摘要
传统路面检测方法在拼接及裂缝识别上效果不佳,与实际路面状况偏差大,无法作为维护和养护依据。为解决此问题,研究了基于三维点云算法的路面拼接及裂缝检测方法。基于三维点云算法,分别对路面拼接及裂缝的平面结构状态进行拟合。通过路面动态测量(PDS)系统进行三维点云下拼接及裂缝单层特征提取。提取拼接及裂缝坐标特征、整合目标特征的分布信息,以提升检测准确度。通过混合损失函数整合拼接及裂缝特征,对计算产生的损失偏差进行计算补偿,以实现路面拼接及裂缝检测。试验结果表明,所提方法的拼接及裂缝检测准确度达到93.76%、数据权重值低于0.6。该方法能够有效提升检测结果的准确度及可靠性,具有较高实际应用价值。
Traditional pavement inspection methods are ineffective in splice and crack identification,and deviate greatly from the actual pavement condition,which cannot be used as the basis for maintenance and repair.To solve this problem,a pavement splice and crack detection method based on three-dimensional point cloud algorithm is researched.Based on the three-dimensional point cloud algorithm,the planar structural state of the pavement splices and cracks are fitted respectively.The single-layer feature extraction of splices and cracks under three-dimensional point cloud is carried out by pavement dynamic survey(PDS)system.The splice and crack coordinate features is extracted and the distribution information of target features is integrated to improve the detection accuracy.The splice and crack features are integrated by hybrid loss function,and the loss deviation generated by the calculation is compensated by calculation to realize the pavement splice and crack detection.The test results show that the accuracy of the proposed method for splice and crack detection reaches 93.76%,and the data weight value is lower than 0.6.The method can effectively improve the accuracy and reliability of the detection results,and has high practical application value.
作者
陈华岳
邱锐
邓怡明
CHEN Huayue;QIU Rui;DENG Yiming(Guangzhou Shiwei Testing Co.,Ltd.,Guangzhou 510288,China)
出处
《自动化仪表》
2025年第1期21-25,共5页
Process Automation Instrumentation
关键词
三维点云
路面拼接
检测方法
坐标特征
检测偏差
Three-dimensional point cloud
Pavement splicing
Detection method
Coordinate feature
Detection deviation
作者简介
陈华岳(1990-),男,学士,工程师,主要研究方向为道路桥梁检测及养护管理,E-mail:Kaitlyestin@163.com;通信作者:邓怡明,男,本科,主要研究方向为道路桥梁检测及养护管理,E-mail:1371912177@139.com。
