A network combining deep residual shrinkage block for infrared and visible image fusion

Qi Wei; Zhan Song; Juan Zhao

doi:10.1117/12.2663015

12 April 2023 A network combining deep residual shrinkage block for infrared and visible image fusion

Qi Wei, Zhan Song, Juan Zhao

Proceedings Volume 12565, Conference on Infrared, Millimeter, Terahertz Waves and Applications (IMT2022); 1256531 (2023) https://doi.org/10.1117/12.2663015
Event: Conference on Infrared, Millimeter, Terahertz Waves and Applications (IMT2022), 2022, Shanghai, China

Abstract

The infrared images depicting the thermal radiation of objects are not affected by objective conditions such as natural environment and climate. Visible image has high spatial resolution, with a lot of details and high contrast. Infrared and visible image fusion takes the advantages of both optical bands to get a fusion image with clearly targets and rich background information. In this paper, we propose a novel deep learning based model by designing a new network structure and loss function. The network consists of an auto encoder and deep residual shrinkage modules. We introduce multiple deep residual shrinkage blocks into encoder to learn adaptive soft threshold parameters for denoising both infrared and visible images, Without affecting the complexity of the model, feature enhancement and extraction are implemented within the network to maximize the retention of practical information, and then the average fusion strategy is used to obtain the fusion features. Finally, the fused image is reconstructed by the decoder. In the design of the loss function, our loss function consists of pixel loss, structural similarity loss and gradient loss together, thus better preserving the texture details and edge information of the image. Experiments are performed on publicly available data sets. Qualitative results exhibit that the fused images obtained by our method are clearer and more natural, and in line with human visual perception. Quantitative results show that our proposed model has achieved the optimal or sub-optimal values compared to the state-of-the-art image fusion algorithms.

Citation Download Citation

Qi Wei, Zhan Song, and Juan Zhao "A network combining deep residual shrinkage block for infrared and visible image fusion", Proc. SPIE 12565, Conference on Infrared, Millimeter, Terahertz Waves and Applications (IMT2022), 1256531 (12 April 2023); https://doi.org/10.1117/12.2663015

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
10 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Image fusion

Infrared imaging

Infrared radiation

Visible radiation

Computer programming

Image enhancement

Convolution

Show All Keywords

Keywords/Phrases

Search In:

Publication Years