| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To provide a technique for efficiently coding a coefficient after quantization in image encoding processing for performing frequency conversion and quantization with a pixel block as a unit. <P>SOLUTION: In an image encoder, image data are supplied to a boundary bit position determination section through a block division section, a sequence transformation section, and a coefficient quantization section. For each coefficient in an aimed block, the number of bits from a bit position of a first "1" in a direction from the most upper bit to the lowest bit to the lowest bit is detected as the number of effective bits, and the number of occurrences for each number of effective bits is counted. In a summation function represented by following formula for variable b (b=0, 1, 2, ...), namely: S(b)=ΣN(Bmax-i) (i=0, 1, ..., b), where the number of occurrences of the number of effective bits B is denoted by N(B) and the maximum number of effective bits of which the number of occurrences is not less than 1 is denoted by Bmax, the maximum integer b of which the relationship with a threshold Th satisfies S(b)<Th is obtained, and a boundary bit position B is determined from b. A coefficient coding section separates respective coefficients into upper and lower bit sections for coding, at the determined boundary bit position. <P>COPYRIGHT: (C)2010,JPO&INPIT</p> |
