This article presents a comprehensive survey, system, and evaluation of video-based lane estimation and tracking for driver assistance. The focus is on the VioLET (Video-based Lane Estimation and Tracking) system, which is designed to enhance driver assistance through robust and accurate lane marking detection. The system uses steerable filters for efficient detection of various lane markings, including circular reflectors, solid lines, and segmented lines, under varying lighting and road conditions. These filters are effective in handling complex shadows, lighting changes, and road surface variations. The VioLET system integrates visual cues such as lane markings and texture with vehicle-state information to improve the accuracy of lane curvature estimation. The system was evaluated using a variety of quantitative metrics across different environmental conditions and driving scenarios. The research also includes an analysis of existing lane detection methods, highlighting their similarities, differences, and applicability under different conditions. The paper emphasizes the importance of selecting appropriate metrics for evaluating system performance and provides a detailed framework for lane-position tracking systems. The VioLET system is designed to be adaptable to various environments and is validated through extensive testing, demonstrating its effectiveness in real-world driving conditions. The contributions of this research include the development of an integrated lane estimation and tracking system, the use of steerable filters for robust lane marking detection, and a comprehensive evaluation of the system's performance under diverse conditions. The study also discusses the challenges and considerations in designing lane detection systems, such as environmental variability, sensor modalities, and the importance of accurate road and vehicle modeling. Overall, the article provides a detailed overview of the VioLET system and its potential applications in improving driver assistance technologies.This article presents a comprehensive survey, system, and evaluation of video-based lane estimation and tracking for driver assistance. The focus is on the VioLET (Video-based Lane Estimation and Tracking) system, which is designed to enhance driver assistance through robust and accurate lane marking detection. The system uses steerable filters for efficient detection of various lane markings, including circular reflectors, solid lines, and segmented lines, under varying lighting and road conditions. These filters are effective in handling complex shadows, lighting changes, and road surface variations. The VioLET system integrates visual cues such as lane markings and texture with vehicle-state information to improve the accuracy of lane curvature estimation. The system was evaluated using a variety of quantitative metrics across different environmental conditions and driving scenarios. The research also includes an analysis of existing lane detection methods, highlighting their similarities, differences, and applicability under different conditions. The paper emphasizes the importance of selecting appropriate metrics for evaluating system performance and provides a detailed framework for lane-position tracking systems. The VioLET system is designed to be adaptable to various environments and is validated through extensive testing, demonstrating its effectiveness in real-world driving conditions. The contributions of this research include the development of an integrated lane estimation and tracking system, the use of steerable filters for robust lane marking detection, and a comprehensive evaluation of the system's performance under diverse conditions. The study also discusses the challenges and considerations in designing lane detection systems, such as environmental variability, sensor modalities, and the importance of accurate road and vehicle modeling. Overall, the article provides a detailed overview of the VioLET system and its potential applications in improving driver assistance technologies.