Research On The Camera Calibration Based On Planar Template And The Related Technology

People understand the world and get the effective information from the outside world through the visual sense.With the continuous development of modern technology and computer vision,it is very common to identify the objects around the world by external equipment.Camera calibration is the premise and basis of computer vision to obtain three-dimensional information,and the calibration results directly affect the accuracy of 3D measurement and the results of 3D reconstruction.Therefore,the research in this field is very hot in recent years.This paper mainly conducts the exploration and Research on related technology of camera calibration method and related parameters,so as to obtain the camera,build a 3D and 2D image plane in the corresponding relationship,provide strong theoretical support for the next step of research in computer vision.In this paper,we focus on the research of camera calibration and the extraction of feature points,and the contents of the research and the results are summarized as follows:Firstly,the extraction of image feature points is studied.In the corner extraction process,Aiming at the principle and deficiency of the classical Harris corner detection algorithm,This article has carried on the comprehensive elaboration to it,Shi-Tomasi corner extraction algorithm is the improvement of Harris corner extraction algorithm,The minimum eigenvalue constraint is used to obtain the strong angle.In this paper,the principle and steps of sub-pixel corner detection are described in detail,and the stability of the algorithm and the length of the algorithm are taken into account.After comprehensive consideration,Shi-Tomasi corner detection operator is used as the initial value of the angle,and then combined with the two polynomial approximation to optimize the corner coordinates,so that the corner has good anti noise performance and robustness.Secondly,The camera calibration method and related theory are studied.This paper presents a simple and fast calibration precision of two plane template calibration method,the calibration method by step and the establishment of a contained radial and tangential distortion model,The method of solving the camera internal and external parameters and the complex distortion model coefficients by using the planar checkerboard calibration object is realized.In this paper,the LM optimization algorithm is used to optimize the refinement of linear initial value,most planar calibration method only considers the main cause of radial distortion based on the distortion,finally joined the radial and tangential distortion in the linear model,the coefficient of the linear transform to solve the distortion and the focal length of camera calibration.Then the LM algorithm is used to optimize all the results.Thirdly,By calculating the absolute error and relative error between the image point and the world coordinate.In addition,the absolute error and relative error of the two cameras before and after calibration are analyzed and evaluated in two aspects.Finally,by comparing the results of two sets of experimental data and calibration error,the accuracy of the calibration algorithm is obtained,the results are satisfactory...