Morphological Research On Underwater Target Echo Geometric Highlight Structure

In present, underwater target detection technology is one of search focuses on underwater sound technology at home and abroad, and a hard problem recognized by academic circle as well, which is studied and analyzed on the basis of acoustic scattering mechanism and the characteristics of target of the underwater target. Underwater target echo geometric highlight structure is a major feature of target echo. However,in the time domain,frequency domain, spatial domain, the processing methods of the echo geometric highlights are always limited by resolution. This is unable to distinguish the geometric highlights of target echo exactly. The paper focuses on the underwater target echo highlight geometric structure under the background of noise,which put forward a new target echo analysis method by combining morphology image processing and time-frequency analysis of signal.The paper mainly introduces the researches about target echo scattering and time-frequency features as well as morphology image analysis. Theoretically, Wigner-Ville Distribution (WVD) of highlight echo and highlight parameters of sun-goal related to target models of the paper are derived. Moreover, the paper induces the basic concepts and theories of morphology image analysis,and also introduces basic operation and transform form two aspects: binary morphology and gray scale morphology.Based on the theoretical basis above, the paper studies a new method: Reconstruction Cross-Component Removal (RCCR). By analyzing the reasons of WVD Cross-Component,combining target echo's features of highlight model, time-frequency features of Linear Frequency-Modulated Signal (LFM) and morphology operation and transform, the paper studies the method of processing target echo based on morphology reconstruction. This method combines Short Time Fourier Transform of time frequency analysis and Wigner-Ville Distribution, takes advantage of morphology reconstruction method to remove Cross-Component produced by multi-component input signals in Wigner-Ville Distribution,constantly fits and recovers the components of output signals so as to analyze the geometric structures of underwater target echo.Simulation results show that the method in the case of changes of component signal chirp rate, energy, number and part of overlap between the component signals is feasible. In order to improve that this method affects underwater sound environment, the paper analyzes the target model by combining pool experiment. The processing results of echo data in this experiment highlight the real highlights of the target echo geometric highlight structure; also suppress cross-term of echo signal WVD and noise. Therefore, it is confirmed that the Cross-Component Removal reconstruction operation of the paper is effective.