An Adaptive Weighted Image Fusion Based on GPU

摘要：

In this paper, we designed an adaptive weighted fusion algorithm of remote sensing image data GPU (Graphics Processing Unit) using the programmability of the GPU, which is a parallel vector processor. The adaptive weighted image fusion algorithm provides a method to get the fusion images by selecting different weights of the ROG (Region of Interested). In this paper the algorithm was implemented on the GPU using fragmentshader, and we use the GLSL (OpenGL shading language), and the time cost was recorded. The adaptive weighted image fusion algorithm that was run on a GPU was compared with the run on a CPU. The result shows the algorithm that was run on a GPU can achieve results much quicker. The result shows that the adaptive weighted image fusion algorithm that runs on a GPU is much faster than the CPU-based algorithm in the case of large data. With the volume of fusion images data getting greater, the advantage of the velocity on GPU is more obvious then on a CPU.

关键词： Weighted fusion GPGPU HVS FBO RTT

作者: Zhang Bao-ming Lu Jun Jiang Ming-fu

作者单位: Remote Sensing Information Engineering Department, Zhengzhou Institute of Surveying and Mapping Zhengzhou, China

会议类型: 国际会议

会议名称: 2010 International Conference on Image Analysis and Signal Processing(2010 图像分析与信号处理国际会议 IASP 10)

会议地点: 厦门

会议语种:英文

页码: 130-134

在线出版日期: 2010-04-12（万方平台首次上网日期，不代表论文的发表时间）