摘要
沥青路面病害图像的分析目前主要依靠人工识别的方法,效率较低。现采用深度学习神经网络,对路况检测车和无人机自动设备采集的大量病害图片进行特征分析、训练和学习,并开发出相应的病害自动识别分析软件,以提高识别精度和效率。
At present,the analysis of asphalt pavement distress image mainly depends on the method of manual recognition,which is not efficient.In this paper,deep learning neural network was used to analyze,train and learn the characteristics of a large number of disease pictures collected by road condition detection vehicles and unmanned aerial vehicle.Afterwards,the corresponding automatic recognition software was developed to improve the recognition accuracy and efficiency.
出处
《上海公路》
2021年第4期16-19,39,M0003,共6页
Shanghai Highways
基金
上海市2019年度“科技创新行动计划”社会发展领域项目(19DZ1203002)。
关键词
路况检测车
无人机
沥青路面病害
深度学习神经网络
road condition detection vehicle
unmanned aerial vehicle
asphalt pavement distress
deep learning neural network
