| 摘要 |
PROBLEM TO BE SOLVED: To prevent the deterioration of motion detection accuracy even when a motion vector is calculated after the number of bits of a picture is reduced by receiving the output data from a quantizer, calculating the motion vector, and detecting and outputting the minimum value. SOLUTION: A quantizer 40 receives the input of pixel data, and quantizes and outputs a reference pixel value (r) and a pixel value (s) to be retrieved. Plural processing units 20 connected in parallel are connected with the output of the quantizer 40, and the processing units 20 receive the output data from the quantizer 40, and calculate and output motion vectors. A minimum value detecting part 30 is connected with the output of the processing unit 20, and the minimum value detecting part 30 receives the output data from the processing units 20, and detects and outputs the minimum value. Thus, instead of using one quantizer for the input part of a motion vector detecting part, a hardware configuration can be simplified by largely simplifying the structure of the processing unit compared with the conventional manner. |
