Motion image decoding apparatus and method reducing error accumulation and hence image degradation
AssigneeSanyo Electric Co., Ltd.
InventorKatsunori Hirase
Life of this patent
Estimating the term…
Abstract
An Hadamard transform coding circuit changes a rounding method for each Hadamard transform block to prevent a rounding operation from providing an error biased in one direction, as seen in a picture in its entirety, for data compression and rounding. More specifically, after Hadamard transform there exist decimal parts 0.0, 0.25, 0.5 and 0.75, and if a numerical value is simply rounded off a probability would be increased that the numerical value is biased to increase in absolute value. Accordingly for 0.5 a rounding operation to provide an integer switches for each block. Thus a motion image decoding apparatus can be provided to reduce error accumulation and suppress flickering, color variation and other similar noticeable degradations.
Claims10
A motion image decoding apparatus for decoding compressed image data obtained through timebase predictive coding effected to compress a motion image, comprising:
- an image reproduction portion receiving said compressed image data to generate reproduced image data;
- an orthogonal transform and compression portion receiving an output from said image reproduction portion to effect orthogonal transform for each predetermined data transform block for data compression, said orthogonal transform and compression portion switching for each said predetermined data transform block a rounding system applied after said orthogonal transform, said switching of the rounding system corresponding to alternating between a first rounding process including adding a first value to or subtracting the first value from a numeral value and dividing result by a second value, and a second rounding process including adding a third value to or subtracting the third value from the numeral value and dividing result by the second value, the third value being equal to two times the first value;
- and
- a storage receiving an output from said orthogonal transform and compression portion for storing therein reference image data for said predictive coding.
A method of decoding a motion image, comprising the steps of:
- generating first reproduced image data based on source image data reproduced from a signal of compressed image data obtained through timebase predictive coding in compression of a motion image, or generating said first reproduced image data based on the reproduced source image data and reference image data;
- switching a rounding system after orthogonal transform for each predetermined data transform block of said first reproduced image data, effecting orthogonal transform coding, and generating second reproduced image data having an amount of data smaller bitwise than said first reproduced image data, said switching of the rounding system corresponding to alternating between a first rounding process including adding a first value to or subtracting the first value from a numeral value and dividing result by a second value, and a second rounding process including adding a third value to or subtracting the third value from the numeral value and dividing result by the second value, the third value being equal to two times the first value;
- storing to a reference image memory said second reproduced image data required to generate said reference image data;
- and
- generating from said second reproduced image data stored in said reference image memory said reference image data corresponding to said first reproduced image data.
Dependencies are inferred from the claim text.