Image processing Occlusion Detection and Handling

Abstract

Occlusion means hiding of an object by another object during multiple human tracking. For multiple object tracking, it is important to
maintain the history of objects before and after occlusions. This paper highlights to fill the missing parts from the past history of a person if available, when occlusion is detected. When dealing with multiple objects tracking, we separate the object state into three parts: Before, during and after occlusion. An Improved Mean Shift Tracking algorithm (IMST) which is special for occlusion target tracking is used. Occlusion can be detected by calculating the centre of mass of both the objects and when the distance between them is zero. By comparing the frames, the occluded part is identified and the missing part is filled from the matched frame when occlusion is detected.