摘要
图像的半自动配准是当前比较先进的图像配准技术。该方法基于初步定位基础,利用模板相关的匹配方法,根据遥感图像之间灰度分布的相关性,选取数量丰富的配准控制点,利用相关系数准则进行图像间的相似性度量,可达到较高的配准精度。实验结果表明,该匹配方法用于遥感图像的配准能够有效提高图像处理的速度以及校正图像的定位精度。
Image automatic registration is a kind of advanced technique of image registration. In this paper, based on the preliminary location and template correlation match, this method is used to select numerous registration control points according to remote sensing image Grayscale distribution＇ s relativity ; it uses correlation coefficient principle to measure the similarity of remote sensing images to obtain higher registration precision. The experiment results show that this method can increase the speed of image processing and improve the location precision.
出处
《无线电工程》
2008年第5期32-34,共3页
Radio Engineering
关键词
模板相关匹配
相关系数
控制点半自动生成
template correlation match
correlation coefficient
semiautomatic generation of control point
