Surpass Assist

In our project, we aim to address the problem of unsafe overtaking of large trucks by using a computer vision solution based on a YOLOV8 model trained on the CoCo dataset for detecting objects. To detect trucks accurately on the road, a bounding box annotation is implemented using the CV2 library and supervision library. The system uses: - A camera on large trucks. - The YOLOv8 object detection model. - The COCO dataset. - Direction detection algorithm . - Supervision library for computer vision functions. The accuracy of the model was tested by creating a demo scenario and running it on Google Colab. Because Colab does not support video playback, an output video was rendered and saved to a shared Google Drive folder for team members to view. The system detects nearby vehicles and objects and measures distances to objects ahead. It also outputs an LED display for vehicles behind the truck: - The green light means: safe to overtake. - The red light means: unsafe, stay behind the truck. Using this solution, drivers can safely overtake trucks by detecting the presence and location of trucks on the road. This could reduce the number of traffic accidents and reduce road congestion caused by unsafe overtaking practices.