Research On Registration Of Rotating Body Fragments Point Cloud In Reverse Engineering

Reverse engineering technology is being widely used along with the development of science and technology.Whether it is the study of foreign advanced technology or independent research and development,reverse engineering technology is extremely important.Not only that,reverse engineering technology is also widely used in cultural relics repair,machinery.Visual and other fields.Although the application field is extensive,the principle is the same.The essence is to use the 3D scanner to collect the surface of the object to obtain the surface point cloud,and further process the obtained point cloud.This paper focuses on the feature extraction and registration of the rotating body debris point cloud.The following are mine main research :Firstly,the point cloud which has the rotating body feature,the radius of rotation is the key factor for the registration of the rotating body point cloud.The relative height of the two point clouds to be registered is determined by the radius of rotation.The calculation accuracy and calculation efficiency of the radius of rotation will directly affect the accuracy and efficiency of the entire registration process.When calculating the radius of rotation,firstly,the obtained three-dimensional point cloud is projected around the Z axis to obtain plane point cloud data on the YOZ plane,and then the main curve of the plane point cloud band is extracted,and the initial cubic B-spline curve is obtained according to the main curve.The initial cubic B-spline is optimized by the SDM algorithm to obtain the final cubic B-spline curve.Finally,the points are evenly taken on the B-spline along the Z-axis direction,and the distance from each point to the Z-axis is calculated as the rotating body debris point cloud.The radius of rotation.It is proved by experiments that the accuracy and efficiency of the registration feature for calculating the radius of rotation is high.In addition,if two rotating body point clouds can be registered,there must be two bus bars on both sides of the common fault line,and the angle between the two bus bars and the broken line is constant at a common plane perpendicular to the axis.Based onthis mathematical feature,this paper proposes a method for obtaining the feature of the central angle based on the cubic B-spline curve obtained.This method avoids the cumbersome calculation of the position of the two intermediate busbars.Starting at a certain point on the B-spline curve,the angle between the point and the farthest point of the point cloud is obtained in the positive and negative directions of the Z-axis.The angle is used to determine whether the point cloud can be registered.The central angle of the circle is then registered using the mathematical feature of the sum of the central angles.The experimental results show that the proposed method has significantly improved the efficiency of calculating the central angle based on the central bus,and effectively improves the accuracy and efficiency of the registration process.Secondly,the point cloud which has no rotation features,this paper adopts a simplified idea.First,the fragment point cloud data is fitted to the least square plane,and then the 3D point cloud data is projected onto the fitted plane.Convert the 3D point cloud data into planar point cloud data,then use the rolling ball algorithm to extract the boundary point cloud boundary points,and fit these points with the cubic B-spline curve,and then further extract the points with obvious features on the curve.The corresponding curvature values ??are calculated and calculated.These curvature values ??form a curvature string.Finally,the curvature string is similarly matched by an algorithm to confirm whether the point cloud to be matched can be registered.Finally,using Visual Studio 2013,PCL point cloud library and Open GL tools,combined with the feature extraction and registration method of the rotating body debris point cloud proposed in this paper,the registration process is visualized to realize the matching of the rotating body debris point cloud quasi.