Research On Target Detection Method In Sea Clutter Environment

As international competition is increasingly gearing up,China promotes to speed up the construction of a marine power.Since the maritime surveillance radar can realize the early warning,monitoring and detection of the large area of the sea,it has great significance and wide application in both military and civilian applications.However,not only the echo signal of the detected target is included in the received signal of the radar,but also the noise signal and the clutter signal.In practical applications,the complex and changeable marine environment,different configurations of radar systems and other factors bring challenges to target detection in the background of sea clutter.If target detection is to be performed effectively in the background of clutter,some corresponding compensation methods and target detection methods should be designed to extract target information.Based on the above research background,the main contents are as follows:1.The properties of sea clutter is studied.Some measured datasets are exploited to study the relevance and amplitude distribution of high-resolution sea clutter high-resolution sea clutter: the longer the time interval between clutter echo signals from the same area,the lower the relevance.Compared with Rayleigh distribution and Weibull distribution,the amplitude distribution of the sea clutters is approximated to the K distribution.2.A target detection algorithm based on sub-band adaptive normalized matching filter is investigated.The normalized matched filter in the traditional method has a mutually restrictive relationship between the estimation accuracy of the clutter covariance and the integration time.By analyzing the energy distribution of sea clutter in the Doppler domain,a filter bank is used to avoid the interference of strong clutter to weak clutter.Then a sub-band adaptive matching detector can be obtained by matching the original received vectors in different sub-bands.Finally,a down-sampling factor is used to improve the stationarity of the sea clutter,which results in a longer integration time and a more accurate estimate of the covariance matrix.3.A sub-band generalized likelihood ratio detector based on amplitude distribution characteristics of sea clutter is inferred.The amplitude distribution model of sea clutter is closer to the K distribution,and the shape parameters and scale parameters of sea clutter in different sub-bands are calculated using the sub-band generalized likelihood ratio detector.Reception decisions are made using the diversity between different sub-bands.At the same signal-to-noise ratio,the detection probability of the sub-band generalized likelihood ratio detector is 0.07 and 0.21 higher than that of the sub-band adaptive matching detector and the traditional adaptive matching detector on average,respectively.4.A target detection algorithm based on support vector machine is analyzed.The detection performance of sea clutter suppression methods based on single time-amplitude information or Doppler features is limited.The joint suppression method using more sea clutter features can achieve effective sea clutter suppression and low-speed detectable targets.detection.In this paper,support vector machine is used to classify relative average amplitude,relative Doppler peak height and relative Doppler entropy in three-dimensional feature space to realize target detection and improve the accuracy of target detection under small signal-to-noise ratio.