An Image Matching Algorithm based on Mutual Information for Small Dimensionality Target
Qihong Yang, Qiang Wei · 2018
To overcome the difficulty of detecting and recognizing the small dimensionality target during precision strike, we propose an image matching algorithm based on mutual information. First, the characteristic of typical small dimensionality target is analyzed, in rough matching stag, Calculate the joint entropy of template image and candidate matching image. then, in precision matching stage, taking the position of maximum mutual information obtained in rough matching as the center, matching pixel by pixel was carried out to obtain the final matching point. Finally, template images was adaptive updated, in military operations, as time changes, the state and the located of the target to be tracked and identified may have changed. Experimental results of some observed optical and infrared images show the method proposed in this paper is verified and can be precisely aligned.