The Research Of Puny Moving Target Deceting And Tracking Interactive Algorithm Based On Acoustic Image

To achieve the goal of tracking target, conventional methods have to complete the detection of targets firstly, and then track the target. But under the background of strong underwater noise clutter, due to the puny target’s low SNR of the received echo, it has few resolution cell. While using CFAR method to detect the received acoustic image, which not only reduces the detection efficiency, increases the false alarm, but also degrades the performance of target tracking. So this paper mainly researches on an interactive algorithm for moving target detection and tracking, intend to increase the efficiency of target detection, thus improving the performance of target tracking, monitor the target in real-time.In the first place, target detection algorithm based on acoustic image is studied. Using FFT beamforming, Chebyshev amplitude weighting and the near field focusing to establish acoustic image; than using sector conversion to transform the coordinate of the image. After that, using image domain CFAR method detect single frame graphics, discussed the size of protection window and different false alarm influence on CFAR, detected acoustic images with different SNR. With the SNR’s decrease, the detection effect is getting worse, to improve the effect, this paper established a multi-frame graphics accumulation target detection algorithm based on CFAR-Hough. Though the algorithm can achieve a good detection of target, parameters of Hough need to be reset at different SNR, so the algorithm has some limitations.Then studying the moving target tracking algorithm and the data association algorithm. Tracking target with Kalman filter; using track gates and Bayesian data association method to determine the proper interconnection relationship between the trace points and the track. Simulation analysis initial different measurement error, clutter density, the gate’s size and target detection probability influence on NNSF, PNNF and PDAF methods’ tracking performance. The simulation found that the probability of target detection affects the system’s tracking performance, if the detection’s probability can improve, then in some extent the target tracking accuracy will be enhanced.Finally, the paper studied moving target detection and tracking interactive algorithm based on acoustic image. Combined CFAR-Hough method with logic method of 2/3 fast tracking initiation to get a good track starting state. Established interactive algorithm block diagram, given algorithm specific processes, simulation analysis the algorithm’s performance. Monitoring the target by interactive algorithm, using target tracking to extrapolate the track, predicting where the target may occur at the next time, these messages can be used as detection unit’s priori information, detecting the area meticulously, this can improve the detection’s efficiency; after that using data association remove false target in the gate, thereby improving the performance of target tracking.