Tracking moving targets

Tracking moving targets is one of the important tasks in autonomous systems such as security and surveillance systems and unmanned aerial vehicles (UAV). When the camera system is stationery, background subtraction is a commonly used technique for segmenting out moving objects in a scene. In this paper we present two filters designed to remove the clutter in the resultant image of background subtraction technique. One is designed for the images with completely static background such as images taken inside a building. The other is capable of removing the clutter created by slight motions in the background such as motion of leaves of a tree due to wind. Background subtraction does not work when the camera system is moving. For this case we present a new technique that is developed using segmentation and coloring technique to isolate the targets of interest from the background. Upon the target of interest is extracted we calculate the position and the velocity of the target using the geometry of the camera system