| 摘要 |
<p>Color digital image data is divided into data sets having an image size most suitable for processing. The data of each image data set is expressed, for each color channel, in a three-dimensional space composed of a plane showing XY ordinates and a Z-axis which is perpendicular to the plane and along which a luminance distribution is shown. This is divided into segments of a given luminance range or into segments in a binary tree. Thus, the data structure is converted to a layered one including from a low-definition layer to a high-definition layer. The data in each layer is compressed and stored. When used, the image data from the lowest layer to a higher layer depending on the use is extracted and an image of a desired definition is reproduced. As necessary, image data in a further higher layer is extracted and combined with the image data previously extracted to reproduce a high-definition or complete image. As a result, a high-definition image can be stored in one image file without any loss, and a reproduced image adapted to the use and the required quality by extracting only necessary layer data is obtained. The data processing efficiency is greatly improved. Hardware resources and network resources can efficiently be utilized.</p> |
