基于自适应遗传算法的图像匹配

为了解决图像匹配中计算速度慢和匹配精度不高的缺陷，提出了一种基于自适应遗传算法(AGA)的匹配方法，该算法与传统遗传算法的不同在于其交叉概率和变异概率随个体的适应度值而变化，避免了后者易陷入局部极值的缺陷，从而增强了算法的快速性和全局收敛性能．图像与模板的相关值是一多峰值函数，模板匹配实质上是多峰值寻优过程．将AGA应用到图像匹配，是以相关值为适应度函数，通过选择、交叉、变异等遗传操作，对遗传个体进行迭代寻优，找出图像中的最佳匹配点．实验结果表明，基于该算法的图像匹配具有运算量小、匹配精确等优点，且算法稳定．

Image matching based on adaptive genetic algorithm

To solve the problem of slow computation speed and low accuracy in image matching, a new approach to image matching using adaptive genetic algorithm(AGA)was proposed. The difference between AGA and standard genetic algorithm(SGA)was that the probabilities of crossover and mutation were varied depending on the fitness values,thus improving the performance in searching speed and global convergence.The correlation value of an image and its template was a multimodal function, so template matching can be seen as an AGA searching optimum point process composed of select, crossover and mutation operators which were varied depending on the correlation value. Experiment show that the approach is simple, reliable and stable in operation, and has high accuracy in matching.