Remote Sensing Images Registration Based on Optimized Harris Operator

摘要：

A semiautomatic match method based-on manual prior knowledge and point feature is proposed in this paper. In this method,the main feature directions and overlapped areas were firstly selected by manual knowledge. Then, the correlation coefficient and Harris operator were applied for coarse registration, and control points with coarse difference were deleted by adopting variance factor test and Baarda data detect method. At last, a dense triangle gird was built, and the precise image match is archived by using facet differential correction. Experimental results show that this method can obtain fast and highly-precies result for multi-source remote image registration.

关键词： registration prior knowledge Harris-Laplace gross error

作者: Cheng Xiran Yang Cunjian

作者单位: Research Center of Remote Sensing and GIS Applications, Key Lab of Land Resources Evaluation and Monitoring in Southwest, Ministry of Education, Sichuan Normal University .Chengdu, China

会议类型: 国际会议

会议名称: 2010 International Conference on Circuit and Signal Processing(2010年电路与信号处理国际会议 ICCSP 2010)

会议地点: 上海

会议语种:英文

页码: 528-532

在线出版日期: 2010-12-25（万方平台首次上网日期，不代表论文的发表时间）