基于代数多重网格的分块融合算法
首发时间:2017-11-28
摘要:根据代数多重网格对图像的结构信息有较好的提取能力,本文提出了一种基于代数多重网格的分块融合算法。首先提取图像粗网格数据,分块进行重建。然后使用滑动窗口,以源图像与重建图像的均方差判定图像的清晰度,对每个像素的来源进行标记,生成决策图。最后对决策图分块,纠正块内来源错误的像素,同时将图像块与其邻域块比较调整,生成最终决策图,将决策图中与之对应的源图像的各个像素放入其中,生成融合图像。实验结果表明,融合图像没有丢失有效信息,最大程度保留了源图像的有效信息。
For information in English, please click here
A block image fusion algorithm based on algebraic multigrid method
Abstract:An image fusion algorithm based on algebraic multi-grid (AMG-Block) method which can be used to extract the image structure information is proposed. At first the rough grid data of the image is extracted to reconstruct the blocks, then the mean square error between the reconstructed image and the original image is utilized to judge the sharpness of the image with the sliding windows tactics. Each pixel which comes from A or B is marked, and a decision map is generated. Finally the decision map is corrected with the assist of the surrounding pixels and the final decision map is obtained. Experiences prove that the AMG-Block method can preserve most of the clear part in the original images.
Keywords: software engineering image fusion algebraic multi-grid sliding window
引用
No.****
同行评议
共计0人参与
勘误表
基于代数多重网格的分块融合算法
评论
全部评论0/1000