论文部分内容阅读
提出了一种基于图象边缘轮廓信息的多源遥感图象匹配定位方法 ,其目的是利用定位精度较高的高分辨率遥感图象对低分辨率图象实现子象素级的匹配定位 .该方法有效地利用了多源遥感图象中共有的区域结构信息 ,将特征匹配和最小二乘影象匹配相结合 ,具有较好的普适性 ,且运算快速、抗噪性能好 .采用该方法进行 NOAAAVHRR图象和 L andsat TM图象、1∶ 10 0万数字地图的边缘图象匹配 ,并应用于 NOAA AVHRR图象的几何精校正中 ,可实现精确的子象素匹配 .
A method of multi-source remote sensing image matching positioning based on edge information of image edges is proposed. The purpose of this method is to use sub-pixel-level matching positioning of low-resolution images with high-resolution remote sensing images with high positioning accuracy. The method effectively utilizes the regional structure information shared by multi-source remote sensing images, combines the feature matching with the least-squares image matching, has good universality, has fast operation and good anti-noise performance, Methods The edge image matching between NOAAAVHRR images and L andsat TM images, 1: 100,000 digital maps, is applied to the geometric refinement of NOAA AVHRR images for accurate sub-pixel matching.