Title: Object detection algorithm combined with dynamic and static for air target intrusion

Authors: Yi Xiao; Faming Shao; Fanjie Meng; Jiqing Luo

Addresses: College of Field Engineering, Army Engineering University of PLA, Nanjing, 210007, China ' College of Field Engineering, Army Engineering University of PLA, Nanjing, 210007, China ' Department of Space Test and Launch, Noncommissioned Officer School, Space Engineering University, Beijing, 102299, China ' College of Field Engineering, Army Engineering University of PLA, Nanjing, 210007, China

Abstract: The infrared detection of target intrusion usually faces the problems of complex background and insufficient brightness. This paper proposed a method combining dynamic detection and static detection, which is mainly composed of adaptive local contrast module (ALCM), interval frame semantic layered (IFSL) module and enhancement network module. ALCM, as a static method, uses adaptive sliding box to traverse the original image to obtain relatively simple background targets. As a dynamic method, IFSL can effectively separate moving targets from the background based on spatial semantic information under dynamic conditions. The function of the enhanced network module is to suppress the background clutter and highlight the target. Compared with the traditional detection method, the SNR of the proposed method is improved by 8.65%, and the computing speed is improved by 7.14%. This strongly proves that the method has high detection rate and detection efficiency in this case.

Keywords: infrared target; static detection; dynamic detection; feature fusion; enhanced network; adaptive local contrast method; interval frame semantic layered; complex background.

DOI: 10.1504/IJSNET.2022.10045455

International Journal of Sensor Networks, 2022 Vol.38 No.2, pp.97 - 112

Received: 19 Mar 2021
Accepted: 20 Mar 2021

Published online: 28 Feb 2022 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article