遥感图像海陆分割的GPU并行加速

摘要

Segmentation algorithm of land and sea is a very important step in remote sensing image target identification.Results of dividing land and sea directly affect the efficiency and accuracy of next target recognition.With the massive growth of remote sensing data, the image processing speed is slower.So the efficiency of segmentation algorithm execution is becoming much more important.The use of CUDA (compute unified device architecture) can transplant the segmentation process from CPU to GPU, accelerating the segmentation algorithm When processing the image date of 2 000×2 000 pixels, experiment proved that this method can complete the segmentation of land and sea in image in 25 ms.And it can be used to build a ground or airborne "real-time" processing platform.%海陆分割是光学遥感图像海上目标识别过程中一个重要步骤,海陆分割的结果直接影响下一步目标识别的效率和正确率;而遥感数据的大量增长,使得图像处理速度变慢,因此分割算法的执行效率变得越来越重要.利用Nvidia开发的统一计算架构CUDA (compute unified device architecture),将海陆分割流程的一系列函数移植到GPU (graphic processing unit)上进行并行处理,能够有效提高算法执行速度.经实验验证最终完成2 000×2 000大小的图片在11 ms内的海陆分割处理.该方法能够满足对于图像数据的传输在25 ms内的处理,适合用于建立地面或者航空搭载的"实时"处理平台.