基于无人机的建筑物变形监测方法研究

陈鲲翔

摘  要

随着我国经济的高速发展，城市土地资源逐渐趋向饱和，地下空间不断被拓展，影响建筑物变形的因素越来越多样化，导致建筑物变形事故频发，因此定期对建筑物的变形监测管理愈显重要。目前建筑物的变形监测方法大多无法直观快速的获取变形信息，同时需要开放的监测环境，但是城市内建筑物的容积率越来越高，开放的监测环境实现较难，因此探索简便新颖的监测方法具有重要意义。

本文在综述筑物变形监测的相关理论和研究现状的基础上，总结了建筑物变形机理并对比分析了现有的监测方法。针对现有理论及技术研究的不足，本文基于无人机图像技术提出了一种新的位移测量方法，运用Pix4Dmapper软件建立点云模型，将区域增长分割原理运用于点云分割中来获取标识点的点云信息，并基于MATLAB针对求解标识点中心点提出了点云包围盒与拟合平面相结合的方法。最后，为了验证本方法的有效性，基于建筑物模型开展了用无人机、全站仪和三维激光扫描仪测量的实验，通过实验结果对比分析了三种方法的误差来源。根据结果可知，基于标识点的无人机图像技术在三维变形监测方面的测量精度接近于三维激光扫描仪的测量精度，其中半径越大的标识点具有更高的竖直位移精度。对比全站仪的测量结果得出该方法主要适用于三、四等测量精度，可用于常规性快速监测。

因此，本文的研究成果以期帮助监测人员更直观快速的获取建筑物的安全状态，可丰富建筑物变形监测的方法，同时还可为无人机图像在工程安全管理中提供一个新的应用角度。

关键词：建筑物变形  监测方法  无人机图像  点云分割  位移计算

Abstract

With the rapid development of China’s economy, urban land resources are becoming saturated, underground space is continuously expanded. Therefore, more and more security incidents caused by building deformation and it is important to regularly monitor the deformation of buildings. At present, the method of building deformation monitoring is always based on traditional means, which is impossible to obtain real-time information quickly and intuitively. At the same time, most of them need an open monitoring environment, but the volume ratio of buildings in cities is getting higher and higher. So it is significance to improve the monitoring technology of building deformation.

Based on the review of the relevant theories and research status of building deformation monitoring, this thesis summarizes the deformation mechanism of buildings and compares the existing monitoring methods. A new displacement measurement method based on UAV image technology was proposed. The Pix4Dmapper software is used to establish a point cloud model, and the regional growth segmentation principle is applied to point cloud segmentation to obtain. The point cloud information of the point is identified, and a method combining the point cloud bounding box and the fitting plane is proposed based on MATLAB for solving the marker point center point. Finally, in order to verify the effectiveness of the method, experiments based on the building model were carried out with UAV, total station and 3D laser scanner. The error sources of the three methods were compared and analyzed through experimental results. According to the results, the measurement accuracy of the UAV image technology based on the marker point is close to 3D laser scanner, and the marker with larger radius has higher vertical displacement accuracy. Compared with the measurement results of the total station, the method is mainly applicable to the measurement accuracy of the third and fourth, and can be used for routine rapid monitoring.

Therefore, the research results of thesis are intended to help the monitoring personnel to obtain the safety status of the building more intuitively and quickly, enrich the method of building deformation monitoring, and provide a new application angle for the engineering safety management of the drone image.

Keywords：Building deformation  Monitoring method Drone image

    Point cloud segmentation  Displacement calculation