Detailed 3D mapping based on image edge-point ICP and recovery from registration failure

This paper presents a method of 3D mapping using a binocular stereo camera with automatic recovery from registration failure. We employ edge points as map element, and apply a variant of ICP algorithm to the inter-frame registration. The method detects failure in the registration due to erratic camera motion or moving objects and recovers from the failure by searching a good image to resume the registration using edge-point SIFT descriptors. In experiments, our method successfully coped with registration failure in 3D mapping in indoor and outdoor environments under various conditions.