Loading patent…
AssigneeInternational Business Machines Corporation
InventorsCesar A. Gonzales, William B. Pennebaker
Estimating the term…
The present invention relates to a predictive coding method and related apparatus which (a) computes a value D representing the difference between an original graylevel value for a current pixel and a predicted value for the current pixel; (b) associates each successive range of D values to a respective quantization level, each quantization level QN having a respective QN value representing the range corresponding thereto; (c) sets a respective QERR value to D for the current pixel, when the QN value is zero; (d) re-maps m QERR values to n (n.ltoreq.m) MERR values; (e) assigns to the value MERR=O those QERR values in which .vertline.QERR.vertline..ltoreq.(+Q1 VALUE)/2 sums the MERR values for successive pixels in a QN=QO region; and (f) adjusts the QN value from the value of QO by the value of Q1 for a pixel at which the summed value reaches a predefined threshold. QN values, adjusted as appropriate, are subjected to entropy coding compression and decompression. In the decompression process, pixel values are reconstructed. According to the invention, the reconstructed values are subjected to a vertical filter in which a pixel having a reconstructed value I.sub.0 is filtered to have a value I'.sub.O =I.sub.0 +[(I.sub.-1 -I.sub.0) .sub.C- +(I.sub.+1 -I.sub.0).sub.C+ ] wherein C+ and C- are respective predefined clamping values, and wherein I.sub.-1 and I.sub.+1 are the respective reconstructed values of pixels adjacently below and adjacently above the current pixel.
In a gray level image data processor which includes (A) a predictive coder which (i) computes a predictive difference value D for the current pixel and (ii) quantizes the difference D to a QN value wherein the number of QN values is less than the number of permitted D values, and wherein ranges of larger D values correspond to QN values of relatively larger magnitudes wherein QN=Q0 corresponds to a region of substantially uniform gay level and (B) an entropy encoder which receives as input and compresses QN value data from the predictive coder, a machine-implemented method for reducing distortion in an image reconstructed from data subjected to compression by predicting coding, the method comprising the steps of:
A predictive coding encoder which supplies the output thereof to an entropy encoder, the predictive coding encoder comprising:
A vertical filter for adjusting the reconstructed value I.sub.0 for a current pixel in an image where the pixel adjacently above the current pixel has a value I.sub.+1 and the pixel adjacently below has a value I.sub.-1, the filter comprising:
Dependencies are inferred from the claim text.