基于机器视觉的燃气管道内环境感知技术研究

发布时间：2018-07-02 08:54

本文选题：机器视觉 + 摄像机标定　；参考：《昆明理工大学》2017年硕士论文

【摘要】：城市燃气管道埋设于地下,由于老化、腐蚀等原因将引起物质泄漏事故,容易造成严重的人员伤亡和经济损失。因此城市地下燃气管道需要定期检查,对受损的区域进行维修。城市地下燃气管道具有数量庞大、结构复杂、管径大小各异和管道外部空间狭小等特点,仅靠人力将无法完成地下燃气管道故障检测。用机器人代替人进入管道实时在线检测管道故障与缺陷是管道检测技术发展的必然趋势。机器人在管道内必须实时在线检测管道内环境参数,为机器人运动导航提供可靠信息。本文结合基金课题,采用辅助激光加单目视觉检测方法,对管道内环境感知进行方法研究。基于机器视觉的燃气管道内环境的感知技术是利用摄像测量学原理、辅助激光加单目视觉技术和图像采集与图像处理技术实现管道内部物理参数测定。首先在现有管道内环境视觉检测技术基础上结合管道机器人工作平台,从理论上提出一种新型的基于辅助激光成像的三维视觉测量方法,并且给出辅助激光加单目视觉三维检测方法的数学原理。其次,将摄像机成像原理与光学三角形理论相结合,推导出基于辅助激光成像的三维视觉测量系统的成像数学模型,实现从被测点图像信息到空间三维坐标的转换。分析摄像机线性参数与像差系数标定方法,并建立三维视觉测量系统结构参数标定模型。然后根据理论模型设计测量系统和参数标定装置,将参数标定算法与迭代运算结合计算出摄像机参数和测量系统的结构参数。最后利用测量系统采集空间被测点图像、计算被测点三维坐标,将系统测量坐标与实测真实坐标相对比验证测量方法的可行性。验证实验后,针对测量系统仍然存在的结构误差提出改进后数学模型。
[Abstract]:Urban gas pipeline buried underground will cause material leakage accidents due to aging and corrosion, which will easily cause serious casualties and economic losses. Therefore, urban underground gas pipelines need regular inspection to repair damaged areas. Urban underground gas pipeline has the characteristics of large quantity, complex structure, different diameter and narrow outer space, etc. It will not be possible to complete the fault detection of underground gas pipeline by manpower alone. It is an inevitable trend of pipeline detection technology to use robot instead of human to detect pipeline faults and defects in real time. In order to provide reliable information for robot navigation, it is necessary for the robot to detect the environmental parameters in the pipeline in real time. In this paper, with the aid of laser and monocular vision detection method, the environmental perception in pipeline is studied. The sensing technology of gas pipeline environment based on machine vision is to use the principle of camera measurement to realize the measurement of internal physical parameters of pipeline by using laser plus monocular vision technology and image acquisition and image processing technology. Firstly, a new 3D vision measurement method based on auxiliary laser imaging is proposed based on the existing vision-detecting technology of pipeline environment and the working platform of pipeline robot. The mathematical principle of the three-dimensional detection method of laser plus monocular vision is given. Secondly, combining the principle of camera imaging with the theory of optical triangle, the mathematical model of 3D vision measurement system based on auxiliary laser imaging is derived, which can transform the image information from the measured point to the three-dimensional coordinate of space. The calibration method of camera linear parameter and aberration coefficient is analyzed, and the calibration model of structural parameters of 3D vision measurement system is established. Then the measurement system and the parameter calibration device are designed according to the theoretical model, and the camera parameters and the structural parameters of the measurement system are calculated by combining the parameter calibration algorithm with iterative operation. Finally, the measurement system is used to collect the images of the measured points in space, calculate the 3D coordinates of the measured points, and compare the measured coordinates with the real ones to verify the feasibility of the method. After the verification experiment, the improved mathematical model is proposed for the structural errors that still exist in the measurement system.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU996.7;TP391.41

【相似文献】