摘要 |
Even in a situation where an image of the rear side of an object cannot be picked up, the shape of the object can be generated with high accuracy. In the case where an object reflected on images is projected from viewing positions of the images to a projection space for every plurality of the images picked up from surroundings of the object, an object shape generating method detects projection regions from an image projecting surface in the projection space up to an image projected surface at a distance of a prescribed length in a depth direction, respectively, and is set to extract a feasibly common portion from each of the detected projection regions. |