基于边缘与角点相结合的目标提取与匹配算法
Target Extraction and Image Matching Algorithm Based on Combination of Edge and Corner

针对红外图像与可见光图像匹配困难的现状，提出一种基于边缘与角点相结合的目标提取与匹配算法。首先考虑红外与可见光图像的差异，采用边缘与角点相结合的方法直接提取红外与可见光图像中的目标区域，剔除无效信息。相似性度量采用傅里叶描述子方法。实验结果表明，该算法通过定位显著性区域，有效提升了算法的执行效率与匹配精度，与傅里叶描述子方法和CSS方法相比，新方法匹配准确精度分别提高了27.9%和17.6%；相比于傅里叶描述子耗时减少了39.1%。
To accurately match infrared and visible image, we propose a target extraction and matching algorithm based on edge and corner. We propose a method to obtain object area to increase the matching precision and speed. The method, considering the difference between infrared and visible images, combines the edge and corner detecting method to extract the target area and excludes background in infrared and visible images. Similarity measure in matching is the Fourier. The experimental results show that our method increases the average accuracy 27.9% in matching and saves 39.1% timing compared with the Fourier descriptor method. At the same time, our method increases the average accuracy 17.6% in matching compared with CSS method