摘要
由于传统Hausdorff距离算法对减少非零均值高斯噪声的干扰不明显,且匹配精度不能满足惯导的要求,因而提出了一种改进的算法分支点的加权Hausdorff距离(Weighted Hausdorff Distance,WHD)算法,并给出了权值的求取公式。方法能有效匹配被非高斯噪声污染的图像,提高景象匹配的精度和速度,增强算法的鲁棒性。并对提出的WHD算法与部分的平均距离算法(PMHD)分别作仿真实验进行比较,证明了前者算法的实用性和有效性。
Due to reducing disturbing of nonzero Gauss noise is not evident with Traditional Hausdorff distance, and matching precision can not be applied to the inertial navigation, an improved approach weighted Hausdorff distance algorithm based on characteristic extraction is proposed, and the weighted formula is given , This approach can effectively match image which is stained by non - gauss noise, and enhance precision and speed of scene matching and robust of algorithm. Experimental results show the efficiency and practicability of this algorithm.
出处
《计算机仿真》
CSCD
北大核心
2010年第2期88-91,共4页
Computer Simulation
关键词
豪斯多夫距离
分支点提取
景象匹配
加权豪斯多夫距离
Hausdorff distance
Bifurcation extraction
Image matching
Weighted Hausdorff distance
