基于区域预测和模板匹配的焊缝跟踪算法及系统研究

发布时间：2018-01-18 08:33

本文关键词：基于区域预测和模板匹配的焊缝跟踪算法及系统研究　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：管道运输是我国石油、天然气的重要运输方式,如何高效、高质量的完成焊接工作是保证运输安全的关键。本文根据运输管道焊缝图像特点设计了一套完整的焊缝自动跟踪焊接系统。本文主要工作如下:1)设计了一套焊缝自动跟踪焊接系统本文设计了一套焊缝自动跟踪焊接系统。本系统主要由系统控制模块,图像采集模块,偏差计算模块,运动控制模块,运动执行模块和焊接模块等六部分组成。摄像机采集的图像数据经图像采集卡传给工控机,工控机通过图像数据计算出焊缝偏差并把偏差传送到运动控制卡,运动控制卡把计算好的偏差数据转换为其对应的脉冲个数并通过导线把脉冲传送给电机控制器,电机控制器通过脉冲控制电机转动最后完成偏差调整。2)给出了一种基于区域预测和模板匹配的焊缝跟踪算法本文给出一种利用灰度特征和梯度方向直方图特征结合的模板匹配算法,本算法可以准确定位焊缝位置;本文给出一种匹配区域预测算法可以减少算法执行时间,显著提高算法的运行速度,实现实时跟踪;本文给出了一种焊缝模板更新策略可以提高系统鲁棒性。本算法可以准确找到焊缝位置并实现实时跟踪,对遮挡,目标外观变化有一定鲁棒性。3)焊缝自动跟踪系统实现本系统在Windows环境下采用C++语言设计了实时跟踪界面,算法调整界面,现场参数设置界面,安装参数设置界面。本系统利用数字量输入输出卡完成跟踪开始、停止,手动调整,系统关闭等功能。
[Abstract]:Pipeline transportation is an important transportation mode of petroleum and natural gas in China. High quality welding work is the key to ensure transportation safety. According to the characteristics of weld image of transportation pipeline, a complete welding automatic tracking system is designed in this paper. The main work of this paper is as follows: 1). A welding seam automatic tracking welding system is designed in this paper. This system is mainly controlled by the system module. The image acquisition module, deviation calculation module, motion control module, motion execution module and welding module are composed of six parts. The image data collected by the camera is transmitted to the industrial control computer through the image acquisition card. The industrial computer calculates the welding seam deviation through the image data and transmits the deviation to the motion control card. The motion control card converts the calculated deviation data to the corresponding pulse number and transmits the pulse to the motor controller through the wire. Motor controller through pulse control motor rotation final error adjustment. 2). This paper presents a seam tracking algorithm based on region prediction and template matching. This paper presents a template matching algorithm which combines gray feature and gradient direction histogram feature. This algorithm can locate the position of weld accurately. In this paper, a matching region prediction algorithm is presented, which can reduce the execution time of the algorithm, improve the running speed of the algorithm, and realize real-time tracking. In this paper, a weld template updating strategy is presented to improve the robustness of the system. The algorithm can accurately find the weld position and achieve real-time tracking and occlusion. Target appearance change has a certain robustness. 3) seam automatic tracking system. This system uses C language to design the real-time tracking interface under Windows environment, and the algorithm adjusts the interface. The system uses digital input and output card to complete tracking start, stop, manual adjustment, system shutdown and other functions.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TE973.3;TP391.41

【相似文献】