This project investigates three advanced image segmentation algorithms—Watershed, Active Contour Snake, and Chan-Vese—for detecting object collisions in 2D simulation environments. Analyzing six diverse video datasets, the study evaluates these algorithms on execution time, segmentation accuracy, and adaptability to dynamic backgrounds and overlapp