基于条纹投影和二维主成分分析的边坡监测方法

Monitoring method of slope based on fringe projection and two-dimensional principal component analysis

针对现有边坡监测方法存在监测点少、分辨率低、测量周期长等问题,为了实时监测煤矿边坡的沉降并有效抑制虚警,提出了一种基于条纹投影和二维主成分分析的边坡监测方法。该方法首先使用三频彩色条纹投影轮廓术快速精确地获取被测边坡的高程信息,然后对其进行二维主成分分析,简化计算量,最后根据被测边坡对应位置不同时刻的欧氏距离来判断边坡是否发生虚警。实验结果表明,该监测方法能有效地监测到边坡的虚警,具有良好的实时性和鲁棒性。

In view of problems of less monitoring points,low resolution,long measurement cycle of existing slope monitoring method,a monitoring method of slope based on fringe projection and twodimensional principal component analysis was proposed in order to real-timely monitor settlement of slope and effectively restrain false alarm.Firstly,the method uses triple-frequency color fringe projection profilometry to obtain height information of observed slope quickly and accurately.Then it uses twodimensional principal component analysis to analyze the height information and simplify computation amount.Finally,it adopts Euclidean distance corresponding to the measured slope position of different times to judge whether the slope false alarm occurs.The experiment results show that the monitoring method can effectively monitor false alarm of slope,and has a good performance of real-time and robustness.