基于AM测度及混合优化算法的图像配准方法

Image Registration Based on AM Measure and Hybrid Optimization Algorithm

介绍了一种基于边缘特征的图像配准方法。首先利用Canny算子提取图像的边缘;然后以AM测度作为配准准则,在优化AM峰值和配准参数的时候,先用遗传算法将配准参数收敛到最优值附近,再使用Powell法进行局部搜索;最后对待配准图像进行刚体变换及双线性插值,从而实现图像配准。实验结果表明,此优化方法不仅提高了计算速度,而且提高了配准精度,配准结果达到了亚像素级水平。

An image registration method based on edge features is presented. Firstly, the edges of image by Canny operator are extrated. Secondly, in order to obtain the maximum AM and registration parameters, the AM measure is used as the similarity measure and a hybrid optimization algorithm based on the genetic algorithm, and Powell＇s method is used as the search technique to find the optimum value. Finally, the rigid transformation and bilinear interpolation in the image prepared for registration is carried out to realize image registration. Experimental results indicate that the optimum method has a faster image registration speed and a better image registration precision. The registration results reach the level of sub-pixel.