In the context of today 's intelligent and unmanned warfare, this paper focuses on the field of infrared target detection and confrontation of unmanned ships. In view of the wide application of infrared intelligent detection technology and its impact on the identification of high-value targets such as unmanned ships at sea, a far-infrared band confrontation sample generation method for unmanned ship targets is proposed. This study intends to use optimization algorithms and other algorithms to generate physical domain infrared countermeasure samples for specific unmanned ship targets to counter the infrared unmanned ship detection system. This paper proposes to first use infrared UAV to collect single-band images of unmanned ships and process them to obtain significant features. Then, based on the machine learning method, the adversarial samples of the single-band image are generated, and the optimization algorithm is used to make the adversarial samples reach the target value for the attack of the single-band detection system of the infrared UAV, and the digital domain adversarial samples of the infrared image in this state are obtained. Then, in the physical domain, a special material that meets the requirements is used to complete the confrontation with the infrared detector, and its multi-position is arranged on the target unmanned ship to complete the interference with the infrared detection system. The success rate of the attack in the digital domain is more than 80%, and the success rate of the attack in the physical domain is not less than 55 %. This paper has a certain reference for the future research on infrared reconnaissance and anti-reconnaissance of unmanned ships.
Research on infrared target detection confrontation of unmanned ship based on optimization algorithm
Eighth Global Intelligent Industry Conference (GIIC 2025) ; 2025 ; Shenzhen, China
Proc. SPIE ; 13685 ; 136850J
2025-07-28
Conference paper
Electronic Resource
English
Target-borne unmanned aerial vehicle for airspace confrontation
European Patent Office | 2022
|Compact intelligent infrared sensor used in unmanned confrontation scene
European Patent Office | 2023
|European Patent Office | 2024
|