[1]王淑青,蔡颖婧.一种基于无监督学习的空间域图像融合方法[J].湖北工业大学学报,2020,(2):52-55+105.
　WANG Shuqing,CAI Yingjing.A spatial Image Fusion Method Based on Unsupervised Learning[J].,2020,(2):52-55+105.
点击复制

一种基于无监督学习的空间域图像融合方法()
分享到：

《湖北工业大学学报》[ISSN:1003-4684/CN:42-1752/Z]

卷:
期数:
2020年第2期
页码:
52-55+105
栏目:
出版日期:
2020-04-30

文章信息/Info

Title:
A spatial Image Fusion Method Based on Unsupervised Learning
文章编号:
1003-4684(2020)02-0052-04
作者:
王淑青1; 2;  蔡颖婧1; 2
1 湖北工业大学湖北省电网智能控制与装备工程技术研究中心，湖北 武汉 430068；
2 湖北工业大学电气与电子工程学院， 湖北 武汉 430068
Author(s):
WANG Shuqing;  CAI Yingjing
School of Electrical and Electronic Engineering, Hubei Univ. of Tech., Wuhan 430068, China
关键词:
多焦点图像融合;  像素活跃度;  深度特征;  决策图
Keywords:
multi-focus image fusion;  pixel activity;  depth characteristics;  decision-making figure
分类号:
TP391.4
文献标志码:
A
摘要:
针对多焦点图像融合问题，提出一种新的无监督深度学习模型。首先，训练一个无监督的编解码器网络来提取输入图像的深层特征，然后利用这些特征和空间频率来测量图像像素活跃度并得到决策图。最后，应用一致性验证方法对决策图进行调整，得出融合结果。该方法的关键在于，只有在景深(DOF)范围内的物体在照片中才有明显的锐度，而其他物体很可能是模糊的。本方法是在深度特征上分析锐度的特征，而不是原始图像。实验结果表明，与已有的16种融合方法相比，该方法在客观评价和主观评价方面均取得了较好的融合效果。
Abstract:
To solve the problem of multifocus image fusion, a new unsupervised deep learning model is proposed. First, an unsupervised codec network is trained to extract the deep features of the input image, and then these features and spatial frequencies are used to measure the pixel activity of the image and obtain the decision graph. Finally, the consistency verification method is used to adjust the decision graph and obtain the fusion result. The key to this method is that only objects within the depth of field (DOF) have significant sharpness in a photograph, while other objects are likely to be blurred. The method of this paper is to analyze the sharpness feature on the depth feature instead of the original image. Experimental results show that compared with the existing 16 fusion methods, this method has achieved better fusion effect in both objective evaluation and subjective evaluation.

参考文献/References:

[1]Stathaki T. Image fusion: algorithms and applications[C].Elsevier,2011. 
[2]Burt P, Adelson E. The laplacian pyramid as a compact image code[J]. IEEE Transactions on Communications,2003, 31(4):532-540. 
[3]Stathaki T. Image Fusion: Algorithms and Applications[M]. Academic Press, 2008.
[4]Li H， Manjunath B，Mitra S. Multisensor image fusion using the wavelet transform[J].Graphical Models and Image Processing,1995,57(3):235-245. 
[5]Lewis J J， OCallaghan R J， Nikolov S G，et al. Pixel- and region-based image fusion with complex wavelets[J].Information Fusion :Special Issue on Image Fusion: Advances in the State of the Art，2007,8(2):119-130.
[6]Nencini F, Garzelli A, Baronti S, et al. . Remote sensing image fusion using the curvelet transform[J].Information Fusion: Special Issue on Image Fusion: Advances in the State of the Art，2007,8(2):143-156. 
[7]Yang B, Li S. Multifocus image fusion and restoration with sparse representation[C].IEEE Transactions on Instrumentation and Measurement，2010,59(4):884-892. 
[8]Li S, Kang X, Hu J,et al. Image matting for fusion of multi-focus images in dynamic scenes[J].Information Fusion，2013, 14(2):147-162. 
[9]Liu Y, Chen X, Peng H, et al. Multi-focus image fusion with a deep convolutional neural network[J]. Information Fusion，2017,36:191-207. 
[10] Prabhakar R. Deepfuse: A deep unsupervised approach for exposure fusion with extreme exposure image pairs[C]. In The IEEE International Conference on Computer Vision (ICCV)，2017. 
[11] Li H, Wu X. Densefuse: A fusion approach to infrared and visible images[C].IEEE Transactions on Image Processing，2019,28(5):2614-2623.
[12] Nejati M, Samavi S, Shirani S. Multi-focus image fusion using dictionary-based sparse representation[J]. Information Fusion，2015,25:72-84.

备注/Memo

备注/Memo:
［收稿日期］ 2019-09-30
［第一作者］ 王淑青(1969-)，女，河北衡水人，理学博士，湖北工业大学教授，研究方向为智能检测与控制，系统分析与集成
更新日期/Last Update: 2020-05-13