Recognition of Partially Occluded Binary Objects using Support Values between Dominant Point Labels

This paper deals with the recognition of partially occluded objects in the work space. The boundary in the binary image is extracted and feature points are detected along the boundary by calculating the curvature. The feature points are used to establish the correspondence between the geometrically similar points of the input and the model image. The pair-wise assignment graph is constructed and unambiguous dominant point labeling is obtained by calculating the support values from neighbouring dominant point labels. An experiment on several sample data is given to show the result of the proposed algorithm.