Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2021, Vol. 47 ›› Issue (7): 289-295,300. doi: 10.19678/j.issn.1000-3428.0058426

• Development Research and Engineering Application • Previous Articles     Next Articles

Meter Pointer Location Algorithm Based on MSER-Otsu and Line Correction

QIN Xuan1, FENG Lei1, LIANG Qinghua1, ZHANG Wei2   

  1. 1. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;
    2. Xuzhou Power Supply Company, State Grid Jiangsu Electric Power Co., Ltd., Xuzhou, Jiangsu 221000, China
  • Received:2020-05-26 Revised:2020-06-30 Published:2020-07-01

基于MSER-Otsu与直线矫正的仪表指针定位算法

秦轩1, 冯磊1, 梁庆华1, 张伟2   

  1. 1. 上海交通大学 机械与动力工程学院, 上海 200240;
    2. 国家电网江苏省电力有限公司 徐州供电公司, 江苏 徐州 221000
  • 作者简介:秦轩(1996-),男,硕士研究生,主研方向为机器视觉、机器人;冯磊,硕士研究生;梁庆华,副教授、博士;张伟,高级工程师。
  • 基金资助:
    国家电网有限公司总部科技项目(5700-202018266A)。

Abstract: When performing on-site inspection tasks,the unattended robots in power distribution stations need to read the meters.To meet the requirement,this paper proposes an algorithm for meter pointer location based on MSER-Otsu threshold segmentation and Hough transform line correction.The pointer region is roughly extracted from the dial image by using the Maximally Stable Extremal Region(MSER) algorithm,and then the iterative Otsu's algorithm is used for its threshold segmentation.The pixels of the pointer,shadow and background are separated,and only the binary image of the pointer is retained.Hough transform is performed on the binary image to pre-locate the center line of the pointer,and subsequently Huber loss is used as the optimization function to correct the line,completing the location of the center line of the pointer.Experimental results show that the proposed algorithm can accurately locate the meter pointer and calculate the pointer deflection angle in complex environment.Compared with the Hough transform algorithm and the Huber loss fitting algorithm,it has higher processing speed and robustness,and can meet the actual needs of inspection robots in power distribution stations.

Key words: square pointer meter, Maximally Stable Extremal Region(MSER), iterative Otsu's method, Hough transform, pointer location

摘要: 针对配电站无人值守作业机器人执行巡检任务时对指针式仪表的识读需求,提出一种结合MSER-Otsu阈值分割与Hough变换直线矫正的仪表指针定位算法。利用最大稳定极值区域算法从表盘图像中粗提取出指针区域,引入迭代Otsu算法进一步对该区域进行阈值分割,将指针、阴影和背景区域像素点分离后仅保留指针的二值化图像,对其采用Hough变换进行指针中心线预定位,再以Huber loss为最优化函数对预定位直线进行矫正,最终完成指针中心线的定位。实验结果表明,该算法能在复杂环境中对仪表指针进行精确定位并计算指针偏转角度,与Hough变换算法和Huber loss拟合算法等相比,其处理速度更快且鲁棒性更高,能满足配电站巡检机器人的实际需求。

关键词: 方形指针表, 最大稳定极值区域, 迭代大律法, Hough变换, 指针定位

CLC Number: