Research On Adaptive Target Detection Algorithm In Complex Background

With the continuous development of radar detection technology,target detection algorithms have been widely used in various fields.However,the echoes of radar practical application scenarios are often filled with a large amount of clutter,especially in the context of multiple neighbouring targets and across different power clutter zones to the detection performance of sea radar is greatly challenged.This thesis first analyses the current state of research on target detection algorithms and introduces the commonly used target detection algorithms and their advantages and disadvantages.Then,the research is focused on the detection of targets in the background of multiple neighbouring targets and across the background of different power clutter areas in the sea scene,and the main research contents are as follows:1.To solve the problem of degraded detection performance at multiple neighbouring target backgrounds and clutter edges,an adaptive constant false alarm detection method based on background discrimination is proposed.The method first determines whether the clutter edge exists by the mean ratio,then estimates the position of the clutter edge using the maximum likelihood criterion,and uses a suitable background power instead for iterative calculation at the target position.The experimental results show that the proposed method can effectively improve the detection performance at multiple neighbouring targets and clutter edges.2.In order to improve the detection performance of multi-neighbourhood target backgrounds in the 2D plane,a 2D constant false alarm detection method based on OS-CFAR improvement is proposed.Firstly,the detection performance of two-dimensional meanlike CFAR and ordered statistical-like CFAR for multi-neighbourhood target background is analysed,and then the selection of reference units is improved based on two-dimensional OS-CFAR to disperse the energy of neighbouring targets by combining reference units,and the theoretical derivation is carried out.The experimental results show that the method can effectively improve the detection probability of sea radar in multiple neighbouring target backgrounds with unknown detection background a priori information.3.In order to improve the detection performance in the sea clutter background,a Res Net-based sea clutter suppression method is investigated so as to improve the signalto-clutter ratio of the target.Firstly,a dataset is constructed using radar-to-sea detection real data,and then a sea clutter suppression model is built on the basis of Res Net50,Res Net101 and Res Net152 structures respectively,and the sea clutter suppression performance of Res Net models with different optimisers and different number of layers is analysed.The experimental results show that the Res Net-based sea clutter suppression method can effectively suppress sea clutter in the context of multiple neighbouring targets and clutter edges.