CURVE SEGMENTATION UNDER PARTIAL OCCLUSION

2-D shape boundary segmentation is required as a fundamental and impor tant stage in the recognition of partially occluded objects. We introd uce here a new segmentation method capable of extracting a controlled number of segments along a smooth boundary curve. This new approach is invariant to similarity transformation, and partial occlusion has onl y marginal influence on the segmentation of the visible part. The basi c concept is to transform the curve into another one which intersects itself. Points or intersection of the new curve are re-transformed to the original curve and serve as endpoints of segments. Properties of t he transform are discussed, and conditions for existence of intersecti on points are given. Simulation results of gray level images are prese nted, and advantages of our method over conventional approaches relyin g on singular points of the curvature-are discussed.