摘要 |
PURPOSE:To form the fractal encoder not to degrade the picture quality of a regenerative image regardless of a reduction effect when performing reduction transformation to an image in a similar area. CONSTITUTION:Transformed image data 14 are generated by performing rotary transformation and reduction transformation inside a picture and in a picture element amplitude direction from a picture element pattern subtraction part 4, picture element interpolation part 7, picture element pattern subtraction part 9, average value separation part 11 and multiplier 13 to second image data 6 in the similar area in similar relation with image data 3 of an encoding target block read from a memory 2 for holding encoding target image data 1. Corresponding to the image data 3 of the encoding target block, vectors 27 and 28 as similar area position information, for which an evaluated value 18 as the product of an error 16 of the transformed image data 14 and a reduction effect amount 35 caused by the reduction transformation is minimum or smaller than a threshold value, and a picture element amplitude direction reduction coefficient 29 as transforming method information are outputted as a code 20 by a code output part 19. |