A Concrete Surface Roughness Measurement System Based On Linear Structural Light

This thesis introduces a system for measuring the roughness of concrete surface using line structured light scanning technology and compares the traditional measurement method with the system measurement method.At the same time,the reliability and stability of the system measurement method are analyzed in depth,and the accuracy of the system measurement results is studied in detail.The measurement system takes Nvidia AGX Xavier as the core processor,and the overall structure,mechanical components and drive circuit of the system are designed and manufactured independently.After in-depth study of the influence of different control schemes among various components on the measurement results,a special control scheme is designed for each part of the system,which greatly improves the measurement accuracy of the system.After in-depth study of the principle of system parameter calibration,this thesis takes Zhang ’ s calibration method as the theoretical basis and uses Python to calibrate.In this thesis,the back projection error of camera calibration results is also analyzed and discussed in detail.In this thesis,the equation for solving the plane where the laser is located is also discussed in depth.Combined with the results of camera calibration,the transformation of spatial coordinates and the calculation of spatial points are carried out.Finally,the laser plane equation is solved by combining the data of multiple calibration images.This thesis discusses the principle of line structured light measurement in detail.When using line structured light for three-dimensional reconstruction,subtracting image background information and preprocessing image can effectively improve the measurement accuracy of the system.In this thesis,the results of the fringe center position extracted by extreme value method and Gaussian fitting method are analyzed and compared in detail.Finally,the Gaussian fitting method is used to solve the subpixel laser fringe center position.In order to accelerate the extraction process of laser stripe center,this thesis also introduces a cluster computing structure and parallel computing structure which can automatically network.Through this computing structure,the whole computing process can be accelerated by 5 to 10 times.Finally,the results and errors obtained by sand filling method and linear structured light measurement are analyzed and discussed in depth.In addition,the influence of integral interval and concrete material on the measurement accuracy is analyzed,and the authenticity and reliability of the system measurement results are demonstrated by combining the results of sand filling method.The average depth was 0.9057 mm when the sand filling method was used for three measurements,and 0.9145 mm when the integral interval was 0.01 mm when the line structured light scanning was used.Finally,the measurement accuracy of the system is about 60μm through statistical analysis of the data.