Design of Calculation Algorithm for Edge Straight Line Accuracy of Metal Plate Based on Point Cloud Data

Metal plate is a common raw material for building structures. If the edge line accuracy of the plate is low, it will lead to problems such as reduced structural assembly accuracy and mechanical properties. At the same time, the traditional detection methods are mostly manual detection, and the level of automation is low. In this paper, the surface of the metal plate is scanned by a 3D line laser sensor driven by a six-degree-of-freedom manipulator to obtain the surface point cloud data of the measured metal plate. Through the design of a point cloud preprocessing algorithm, it can identify and filter out a variety of outliers in the point cloud data. In the face of different types of outliers, different identification strategies can be adopted. After the preprocessing operation is completed, the edge data is identified and recorded by Sobel algorithm, and the recorded data points are used as fitting points. Based on the spatial least squares method, the least squares midline is fitted, and the straight line accuracy of the metal plate edge line is finally calculated. The correctness of the straight line accuracy calculation algorithm is verified by measuring the linear accuracy of the linear guide. The experimental part calculates the linear accuracy of the edge line of the aluminum plate. The magnitude of the calculation result is 0.1mm.Considering that the installation error during the measurement cannot be eliminated, the calculation result is within the acceptable range.