ARCHIVES

Lane detection is a fundamental component of Advanced Driver Assistance Systems (ADAS) and autonomous driving applications, where real-time performance, accuracy, and computational efficiency are critical. This paper presents a deterministic real-time lane detection framework based on Region of Interest (ROI) gated edge refinement to achieve robust lane extraction under varying road and illumination conditions. The proposed method reduces computational overhead by dynamically restricting image processing to a predefined ROI corresponding to the probable lane region. Edge features are extracted using gradient-based operators and further refined through adaptive filtering and morphological enhancement to suppress noise and irrelevant road artifacts. A deterministic processing pipeline ensures predictable execution latency, making the system suitable for embedded automotive platforms with strict real-time constraints. The refined edge map is subsequently processed using probabilistic Hough transformation to accurately identify lane boundaries. Experimental evaluation on urban and highway driving datasets demonstrates improved detection stability, reduced false positives, and faster processing speed compared with conventional full-frame edge detection techniques. The proposed approach achieves reliable lane localization while maintaining low computational complexity, making it an efficient solution for real-time intelligent transportation and autonomous navigation systems