计算机工程与应用Issue(5)：172-177,6.DOI:10.3778/j.issn.1002-8331.1303-0498
一种抗视角变换的改进SIFT景象匹配算法
Improved SIFT algorithm on scene matching for anti-viewpoint chang-ing
摘要
Abstract
To overcome the great viewpoint changing between the real-time image which is acquired by the seeker and the reference image of target area during the electro-optical imaging guidance, an improved SIFT algorithm on scene matching is proposed to bridge the gap of the original one, which may lead to failed scene matching for the lack of correct matched pairs. It increases matched pairs by extracting viewpoint changing resistant points which got from the affine transformation resistant regions extracted in the Gaussian scale-space. Experimental results show that the proposed algo-rithm is robust to viewpoint changing:it can acquire nearly 5 times more correct rate of matching than the original one with the viewpoint ranges from 50° to more than 60° and can satisfy the practical need.关键词
尺度不变特征变换(SIFT)/抗仿射变换区域/视角变换/景象匹配Key words
Scale Invariant Feature Transform(SIFT)/affine transformation resistant region/viewpoint-changing/scene matching分类
信息技术与安全科学引用本文复制引用
苏可心,韩广良,孙海江..一种抗视角变换的改进SIFT景象匹配算法[J].计算机工程与应用,2015,(5):172-177,6.基金项目
国家自然科学基金（No.61172111）。 （）
