龙源期刊网 http://www.qikan.com.cn 图像相似度计算算法分析 作者:王朝卿沈小林李磊 来源:《现代电子技术》2019年第09期 摘 ;要:针对灰度直方图提取算法在计算图像相似度时,受颜色分布等外界因素干扰较大的问题,提出基于特征点匹配的SIFT算法。其可通过构建尺度空间提取特征关键点,求解匹配度来弥补传统算法在计算图像相似度时的局限性。实验结果表明,相比于传统算法,SIFT算法能够通过匹配更多的特征点,从而更好地计算图像的相似度;对于一组相似图片,通过SIFT算法能提取出308个特征点,图片相似度可达63%。 关键词:图像识别; 图像相似度; 灰度直方图; 特征点匹配; 关键点; 尺度空间 中图分类号: TN911.73⁃34; TP391.4 ; ; ; ; ; ; ; ; ;文献标识码: A ; ; ; ; ; ; ; ; 文章编号:1004⁃373X(2019)09⁃0031⁃04 Analysis on image similarity calculation algorithm WANG Chaoqing1, SHEN Xiaolin1, LI Lei2 (1. School of Electrical and Control Engineering, North University of China, Taiyuan 030051, China; 2. School of Software Engineering, University of Science and Technology of China, Hefei 230000, China) Abstract: A SIFT algorithm based on feature point matching is proposed to solve the problem that the gray histogram extraction algorithm is heavily disturbed by external factors such as color distribution while calculating image similarity. It can make up for the limitation of traditional algorithm in calculating image similarity by constructing scale space, extracting feature key points and solving matching degree. The experimental results show that, in comparison with the traditional algorithm, the SIFT algorithm can accurately calculate the image similarity by means of matching more feature points; can extract 308 feature points for a group of similar images, and the image similarity can reach up to 63%. Keywords: image identification; image similarity; gray histogram; feature point matching; key point; scale space 0 ;引 ;言