基于二维连续小波变换的颗粒物数字粒径方法研究  

Digital grain size estimation based on two-dimension continuous wavelet approach

作　　者：郭立恒 刘海江[2] GUO Liheng;LIU Haijiang(Ocean College,Zhejiang University,Hangzhou 310058,China;College of Civil Engineering and Architecture,Zhejiang University,Hangzhou 310058,China)

机构地区：[1]浙江大学海洋学院,浙江杭州310058 [2]浙江大学建筑工程学院,浙江杭州310058

出　　处：《海洋工程》2018年第5期38-48,共11页The Ocean Engineering

基　　金：国家重点研发计划项目(2017YFC0405401);国家自然科学基金(11632012;51761135015)

摘　　要：对于海岸工程的研究而言,测定泥沙、砾石、卵石等颗粒物的中值粒径具有重要的研究意义。近年来,许多学者逐渐将统计学方法应用于分析泥沙等颗粒物的粒径大小及其级配分布,小波变换是该研究领域时新的统计计算数学方法之一。由于一维连续小波变换(1D-CWT)在估算颗粒数值粒径方面存在一定的局限性,本研究基于二维连续小波变换(2D-CWT),提出了新的颗粒物粒径统计学分析方法,用于估算筛分良好的颗粒物中值粒径。以玻璃珠、卵石块和砾石三类不同性质的颗粒物构建数字粒径实验分析样本,对新方法进行了测试和验证。研究表明基于2D-CWT的数字粒径分析方法能够快速、便捷地估算出筛分良好的颗粒物的中值粒径,且计算结果准确、稳定。相较于基于1D-CWT的数字粒径分析方法,2D-CWT方法在颗粒物数字粒径分析方面有着更高的计算成功率和更广的适用性。Determination of median diameter of sediment particles is of great significance to coastal engineering.Recently,various statistical approaches have been used to determine the digital grain size and associated parameters,among which the wavelet transform technique is becoming an up-to-date one.Considering that the one-dimensional continuous wavelet(1D-CWT)approach has certain limitations in digital grain size estimation,this study proposed a new method to estimate the medium grain size of well-sorted particles on the basis of two-dimensional continuous wavelet(2D-CWT)algorithm.Subsequently,the 2D-CWT method was tested by estimating the medium grain size of various particle samples consisting of white glass beads,cobble stones and gravels.It demonstrates that the newly proposed 2D-CWT approach can estimate the median diameter of sample grains promptly and easily,and the corresponding results are accurate and robust.Compared with the 1D-CWT approach,digital grain size estimation using the 2D-CWT method shows a higher successful rate in medium grain size estimation and a more extensive range for application.

关 键 词：颗粒物数字粒径 二维连续小波变换 图像分析 中值粒径 筛分良好颗粒物 

分 类 号：P751[交通运输工程—港口、海岸及近海工程]