Font Size: a A A

Detection Of The Sea Target On The Condition Of Different Grazing Angle

Posted on:2014-06-30Degree:MasterType:Thesis
Country:ChinaCandidate:F WangFull Text:PDF
GTID:2268330422951501Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
With the development of the modern electronic technology, the research on thedetection of sea targets plays a highly significant role for the development of thecountry’s military development.Thus, the study of sea clutter characteristics and thedetection of sea targets on condition of different grazing angles have becomeinternational focus.The sea clutter is no longer subjected to the Rayleigh distribution on conditionof high resolution radar, which also has severe tailing effects. In this paper, the seaclutter is modeled in two aspects. Firstly, the sea clutter subjected to the Kdistribution is generated by the method of spherically Invariant Random Process(SIRP), which has a good consistence with the theoretical value. The generated seaclutter is verified by the real experimental data of IPIX radar. It is found that thecharacteristics of sea clutter are subject to the K distribution on condition of lowgrazing angles. Secondly, sea clutter data is generated by the Monte Carlo methodand the Kirchhoff Approximation (KA) based on the PM spectrum. This method hassome defects. This study adopts the traditional Weibull distribution model oncondition of moderate grazing angles.The performance of target detection and recognition can be effectivelyimproved by the polarization information. Firstly, the study explores three kind ofpolarized adaptive detectors. It is found that the polarized adaptive P-Wald detectorhas a good performance in detection. The detector’s performance is also analyzedunder different circumstances. The more number the auxiliary units have, whichcould produce more accurate estimation of the covariance, the better performance ofthe detection we get; if the shape parameter of the sea clutter is large, the detectioncurve will be steep; for the moving targets, the performance of the detector is better.Secondly, with the absence of training data, the P-Wald detector based on the ARmodel is applied in two polarization channels. The detector can reduce thecomputation quantity,while keep a good performance. These detectors are mainlyused at small grazing angles. At the middle grazing angle, the sea clutter subjectedto Weibull distribution is normalized, thus it presents the features of CFAR. Thecell-averaging CFAR detector is adopted and the detection curve is built. We alsoanalyze the characteristics of constant false alarm rate of the detection.
Keywords/Search Tags:target detection, grazing angles, sea clutter modeling, CFAR, polarization detection
PDF Full Text Request
Related items