基于人工网格控制点匹配的摄像机标定  

Camera Calibration Based on Control Point Matching of Artificial Grid

作　　者：董三锋 DONG San-feng(School of Mechanical and Electrical Engineering,Xi'an Traffic Engineering Institute,Xi'an 710300,Shaanxi)

机构地区：[1]西安交通工程学院机械与电气工程学院,陕西西安710300

出　　处：《西安交通工程学院学术研究》2022年第3期42-47,共6页Academic Research of Xi'an Traffic Engineering Institute

摘　　要：数字图像处理技术应用于测量领域,可以产生一些虚拟的测量工具,这其中必然需要对测量的核心部件摄像机进行标定。标定方法可以分为传统标定方法和自标定方法两大类,传统标定方法过程复杂不易计算,自标定方法鲁棒性不是很好,本文采用人工网络控制点匹配的摄像机标定,获取15对理想值和实际值坐标对,通过线性和非线性两种模型反求畸变系数,获取两个模型的矫正后坐标对与实际值坐标对对比,实验与标定结果表明发现三次多项式摄像机系统畸变模型性能可以获取0.8个像素以下的绝对误差精度。The application of digital image processing technology in the field of measurement can produce some virtual measurement tools,which must be calibrated to the camera,the core component of the measurement.The calibration methods can be divided into two categories:traditional calibration methods and self-calibration methods.The traditional calibration method is complicated and difficult to calculate,and the self-calibration method is not very robust.In this paper,the camera calibration with artificial network control points matching is used to obtain 15 pairs of ideal and actual coordinate values,and the distortion coefficients are obtained by linear and nonlinear models.The corrected coordinate pairs of the two models are obtained and compared with the actual value coordinate pairs.The experimental and calibration results show that the cubic polynomial camera system distortion model performance can obtain absolute error accuracy below 0.8 pixels.

关 键 词：数字图像处理技术 摄像机标定 人工网格 绝度误差 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]