| 摘要 |
PURPOSE:To attain high efficient coding at a low cost with simple configuration by dividing 3-dimension volume data as sets of cubes or rectangular parallelpipeds, selecting adaptively a section suitable for orthogonal transformation and applying linear or 2-dimensional orthogonal transformation to the section. CONSTITUTION:Three-dimension volume data are segmented to sets of cubes or rectangular parallelpipeds by a data segmentation section 1, and a section selection section 2 adaptively selects a section suitable for orthogonal transformation of picture data of the cube. Furthermore, a selection section 4 adaptively selects which of linear and 2-dimensional orthogonal transformation is to be executed for the section. Then the section selected by the selection section 2 is scanned by a scan converter 3 and an orthogonal transformation device 6 applies linear/2-dimensional orthogonal transformation to the section. Prior to the orthogonal transformation, a prediction section 5 takes matching of the data with data already coded to implement prediction. The data after orthogonal transformation by the orthogonal transformation device 6 are quantized by a quantizer 7 and coded by a coder 8 and the result is outputted. As a result, the coding is implemented with high efficiency at a low cost with simple configuration. |
