COHERENCE MODULATED ACOUSTIC SPECKLE INTERFEROMETRY

埋在沿海岸线海床上的复杂水雷对附近的任何船只构成威胁。它们的危险主要来自于它们的伪装，因为它们出现在沙子里很容易被误认为是岩石、珊瑚，甚至是人造材料。尽管使用了大量的海底测绘技术，但要探测到这样的地雷，必须有很高的误报概率。本报告设计并详细介绍了一种表征这些物体以便进行检测的方法。散斑干涉测量是一种光学过程，可以分析物体的粗糙度。这项技术需要将物体暴露在激光下，如果它的反射散射并干扰到物体本身，就可以对物体的表面纹理进行表征。然而，由于介质的不均匀性，光学识别方法在海洋环境中是不可靠的。例如，小颗粒，密度和盐度的波动，或海洋生物。然后将该技术转化为声学环境，其中这种不均匀性较少成为障碍。这个想法是将水下探测仪发出的声波发送到目标物体，并使用线性传感器阵列记录其反射。阵列上振幅的变化，特别是由于信号的部分相干性引起的干扰图，应该允许对反射表面的结构进行分类。该技术设计中面临的一个挑战是确定区分给定表面所需的源信号。为了对物体表面的精细细节进行分类，可能需要一个具有形状谱的信号，为此需要对其参数进行研究和确定。除此之外，还存在开发声学表面反射软件模型的问题，特别是当表面不规则且声源具有复杂的光谱结构时。这份文件描述了这个项目的研究计划，包括可以从中得出的当前工作。这包括之前在博士学位的直接前身所做的工作。计划概述了开发软件工具，这对大部分研究都很重要，并设计了基于类似技术的干涉图确定表面粗糙度的算法。

Sophisticated mines buried in the sea bed on littoralcoast lines pose a threat to any vessel in the vicinity.Their danger comes largely with their camouflage, astheir presence in the sand could easily be mistaken fora rock, coral or even man made material.Despite elo-quent seabed mapping techniques, it is difficult to detectsuch a mine without having a high probability of falsepositives.A method of characterising these objects inorder for detection to be made possible is devised anddetailed in this report.Speckle interferometry is an optical process that al-lows the analysis of an object's roughness.The tech-nique involves exposing the object to a laser, and if its re-flection scatters and interferes with itself, a characterisa-tion can be made on its surface texture.However, opticalmethods of identification are unreliable in ocean envi-ronments due to inhomogeneities in the medium.For ex-ample, small particles, fluctuations in density and salin-ity, or sea life.The technique is then translated into anacoustic context, where such inhomogeneities are lessof an obstacle.The idea is to send a sound wave froman underwater sounder to a target object and record itsreflection using a linear array of sensors.The variationin amplitude over the array, particularly the interferencepatterns due to the partial coherence of the signal, shouldallow a classification of the structure of the reflectingsurface.A challenge which is faced in the design of this tech-nique is determining the source signal required for thedistinction of given surfaces.For the classification offiner details on the object surface, a signal with a shapedspectrum may be necessary, for which the parametersneed to be investigated and determined.Further to this,there is the problem of developing a software model foracoustic surface reflections, particularly when the sur-face is irregular and the source has a complex spectralstructure.This document describes the plan of research for thisproject including the current work from which can bedrawn.This includes previous work done on the topicas a direct predecessor to the Ph.D. The plans are out-lined for developing the software tool, which will beimportant for much of the research, and designing thealgorithms for determining surface roughness from theinterferograms based on similar technologies used formapping.