In this project, we detect the accident between two vehicles. For the purpose of detection, we use certain algorithms such as YOLO and CNN (Convolutional Neural Networks). This project is mainly built upon the combination of Object Detection and Tracking System (which is also abbreviated as ODTS) and RCNN algorithms which are mainly used for detecting purposes. In this project, a video is given as input where the coded algorithm completely separates the video into frames and analyzes the-each frame and detects the crash in the frames, and gives those frames as an output. In addition to accident detection, YOLO also detects the surroundings such as persons, cars, trucks, etc. The main point of using YOLO is to detect the surroundings and the crash (with the help of CNN).