Real-Time Collision Warning System Based on Computer Vision Using Mono Camera

This paper aims to help self-driving cars and autonomous vehicles systems to merge with the road environment safely and ensure the reliability of these systems in real life. Crash avoidance is a complex system that depends on many parameters. The forward-collision warning system is simplified into four main objectives: detecting cars, depth estimation, assigning cars into lanes (lane assign) and tracking technique. The presented work targets the software approach by using YOLO (You Only Look Once), which is a deep learning object detector network to detect cars with an accuracy of up to 93%. Therefore, apply a depth estimation algorithm that uses the output boundary box's dimensions (width and height) from YOLO. These dimensions used to estimate the distance with an accuracy of 80.4%. In addition, a real-time computer vision algorithm is applied to assign cars into lanes. However, a tracking proposed algorithm is applied to evaluate the speed limit to keep the vehicle safe. Finally, the real-time system achieved for all algorithms with streaming speed 23 FPS (frame per second). © 2020 IEEE.