基于GPU的GeoEye-1影像正射纠正  

Ortho-rectification of Geo Eye-1 Image Based on GPU

作　　者：颜源[1] 张靖[2] 李紫凝 王洁茹[1] 

机构地区：[1]长江航道测量中心,湖北武汉430010 [2]武汉大学,湖北武汉430079

出　　处：《地理空间信息》2016年第4期70-72,11-12,共3页Geospatial Information

基　　金：国家自然科学基金资助项目(41301367)

摘　　要：在介绍和分析GPU并行编程方法的基础上,给出了在GPU上进行Geo Eye-1影像正射纠正的算法过程。利用NVIDIA Quadro 600显卡对31 985像素×15 251像素的3波段Geo Eye-1影像进行正射纠正,其总体速度比没有加速的算法提升40%以上。试验结果表明,采用GPU并行处理技术可以提升遥感影像处理效率。According to the introducing and analysing GPU parallel programming method, the Geoeye-1 image ortho-rectification algorithm based on GPU was presented in this paper. We used the NVIDIA Quadro 600 graphics card to ortho-rectify a three bands Geoeye-1 image with 31 985 pixels×15 251 pixels. The result shows that the speed can be increased by more than 40% compared with normal method. So GPU parallel programming method is a good choice to improve the remote sensing image processing speed.

关 键 词：正射纠正 并行计算 GPU CUDA GeoEye-1 

分 类 号：P231[天文地球—摄影测量与遥感]