| 摘要 |
<p>A color digital image data set is divided into subsets of an image size of an optimization unit. Each image data subset of each color channel is expressed in a three-dimensional color space composed of an X-Y coordinate plane and a luminance distribution along Z-axis perpendicular to the X-Y plane. This is divided into plural steps in a given luminance range or divided like a binary tree. Thus the data is converted into data of hierarchical structure of from low to high definition layers to create raster data. A data set of data representing an outline and a data set of data representing the color inside the outline are created from each layer data set. When data is used the outline data on each layer is superposed on the color data on the layer from the lowest layer to the layer of the definition corresponding to the use so as to create an artificial image containing the outline data of a desired quality. Therefore, a raster data set is efficiently converted to an artificial image having a definition corresponding to the use and requirement from one raster data set, and the image quality is effectively and readily controlled.</p> |
