Fusion of Intensity and Range Image Data for 3D Object Recognition

This paper describes a rules-based method to fuse the intensity and range image data obtained simultaneously from a laser scanner which measures the range of an object and its itensity for 3D object recognition.The intensity and range image information are commonly used as input data for intelligent robot vision.The purpose of this work is to show how the data are fused together to positively recognize 3D objects.The fusion is on two levels (edge detection and curvature analysis),and includes:extraction of edges and their attributes, consistency of the edge attributes,estimation and analysis of the curvature-type.