A Review of Deep Learning-Based Lane Detection Methods

Read full text Summarise with VIRAT
Authors: Ting Liu, Megat Norulazmi Megat Mohamed Noor, Mohammad Faizuddin Bin Md Noor
Cited by: 0
References: 12

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Lane detection serves as a critical environmental perception application designed to identify lane markings utilizing onboard cameras or LiDAR systems. In recent years, concurrent with the advancement and deployment of computer vision applications, lane detection tasks have garnered substantial attention, leading to the emergence of a diverse array of lane detection methodologies. This article presents a comprehensive overview of lane detection techniques predicated on the acquisition of two-dimensional images via onboard cameras. Initially, it outlines the fundamental task of lane detection. Subsequently, it introduces datasets pertinent to lane detection. Following this, it delineates both traditional lane detection approaches and those grounded in deep learning. Ultimately, it provides an in-depth discussion of deep learning-based methods. Lane detection methodologies leveraging deep learning can be systematically categorized into four primary types: segmentation-based approaches, detection-based approaches, parameter curve-based approaches, and key point-based approaches.

Fields of Study & Subjects

Autonomous Vehicle Technology and Safety Advanced Neural Network Applications Infrastructure Maintenance and Monitoring

Publication details

Journal / source
Al lnnovations and Applications
Volume
2
Issue
2
Pages
1–17
Publication date
27 March 2026
Language
EN