FIELD: virtual image engineering. SUBSTANCE: method for compressing digital data of synthesized scene image sequences which is object of script involves following steps: data processing for modeling given scene basing on mathematical data, formation of image to produce synthesized image based on modeling made and on division of this synthesized image into image blocks, and differential coding of current image block basing on blocks of other synthesized image, this block being determined basing on one movement vector to provide for differential block. Movement vector is computed using mathematical data taken from script of synthesized images determining explicit movement of various scene-forming objects, this scene being formed by given sequence. Method is implemented by appropriate device that has digital data processing unit 1, reproduction computing unit 2, and movement interface unit 6, as well as circuit 3 dividing images into blocks, subtracter 4, mode selecting circuit 5, movement compensation unit 7, and coding circuit 8-10. EFFECT: optimized characteristics of image compression and quality for given data stream; enhanced efficiency of using movement information. 13 cl, 2 dwg
