Study On Algorithms For Calibration And Inclusion Testing In 3D Information Obtaining Technology

With the development of science and technology, people need to obtain the 3D surface information of body quickly and accurately in many fields, to reconstruction 3D model of body, it need to get 3D surface data of body first.3D scanning system is the main way to collect surface data of body. This system related to mechanism,optics,electronics,control,computer vision,image processing,computer graphics,software engineering. An overview for the current status, technique points and future development trend of the 3D information technology in the world is introduced.Firstly, study on the camera calibration. The calibration of camera lens plays an important role in the field of computer vision, its relevant parameters can be classified into intrinsic and extrinsic ones. The intrinsic parameters are the lens center coordinates, focus, and distortion error, the extrinsic parameters are the position and orientation of camera lens in space. These parameters should be calibrated by experiments in most cases, the precision of 3D object reconstruction is mostly relying on camera lens calibration.In traditional photogrammetry camera lens calibration, a calibration 3D physical model block with many control points is needed, the coordination of these points are measured high precision of three dimensional coordinates. These points should be evenly distributed in a three dimensional space for calibration precision, and it is difficult to make the model block especially big model. In this paper, a calibration plate was made and the plate was moved in its normal direction, parallel, no rotation, to instead of a calibration model block. it is implemented to provide large area 3D control points with variable Z values, the moving plate is a 3D model, moreover, it could increase control points significant, the plate is much easier to make than traditional 3D physical model.Combine the traditional photogrammetry calibration with neural network to overcome instability and computation complexity due to nonlinearity factors such as aberration. calibration parameters for computer vision are obtained based on neural network, and the unknown point world coordinates calculated with a five-step method. Experimental results show that the proposed neural network based method with moving calibration plate can provide high accuracy and stability.Secondly, study on calibration parameters correction and off-centering correction. Calibration parameters may change after some times, so these parameters must be correction. For scanning system, off-centering problem exist always, so it need correction.Thirdly, study on inclusin testing of point-line-polygon, algorithm for checking point in polygon is present, Vertor product algorithm:For testing the relationship of a point and a polygon, the direction of the polygon was defined, vector product of the test point and polygon was defined, discriminant function was defined according to the vector product. Calculate the vector product and discriminant function. The decision of the point in polygon testing was based on the discriminant function. Experiment result indicates that this method was simple, robust and fast.The relationship test of two polygons is algorithm for polygon in polygon. every peak of two polygons was arranged a number in same direction, crossing points of two polygon were calculated, The edges were divided into some segments by crossing points, every segments located in another polygon were found out, segment that can link with each other was found out, it is done continuously till these segments closed, it is a new polygon, the common part of this two polygons.