基于ROS的相机标定技术分析

An Technical Analysis of Camera Calibration Based on ROS

在机器视觉中使用相机获取物体的图像信息时,由于图像信息与实际物体存在误差,导致机器视觉系统不能准确还原出物体的三维信息,进而导致机器人目标检测、路径规划的失败。通过建立图像信息和实际物体间的参数模型,并使用相机标定技术确定这些参数,可以使机器视觉系统输入准确的结果。本文介绍了相机的成像模型,分析了不同相机标定方法的优缺点,采用张氏标定法,在ROS平台下对相机进行了标定,获得了相机的内外参数。本方法简单、高效,对硬件要求低,便于在工业机器人系统中实施,提高了机器视觉系统的精度,具有重要的研究意义。

When obtaining image through camera based on machine vision technology,the 3D information of an actual object cannot be reconstructed accurately.This often leads to the failure to implement target detection and path planning of a robot.The inputs of the machine vision system tend to be accurate by establishing a parameter model between image information and actual object,which can be determined by using camera calibration technology.This paper introduces the camera imaging model and analyzes the demerits of different camera calibration methods.Using Zhang's calibration method,by calibrating the camera based on ROS platform,the external and internal parameters of the camera can be obtained.This method,which has important significance to improve the accuracy of an machine vision system,is simple,efficient and low hardware requirement.