3D Point Cloud Surface Reconstruction Based On Kinect Depth Data

Three-dimensional reconstruction is an important research area of computer machine vision,with the acceleration of computer science and technology to enhance the rapid development and testing technology,3D reconstruction have been widely used in computer animation,augmented reality,virtual reality,mixed reality,and real-time positioning map building,human-computer interaction of various engineering fields.Mainly refers to the three-dimensional reconstruction by computer graphics digital means,the objective world of three-dimensional solid weight of the computer now.The core issue is to obtain three-dimensional reconstruction of the different perspectives of the object scene information,and get the information obtained valid registration fusion,in computer graphics is to solve spatial coordinate transformation between multiple perspective.With the development of 3D laser scanning hardware and software technology equipment,surface reconstruction 3D point cloud data has become a hot topic of computer graphics.Point cloud data obtained by conventional laser scanning accuracy is high enough though,but a laser scanner expensive,this paper,Microsoft launched depth camera somatosensory equipment Kinect,the use of open-source PCL(Point Cloud Library)point cloud library on the computer simply and efficiently to achieve a three-dimensional surface reconstruction of the object point cloud.In summary,this paper Kinect depth camera to obtain a three-dimensional point cloud input device,the point cloud point cloud data relates to three-dimensional surface reconstruction of the pipeline acquisition,pre-processing point cloud point cloud registration,and point cloud surface triangulation techniques research and related theoretical analysis,and ultimately the three-dimensional surface based on Kinect depth camera reconstitution experiments somatosensory processing platform.With the development of 3D laser scanning hardware and software technology equipment,surface reconstruction 3D point cloud data has become a hot topic of computer graphics.Point cloud data obtained by conventional laser scanning accuracy is high enough though,but a laser scanner expensive,this paper,Microsoft launched depth camera somatosensory equipment Kinect,the use of open-source PCL point cloud library on the computer simply and ultimately a three-dimensional surface reconstruction software platform based on the experimental treatment somatosensory Kinect depth camera.The main work of this paper is as follows:First Introduces the research background of the 3D point cloud surface reconstruction technology and related surface technology and research status,structure and depth of the sensor Kinect hardware,Kinect depth sensor three native data that RGB color data,depth data,skeletal data Get principle summarized.Secondly the PCL point cloud library,including the overall organizational structure and associated processing algorithms module PCL point cloud library,access to information within the depth of the scene the camera body through Kinect,through coordinate transformation to convert the information into a depth consistent with the actual three-dimensional model spatial point clouds.Then he made a general description of the current three-dimensional surface reconstruction.Finally,with PCL final point cloud library to discuss the relevant point cloud data processing,a three-dimensional point cloud registration,and surface reconstruction.GridProjection projection mesh surface reconstruction,fast triangulation,three methods of reconstruction Poisson surface reconstruction performance results were analyzed.Experiments show that,through the surface point cloud library PCL reconstruction results in line with expectations,This paper set forth the method enables users to use Kinect and personal computers in the real world objects digital display.In this paper,the development of software systems user-friendly,easy to operate,although Kinect obtain data accuracy is not high,the details of the point cloud library PCL surface reconstruction is not perfect,but not the precise application requirements to meet the basic needs.