Interval recursive least-squares filtering with applications to video target tracking

This paper focuses on applying an interval recursive least-squares (RLS) filter to a video target tracking problem. This is to circumvent the potential limitation of a RLS filter due to its sensitivity to variations in filter parameters and disturbances to state observations. Such sensitivity can make the solutions invalid in practical problems. In particular, in the application of video target tracking using a RLS filter, inaccurate parameters in the affine model may result in noticeable deviations from true target positions sufficient to lose a target. An interval RLS filter is proposed to produce state estimation and prediction in narrow intervals. Simulations show that an interval RLS filter is robust to state and observation noise and variations in filter parameters and state observations, and it outperforms an interval Kalman filter. Using an interval RLS filter, a video target tracking algorithm is developed to estimate the target position in each frame. The proposed tracking algorithm using an interval RLS filter is robust to noise in video sequences and errors in the affine models, and outperforms that using a RLS filter. Performance evaluations using real-world video sequences are provided to demonstrate the effectiveness of the proposed algorithm.