摘要
提出了可以保持源图像特征和细节信息的基于结构张量的变分多源图像融合算法.首先叙述基于结构张量的融合梯度场,然后测量每幅源图像的特征图,根据特征图为源图像的每个梯度构造一个权值,将携带明显特征的梯度在融合的梯度场中凸显出来,从而使源图像的特征和细节得到保持,最后应用变分偏微分方程理论从目标梯度场重建出融合的图像.实验结果表明,本文算法融合图像的灰度平均梯度和信息熵均高于小波变换算法、塔分解法和直接梯度融合算法,视觉效果上,融合图像很好的保留了源图像的特征和细节,为图像目标检测和识别提供了高质量的图像信息.
This article describes the variational multi-source image fusion using the structure tensor algorithm, which can keep the image features and details very well. We first narrative the fusion gradient field based on structure tensor, then measure the characteristic graphs of each source image, and thus construct a weight value for the source image gradient according to the characteristic graph. Gradients with high image features are highlighted in the fusion gradient field, and thus image features in the sources are well preserved. By using variational partial differential equation, the fusion image is reconstructed from the target gradient field. From the actual experimental results, the average gradient value and entropy of the fused image are found to be higher than those obtained by using the wavelet transform algorithm, tower decomposition algorithm, and direct gradient fusion algorithm, and the visual effect of the fusion image is good enough to retain the feature of source images and details in it. Therefore, it can give qualified image information for target detection and identification.
出处
《物理学报》
SCIE
EI
CAS
CSCD
北大核心
2013年第21期156-162,共7页
Acta Physica Sinica
基金
国家自然科学基金(批准号:05891JM50)资助的课题~~
关键词
图像融合
梯度场
结构张量
变分法
image fusion, gradient field, structure tensor, variational method
