红外技术, 2017, 39 (3): 201, 网络出版: 2017-04-10
DWT、NSCT和改进PCA协同组合红外偏振图像融合
Infrared Polarization Image Fusion Using the Synergistic Combination of DWT, NSCT and Improved PCA
摘要
为充分保留红外光强和偏振图像细节、强度等信息,综合多算法的优势性能,提出一种 DWT、 NSCT和改进 PCA的多算法协同组合融合新方法,在考虑 3种算法互补协同关系基础上,充分保留源图重要目标和细节信息。首先,用离散小波变换(DWT)将源图分解为高低频分量,低频用非下采样轮廓波变换(NSCT)再次分解;其次,对主成分分析法(PCA)进行权值改进,分块融合 NSCT分解所得低频分量;然后,提出“相关系数-局部能量-局部标准差”规则融合 NSCT分解所得高频,用 “层内对比度”规则融合 DWT分解所得高频;最后,NSCT逆变换重构所得图像作为 DWT低频融合图,再用 DWT逆变换获得最终融合图像。实验结果表明,所提方法在视觉效果、细节层次及保留等方面比单一或简单组合方法更具优势,对不同场景适应性较强。
Abstract
In order to retain the details and the intensity information of infrared intensity image and polarization image adequately and integrate prominent performance of different algorithms, a novel synergistic combination fusion method of infrared polarization images based on DWT, NSCT and improved PCA is proposed, which can fully preserve important objectives and details of source images after considering complementary relationship of the three fusion algorithms. Firstly, the source images are decomposed into high and low-frequency components with DWT. The low-frequency coefficients are decomposed again with NSCT. Secondly, the weights generated by PCA are improved to fuse the low-frequency coefficients obtained by decomposing of NSCT block-by-block. Then, a Correlation coefficient, Local energy and Local standard deviation rule is put forward to fuse the high-frequency coefficients obtained by decomposing of NSCT. And the high-frequency coefficients obtained by decomposing of DWT are fused with the Layer Contras rule. Finally, the fusion image obtained by inverse NSCT is used as the low-frequency fusion image of DWT. And the final fusion image is obtained using inverse DWT. Experimental results show that the proposed method has more advantages than the single or simple combination fusion method in the visual effect, levels of detail and details preservation and has strong adaptability to different scenes.
杨风暴, 董安冉, 张雷, 吉琳娜. DWT、NSCT和改进PCA协同组合红外偏振图像融合[J]. 红外技术, 2017, 39(3): 201. YANG Fengbao, DONG Anran, ZHANG Lei, JI Linna. Infrared Polarization Image Fusion Using the Synergistic Combination of DWT, NSCT and Improved PCA[J]. Infrared Technology, 2017, 39(3): 201.