LANE DETECTION IN AUTONOMOUS DRIVING USING ENHANCED PREPROCESSING AND AI TECHNIQUES

Abstract

Accurate lane detection plays a vital role in the development and deployment of autonomous driving systems. It serves as the backbone for vehicle localization, navigation, and decision-making processes. In this paper, we propose an integrated lane detection framework that combines advanced image preprocessing methods with state-of-the-art machine learning techniques to address the limitations of traditional approaches. The methodology includes a series of image transformations, such as color space conversion, edge enhancement, region of interest (ROI) extraction, and perspective warping to enhance lane visibility. These preprocessing steps are followed by the deployment of deep learning models—namely convolutional neural networks (CNNs) and transformer-based architectures—for semantic lane segmentation and curve fitting. The proposed system is tested on widely-used datasets under various challenging environmental conditions, including shadows, low light, and occlusions. Our results highlight the improvements in detection accuracy and processing efficiency, showing potential for real-time autonomous applications.