一种基于点特征与Freeman链码的红外与微光图像配准算法研究

Investigation on Infrared and Low Light Level Image Registration Algorithm Based on Point Feature and Freeman Chain Code

由于红外与微光图像成像原理的不同,成像特征的巨大差异,研究提出了一种基于点特征与Freeman链码的红外与微光图像配准算法。目的是解决红外与微光图像配准中特征点提取较复杂、特征匹配难的问题;采用优化的Harris角点检测算法进行特征点提取,结合环形灰度区域、RSTC不变矩和Freeman链码对红外与微光图像进行特征点匹配。实验结果表明该算法能够提取出有效的匹配点,能够有效地解决红外与微光图像配准中遇到的视场不统一、旋转、平移问题。

The paper puts forward a registration algorithm of infrared and LLL image based on the point feature and the Freeman chain code, as a result of different principle and features of infrared and low light level images. This algorithm has been designed to solve the problem of complex feature point extraction and difficult feature matching of infrared and low light level image registration. This algorithm extracts feature points by optimizing the Harris corner detection algorithm and matches feature points combining with the annular gray region, RSTC moment in variants and Freeman chain code. Experimental results show that such algorithm can extract the matching point effectively, meet the effective solution in the field that is not uniform, rotation, translation of infrared and low light level image registration problems.