首页|基于YOLOv5和数字图像处理的指针式压力表识别方法

基于YOLOv5和数字图像处理的指针式压力表识别方法

扫码查看
针对指针式压力表自动检测读数时,受复杂背景环境等因素影响造成仪表图像检测难、读数困难和误差大等问题,提出基于YOLOv5和数字图像处理技术相混合的指针式压力表读数识别方法,通过YOLOv5网络实现复杂背景环境下表盘的检测与提取,再通过数字图像处理技术实现表盘圆心和指针直线的高精度定位,最后通过CRNN网络确定仪表量程.实验结果表明,该方法可以在复杂的环境中较为高效准确地读取指针式压力表的示数,读数的平均基本误差为0.32%,较现有方法显著提高,证明了该方法具有较好的实际应用价值.
Identification Method of Pointer Pressure Gauge Based on YOLOv5 and Digital Image Processing
In view of the difficulties in meter image detection,reading difficulties and large errors caused by complex background environment and other factors during automatic reading detection of pointer pres-sure gauge,a reading recognition method of pointer pressure gauge based on the mixture of YOLOv5 and digital image processing technology was proposed.The detection and extraction of dial under complex back-ground environment was realized through YOLOv5 network.Then the digital image processing technology is used to realize the high-precision positioning of the dial circle center and the pointer line,and finally the measuring range is determined by CRNN network.The experimental results show that this method can read the indicator of pointer pressure gauge efficiently and accurately in complex environment,and the average basic error of reading is 0.32%,which is significantly higher than the existing method,and proves that this method has good practical application value.

pointer pressure gaugeYOLOv5digital image processingreading recognition

王明、闫泽恩、王思瑶、张贵斌、高腾、庞桂兵

展开 >

大连工业大学机械工程与自动化学院,大连 116034

空装驻西安某军代表室,西安 710065

指针式压力表 YOLOv5 数字图像处理 读数识别

2024

组合机床与自动化加工技术
大连组合机床研究所 中国机械工程学会生产工程分会

组合机床与自动化加工技术

CSTPCD北大核心
影响因子:0.671
ISSN:1001-2265
年,卷(期):2024.(12)