Discover whole part relationship between objects in video

This Paper develops a technique that efficiently discovers and analyzes the relationship between objects in captured video. The proposed framework detects and track the whole objects to extract temporal information from frame to frame which helps to find relation between objects by performing matching between objects in different frames. We use different techniques to implement these methods which were evaluated in order to assess how well they can detect moving regions, track objects and find relations. The final result of the algorithm is analyze by finding false positive and negative number of object detected which have relationship among each other in different frames.