基于模极大值特征提取的超声波海底沉积物分类识别研究

【作者】 聂双双；

【导师】 邓跃红；

【作者基本信息】 中南大学 ， 机械设计及理论， 2008， 硕士

【摘要】 随着陆地矿产资源的日趋枯竭,大洋钴结壳成为21世纪最具有商业开采价值的战略资源,世界主要发达国家已经开始了开采研究工作。为加快我国深海钴结壳开采的步伐,在国家自然科学基金项目“深海钴结壳微地形监测技术与最佳采集深度建模研究”的资助下,本论文对海底沉积物的识别与分类做了相关研究。根据模极大值在特征提取中的应用,本文提出了一种基于模极大值的超声波回波探测沉积物分类方法。利用超声波探测不同的沉积物得到不同的回波波形,运用互相关算法找出真实回波起点,截取回波,再对回波进行小波变换,提取模极大值特征,然后将模极大值特征利用最佳鉴别矢量法进行特征压缩,最后送入分类器进行分析。在此理论基础上,根据自然界中沉积物的分类标准,由泥、砂、砾三种组分通过不同的配比组成14种不同的海底沉积物,对应14种不同的回波,得到的模极大值的特征是不同的。通过水下回波实验建立了一套沉积物的分类识别模型。运用所提出的方法,对建立的模型进行实验验证。通过采集湘江河床的沉积物作为样品,进行水下回波实验,并将样品进行筛分称重,得到真实的沉积物类型,将实验得到的分类结果与筛分得出的结果进行对比,得到正确识别率在80%左右。由此可以看出,本论文所提出的沉积物分类方法及建立的分类识别模型是正确的、可靠的。本论文的研究为海底沉积物的识别与分类提供了良好的理论基础,对海底地貌及其性质的识别具有实际意义,为我国的深海采矿事业提供有效的技术支持。更多还原

【Abstract】 Along with land mineral resource used up, Oceanic cobalt crust resource has been a commercial foreground strategical resource in 21 century and the head developed countries in the world have been doing work of investigation and exploitation. In order to quicken the step of exploitation of seabed cobalt crust in our country, and imbursed by the national natural science fund item "Study on the Abyssalbenthic Cobalt-rich Crusts Tiny terrain Detecting Technology and the Best Collection Deepness Model", the investigation was done of the identification and classification of seabed sediment in the dissertation.A classification method was proposed based on the application of modulus maxima in feature extraction. Different echo is gained to detect different sediment. The jumping-off was found out making use of the correlation principle. Intercept echo. Then do wavelets transform and pick out the modulus maxima feature. They were compressed with optimal set of discriminant vectors. At the end, analysis was done in the classifier. Based on above-mentioned theory, according as the standard of nature sediment, 14 sediment was made of mud、sand and gravel according to different proportion. The modulus maxima feature is different corresponding different echo. A classification and identification model was found on experiments. In order to validate the model, we sampled sediment from the XiangJiang riverbed. The experiment was done to the samples. A portion of the samples were griddled and weighed. Compared the two results, the exactness identification rate was about 80 percents. Look from the results, this method and model are correct and reliable.The research affords favorable theoretic value for the identification and classification of seabed sediment. It also has realistic significance for the seabed physiognomy and characters. And the research furnishes effective technique sustain for the country abyssalbenthic mining.更多还原