Font Size: a A A

Research On Detection Of Ship Target From SAR Imagery

Posted on:2010-10-21Degree:MasterType:Thesis
Country:ChinaCandidate:K TanFull Text:PDF
GTID:2178360278456945Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
Synthetic Aperture Radar (SAR), a kind of high-resolution imaging radar in range and azimuth, not only can capture the data of all-weather, multi-polarization, multi-frequency and multi-angle, but also has the ability of penetration through smog, cloud, rain and vegetation. With fast development in recent years, SAR has been widely applied in many fields, such as battlefield reconnaissance, aero-photographing, terrestrial remote sensing, terrestrial resource exploration, mapping, disaster forecast, etc.Ship detection and surveillance is the traditional task of the seacoast country all over the world, which has great application in military and civilian domain. It is of great significance for developing ship detection in SAR imagery in virtue of broad marginal sea and abundant ocean resources in our country. This article mainly pays attention to studies on ship detection in SAR imagery.The detection in SAR imagery is to find out the target echo in the environment clutter, includes not only the target itself, but also the environment. Thus, the sea-clutter statistical models are introduced.The derivation and parameter estimation for several kinds of SAR imagery clutter distributing models are summarized; the assessment criteria of statistical modeling for clutter and relevant emulation methods for some clutter are presented; then true SAR imagery section is fitted by cluter's histogram through experiments.Guided by the conclusion of statistical modeling to sea clutter, the core problem of the target checks for ships and warships is mainly studied. At first, the affecting factors on target detection of ships analyzed; secondly, global threshold CFAR and the two-parameter CFAR is compared, and improved adaptive CFAR based on local clutter statistical models is presented; finally, Detection algorithm based on Gamma kernels is introduced, in the case of complex clutter, is applied in the ships and warships target detection, also the result is given.According to the characteristics of the ship targets in SAR imagery, the problems of ship target parameters estimation is studied. The estimation results using different methods are compared, and the experiment results are displayed.At last, the full contents of this article are summarized and the interesting areas for future research on ships detection are prospected.
Keywords/Search Tags:Synthetic Aperture Radar Image, Ship, Constant False Alarm Rate (CFAR), Statistical Model, Target Detection, Quadratic Gamma Kernel
PDF Full Text Request
Related items