摘要
高速公路是重要的基础建设项目,针对传统公路施工状态评估效率低下问题,本文提出一种基于深度学习的施工状态快速评估方法。选取高分辨率无人机影像进行实验,首先根据施工状态对目标进行标注与切分,制作深度学习数据集,输入DeepLabV3深度学习网络中进行训练,构建出施工状态分类模型,进而对研究区进行分类。两个研究区的实验证明,本文方法状态识别精度优于同类方法,基于本文方法能够实现公路施工宏观状态准确快速评估。
Highway construction is an important part of national infrastructure construction.This paper proposed a rapid evaluation method of construction status based on deep learning by using UAV images.High-resolution UAV images were applied in the experiment.All the images were labeled and classified according to the construction status which we called learning data set.The learning data set was input into the DeepLabV3 to obtain the classification model.Experiments in two research areas show that the state recognition accuracy of the proposed method is better than that of similar methods.Based on the proposed method,the macro state of the highway construction work can be quickly evaluated.
作者
李圣明
刘亚萍
明洋
孙杰
Li Shengming;Liu Yaping;Ming Yang;Sun Jie(CCCCC Second Highway Consultants Co.,Ltd.,Wuhan 430056,China;School of Information Engineering,China Geoscience University,Wuhan 430074,China)
出处
《工程勘察》
2023年第6期55-60,共6页
Geotechnical Investigation & Surveying
基金
武汉市科技计划项目(2019030703011506)
中交第二公路勘察设计研究院有限公司科技研发项目(KJFZ-2017-046,KJFZ-2018-043)。
关键词
施工状态
DeepLabV3
监督分类
遥感影像分类
工程进度评估
construction status
DeepLabV3
supervised classification
remote sensing image classification
construction status evaluation