Computer Engineering and Applications ›› 2008, Vol. 44 ›› Issue (4): 242-244.

• 工程与应用 • Previous Articles     Next Articles

Research on vision-based localization system underwater of condenser tubes

CAI Yu-lian,WANG Yao-nan,ZHANG Zhi-guo   

  1. Key Laboratory of Intelligent Automation Technology,Hunan University,Changsha 410082,China
  • Received:2007-07-03 Revised:2007-08-14 Online:2008-02-01 Published:2008-02-01
  • Contact: CAI Yu-lian

冷凝管水下视觉定位系统研究

蔡玉连,王耀南,张志国   

  1. 湖南大学 智能自动化重点实验室,长沙 410082
  • 通讯作者: 蔡玉连

Abstract: According to the work environment and conditions of the robot,and aiming at the limitation of localization with coordinates input,machine vision technique is added to the paper,system feasibility is discussed,an embedded machine vision system based on DSP is designed,structure and principle of illumination mode,image gathering module and processing module are also introduced in detail.Related to the system application and algorithm real-time capability,the circle center of tube is localized using the maximum square error method and gravitational center method.Experimental results of the vision system have proved that the robot’s integration,speed and precision of positioning are increased to some extend.

Key words: condenser tubes, machine vision, DSP, image processing, localization

摘要: 根据机器人水下工作环境和工况要求,针对冷凝管清洗设备坐标输入法定位不准的问题引入了机器视觉技术,分析了系统的可行性设计,设计了一套基于DSP的嵌入式机器视觉系统,并详细叙述了照明方式、图像采集模块、图像处理模块的组成及工作原理,最后结合系统应用要求和算法的实时性,通过最大方差法及重心法完成了管口圆心的定位。该视觉系统取得良好的实验结果,能在一定程度上提高整个机器人系统的集成度、定位的快速性与精确度。

关键词: 冷凝管, 机器视觉, DSP, 图像处理, 定位