基于双层并行PCNN和粗集理论的图像融合  

New image fusion algorithm based on PCNN and rough set

作　　者：张利强[1] 李毅[1] 

机构地区：[1]四川大学计算机学院,成都610065

出　　处：《计算机工程与应用》2011年第31期179-181,223,共4页Computer Engineering and Applications

摘　　要：为了能更好地进行多传感器图像融合,提出了一种基于双层并行PCNN和粗集理论的图像融合方法。该方法首先对两幅图像去噪,将一幅图像作为主PCNN网络的输入,另一幅图像作为从PCNN网络的输入,计算每幅图像的清晰度,分别将每幅图像的清晰度矩阵送入主从PCNN网络处理,然后根据粗集理论对原始图像分类,最后生成融合图像。该方法不仅能保留原图像信息,而且得到的融合图像清晰度高、对比度大。仿真实验结果以及与其他融合算法的比较,表明该算法的有效性和优越性。In order to better carry out multi-sensor image fusion,a novel algorithm based on double Parallel Pulse Coupled Neural Network（PCNN） and rough set for image fusion is proposed.In the method,two original images are denoised firstly,one of the images is chosen randomly as the input to the main PCNN network,and the other original image as the input to the subsidiary one.The clarities of the two original images are calculated,and the two matrixs of the clarity are processed by the main and subsidiary PCNN network seperatedly.Then the original image pixels are classified based on rough set theory.Finally,a fusion image is created according to the classified results.This method not only retains the information of original images,and the obtaining image has larger clarity and higher contrast.Simulation results and comparison with other fusion algorithms show the effectiveness and superiority of the method.

关 键 词：图像融合 双层并行PCNN 粗集理论 清晰度 主PCNN 从PCNN 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]