Tracking object under intense disturbance based on adaptive histogram with an active camera

A novel method based on adaptive histogram is proposed for follow-up tracking moving object under intense disturbance with an active camera. The Mean Shift algorithm shows an excellent performance on robust and fast object tracking, however, it is prone to fail when facing intense disturbance from background. This paper addresses to solve these problems. At first how the number of dimensions and the number of bins of histogram affects the target representation is analyzed, then an adaptive selection strategy for dimensions number and bins number is proposed depending on the intensity of disturbance from background. To follow-up tracking moving object with an active camera, a closed loop control model based on speed regulation is proposed to drive a PTZ camera to center the target. The results of experiments show that the active camera can follow-up track moving target stably, even when encountering large area intense disturbance from background. The algorithm is computationally efficient and can run in realtime speed.