Research on infrared dim and small target detection algorithm based on frequency domain difference
Leihong Zhang,Haiqing Miao,Jian Chen,Weihong Lin,Runchu Xu,Dawei Zhang
DOI: https://doi.org/10.1080/01431161.2022.2075245
IF: 3.531
2022-06-02
International Journal of Remote Sensing
Abstract:The high accuracy of infrared dim and small target detection in complex backgrounds is of great relevance for infrared identification and tracking systems. Traditional infrared dim and small target detection methods suit scenes with a single and homogeneous continuous background. However, human vision system methods suffer from an undetectable or high false alarm rate in complex scenes with dim small targets. To address this shortcoming, this paper proposes an infrared dim and small target detection algorithm based on frequency domain differencing (FDD). The proposed algorithm consists of a spectrum residual module and a Gaussian greyscale difference module. Firstly, the target enhancement image is constructed by using the spectrum residual module to highlight small targets and suppress background noise. Secondly, the local contrast of the image is enhanced by the Gaussian grayscale difference module, which accurately depicts the edge information of small targets and locates them. Finally, the target enhancement image and Gaussian grayscale difference image are fused to detect infrared dim and small targets. Experimental results show that the proposed algorithm has higher accuracy under the evaluation metrics of local signal-to-noise ratio gain (LSNRG), signal-to-clutter ratio gain (SCRG) and background suppression factor (BSF). At the same time, compared with other algorithms, the detection rate of the proposed algorithm is higher for infrared dim and small targets in complex scenes. Code is available at https://github.com/m156879/FDD-module.
imaging science & photographic technology,remote sensing