基于OpenCV的数字仪表字符提取及快速识别研究  被引量:4

Research on Digital Instrument Character Extraction and Fast Recognition Based on OpenCV

在线阅读下载全文

作  者:张银 吕彦明 钱云杰 ZHANG Yin;LYU Yanming;QIAN Yunjie(School of Mechanical Engineering,Jiangnan University,Wuxi 214122,China)

机构地区:[1]江南大学机械工程学院,江苏无锡214122

出  处:《自动化仪表》2023年第5期46-50,55,共6页Process Automation Instrumentation

摘  要:为了实现工业仪表的自动化识别检验、降低人力消耗,应用机器视觉和图像处理技术对数字仪表自动读数识别进行研究。对采集的字符图像进行一系列预处理,包括灰度化以及采用双边滤波算法对含有背景的图像进行图像降噪预处理。提出了1种针对数字仪表字符区提取及快速识别的算法。基于透视变换以及轮廓提取函数实现仪表字符区的倾斜校正及快速提取,并对提取的字符进行形态学处理。提出了1种改进的穿线法对未进行字符分割处理的图片进行识别,并针对字符1的特征提出了快速识别方法。该算法省去了字符分割的时间,提高了字符识别效率。试验证明,算法对数码管的识别正确率达到99.3%、单幅图片识别时间为6.95 ms,可有效应用于计量、检验等数显仪表识别场合,提高检验效率。In order to realize the automatic identification and inspection of industrial meters and reduce human consumption,machine vision and image processing techniques are applied to investigate the automatic reading recognition of digital meters.A series of preprocessing is performed on the acquired character images,including grayscale and image noise reduction preprocessing for images containing backgrounds using bilateral filtering algorithms.An algorithm is proposed for the extraction and fast recognition of character areas of digital meters.Based on perspective transformation and contour extraction function to achieve tilt correction and fast extraction of instrument character areas,and morphological processing of the extracted characters.An improved threading method is proposed to recognize the images without character segmentation processing,and a fast recognition method is proposed for the characteristics of character 1.The algorithm eliminates the time of character segmentation and improves the efficiency of character recognition.The test proves that the algorithm achieves 99.3%correct recognition rate for digital tube and 6.95 ms recognition time for a single picture,which can be effectively used in the recognition of digital display instruments such as metrology inspection to improve the inspection efficiency.

关 键 词:图像处理 机器视觉 数字仪表 穿线法 自动识别 OPENCV 透视变换 轮廓提取 

分 类 号:TH-701[机械工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象