Real Time Corrected Background Weighted Histogram Based Mean Shift Tracking

Michael Daniel Nachipyangu

Abstract: The real time target tracking provides the capability for the user to choose object s/he want to track as it moves in real time. The Corrected Background Weighted Histogram (CBWH) based mean shift Tracking is used on designing the algorithm for tracking. The algorithm used basis on reducing the effects caused by the changes on background which in turn affect the tracking efficiency. Whenever the system starts the target model is made and the background models too, when there is a change on background of the target object the system updates it accordingly. This algorithm also use color as its feature on tracking. The system is well designed with the Graphical user interface enabling even the user with less knowledge being able to operate the system.

Keywords: Corrected Background Weighted Histogram CBWH, Background updating model, mean shift, tracking and MatLab