交叉双边滤波和视觉权重信息的图像融合

发布时间：2018-01-30 19:18

本文关键词： 图像融合多层分解视觉权重交叉双边滤波器　出处：《仪器仪表学报》2017年04期 　论文类型：期刊论文

【摘要】：图像融合技术是图像分析领域重点研究内容之一,为了更好地保留原图像中的细节信息,提高融合图像的对比度,提出了基于视觉权重图的多尺度图像融合方法。首先,利用可变参数的交叉双边滤波器对两幅待融合图像进行多尺度分解;然后,在每个分解层分别计算相应的视觉权重图,并针对不同分解层赋予不同的权重值;最后,综合这些结果生成融合图像。由于对原始图像的分解没有采用下采样和上采样操作,因此不会损失图像中的信息,且克服了传统像素级融合方法中融合图像模糊、对噪声敏感等不足。通过4种定量分析实验表明,在多种模式的图像融合应用中,本方法优于其他5种对比方法,融合时间小于0.2 s。融合后图像细节信息、对比度得到增强,同时降低处理时间。
[Abstract]:Image fusion technology is one of the most important research contents in the field of image analysis. In order to better preserve the details of the original image and improve the contrast of the fused image. A multi-scale image fusion method based on visual weight graph is proposed. Firstly, two images to be fused are decomposed by a cross-bilateral filter with variable parameters. Then, the corresponding visual weights are calculated in each decomposition layer, and different weights are assigned to different decomposition layers. Finally, the fusion image is generated by synthesizing these results. Since the decomposing of the original image does not use down-sampling and up-sampling operations, the information in the image will not be lost. It overcomes the shortcomings of traditional pixel level fusion method, such as image blur and sensitivity to noise. Through four quantitative analysis experiments, it is shown that this method can be used in multi-mode image fusion. This method is superior to the other five comparison methods, and the fusion time is less than 0.2 s. After fusion, the image details are enhanced, the contrast is enhanced, and the processing time is reduced.
【作者单位】：海军航空工程学院;中国国防科技信息中心;
【基金】：国家自然科学基金(61303192)项目资助
【分类号】：TP391.41
【正文快照】： 1引言图像融合是把多种传感器获得的多幅同一场景图像综合成一幅图像的先进图像处理技术,其目的是集成多个源图像中的互补信息,增加图像理解的全面性。它在计算机视觉、医学图像以及军事应用等领域中有着广泛的应用前景。目前,图像融合方法大多基于多尺度变换,常用的方法有拉

【相似文献】