海量遥感影像分块逆映射快速坐标转换  

Massive Remote Sensing Image Fast Coordinate Transformation Algorithm of the Inverse Mapping Blocking Method

作　　者：范铀 伍超云 王琨[2] 范冲[2] 

机构地区：[1]广东南方数码科技股份有限公司,北京100055 [2]中南大学地球科学与信息物理学院,湖南长沙410083

出　　处：《测绘通报》2017年第4期53-57,共5页Bulletin of Surveying and Mapping

基　　金：国家973计划(2012CB719904)

摘　　要：提出了一种海量遥感影像分块逆映射快速坐标转换算法,并针对分块窗口的形状和大小对处理过程所占内存与处理效率造成的影响作了试验与分析并得出结论。该算法使得处理过程所占内存不受原影像与处理后影像的尺寸大小所限制,并且通过与主流遥感影像处理软件对比试验的结果可以看出,该算法可以完成超大幅遥感影像与分幅遥感影像的坐标转换处理,处理分幅影像时有效地避免了影像间的接缝,同时处理效率达到主流的专业遥感影像处理软件水平。This paper proposes massive remote sensing image coordinate transformation algorithm of the inverse mapping blocking method, analyzes the influence from the geometric shape and size of the blocks to the memory usage and efficiency of coordinate transforming processes with experiments then draws conclusions. This algorithm makes the memory usage of coordinate transforming processes independent from the size of the original image or the transformed image. By experiments contrast with main current remote sensing image processing software, it is observed that this algorithm is able to achieve the coordinate transforming processes for huge remote sensing images and subdivided images, in the process of subdivided images coordinate transform this algor/thm is able to avoid the seams between images, and the processing efficiency of this algorithm achieves the level of main current remote sensing image processing software.

关 键 词：关键词:逆映射 分块 大幅影像 分幅影像 坐标转换 

分 类 号：P237[天文地球—摄影测量与遥感]