本文提出了一种基于图像配准的自动目标识别算法,图像配准算法采用基于归一化互信息相似性判据,并采用模糊自适应粒子群优化算法作为搜索策略。在图像精确配准的基础上,通过图像间的相互转换,间接实现了目标的准确识别。仿真试验结果表明,该方法可以实现复杂背景下目标的准确识别。

-This paper presents a novel image registration algorithm for automatic target recognition, image registration algorithm based on normalized mutual information similarity criterion, and the fuzzy adaptive particle swarm optimization algorithm as search strategy. In the image on the basis of accurate registration, through the conversion between images, the accuracy of indirect recognition to achieve the target. Simulation results show that the complex background can accurately identify targets.