基于目标提取与引导滤波增强的红外与可见光图像融合

发布时间：2019-03-13 18:13

【摘要】：为了使融合结果突出目标并发掘更多细节,提出了一种基于目标提取与引导滤波增强的红外与可见光图像融合方法。首先对红外图像依据二维Tsallis熵和基于图的视觉显著性模型提取目标区域。然后对可见光与红外图像分别进行非下采样Shearlet变换(NSST),并对所得低频分量进行引导滤波增强。由增强后的红外图像和可见光图像低频分量基于目标提取的融合规则得到融合图像的低频分量,高频分量则根据方向子带信息和取大来确定。最后经NSST逆变换得到融合图像。大量实验结果表明,本文方法在增强融合图像空间细节的同时,有效突出了目标,并且在信息熵、平均梯度等指标上优于基于拉普拉斯金字塔变换、基于小波变换、基于平稳小波变换、基于非下采样Contourlet变换(NSCT)、基于目标提取与NSCT变换等。
[Abstract]:In order to make the fusion result highlight the target and discover more details, an infrared and visible image fusion method based on target extraction and guided filtering enhancement is proposed. Firstly, the target region is extracted from infrared image based on two-dimensional Tsallis entropy and visual salience model based on graph. Then non-downsampling Shearlet transform (NSST), is performed for visible and infrared images, and the low-frequency components obtained are enhanced by guided filtering. The low frequency component of the fusion image is obtained from the low frequency component of the enhanced infrared image and the visible image based on the fusion rule extracted from the target, and the high frequency component is determined by the direction sub-band information and the enlargement. Finally, the fusion image is obtained by inverse NSST transform. A large number of experimental results show that the proposed method not only enhances the spatial details of fusion images, but also highlights the target effectively, and it is better than Laplace pyramid transform and wavelet transform in information entropy, average gradient and other indexes. Based on stationary wavelet transform and non-subsampled Contourlet transform, (NSCT), is based on target extraction and NSCT transform.
【作者单位】：南京航空航天大学电子信息工程学院;南京信息工程大学江苏省大数据分析技术重点实验室;浙江工业大学浙江省信号处理重点实验室;广西师范大学广西多源信息挖掘与安全重点实验室;成都理工大学国土资源部地学空间信息技术重点实验室;中国地质科学院矿产资源研究所国土资源部成矿作用与资源评价重点实验室;
【基金】：国家自然科学基金(61573183) 南京信息工程大学江苏省大数据分析技术重点实验室开放基金资助(KXK1403) 浙江省信号处理重点实验室开放基金(ZJKL_6_SP-OP2014-02) 广西多源信息发掘与安全重点实验室开放基金(MIMS16-01) 成都理工大学国土资源部地学空间信息技术重点实验室开放基金(KLGSIT2015-05) 国土资源部成矿作用与资源评价重点实验室开放基金(ZS1406) 江苏高校优势学科建设工程
【分类号】：TP391.41

【相似文献】