| 基于数字图像处理的颗粒流厚度动态提取方法研究 |
| |
| 引用本文： | 吴越,李坤,程谦恭,王玉峰,龙艳梅,姜润昱,宋章,刘毅.基于数字图像处理的颗粒流厚度动态提取方法研究[J].水文地质工程地质,2021,48(4):151-159. |
| |
| 作者姓名： | 吴越 李坤 程谦恭 王玉峰 龙艳梅 姜润昱 宋章 刘毅 |
| |
| 作者单位： | 1.西南交通大学地质工程系，四川 成都 610031 |
| |
| 基金项目： | 第二次青藏高原综合科学考察（2019QZKK0906）；国家自然科学基金项目（41530639；41761144080；41877226；41877237） |
| |
| 摘 要： | 颗粒流厚度及其演化趋势是碎屑流物理模型试验中重点关注的要素。目前试验中颗粒流厚度的监测主要有传感器监测、机械原件测量、人工测读等方法。随着计算机跨学科的应用及计算机数字图像处理技术的成熟，越来越多的数字图像处理技术被应用于工程地质领域。以颗粒流斜槽试验为依托，基于自适应中值滤波、图像二值化、图像腐蚀及种子填充等数字图像处理方法，对高速相机所采集的颗粒流图像序列进行处理分析并编制了相关程序，实现了连续提取颗粒流运动过程中的厚度值。分析结果表明:基于数字图像处理方法提取的颗粒流厚度在颗粒流主体区段与实测厚度值相吻合，在颗粒流尾部由于颗粒离散会存在一定的偏差，主要是由于部分三维空间中的颗粒在二维图像中呈现出重叠的形式，造成颗粒连续的假象。总体而言，通过该方法获取的颗粒流厚度值在一定条件下具有较高的精度，相比于其他方法具有效率高、获取参数多、采样频率高、扰动低等优势，可作为颗粒流试验中流态参数动态获取的常规方法之一。
|
| 关 键 词： | 颗粒流 厚度 斜槽试验 高速相机 数字图像处理 流动型滑坡 |
| 收稿时间： | 2020-07-14 |
A study of the dynamic extraction method for granular flow thickness based on digital image processing |
| |
| Institution: | 1.Department of Geological Engineering, Southwest Jiaotong University, Chengdu, Sichuan 610031, China2.State-Province Joint Engineering Laboratory of Spatial Information Technology for High-Speed Railway Safety, Chengdu, Sichuan 611756, China3.Key Laboratory of High-Speed Railway Engineering, Ministry of Education, Chengdu, Sichuan 610031, China4.China Railway Eryuan Engineering Group Co. Ltd., Chengdu, Sichuan 610031, China |
| |
| Abstract: | The granular flow thickness and its evolution trend are key factors in analyses of the physical model experiments of avalanches. At present, the measurements of granular flow thickness mainly include sensor monitoring, mechanical measurement and manual measurement. With the development of computer technology, digital image processing methods have been applied in more and more subjects including engineering geology. In this paper, a series of granular chute flow experiments are conducted for the extraction of granular flow thicknesses. Based on the digital image processing methods including adaptive median filter, image binarization, image erosion and seed filling, the granular flow image sequence recorded by the high-speed camera is processed and analysed with related codes compiled, thereby achieving continuous extraction of flow thickness during the granular flow propagation. The results show that the granular flow thicknesses extracted with the digital image processing methods are consistent with the measured true thicknesses in the main bodies of the granular flows. Some deviations appear in the tail of granular flow accompanied by significant particle dispersions. The deviations are mainly attributed to the fact that some particles in the three-dimensional space show overlapping phenomenon in the two-dimensional image, which results in the illusion of granular continuity. In general, the granular flow thicknesses obtained with the digital image processing methods is of a high accuracy in the densely packed granular flows. Compared with other methods, this method is of higher efficiency, higher sampling rate and lower disturbance on granular flows. In addition, additional parameters, such as velocities and displacement, can be obtained simultaneously. Therefore, this method can be used as one of the conventional methods to obtain kinematic parameters in granular flow experiments. |
| |
| Keywords: | |
|
| 点击此处可从《水文地质工程地质》浏览原始摘要信息 |
| 点击此处可从《水文地质工程地质》下载免费的PDF全文 |
|
