The Fusion And Supplement Of Three-Dimensional Color Point Cloud Data

Three-dimensional ranging technology has been widely applied to get the coordinates of three-dimensional objects and color information In recent years,with the development of high speed and high resolution electronic sensors,three-dimensional ranging technology research and commercial application made great progress,mainly applied in reverse engineering,quality detection,cultural relics protection,machinery manufacturing and other categories.Data fusion and supplement has important significance in three-dimensional ranging technology.Three-dimensional color point cloud data fusion is the key to three-dimensional ranging technology.Through multi-sensor data fusion,we will be unified target position information collected by laser and image information collected by camera under the same coordinate system,in order to generate three-dimensional color point cloud data,realize the fusion of color and position information and the reconstruction of real scenes.Data fusion is the process of camera calibration,which refers to finding the corresponding relation by the space coordinates and the image pixel coordinates of the target feature points,according to the mathematical model of the imaging system.We can obtain the geometric mapping between the laser range finder and the camera,which is determined by the internal parameters of the camera and the external calibration of the laser rangefinder and camera.In view of the existing laser rangefinder type,two calibration methods are put forward.For 2D laser range finder,this paper proposes a calibration method based on boundary constraint and mean approximation,calibrates data by line fitting,the dual strategy of boundary constraint and average approximation is used to calculate the center of the hole,and to establish a data connection between laser and camera measurement.For 3D laser range finder,this paper proposes a calibration method based on geometric constraints,combines the constraints of point to plane and line to plane,to establish data correlation between laser measurement and camera measurement.In three-dimensional ranging technology,multiple modal data fusion is the first step,in order to achieve interactive application between spatial coordinates and color of the target to lay a solid foundation,we can complete the depth of the target information and image information fusion and collect three-dimensional color point cloud data.But in the threedimensional color point cloud data in the process of acquisition,because hardware resolution limit and environmental barriers,inevitably existing data deficiency,the resolution of the image resolution is far higher than that of laser spot.In order to make full use of the extra pixel information in memory,data interpolation algorithm is put forward,according to the number of pixels to realize line interpolation of point cloud,interpolation by line method is adopted on the same frame data to realize real-time supplement,and analyze the interpolation of threedimensional color point cloud data.This method can improve not only the utilization of image information and the visualization effect of 3D reconstruction,but also the real-time performance of the missing three-dimensional color point cloud.