• Shreeja Kale

Accurate perception of road conditions is a fundamental requirement for intelligent transportation systems and advanced driver assistance applications. Reliable assessment of both road surface quality and lane markings is particularly challenging in real-world scenarios due to variations in lighting, weather conditions, occlusions, and the heterogeneous nature of road infrastructure, especially in developing regions. This work presents an integrated vision-based framework for simultaneous road quality classification and lane detection using machine learning and image processing techniques. The proposed system employs a convolutional neural network to classify road surface conditions into multiple quality categories, while lane detection is achieved through robust edge detection and line extraction methods. To enhance adaptability in dynamic environments, a confidence-driven self-learning mechanism is introduced, allowing the system to automatically collect high-confidence predictions and incrementally retrain the model. This approach reduces dependency on manually labeled data and improves long-term performance. Experimental evaluation conducted on continuous road video streams demonstrates effective lane localization and reliable road quality classification with real-time performance. Performance metrics such as accuracy, intersection-over-union, mean absolute error, and frame processing rate validate the effectiveness of the proposed approach. The results indicate that the system is suitable for deployment in real-world traffic monitoring and driver assistance scenarios, offering a scalable and adaptive solution for intelligent road analysis.

"Enhanced Lane Detection System for Autonomous Vehicles Using Advanced Computer Vision", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 2, page no.b509-b516, February-2026, Available :http://www.jetir.org/papers/JETIR2602171.pdf

"Enhanced Lane Detection System for Autonomous Vehicles Using Advanced Computer Vision", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 2, page no. ppb509-b516, February-2026, Available at : http://www.jetir.org/papers/JETIR2602171.pdf