A PDAF_AI Based Technique For Tracking Of Dim Moving Point Target In Image Sequences

Usually, the people mainly concentrate to the tracking system research of the short-distance big target and high signal/clutter-to-noise ratio (SCNR) in the image sequences. In recent years, the unceasing enhancement of military requests in defense system performance, which make the people have the strong interest in the detection and tracking technology of the long-distance range and low SNR dim point target, enables this topic to become one of present research hot spots.The present paper has studied one of probabilistic data association filter (PDAF) tracking technologies based on the target initial location, velocity and amplitude information (AI).This technology is conditioned on successful detection of a target by multi-frame detection scheme such as track-before-detect (TBD). TBD detector transfers information to tracker, which includes the target's initial location, velocity, and amplitude. After the tracker receives the information, one can predict the possible target area in the next frame with Kalman filter. Then the CFAR detection procedure takes place in that area. Due to the low SNR and false-alarm rate, there may be too many false targets. We can use data association technology and target's initial information to decide a true target, so that achieve the goal of tracking.The present paper infers the formulation of probabilistic data association filter, and has carried on the simulation with MATLAB to its algorithm, obtaining the important theoretical analysis and the experimental result.