Addressing Vehicle Safety and Platooning Using Low-Cost Object Detection Algorithms

Object detection is a computer vision-based technology that is very pertinent to the study of automation. You only look once (YOLO) is an object detection algorithm used to detect objects belonging to certain classes in images and videos. YOLO provides an increased speed and accuracy of object detection over conventional solutions. This research paper compares two versions of YOLO: YOLOv5 and YOLOv3, in terms of precision and recall, by training the models from the ground up from a dataset containing vehicle images. These results have been integrated with a single-frame lane detection pipeline and were evaluated over two different architectures: an i5 11th gen processor and Nvidia GTX 1050 graphics card. The best possible solution based was able to achieve an average speed of 12 frames per second for the entire pipeline.