Abstract: Aiming at the problem of slow speed and noise-sensibility in image matching，the paper suggests a GPSO approach to image matching，which is based on grey theory and particle swarm optimization.Firstly，several matching positions and updating speeds are acquired by the initialization of the particle swarm.Secondly，a referential sequence and a comparative sequence are separately constructed by the histogram information of the template image and the current searching subimage.Thirdly，based on the grey relational degree between the two sequences，a fitness function is designed.And then，guided by the personal experience and the global experience，all of the particles concurrently approach to the best matching position generation by generation.The experimental results indicate that the algorithm not only obtains the precise position，but also obviously increases the matching speed and the robustness.

摘要： 针对图像匹配中速度慢、抗噪性差等问题，提出一种基于灰色理论和粒子群优化的快速图像匹配算法——GPSO算法。该算法首先通过粒子群初始化，获得待匹配的多个初始位置和更新速度；然后，利用模板图和当前搜索位置子图的直方图信息，形成参考序列和比较序列，设计基于两类序列间灰色关联度的适应度函数。在此基础上，各粒子根据个体经验和社会经验，利用群体智能的高效并行寻优能力，逐代逼近最佳匹配位置。实验显示，本算法在保证了一定匹配精度的情况下，明显提高了匹配速度和鲁棒性。