SEHRNet: A lightweight, high‐resolution network for aircraft keypoint detection

• Published in: IET image processing Vol. 18; no. 9; pp. 2476 - 2489
• Main Authors: Zhang, Zhiqiang, Zhang, Tianxiong, Zhu, Xinping, Li, Jiajun
• Format: Journal Article
• Language: English
• Published: Wiley 01-07-2024
• Subjects: aircraft; computer vision; convolutional neural nets; image processing; pose estimation
• ISSN: 1751-9659; 1751-9667
• DOI: 10.1049/ipr2.13111

Current research on apron conflict detection is often limited to the interaction between the aircraft as a whole and other objects, making it difficult to accomplish targeted identification of vulnerable and high‐cost aircraft components. However, the implementation of detailed aircraft identification is of great significance to enhance the safety of airport surface operations. Based on the excellent performance of High‐Resolution Network (HRNet) in keypoint detection, a lightweight end‐to‐end keypoint detection network, namely Squeeze and Excitation High‐Resolution Network (SEHRNet), is proposed in this paper to solve the problems of HRNet's slower computation and more redundancy. First, the errors arising from coordinate transformations in the coding and decoding process are solved by an improved feature map coding and decoding process. Second, replace the BasicBlock in HRNet with the Depthwise separable convolutions based on the Squeeze‐and‐Excitation Networks, which drastically cuts the computational cost of the network. Third, the improved Bottleneck module is used to further enhance the capability of feature extraction. Experimental results prove that, based on the aircraft keypoint detection dataset, the SEHRNet proposed in this paper shows stronger applicability compared to the current mainstream networks. Compared with the original HRNet, the improved network has higher accuracy, faster speed, and lighter model for aircraft keypoint detection.
There are few studies on aircraft key component identification. Inspired by the network structure of High‐Resolution Network, this paper proposes targeted improvements by combining the characteristics of apron operation and the physical features of aircraft key components. The experimental results show that the algorithm in this paper can realize the real‐time high‐precision detection of key components of aircraft in the apron gate.