张桂梅, 蔡报丰. 基于内距离形状上下文特征的形状匹配研究[J]. 南昌航空大学学报(自然科学版), 2016, 30(2): 1-7. DOI: 10.3969/j.issn.1001-4926.2016.02.001
引用本文: 张桂梅, 蔡报丰. 基于内距离形状上下文特征的形状匹配研究[J]. 南昌航空大学学报(自然科学版), 2016, 30(2): 1-7. DOI: 10.3969/j.issn.1001-4926.2016.02.001
ZHANG Gui-mei, CAI Bao-feng. Shape Matching Based on Inner-distance Shape Context[J]. Journal of nanchang hangkong university(Natural science edition), 2016, 30(2): 1-7. DOI: 10.3969/j.issn.1001-4926.2016.02.001
Citation: ZHANG Gui-mei, CAI Bao-feng. Shape Matching Based on Inner-distance Shape Context[J]. Journal of nanchang hangkong university(Natural science edition), 2016, 30(2): 1-7. DOI: 10.3969/j.issn.1001-4926.2016.02.001

基于内距离形状上下文特征的形状匹配研究

Shape Matching Based on Inner-distance Shape Context

  • 摘要: 场景中目标存在肢体变化是形状匹配中的难题,也是制约模式识别发展的瓶颈。本文从目标的轮廓特征出发,提出了一种基于内距离(inner-distance)形状上下文的形状匹配方法。首先定义了内距离的度量方法,并使用内距离代替形状上下文中的欧氏距离度量,得到了一种新的形状描述子即内距离形状上下文特征描述子,然后通过动态规划算法完成形状匹配。由于内距离对目标肢体变化和非刚性变化具有良好的鲁棒性,且相比欧氏距离对于复杂目标形状具有更好的识别力,因此本文的方法能更好地处理目标的肢体变化和非刚性变化。实验表明该方法对目标存在肢体变化和非刚性变化取得了较好的匹配效果。

     

    Abstract: Shape matching of complex shapes with articulation is a difficult problem, it is also an important problem in pattern recognition. Considering the contour of shape, this paper proposes a shape matching method which based on the inner distance shape context. First, we define a distance measure called inner-distance, which is insensitive to shape articulation and non-rigid deformation,it is often more discriminative than the Euclidean distance for complex shape. Then the Euclidean distance of shape context is directly replaced by the inner-distance. Finally, we use dynamic programming based method for shape matching. The inner-distance is insensitive to shape articulation and non-rigid deformation,it is often more discriminative than the Euclidean distance for complex shape, thus the method can effectively deal with shape with articulation and non-rigid deformation. Experimental results show that the proposed method can achieve good results for shape matching with articulation and non-rigid deformation.

     

/

返回文章
返回