Research On Indoor 3D Measurement Technology Based On 2D Single Line Lidar

In recent years,Automatic three-dimensional measurement of indoor geometric parameters is a potential demand for the rapid development of China’s real estate industry.It is an inevitable trend to transition from transform contact measurement to non-contact three-dimensional measurement represented by optical measurement.The problem of high randomness and low degree of automation in manual measurement can be solved by building a low-cost indoor three-dimensional measurement platform with single-line lida.In this paper,the basic indoor measurement platform with two-dimensional single-line lidar as the core to realizes the automatic acquisition of indoor three-dimensional point cloud data,and measures the geometric parameters of indoor doors,windows and walls The main work is as follows.Firstly,the basic platform of indoor measurement system with two-dimensional lidar as the core was built,and the software driver of the basic platform was written,The influence of measurement distance and measurement angle on measurement error was analyzed.Based on the requirement analysis of indoor three-dimensional measurement,the hardware platform design of indoor scanning system was completed.Under the requirements of the system scheme,the overall construction of the single-line lidar indoor scanning system was carried out,and the communication of the lidar,the control of the stepper motor,the realization of the three-dimensional scanning of the single-line lidar were completed.At last,the software design of the scanning system was completed.Through the error analysis of the measurement precision of the basic platform,the validity of the measurement precision of the indoor three-dimensional measurement system was verified.Secondly,based on the single site cloud data obtained and preprocessed,the target image recognition and three-dimensional parameter measurement including Door&window and walls were carried out.The local feature point clouds of three-dimensional doors and windows were converted into two-dimensional images.The target detection algorithm based on neural network was used to recognize Door&window.Then,the two-dimensional image containing the recognition information of Door&window was transformed into the three-dimensional feature points of the point cloud of Door&window.In the end,the function of indoor three-dimensional point cloud to automatically detect the accurate location of local Door&window was realized.The measurement method of point cloud slice was adopted to calculate the specific three-dimensional information of local doors,windows and walls.Finally,through the analysis of the location of multiple sites to coordinate the transfer sites to achieve the point cloud Mosaic function and software verification,and calculate the indoor three-dimensional information.According to the principle of indoor scanning station distribution,Plan appropriate sites and use the basic indoor measurement platform to capture the cloud of each site.The transfer station measurement method is adopted to align the coordinate systems of each station.The indoor multi-station cloud Mosaic is completed and verified by the self-designed automatic Mosaic software.Then,the random sampling consistency algorithm is used to segment the walls in the point cloud,and the relationship between the surfaces is used to calculate the high-precision wall contour.Finally,the indoor three-dimensional measurement task is completed.The system verifies the feasibility of the relevant methods and techniques,and the indoor three-dimensional measurement results are accurate and reliable,which provides an effective reference for the establishment of indoor three-dimensional measurement system of single-line lidar.