Extraction of Features from a Bird's-Eye View for Path Detection
Ahmed, A., ElKattan, A., Omar, A., Abdelaal, H. (2023). Extraction of Features from a Bird's-Eye View for Path Detection. EKB Journal Management System, 6(1), 1-14. doi: 10.21608/ijaiet.2023.234712.1001
Abdelmoty Ahmed; Ahmed ElKattan; Ahmed Omar; Hammam Abdelaal. "Extraction of Features from a Bird's-Eye View for Path Detection". EKB Journal Management System, 6, 1, 2023, 1-14. doi: 10.21608/ijaiet.2023.234712.1001
Ahmed, A., ElKattan, A., Omar, A., Abdelaal, H. (2023). 'Extraction of Features from a Bird's-Eye View for Path Detection', EKB Journal Management System, 6(1), pp. 1-14. doi: 10.21608/ijaiet.2023.234712.1001
Ahmed, A., ElKattan, A., Omar, A., Abdelaal, H. Extraction of Features from a Bird's-Eye View for Path Detection. EKB Journal Management System, 2023; 6(1): 1-14. doi: 10.21608/ijaiet.2023.234712.1001

Extraction of Features from a Bird's-Eye View for Path Detection

International Journal of Artificial Intelligence and Emerging Technology
Volume 6, Issue 1, June 2023, Page 1-14  XML PDF (778.15 K)
Document Type: Original Article
DOI: 10.21608/ijaiet.2023.234712.1001
View on SCiNiTO View on SCiNiTO
Authors
Abdelmoty Ahmed1; Ahmed ElKattan2; Ahmed Omar3; Hammam Abdelaal4
1College of Computer Science, Nahda University in Beni Suef (NUB)
2Surrey Business School, University of Surrey, Guildford, UK/ Faculty of Economics and Business, Complutense University of Madrid, Madrid, Spain
3Department of Computer Science, Faculty of Science, Minia University, Egypt
4Department of Information Technology, Faculty of Computers and Information, Luxor University
Abstract
Path detection has emerged as a complex challenge that has garnered significant attention within the computer vision community for a considerable span. The foundational problem of recognizing routes has evolved into a substantial impediment in the realm of computer vision. While a variety of Deep Learning methods have been employed for route identification, the focus has often centered on feature generation. Nonetheless, advanced Deep Learning techniques have exhibited proficiency in discerning these attributes. Yet, the complete implementation of these approaches to efficiently and accurately identify lanes is still pending. In this paper, we introduce an innovative perspective to address this issue. A distinctive approach to route preprocessing, coupled with a bird's-eye view, is presented. The primary aim involves the selection of the bird's-eye view area based on the processed image. Subsequently, an HSL (hue, saturation, and lightness) color transformation is applied to extract white features, along with a unique preliminary edge feature detection method. The lane is then identified using this new preprocessing strategy. It is noteworthy that numerous prior works have drawn upon the same standard datasets.
Keywords
Path Detection; leveraging preprocessing techniques; Bird' s-Eye View
Statistics
Article View: 299
PDF Download: 189