Object recognition using shape-from-shading

This paper investigates whether surface topography information extracted fr om intensity images using a recently reported shape-from-shading (SFS) algo rithm can be used for the purposes of 3D object recognition. We consider ho w curvature and shape-index information delivered by this algorithm can be used to recognize objects based on their surface topography. We explore two contrasting object recognition strategies. The first of these is based on a low-level attribute summary and uses histograms of curvature and orientat ion measurements. The second approach is based on the structural arrangemen t of constant shape-index maximal patches and their associated region attri butes. We show that region curvedness and a string ordering of the regions according to size provides recognition accuracy of about 96 percent. By pol ling various recognition schemes. including a graph matching method. we sho w that a recognition rate of 98-99 percent is achievable.