Algorithm 1: Proposed Tracker at time step t

Input: Image Sequence of n Frames. Position of Target at First Frame. Output: Target Position for each frame in Image Sequence. for frame 1 to n frames. (1)Calculate context prior model using (3). (2)Calculate confidence map using (11). (3)Calculate target center. (4)Calculate translation correlation using (17). (5)Calculate the maximum value of response map. (6)if response map < threshold (7)new position = Kalman prediction (8)end (9)Calculate Kalman gain using (20). (10)Estimate position for next frame using (21). (11)Estimate error covariance using (22). (12)Calculate APCE using (23). (13)Update model using (24). (14)Calculate scale correlation using (17). (15)Update translation and scale model using (15) and (16). (16)Update context prior model using (3). (17)Update spatial context model using (9). (18)Update spatio-temporal context model using (12). (19)Calculate the target position for each frame. (20)Draw a rectangle on the target in each frame. End