免疫粒子群优化算法在图像融合中的应用

Application of immune particle swarm optimization algorithm in image fusion

提出了一种基于图像分块的小波多聚焦图像融合方法,并将免疫粒子群优化搜索策略应用于多聚焦图像融合子块寻优中。将图像子块作为粒子,以寻求最优组合分块形成的融合图像。利用两种评价参量,即信息熵和交叉熵进行不同图像融合方法的分析及效果评价,实验结果表明,其融合性能优于对图像只进行分块而不作小波分解的融合方法和只作小波分解而不进行分块的融合方法,该方法既能消除块痕迹,又能节约运算量,取得了很好的融合效果。与标准粒子群相比,免疫粒子群的收敛性能和达优率更好。

Image fusion method based on wavelet transform and swarm optimization search algorithm is developed for fusion of two block segment of the images is proposed.An immune particle spatially registered multi-focus images.In order to get the optimal image,the size of block is defined as particle.Two evaluation criteria such as information entropy and cross entropy are used on the analysis and effect evaluation of different fused images.The performance of this method is better than the performance of the existed fusion method.It can eliminate the blocking artifacts of the fusion image and save the time of fusion.The experiment results show that the method is an effective method.W-PSO is tested for performance comparison with IPSO,and the result show IPSO has the higher efficiency than W-PSO.