自动图像校正原理与方法及在火车导轨图像测量中的应用  

Principle and Method of Automatic Calibration in Digital Image Processing of Rail Wear

作　　者：吴福华[1] 余学才[1] 吴柯庆[1] 张蕾[1] 

机构地区：[1]电子科技大学光电信息学院,四川成都610054

出　　处：《计量学报》2013年第1期7-10,共4页Acta Metrologica Sinica

摘　　要：提出了一种操作简便的隐参数校正方法。通过校正板上特征点的世界坐标和对应的畸变图像中的图像坐标求出两坐标系之间的转换矩阵元素——隐参数，再由隐参数将被测图像还原到世界坐标系中，得到相当于摄像机在与测量平面垂直并且镜头不存在任何畸变的情况下拍摄得到的图像，并进行图像测量。将此方法应用于火车导轨磨耗图像测量中，现场测试的结果表明，图像校正精度可达0．2mm，测量系统精度达到0．5mm，符合标准要求。A simple implicit parameter calibration method is proposed. The transition matrix element which is also called implicit parameter is obtained from the relation between the calibration board feature points world coordinates and its corresponding distorted image coordinates. Then the measured image is restored to the world coordinate system for measuring which can be seen as the image obtained from the case that the camera perpendiculars to the measured plane and under the case of no distortion in the lens. This method is used in the rail wear image measurement. After field test, shows that the image calibration accuracy reaches 0. 2mm and accuracy of the measurement system is better than 0. 5mm which meets the requirements of the standard.

关 键 词：计量学 钢轨磨耗 图像传感器校正 图像测量 隐参数定标 

分 类 号：TB92[一般工业技术—计量学]