Edge Detection to Enhance Definite Range Image Segmentation by Clustering Algorithms
In this paper, the authors considered the problem of generating correct topological information of objects from edge detection to enhance range image segmentation by clustering algorithms. In order to fulfill this goal, it is necessary to obtain reliable edge information. Then, the edge map can supply required information to the clustering algorithm based on features extracted from the topology of the observed objects, leading to important improvements. The approach contributes to extent the automation level of an entire vision system under, development, enhancing its reliability.