Temporal Differencing and Statistical Method and we also discuss various related information. Among this method Background Subtraction is better due to its accuracy and effectiveness. Temporal Differencing method is quite good but this method fails to extract the pixel of foreground object when object has uniform texture or move slowly. But when object stop this method fails completely. We also studied Object Tracking in this paper. Here point tracking, kernel tracking and silhouette tracking have been studied. In point tracking method, to track the object state we have used statistical filtering method. Another filtering method such as Kalman filter and particle filter are also good method. In kernel tracking method mean shift technique is used now a day. And last in silhouette tracking method, object is tracked using silhouette by calculating the object region in consecutive frames.