| With the rapid development of my country's economy and the acceleration of urbanization,buildings with super-high,super-scale and complex structures have come up like mushrooms after a rain.In order to ensure the safety of their entire life cycle during construction and operation after completion,they are to be carried out on a regular basis monitor.However,the deformation of such multifarious buildings is uneven,both integral and partial.Deformation monitoring methods represented by total stations and GNSS precision positioning technology mainly use discrete point measurement,and it is difficult to reflect the deformation laws of the overall and local structures of complex buildings with points and planes.3D laser scanning,drone tilt photogrammetry,close-range photogrammetry and other technologies can quickly acquire 3D spatial data of the target and establish a 3D real scene model,which can truly and comprehensively restore and reflect the full picture of the target,and has an intuitive expression of the overall space of the object information capacity.After the deformation monitoring is combined with the three-dimensional spatial data and the real scene model,not only can we have an intuitive and comprehensive understanding of the deformation of the target,but also the deformation results and trends can be correlated with the monitoring points to improve the relevance and integrity of the monitoring,and make the monitoring results more three-dimensional and intuitive.However,when the target is a composite structure,the single technology is limited by the observation mode,and it is difficult to achieve the complete 3D spatial information acquisition of the target and the reconstruction of the real scene model.It is urgent to be used multi-source data fusion technology to make up for the above shortcomings in order to better integrate the 3D space.The datum is applied to the field of building deformation monitoring.This article focuses on the combination of 3D laser scanning technology and UAV tilt photogrammetry technology,supplemented by digital camera close-range photogrammetry,organically integrates these several technical means,exerts their respective advantages in 3D reconstruction,and establishes a comprehensive and complete,A three-dimensional real scene model with well expressed details,and this fusion point cloud data is applied to the deformation monitoring of buildings to realize the transformation from traditional point monitoring to "shape monitoring" and "body monitoring".The main research contents are as following:(1)Introduced the measurement working principle,system composition and classification method,point cloud data collection and processing process of 3D laser scanning technology.Based on the analysis of the main factors affecting the accuracy of 3D laser scanning,the two factors,scanning distance and laser incident angle,are the principal research objects,and the law of the two factors affecting the scanning accuracy is studied.The design experiment is further verified,and the deformation monitoring is given.Reasonable scanning distance and angle range when working in the field.The designed simulation deformation experiment further embodies the advantages of the application of 3D laser scanning point cloud to deformation monitoring.(2)Taking UAV tilt photography and close-up photography as the research objects.The methods and procedures of air-ground image data collection and real scene reconstruction are given.Firstly,based on the analysis of drone tilt photogrammetry technology and its application benefits,a detailed process of internal and external operations of drone tilt photography is given.A reliable teaching building of our school is selected as the object,and the drone tilt photogrammetry technology is used for data.Acquisition,3D model reconstruction,and accurate assessment of the model.In order to make up for the shortcomings of drones and terrestrial 3D laser scanners,obtain richer data and better grasp the details,use digital cameras to enrich the data,give close-range photogrammetry and use digital cameras for operations that need to be paid attention to Some rules,and the use of mobile phone cameras to perform three-dimensional reconstruction of a stone relief.(3)Research on the theory and strategy of data fusion from the two levels of image data and point cloud data,and choose Zhang Zongheng Gate Building of our school during the Republic of China period as the experimental object to verify.Due to the incomplete data collection,when a single data source is modeled,there will be model holes,fraying,and distortion brought about by the missing data.Multiple data source fusion methods proposed in this article can completely obtain gatehouse data and establish a real scene compared with a single data model.The three-dimensional model is more complete and beautiful,and the texture details are well expressed.Finally,fusion data and model accuracy are verified at the point cloud and model level.(4)By designing two simulated deformation tests,it is verified that the three-dimensional point cloud data can be implemented to deformation monitoring and analysis and can reflect the integrity and unevenness of deformation.Taking an overpass inspection and appraisal project as an example,the application of 3D laser scanning technology in deflection and deformation inspection is verified.Use the fusion point cloud data to monitor and analyze the deformation of the teaching building,the drone tilt photography to collect the image and texture information of the top and middle and upper parts of the building,and the three-dimensional laser scanner to collect the middle and near-ground data of the building to fuse the two kinds of data.Processing is to obtain complete and comprehensive point cloud data of the building.By visualizing the comparison and analysis of the two-phase point cloud superposition,the deformation visualization cloud map and the overall deformation of the building are obtained.Finally,the accuracy and feasibility of this method are checked by comparison with traditional monitoring methods. |
PDF Full Text Request