摘要
针对传统匹配算法中匹配精度与匹配速度不能同时兼顾、寻优能力差的问题,提出一种新的分层匹配算法。第一层采用改进的遗传算法:将边缘算子提取的边缘图像作为遗传操作对象,减少匹配计算量;选择算子采用最佳保留策略,有效降低误匹配率,提高了算法的准确性;交叉算子采用自识别交叉算子,避免局部收敛问题,提高了算法的全局性。第二层在第一层的基础上进行区域匹配,缩短匹配时间,提高了算法的匹配效率。实验表明提出的算法能够提高算法的匹配精度,同时缩短算法的匹配时间,增强了算法的寻优能力。
A new hierarchical matching algorithm is proposed to solve the problems that the matching accuracy and speed cannot be considered simultaneously and the searching ability is poor in the traditional matching algorithm.In the first layer,the improved genetic algorithm is used:the edge image extracted by edge operator is used as the object of genetic operation to reduce the amount of matching computation;the selection operator adopts the best reservation strategy to effectively reduce the mismatch rate and improve the accuracy of the algorithm;the crossover operator uses self-identification crossover operator to avoid the local convergence problem and improve the global nature of the algorithm.The second layer performs region matching on the basis of the first layer,which shortens the matching time and improves the matching efficiency of the algorithm.Experiments show that the proposed algorithm can improve the matching accuracy,shorten the matching time and enhance the optimization ability of the algorithm.
作者
王金鹤
衣家欣
孟凡云
WANG Jin-he;YI Jia-xin;MENG Fan-yun(Computer Engineering Institute,Qingdao University of Technology,Qingdao Shandong 266033,China)
出处
《计算机仿真》
北大核心
2019年第12期224-227,共4页
Computer Simulation
关键词
图像匹配
遗传算法
区域匹配算法
Image matching
Genetic algorithm
Region matching algorithm
