Research on Infrared and Visible Image Fusion Based on Improved Generative Adversarial Network

Citation Author(s):
Qianying
Wang
Haiyan
Xie
Huimin
Qu
Submitted by:
qianying wang
Last updated:
Mon, 11/25/2024 - 07:41
DOI:
10.21227/wx1q-dt81
License:
0
0 ratings - Please login to submit your rating.

Abstract 

The file contains merely a small portion of our research results. The particular picture presented showcases the outcomes that were achieved through our fusion method when applied to the TNO test set. The "ir", which is the abbreviation for infrared light image, is characterized by being rich in thermal radiation information. However, it unfortunately has a rather low spatial resolution. On the other hand, the "vis", representing the visible light image, is abundant in scene information. But it has a drawback in that the human targets within it are not so distinctly visible. The "our" indicates the fused image generated by our method. This fused image possesses rich texture details and, most importantly, the targets are clearly and distinctly visible, making it highly valuable for further analysis and various applications.

Instructions: 

The file contains merely a small portion of our research results. The particular picture presented showcases the outcomes that were achieved through our fusion method when applied to the TNO test set. The "ir", which is the abbreviation for infrared light image, is characterized by being rich in thermal radiation information. However, it unfortunately has a rather low spatial resolution. On the other hand, the "vis", representing the visible light image, is abundant in scene information. But it has a drawback in that the human targets within it are not so distinctly visible. The "our" indicates the fused image generated by our method. This fused image possesses rich texture details and, most importantly, the targets are clearly and distinctly visible, making it highly valuable for further analysis and various applications.

Comments

fusion results of our method

Submitted by qianying wang on Mon, 11/25/2024 - 07:44