In the military and civil fields, detecting small aircraft is of great significance. In recent years, the rapid development of LiDAR technology has made it possible to detect small aircraft at long distances. However, the scale change and attitude change make the detection difficult. Therefore, a detection network of multi-attitude small aircraft based on LiDAR anchorfree is proposed in this paper. The network structure is improved on the basis of the CenterNet network; using the encoder-decoder network structure, the extended convolutional module is designed to improve the receptive field and obtain the multi-scale information of the object. The IOU sensing branch is added to the detection header of the network to improve the localization accuracy of the object. The experimental results show that the detection accuracy of the improved network on the self-built simulation data set is 2.12% higher than that before the improvement and finally reaches 92.45%. Therefore, using this method can effectively improve the detection accuracy of the object.
Multi-attitude small aircraft object detection based on Gm-APD LiDAR
Advanced Fiber Laser Conference (AFL2023) ; 2023 ; Shenzhen, China
Proc. SPIE ; 13104
18.03.2024
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Abnormal attitude detection system and method for multi-rotor aircraft
Europäisches Patentamt | 2020
|Panoramic 3D LiDAR-based Object Detection
IEEE | 2023
|Small electric actuating mechanism for aircraft attitude control
Europäisches Patentamt | 2023
|