学术论文

      基于ORB和改进Hough变换的指针仪表智能识读方法

      Intelligent Recognition Method for Pointer Meters Based on ORB and Improved Hough Transform

      摘要:
      为了满足变电站巡检机器人对电站指针仪表的智能识读要求,提出一种基于ORB和改进Hough变换的指针仪表识读方法,描述了该算法中仪表区域识别、二值化处理、骨架提取和读数获取等环节,利用C++在计算机上进行算法实现、验证.结果表明该算法满足读数精度要求、具有较好的实时性,拥有较高的可靠性和应用价值.
      Abstract:
      In order to meet the requirement of intelligent recognition of pointer meters ,which was adapted to the equipment inspection robot ,a recognition method for pointer meters based on ORB and improved Hough transform was proposed .The major steps of the algorithm including detecting of meter device region ,image binarization,acquirement of image skeleton and reading recognition were described .The algorithm was compiled in C++and verified on PC.The results show that the algorithm is real-time and reliable .The algorithm meets the requirement of accuracy and has higher practicability .
      作者: 朱柏林 [1] 郭亮 [2] 吴清文 [2]
      Author: ZHU Bo-lin [1] GUO Liang [2] WU Qing-wen [2]
      作者单位: 中国科学院长春光学精密机械与物理研究所,吉林长春 130033;中国科学院大学,北京 100049 中国科学院长春光学精密机械与物理研究所,吉林长春,130033
      刊 名: 仪表技术与传感器 ISTICPKU
      年,卷(期): 2017, (1)
      分类号: TP242.6
      在线出版日期: 2017年3月31日
      基金项目: 中国科学院青年创新促进会资助项目