Research On Multi-Frame Accumulation Detection Methods For Weak Target In Sea Surface Radar Images

To maintain marine safety and detect the marine environment,radar remote sensing technology is used to obtain maritime information.For the radar image generated by the imaging radar in the sea scene,it is necessary to use an effective target detection method to find the target that users pay attention to.However,in practical applications,there are always weak targets with weak echo signal amplitude,small reflection cross section and fluctuating and unstable characteristics,which puts forward higher requirements for existing target detection technology.Therefore,it is urgent to study how to adapt to weak targets and make effective use of radar image information to realize fast detection and accurate extraction of concerned targets.This thesis mainly analyzes the characteristics of weak targets and clutter background,and conducts research on multi frame accumulation detection methods to address the phenomenon of information omission in multiple frames of unstable weak single targets;To address the issue of interference in target detection caused by the accumulation envelope expansion of weak multiple targets,research on effective detection methods for multiple targets has been carried out.And the main tasks and innovations are as follows:1.The weak single target with fluctuating and unstable characteristics is studied,and a multi frame joint accumulation detection method is proposed.This method uses the dynamic programming method optimized by the bionic memory model to accumulate the collapse energy caused by the target loss in the image,enhance the correlation between multi frame images,effectively use the radar image information,and ultimately improve the detection performance of weak targets.2.The accumulation envelope expansion caused by the presence of weak multiple targets resulted in mutual interference in the detection of targets close to each other,and an effective detection method for weak multiple targets was proposed.This method utilizes pseudospectral information to assist in processing,and combines with a biomimetic memory model based method to focus on cell energy,improve the accumulation efficiency of multiple frame images,and ultimately reduce the probability of missed detection and false detection.The simulation results show that the above methods can effectively improve the detection ability of weak targets with unstable sea surface,widely adapt to the application scenarios of weak multi-target detection,and realize the reliable detection of the concerned targets.