Research On Methods Of Target Detection On Sea Surface Based On Time-frequency Analysis

Target detection of sea clutter has always been a difficult problem in the field of radar signal processing.The main reasons are that the physical scattering mechanism of the sea clutter is extremely complex,it is difficult to establish the sea clutter model,the target is weak,and the target is sparse relative to the sea clutter.To improve the ability of target detection under the background of sea clutter,the target detection method that based on time-frequency analysis is studied and analyzed by using the X-band measured sea clutter data published in the “Sea-detecting X-band Radar and Data Acquisition Program” on the website of Radar Journal in 2021 in this paper.The main research work of this paper is as follows:In the first part,introduce the basic theory related to sea clutter.Firstly,the principles of physical scattering model,amplitude distribution model and power spectrum model of sea clutter are introduced.Then,the measured sea clutter data under different sea conditions are used for model fitting,and the fit of each model is compared.Finally,the spatiotemporal correlation is introduced,and the correlation characteristics of the measured sea clutter data under different sea conditions are analyzed.In the second part,the methods based on time-frequency analysis are studied.Firstly,three typical time-frequency analysis methods are introduced,and using the measured sea clutter data under different sea conditions to compare and analyze the time-frequency domain characteristics.Secondly,target detection is performed on the measured sea clutter data under different sea conditions,and the detection performances of time-frequency analysis methods are compared and analyzed.The experimental results show that the traditional time-frequency analysis methods cannot detect the target under the condition of low signal-to-clutter ratio,but can detect the target under the condition of high signal-to-clutter ratio,with a high false alarm.Finally,the basic principles,characteristics and detection performance of the time-frequency analysis methods are summarized and compared to provide reference for the application in practical engineering.In the third part,for the problem that the wavelet basis function is not adaptive in wavelet transform,a tunable Q-factor wavelet transform algorithm is introduced.The algorithm selects the parameters matching the target according to the difference of oscillation characteristics of the target and the sea clutter;uses the sparsity of the wavelet coefficients to reconstruct the target,compares it with the adaptive threshold,and judges whether there is a target;does experiments on the measured sea clutter under different conditions.The experimental results show that the tunable Q-factor wavelet transform algorithm can detect the target,but there is a false alarm.In the fourth part,to further improve the performance of target detection,an improved tunable Q-factor wavelet transform algorithm is proposed.Considering the energy of sea clutter is strong,more accurate reconstruction information can obtain by matching sea clutter.Therefore,the algorithm no longer reconstructs the target,but the sea clutter;selects parameters that match the oscillation characteristics of the sea clutter;reconstructs the sea clutter by using the sparsity of the coefficients of the sea clutter,compares the difference between the original echo and it with the adaptive threshold,and judges whether there is a target;does experiments on the measured sea clutter under different conditions.The experimental results show that the improved tunable Q-factor wavelet transform algorithm can detect the target,and have almost no false alarm;Finally,the basic principle and detection performance of the tunable Q-factor wavelet transform algorithm and the improved tunable Q-factor wavelet transform algorithm are summarized and compared.