Ship Targets Recognition Reserch Based On High Range Resolution Profiles

As the high-resolution radar technology becomes mature gradually, RATR(Radar Automatic Target Recognition) has played an important role in sea battle, port management and marine search and rescue. The HRRP(High Resolution Range Profiles), which can indicate the distribution of target’s scattering centers along the radial direction of the radar, has attracted many researchers’ attention because it can be easily obtained and processed. Target recognition algorithms based on HRRP are mostly focused on aircrafts, missiles and other air targets, while the ships’ HRRPs are mainly interfered by the sea clutter, which has different characteristics from Rayleigh clutter of low-resolution radar. Numerous researches show that, the distribution of high-resolution radar sea clutter can be approximated by the K distribution model. As a result, this thesis has done research on the recognition of ship targets based on HRRPs. The main works are as follow:1. Discussed the fundamental of dechirping based on scattering centers model of radar targets and did computer simulations to obtain the HRRPs of three classes of ship targets. In this part, three sensitive natures of the HRRPs of ship targets were studied and the way to overcome them was figured out too. The above-mentioned theory was tested and verified by simulation experiments.2. Introduced the K distribution model of sea clutter and studied its influence on HRRPs of ship targets. The effect of the shape parameter and scale parameter to feature extraction was studied and contrasted with the Rayleigh clutter under the same SCR(Signal-to-Clutter Ratio).3. Discussed several classic methods of feature extraction and classifier design in pattern recognition and the performance of each method was obtained by processing simulation data. Different combinations of feature extraction methods and classifiers were studied when did research on ship targets’ recognition based on HRRPs. The Monte Carlo experiments were adopted to compare the performance and computational efficiency of the above-mentioned algorithms under the K distribution sea clutter.4. Two different sea clutter suppression technologies based on AR(Auto Regression) model and SVD(Singular Value Decomposition) each were studied and applied to the ship targets recognition based on HRRPs under low SCR. Theoretical analysis and simulation experiments were done to contrast the performance of the two methods and the conclusion was that, the sea clutter suppression technologies based on AR model has more advantage against the one based on SVD.The sea clutter will decrease the difference of features extracted from HRRPs of different classes of Ship targets. By the application of sea clutter suppression technologies, the performance of ship targets recognition based on HRRPs under low SCR will get better.