Abstract: It is difficult to recognize objects when they are flexibility change and partial occlusion. The traditional algorithm based on shape matching can’t be effectively used to recognize them. To solve this problem, a new method for recognition flexibility change and partial occlusion objects combined skeleton and histogram is proposed. The combine-points of skeleton are detected, and the shape of objects is divided into several regions based on combine-points; The invariant of sub-region is established and construct the invariant features in a histogram; A new similarity measure function is built for the histogram; The similarity of both model and target is computed to judge whether two shapes are match. In this paper, skeleton combine-points are extracted, compared with skeleton end-points, this method is more stable and the matching accuracy has also improved. This algorithm uses sub-region description, which can effectively recognizes objects of flexibility change and partial occlusion. Theoretical analysis and experimental results show that the algorithm is effective.

Key words: skeleton, combine point, statistics histogram, shape matching

摘要： 目标的柔性变化和局部遮挡是形状匹配的主要困难，传统的形状匹配算法对这类问题常不能奏效。针对该问题，提出了一种新的接合骨架和统计直方图匹配柔性变化和局部遮挡目标的方法。算法首先提取骨架接合点，根据骨架接合点将目标形状分块；基于每个骨架接合点构造各个子块的不变特征并将其直方图化；再构建了一种新的相似性度量函数；通过对骨架接合点的匹配实现目标形状的匹配。由于算法是提取骨架接合点，较骨架端点提取更稳定，从而能提高匹配精度；此外算法采用分块描述匹配的思想，所以能有效处理局部遮挡和柔性变化。理论分析和实验结果均表明算法是有效的。

关键词: 骨架, 接合点, 统计直方图, 形状匹配