Research On Sea Surface Image Background Modeling And Target Detection Method Of Airborne Radar

The maritime target detection from airborne radar image aims at detecting the target of interest from the sea clutter background,locating the target position and achieving the perception of the regional situation information,which has significant application value in coastal monitoring,anti-submarine reconnaissance,regional escort and salvage.However,the complexity of the maritime conditions and the diversity of target types make the target detection from airborne radar image be faced with many difficulties.The modeling of sea clutter background,and the stable target detection from sea clutter background are the keys to achieving high-performance target detection from airborne radar image,which are also the international frontier subjects that need to be solved urgently.To address these problems,this dissertation has conducted theoretical analysis,method research,simulation verification and measured data verification.The main innovations include:1.The model of sea surface image background in airborne radar synthetic aperture mode is established,and a fast spectral domain generation method of sea surface synthetic aperture radar(SAR)image is proposed,which can realize effective simulation of sea surface SAR image and reduce the computational burden of joint solution between sea surface and target.2.The model of sea surface image background in airborne radar scanning mode is established,which can accurately reflect the non-Gaussian and space-time correlation characteristics of the image background.Based on integral equation method,a method of solving the parameters of sea clutter model is proposed,which can realize the simulation of the compound K-distributed sea clutter background,and lay a foundation for the construction of the target detection method in scan mode.3.Based on the proposed local contrast variance weighted information entropy,a fast target detection method is presented.The method can effectively reflect the relationship between the target candidate region and the surrounding regions,and efficiently realize the robust maritime target detection from the SAR image with strong sea clutter background and non-uniform sea clutter background.4.A multi-frame target detection method is presented based on spectral-residualbinary-entropy-based particle filter,which can reduce the computational burden of likelihood ratio calculational and improve the efficiency of undulating target energy accumulation.The proposed method can achieve robust target detection of moving targets from the medium and low signal-to-clutter ratio scanning observation area without depending on the prior information of the sea clutter background.The above-mentioned models of sea surface image background and target detection methods have been verified by simulation and measured data.The results show that the proposed models of sea surface image background can effectively reflect the regular of sea surface radar image,and the proposed target detection methods can achieve fast target detection with high probability.