Abstract

For 3-D video applications, dense depth maps are required. We present a segment-based structure-from-motion technique. After im- age segmentation, we estimate the motion of each segment. With knowl- edge of the camera motion, this can be translated into depth. The optimal depth is found by minimizing a suitable error norm, which can handle occlusions as well. This method combines the advantages of motion es- timation on the one hand, and structure-from-motion algorithms on the other hand. The resulting depth maps are pixel-accurate due to the seg- mentation, and have a high accuracy: depth difierences corresponding to motion difierences of 1/8th of a pixel can be recovered.