Advanced Pedestrian Monitoring for Adas Using Lidar and Cameras

This study introduces a novel pedestrian monitoring methodology that integrates LiDAR point cloud data with camera systems to enhance road safety in autonomous driving environments. By projecting LiDAR data onto camera images, the proposed system enables precise distance estimation for effective collision prevention. Advanced statistical techniques and the RANSAC algorithm are employed to extract ground information and generate accurate distance maps. Validation using the KITTI dataset demonstrates promising performance in short- to midrange distance estimation, achieving accuracy levels of 98.2% to 99.7% within 0-10 meters (5m threshold). Addressing the limitations of radar-based Advanced Driver Assistance Systems (ADAS), this research presents a comprehensive framework for pedestrian detection, tracking, and distance estimation. The findings establish a solid foundation for integrating LiDAR and camera technologies into autonomous vehicles, contributing to safer and more efficient driving systems.