A fusion method for infrared and visible images based on iterative guided filtering and two channel adaptive pulse coupled neural network. (2nd January 2021)

• Record Type: Journal Article
• Authors: Fan, Qiufeng; Hou, Fanbo; Shi, Feng
• Is Part Of: International journal of image and data fusion. Volume 12:Number 1(2021)
• Page Start: 23
• Page End: 47
• Publication Date: 2021-01-02
• Subjects: Iterative guided filtering -- multi-scale decomposition -- two-channel adaptive pulse coupled neural network -- image fusion; Image processing -- Periodicals; Multisensor data fusion -- Periodicals; Multisensor data fusion; Periodicals; 621.36705
• DOI: 10.1080/19479832.2020.1814877
• Languages: English
• ISSNs: 1947-9832

Abstract

ABSTRACT: In order to make full use of the important features of the source image, an infrared and visible fusion method based on iterative guided filtering and two-channel adaptive pulse coupled neural network is proposed. The input image is decomposed into basic layer, small scale layer and large scale layer by an iterative guide filter. The base layer is fused by combining pixel energy and gradient energy. Then we fuse the large scale layer and small scale layer via two-channel adaptive pulse coupled neural network. The fused image is obtained by the inverse mixing multi-scale decomposition method. Experimental results show that compared with other multi-scale decomposition methods, the proposed method can better separate spatial overlapping features, and preserve more detailed information in fused image, effectively suppress artefacts.