首页 | 本学科首页   官方微博 | 高级检索  
     

基于摄影法的红黏土裂隙试验研究
引用本文:熊俊豪,刘宝臣,刘磊,颜荣涛,张炳晖. 基于摄影法的红黏土裂隙试验研究[J]. 地下空间与工程学报, 2020, 16(2): 381-377
作者姓名:熊俊豪  刘宝臣  刘磊  颜荣涛  张炳晖
基金项目:国家自然科学基金(11562007);桂科能基金(16-J-21-6)
摘    要:准确提取红黏土裂隙特征,有助于进一步研究裂隙对红黏土工程性质的影响。本文以桂林红黏土试样为试验研究对象,获取红黏土表面裂隙数字图像后,通过比对不同方法,发现直方图阈值法适用于红黏土的裂隙图像分割。通过去除杂点、填充细小孔洞,将裂隙数字图像转化为二值图像,最后进行骨骼化和消刺处理,可得裂隙骨骼图像。研究发现试样在第5次干湿循环后裂隙发育均已趋稳定;初始含水率为25%的试样在干湿循环作用下发育出线状裂隙;初始含水率为28%的试样则发育出网状裂隙;干湿循环次数与裂隙率和裂隙总长度具有良好的相关性,可用二次函数拟合。

关 键 词:红黏土  裂隙  图像处理  相关性  
收稿时间:2019-10-24

A Method for Crack Image Processing of Red Clay
Xiong Junhao,Liu Baochen,Liu Lei,Yan Rongtao,Zhang Binghui. A Method for Crack Image Processing of Red Clay[J]. Chinese Journal of Underground Space and Engineering, 2020, 16(2): 381-377
Authors:Xiong Junhao  Liu Baochen  Liu Lei  Yan Rongtao  Zhang Binghui
Abstract:Accurate extraction of red clay crack features contributes to further investigate the influence of cracks on engineering properties of red clay.This paper takes Guilin red clay sample as research object,after obtaining red clay surface crack photograph for the digital image.By comparing different methods,which found that histogram threshold method is suitable for the image segmentation of red clay crack.Wiping off miscellaneous points and filling tiny holes,the red clay crack digital images into binary image,finally,the image of the crack skeleton is obtained by skeletal treatment and removing burrs treatment.Research finding the crack development of samples is stable after the 5th dry-wet cycle;the linear crack is developed in the samples with initial water content radio of 25% under the dry-wet cycle,and the reticular crack is developed in samples with an initial water radio of 28%;the times of dry-wet cycle has a strong correlation with the crack rate and the total length of crack,which can be fitted by quadratic function.
Keywords:red clay  crack  image processing  correlation  
本文献已被 CNKI 等数据库收录!
点击此处可从《地下空间与工程学报》浏览原始摘要信息
点击此处可从《地下空间与工程学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号