摘要
三维成像载荷是一种能够同时采集激光雷达数据以及CCD影像,并快速融合生成三维立体影像的新型载荷,其所获取的激光点云具有规则格网分布的特点,适合采用并行的方式进行处理。结合计算统一设备架构(CUDA)对三维成像载荷快速融合处理算法核心部分进行了针对性的并行方案设计,实现了三维影像的高程值内插与平面坐标解算的快速处理,与基于CPU的串行执行模式相比,基于CUDA的三维成像载荷快速融合处理方法可显著提高计算效率,并行计算加速比可达约12倍,可支持成像区域三维影像的实时生成。
The 3D Imaging Payload is a new type of load which can get laser point clouds and CCD images at the same time.And it can generate three-dimensional images by fast fusion.The laser point clouds acquired by the 3D Imaging Payload have a regular grid distribution and are suitable for parallel processing.Based on those characteristics and Compute Unified Device Architecture（CUDA）, a parallel scheme of fast fusion has been designed and implemented.The parallel scheme uses GPU to realize the interpolation of the eleva- tion value and calculation of the plane coordinate by collinearity equation.Compared with serial processing, the parallel scheme of fast fusion can significantly improve the computational efficiency and the speedup can reach at 12 times.And it also can support the real-time generation of the three-dimensional images.
出处
《遥感技术与应用》
CSCD
北大核心
2017年第3期443-448,共6页
Remote Sensing Technology and Application
基金
国家863计划项目“主被动一体化迅捷多维成像技术”(2015AA123801)
