文章摘要
雷雨萌,陈祖煜,于沭,温彦锋,王玉杰,李炎隆.基于深度阈值卷积模型的土石料级配智能检测方法研究[J].水利学报,2021,52(3):369-380
基于深度阈值卷积模型的土石料级配智能检测方法研究
Intelligent detection of gradation for earth-rockfill materials base on Deep Otsu Convolutional Neural Network
投稿时间:2020-07-13  
DOI:10.13243/j.cnki.slxb.20200499
中文关键词: 土石料级配  深度阈值卷积模型  筛分试验  细小黏连颗粒  图像识别  土石坝
英文关键词: gradation of earth-rockfill materials  DO-CNN  artificial screening trials  small touching particles  image recognition  rockfill dam
基金项目:国家重点研发计划项目(2018YFC0407103);中国水利水电科学研究院基本科研项目(GE0145B0112017)
作者单位E-mail
雷雨萌 西安理工大学 省部共建西北旱区生态水利国家重点实验室, 陕西 西安 710048  
陈祖煜 西安理工大学 省部共建西北旱区生态水利国家重点实验室, 陕西 西安 710048
中国水利水电科学研究院 岩土工程研究所, 北京 100048 
 
于沭 中国水利水电科学研究院 岩土工程研究所, 北京 100048 yushu@iwhr.com 
温彦锋 中国水利水电科学研究院 岩土工程研究所, 北京 100048  
王玉杰 中国水利水电科学研究院 岩土工程研究所, 北京 100048  
李炎隆 西安理工大学 省部共建西北旱区生态水利国家重点实验室, 陕西 西安 710048  
摘要点击次数: 106
全文下载次数: 67
中文摘要:
      土石料级配合理性直接影响土石坝的力学特性与抗渗性能,当前级配检测主要依靠人工筛分试验,无法实现大规模快速检测。传统图像识别算法不符合土石料级配检测的精度要求,深度学习图像识别算法需要海量人工标记的训练样本,难以满足工程实际需求。本文结合传统图像识别中基于最大类间方差的边缘检测算法与卷积神经网络深度学习模型,研究了土石料图像识别与级配数据智能分析相结合的级配检测方法,建立了可实现级配快速检测的深度阈值卷积模型(Deep Otsu Convolutional Neural Network,DO-CNN),提高了级配检测的精度,并以灰岩石料为典型样本,通过18组标准筛分试验获取土石料级配数据及图像,进行模型训练与验证。结果表明:与仅使用基于最大类间方差法的边缘检测模型相比,DO-CNN模型能够极大提高级配检测的准确率与稳定性,实现基于图像的土石料级配快速检测。对于5 mm以下细小土石料颗粒,模型识别精度同样较高。
英文摘要:
      The mechanical properties and impermeability of rockfill dam can be directly affected by rationality of earth-rockfill materials gradation. At present, the gradation testing mainly relies on artificial screening, which cannot achieve large-scale rapid detecting. Traditional image recognition algorithms do not meet the accuracy requirements of earth-rockfill materials gradation detecting. Image recognition based on deep learning require massive manual labeled samples, which is difficult for engineering. In this paper, c ombining with the edge detection algorithm based on Otsu in traditional image recognition and the deep learning model of convolutional neural network, a Deep Otsu Convolutional Neural Network model is established to realize rapid detection of gradation by integrating the image recognition of earth-rockfill materials and intelligent analysis of the gradation data. Using limestone as a typical sample, the gradation data and images are obtained through eighteen groups of standard screening trials for model training and verification. The results show that the DO-CNN model can greatly improve the stability and accuracy of gradation detecting compared to using the edge detection model based on Otsu algorithm,and realizes the rapid detection of earth-rockfill materials gradation based on image. For the small earth-rockfill materials under 5mm, the model also maintained high accuracy.
查看全文   查看/发表评论  下载PDF阅读器
关闭