高斯形状上下文描述子  

Gaussian shape context descriptors

作　　者：冯祥卫 冯大政[1] 侯瑞利 

机构地区：[1]西安电子科技大学雷达信号处理国家重点实验室,陕西西安710071

出　　处：《西安电子科技大学学报》2016年第4期45-50,共6页Journal of Xidian University

基　　金：国家自然科学基金资助项目(61271293)

摘　　要：利用高斯窗函数,构建了一种新的形状上下文描述子,并将该描述子应用到形状配准当中.高斯形状上下文以参考点为中心建立一组高斯窗函数,根据中心点与其他采样点之间的欧氏距离,利用尽量多的采样点计算各个窗函数的值并组成形状描述子,描述点与点之间的位置关系.通过比较形状描述子之间的相似性,能够较为准确地找到两个形状间点与点的对应关系.实验表明,使用高斯形状上下文作为形状描述子,比传统的形状上下文能够更加准确地找到点与点的对应关系,使配准算法快速收敛,具有良好的抗噪性和鲁棒性.The shape context （SC） descriptors based on the Gaussian window （GSC） are proposed. The GSC descriptors can be used to match shapes. Encireling the reference point, the GSC sets up a group of Gaussian window functions. According to the Euclidean distance between the center points and the sampling points, the GSC describes the spatial relations between sampling points based on the shape descriptors calculated by Gaussian window functions by the use of as many sampling points as possible. More accurate correspondences between two shapes are determined by comparing the similarities of the shape descriptors. Experimental results show that the GSC can find more accurate correspondence relations and provide a good initial value. The proposed descriptors can also make the matching algorithm converge fast, while keeping a good anti-noise and robust performance compared with SC.

关 键 词：形状上下文 高斯窗 形状描述子 形状配准 鲁棒性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]