The aim of this project is to create a computer vision system that can classify car crashes or collisions. This system will mainly use a pre-trained VGG16 model to examine video footage and images from different traffic situations. The ultimate purpose is to improve road safety by accurately classifying potential collisions under various circumstances.