Research On The Algorithm Of Underwater Target Active Detection Based On Matched Filter

As the development of human society becomes more and more dependent on the ocean,the marine cause,especially the issue of coastal defense,has attracted great attention from countries around the world.The sonar system has also played an increasingly critical role in it.People's requirements for the sonar system are no longer limited to finding targets,but need to obtain the position,attitude and type of the targets while discovering the targets.At the same time,with the development of new combat equipment such as frogmen and underwater vehicles,the detection of such small underwater targets has become more difficult.Therefore,in this context,it is of great significance to study high-performance detection algorithms and explore their feasibility and reliability in engineering practice.As the most commonly used signal detector,the matched filter is a "best" linear filter based on the maximum output signal-to-noise ratio under a white noise background(ideally).Therefore,based on the matching filter theory,this paper studies the post-processing algorithm of the matched filter for underwater moving targets and the purpose of improving the processing performance of the matched filter.Focused on the frequency domain adaptive line spectrum enhancement technology based on matched filtering,under low signal-to-noise ratio,the algorithm reflects more than the classic matched filter and the deconvolution matched filter introduced as a comparative study.Excellent detection performance,effectively suppresses noise,and further improves system gain.In addition,after the signal is transmitted through multiple channels underwater,the system gain of the matched filter is reduced.At the same time,due to the effect of multi-path effects,other relevant peaks will be generated at the output of the matched filter,resulting in long processing gain and range resolution.Lower than theoretical predictions.Aiming at this problem,this paper studies the environment-modified matched filter.By modifying the reference signal of the matched filter,the correlation between the signal echo and the transmitted signal is improved,and the problem of processing performance degradation when the matched filter is affected by the multipath effect is further improved.At the same time,the accuracy of channel estimation has a direct impact on the detection performance of the environmentally modified matched filter.Therefore,this paper further compares two channel estimation algorithms,and chooses a second-order cone programming algorithm with lower error at low signal-to-noise ratio to apply to the environment-modified matched filter.Finally,in this paper,the azimuth history map obtained by the experimental processing is used to extract moving targets in the opencv environment by computer vision,so that the target and the target's motion trajectory are displayed on the screen more clearly and intuitively.