Paper
6 November 2023 Multi-scale local intensity and gradient networks for infrared small target detection
Zehao Zhao, Yukun Ma, ZhiQiang Chen, Wang Hao, Yaohong Chen
Author Affiliations +
Proceedings Volume 12921, Third International Computing Imaging Conference (CITA 2023); 129214E (2023) https://doi.org/10.1117/12.2691798
Event: Third International Computing Imaging Conference (CITA 2023), 2023, Sydney, Australia
Abstract
Robust small infrared target detection has becoming increasingly important in security and defense applications. Current detection methods can be divided into two parts. The performance of the model-driven methods is limited by the prior assumption parameters, while data-driven methods perform poor in semantic feature extraction due to the lack of the shape characteristic. In this paper, we proposed to embed a model-driven module into a convolution network, which has clear advantages on detection performance and running time. Our method consists of a backbone based on dilated convolution, a multi-scale local intensity and gradient module, and a segmentation head. Experimental results using extended SIRST dataset demonstrated that the proposed method achieves a good balance between the detection performance and running time.
© (2023) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zehao Zhao, Yukun Ma, ZhiQiang Chen, Wang Hao, and Yaohong Chen "Multi-scale local intensity and gradient networks for infrared small target detection", Proc. SPIE 12921, Third International Computing Imaging Conference (CITA 2023), 129214E (6 November 2023); https://doi.org/10.1117/12.2691798
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Target detection

Small targets

Infrared imaging

Convolution

Semantics

Performance modeling

Feature fusion

Back to Top