基于布谷鸟算法的数字散斑相关方法优化

Optimization of digital speckle correlation method based on cuckoo algorithm

将布谷鸟算法引入到数字散斑相关方法的搜索中,通过设定隔离边界对数字散斑图像分块,在每一个子块上利用布谷鸟算法进行局部搜索,在此基础上利用爬山算法进一步搜索峰值。全局搜索可以设定一个较弱的收敛条件,快速搜索到相关系数分布单峰上,充分利用了布谷鸟算法的全局搜索能力与爬山算法的快速收敛特点;最后通过模拟散斑图片进行了相关仿真,并通过实际人工散斑进行了相关实验,表明该算法能有效稳定测量刚体的位移量。

The cuckoo algorithm(CS algorithm)is introduced into the search of digital speckle correlation method(DSCM).By setting the isolation boundary,the digital speckle image is divided into blocks.The cuckoo algorithm is used for local search on each sub-block.On this basis,the mountain climbing algorithm is used to further search the peak value.Global search can set a weak convergence condition and quickly search the single peak of correlation coefficient distribution,making full use of the global search ability of cuckoo algorithm and the fast convergence characteristics of mountain climbing algorithm.Finally,the simulation of speckle images is carried out,and the related experiments are carried out through the actual artificial speckle,which shows that the algorithm can effectively and stably measure the displacement of rigid body.