| 摘要 |
A range image is input, a plurality of range images having different resolutions are generated from the input range image, normal vectors are calculated for each of the plurality of generated range images, an edge map is formed for each of the plurality of range images based on the calculated normal vectors, a single edge map is formed by synthesizing the plurality of formed edge maps, and polygon data for describing the surface shape of the object represented by the input range image are formed based on the synthesized edge map. It is thereby possible to form polygon data having high precision which represent characteristics of the surface shape of the object in an excellent manner with a small amount of data. <IMAGE> |
