期刊
OPTICS COMMUNICATIONS
卷 354, 期 -, 页码 299-313出版社
ELSEVIER SCIENCE BV
DOI: 10.1016/j.optcom.2015.05.020
关键词
Compressive sensing; NSCT; Dual-layer PCNN; CoSaMP
类别
资金
- National Natural Science Foundation of China [61374135, 61203321]
- China Postdoctoral Science Foundation [2012M521676]
- China Central Universities Foundation [106112015CDJXY170003]
- Chongqing Special Funding in Postdoctoral Scientific Research Project [XM2013007]
- Chongqing Graduate Student Research Innovation Project [CYB14023]
Image fusion can integrate complementary and relevant information of source images captured by multiple sensors into a unitary synthetic image. The compressive sensing-based (CS) fusion approach can greatly reduce the processing speed and guarantee the quality of the fused image by integrating fewer non-zero coefficients. However, there are two main limitations in the conventional CS-based fusion approach. Firstly, directly fusing sensing measurements may bring greater uncertain results with high reconstruction error. Secondly, using single fusion rule may result in the problems of blocking artifacts and poor fidelity. In this paper, a novel image fusion approach based on CS is proposed to solve those problems. The non-subsampled contourlet transform (NSCT) method is utilized to decompose the source images. The dual-layer Pulse Coupled Neural Network (PCNN) model is used to integrate low-pass subbands; while an edge-retention based fusion rule is proposed to fuse high-pass subbands. The sparse coefficients are fused before being measured by Gaussian matrix. The fused image is accurately reconstructed by Compressive Sampling Matched Pursuit algorithm (CoSaMP). Experimental results demonstrate that the fused image contains abundant detailed contents and preserves the saliency structure. These also indicate that our proposed method achieves better visual quality than the current state-of-the-art methods. (C) 2015 Elsevier B.V. All rights reserved
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据