Feature Extraction And Tracking Of Underwater Moving Targets Based On Transform Domain

With the development of modern sonar technology, the radiated noise of underwatermoving target has continuously reduced, the concealment performance of underwater hasimproved significantly, so the feature extraction of underwater moving targets and trackingplays an important role in the research subject of the processing of passive sonar signal.Underwater moving targets in variable speed generated by the radiation noise has thecharacteristics of transient signals, has yet to become the main objects of noise and vibrationreduction. This paper uses the method of signal processing and image processing forunderwater variable motion target radiation noise signal analysis.The main research works of the paper are as follows:First of all, this paper analyzes the basic properties of variable speed moving underwatertarget radiated noise, and according to the practical mathematical model is established, getsimulation signal which is more close to the measured signal.Then, the paper studied the application of the time-frequency analysis technique andimage preprocessing technique in the treatment of radiated noise signal of underwater variablespeed moving target. Studied the transform domain feature extraction algorithm, aiming at theshortcomings of the traditional Hough transform algorithm, in this paper, the algorithm isimproved. Feature extraction algorithm is proposed by VC visual programming. Use thesoftware for algorithm simulation and experiment data processing. Verified the feasibility ofthe algorithm in feature extraction.Finally, studied the threshold segmentation algorithm in image processing, and thealgorithm was applied in line spectrum tracking of underwater moving targets. Verified thefeasibility of the algorithm by simulation and experiment data processing.The paper aims at improving the passive sonar warning capabilities and providingtheoretical basis and technical support by the study of this paper.