| 摘要 |
PURPOSE: A method of compressing an image for image decoding based on a motion compensation is provided to set up a size of a pel block to define a bundle and to supply a classified differential coding(CDC) algorithm for retrieving values between adjacent pels and change patterns of the values, so as to reduce a calculation complexity in an image compression coding process and to satisfy a random access capacity of a memory. CONSTITUTION: Images are divided into predetermined sizes of sub block units to be compressed. The divided images are defined as a bundle. Bundle types reflect image characteristics from difference values between adjacent pels of the bundle. The bundle types are identified. Each of the identified bundle types includes a bundle type code(Tc). Bits are allocated with more than one combinations of an anchor code(Ca), a prediction code(Cp), a slope code(Cs), and a slope compensation code(Cc). The allocated bits are coded. |
