改进PSO算法在图像配准中的应用

图像配准是目前图像处理中的热点,很大程度上因为它是图像融合的前提,具有重要的研究价值.粒子群算法是一种基于群体搜索策略的自适应随机算法,由于该算法实现简单、参数较少等特点被广泛使用.针对标准PSO易陷于局部最优的问题,对标准PSO进行了改进.实验过程中,参考图像和浮动图像的相似性测度采用的是峰值信噪比(PSNR).通过实验结果对比,改进后的算法由于克服了容易陷入局部最优解的问题从而提高了配准精度.

Improved PSO algorithm in image registration

Abstract: Image registration is the premise of image fusion and also hot spot in the current image processing. It has important value. The algorithm of PSO with an adaptive randomized algorithm based on swarm search strategy is widely used because the algorithm is simple and with fewer parameters. This paper was improved the standard PSO because the standard PSO easily trapped in local optimal problems. In this experiment, the PSNR is similarity measure as the reference image and the floating image. Experimental results show that the method overcomes the problem of easily fall into local optimal solution with high precision.