基于边缘强度匹配的图像融合并行算法的研究

摘要

图像融合是图像理解和计算机视觉领域中的一项重要课题。随着图像规模的增大，图像融合面临着处理速度的挑战。最近几年，GPU面向通用计算应用得到了快速发展。本文基于GPU编程模型和硬件特性，深入研究了基于边缘强度的图像融合算法，提出了该算法的并行模型。实验结果表明，该方法有效地综合了源图像中的重要信息，融合图像边界清晰，得到较好视觉效果和较优的评价指标，执行速度与CPU上相比提高了3个数量级。%Image fusion is an important topic in the field of image understanding and computer vision. Image fusion is facing a challenge on the processing speedwith the increased image scale.In recent years, GPU for general computing applications have been developing rapidly.Firstly,this article put forward the algorithm of parallel modelbased on edge strength. The experimental results show that the proposed algorithm can not only effectively combine the important information in the original images and image fusion clear boundaries, but also get better visual effect and better evaluation index. As the image size increases, it can increase the speed of execution up to 3 orders of magnitude compared with the CPU.