A Novel Detection Approach for Quasi-Convex Quadrilateral Aircraft Covers
IEEE Access · 2026 · 14 · 34151-34161
DOI: 10.1109/access.2026.3669209 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Wande Yue und 4 weiteren Autor:innen. Die Arbeit nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Key Research and Development Program of Shaanxi.
Abstract (Originalsprache)
To address the challenge of rapid and efficient detection of aircraft cover plate target images resembling quadrilaterals, this paper introduces the concept of quasi-convex quadrilaterals and presents a corresponding target image processing and detection approach. Specifically, the definitions, properties, and determination methods of quasi-convex quadrilaterals are meticulously studied and expounded. First, Fourier transform-based noise reduction is applied to contours that meet the definition of quasi-convex quadrilaterals. Then, curvature is used to identify segmentation points and four quasi-straight edges. Subsequently, the least-squares method is employed to fit these four quasi-straight line segments, generating the adjoint quadrilateral of the quasi-convex quadrilateral. The vertices of the adjoint quadrilateral act as control points for distortion correction of the quasi-convex quadrilateral, and template matching is adopted to detect the contour of the target aircraft cover plate. Experimental validation is carried out on sample photos of three subtypes of quasi-convex quadrilateral aircraft cover plate contours. The proposed approach achieves a 98.81% True Positive Rate (TPR), outperforming both the standard Hough Transform (93.25% TPR) and improved variants (95.63% TPR). It maintains a low False Positive Rate (4.37% FPR) while reducing computational time by approximately 26.5% compared with standard HT. Additionally, compared with the lightweight CNN model MobileNet Single Shot MultiBox Detector (MobileNet-SSD), the proposed method exhibits advantages in low-resource scenarios (no GPU, no training dependency) in terms of detection accuracy and inference efficiency. The algorithm’s linear time complexity (O(N)) ensures scalability for high-resolution industrial imaging applications, making it particularly suitable for real-time aircraft maintenance systems.
Förderung
- Key Research and Development Program of Shaanxi
Zitieren
APA
Yue, W., Liu, Z., Wang, K., Hu, Z., & Wang, H. (2026). A Novel Detection Approach for Quasi-Convex Quadrilateral Aircraft Covers. IEEE Access, 14, 34151-34161. https://doi.org/10.1109/access.2026.3669209
Vancouver
Yue W, Liu Z, Wang K, Hu Z, Wang H. A Novel Detection Approach for Quasi-Convex Quadrilateral Aircraft Covers. IEEE Access. 2026;14:34151-34161. doi:10.1109/access.2026.3669209
BibTeX
@article{Yue2026,
author = {Yue, Wande and Liu, Zhouzhou and Wang, Ke and Hu, Zenglin and Wang, Haipeng},
title = {A Novel Detection Approach for Quasi-Convex Quadrilateral Aircraft Covers},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {34151--34161},
doi = {10.1109/access.2026.3669209}
}
EUVIMED