STRUCTURE RECOVERY FROM MULTIPLE IMAGES BY DIRECTLY ESTIMATING THE INTERSECTIONS IN 3-D SPACE

This paper presents a new approach to the recovery of 3-D structure fr om multiple pairs of images from different viewpoints. Searching for t he corresponding points between images, which is common in stereopsis, is avoided. Extracted edges from input images are projected back into 3-D space, and their intersections are calculated directly. Many fals e intersections may appear, but if we have many pair images, true inte rsections are extracted by appropriate thresholding. Octree representa tion of the intersections enables this approach. We consider a way to treat adjacent edge piexels as a line segment rather than as individua l points, which differs from previous works and leads to a new algorit hm. Experimental results using both synthetic and actual images are al so described.