期刊名称：International Journal on Smart Sensing and Intelligent Systems
印刷版ISSN：1178-5608
出版年度：2014
卷号：7
期号：1
页码：321-336
出版社：Massey University
摘要：Image-matching is a basic task in image procession used to (geometrically) match two or more images such as different times, different sensors and different view points. Improved image-matching technology has been widely applied in medical image, remote sensing image, computer vision military and so on. What put forward at the present can be most divided into matching method based on gray and features. The article presents a scheme that is based on subtraction images matching with dynamic selection of templates, which overcomes these more sensitive disadvantages like gray, rotate, deformation and occlusion to achieve accurate calibration for different images
