摘要
多传感器数据融合技术最近在图像处理领域得到广泛的重视和应用 .传统图像处理是典型的二维信号处理 ,图像数据融合也属此范畴 .为了融合来自同一景物的多幅带有不同噪声的图像 ,本文提出了一种基于一维自组织特征映射神经网络的图像融合算法 ,利用等灰度图法来确定自组织映射神经网络的聚类数 ,并用一维的像素特征量直接进行二维图像信号处理 .仿真试验给出的详细结果表明了该图像融合算法的有效性 .
Multisensor data fusion has played an important role in image processing recently.Traditional image processing is a typical two dimensional signal processing.The image data fusion belongs to this area as well.For some images from the same scene with different kinds of noise,this paper presents a new kind of image data fusion algorithm based on the one dimensional self organizing feature map neural network.This method uses equal gray density figure to determine the number of clusters,and utilizes one dimensional pixel value to perform two dimensional image processing.Simulation results demonstrate the effectiveness of the proposed image data fusion method.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2000年第9期74-77,共4页
Acta Electronica Sinica
关键词
自组织
算法
神经网络
图像数据融合
image filter
one dimensional self organizing feature map neural network
image data fusion
