Infrared Dim Target Tracking Algorithm Search

With the development of disguise technology, far distance small target detection and recognition relies more and more on negative detection technology, thus the Infrared detection technology becomes popular. With the situation that the target is hard to divide during complex environment in Infrared warning system, the false alarm rate is relatively high, the moving target is hard to track robustly, this paper focus on such aspects as follow.(1) The image characteristics of small Infrared target is analyzed to build the imaging model and moving model for target as well as the background model and noise model. Based on such models, the judge criterion for each algorithm is proposed.(2) Some familiar spatial and frequency domain background inhibition methods are analyzed. A new background inhibition method named Divide Target Area method is proposed. The contrast of image processing result and SCR proved the outstanding capability of the proposed algorithm in inhibiting background.(3) The difference between multi-frame detection and multi-frame searching is analyzed. A new searching algorithm based on target moving characteristics is proposed. Simulate experiments proved the better searching performance of the proposed algorithm against normal searching algorithms in multi-noise environment. The proposed algorithm could debase the false alarm rate efficiently. The number of practical searching frames is discussed in the end.(4) The Kalman Filtering tracking method is discussed. Fuzzy analyzing method is removed in tracking algorithm to predict the coming target’s position. With the support of Infrared Warning System, experiments proved that the prediction error is small enough to tracking the moving target efficiently in the proposed algorithm.The main innovation of this paper can be summarized as follow:1. A new efficient background inhibition algorithm is proposed. This algorithm is proved to achieve a relative high level SCR with the number of7.680against normal algorithms.2. A new target searching algorithm is proposed. The proposed algorithm is proved to be able to keep the false alarm rate at5%while the noise density is up to1%.3. Fuzzy prediction method is introduced to predict target position. The proposed algorithm exist a0.0139°error when predicting a moving target with angle accelerator of3.9°/s2in a warning system with scene of1°×1.3°...