基于多尺度方向引导滤波和卷积稀疏表示的红外与可见光图像融合

发布时间：2018-04-05 00:04

本文选题：图像处理　切入点：图像融合　出处：《光学学报》2017年11期

【摘要】：基于引导滤波和非下采样方向滤波器,提出了一种多尺度方向引导滤波图像融合方法,该方法兼具边缘保持特性和方向信息提取能力,能够有效提取源图像的有用信息。所提方法对源图像进行多尺度方向引导滤波,得到了包含低频近似部分和强边缘部分的低频分量,而后通过高斯低通滤波将其进行有效分离,分别应用基于卷积稀疏表示和区域能量自适应加权平均的融合规则;对高频细节方向分量应用显著性与引导滤波相结合的融合规则,以保持空间一致性,得到了相应的高频细节融合分量。结果表明,所提方法能更好地提取源图像的目标特征信息,保留丰富的背景信息,客观评价指标优于现有方法,融合结果具有更好的主观视觉效果。
[Abstract]:Based on the guided filter and the non-downsampling directional filter, a multi-scale direction-guided filtering image fusion method is proposed. This method has both edge holding characteristics and directional information extraction ability, which can effectively extract the useful information of the source image.The proposed method uses multi-scale direction-guided filtering to obtain the low-frequency components including the low-frequency approximation part and the strong edge part, and then separates them effectively by Gao Si low-pass filter.The fusion rules based on convolution sparse representation and regional energy adaptive weighted average are applied respectively, and the fusion rules combining salience with guided filtering are applied to the direction components of high frequency details to maintain spatial consistency.The corresponding high frequency detail fusion component is obtained.The results show that the proposed method can better extract the target feature information of the source image, retain abundant background information, and the objective evaluation index is better than the existing methods, and the fusion result has better subjective visual effect.
【作者单位】：军械工程学院军械技术研究所;
【基金】：总装人才战略工程专项资助基金(ZZ[2013]714号)
【分类号】：TP391.41

【相似文献】