT domain if the quantization errors are symmetric with respect to each reconstruction level. Conventional quantizers, which simply round data values to the nearest integral multiple of the step size, exhibit this desired behavior for, odd step sizes. The invention provides this feature also for even step sizes. Even step sizes, particularly powers of 2, are often used in low-cost quantizers because it reduces the quantization process to simple shift operations.
REFERENCES:
patent: 5663763 (1997-09-01), Yagasaki et al.
patent: 5790705 (1998-08-01), Anderson et al.
patent: 6477276 (2002-11-01), Inoue et al.
patent: WO9838800 (1998-09-01), None
Ahn Y. et al: “Quantization Error Analysis of the RADIX-4 Signed-Digit Arithmetic” Midwest Symposium on Circuits and Systems, US, New York, IEEE 1998, pp. 1407-1410.
“An MPEG Decoder With Embedded Compression for Memory Reduction”, IEEE Transactions on Consumer Electronics, vol. 44, No. 3, Aug. 1998.
Koninklijke Philips Electronics , N.V.
Russell Gross
Wu Jingge
LandOfFree
Method and arrangement for quantizing data 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 arrangement for quantizing data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and arrangement for quantizing data will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3160347