摘要
为了高精度测量家具板件孔位,提出一种家具板件圆孔位置的机器视觉在线检测方法,为圆形孔位在线测量技术的进步提供支撑。图像经过像素值和距离相关滤波函数去噪处理、自适应二值化处理后,通过本文提出的二次孔定位算法进行处理,通过微调孔位逼近真实圆孔,解决常见畸变孔的位置提取。采用本文算法的试验结果表明,1022块家具板件上4056个圆孔位置检测5次的平均值为100.05 mm,重复标准差为0.041 mm;在±0.5 mm的误差范围内,孔心X坐标、Y坐标的人机测量相符度分别为99.8%、92.9%。
In order to measure the hole position in the furniture panel with the high precision,a machine vision onset detection method was proposed in the paper.After the image denoising which is done by the pixel value filter and distance correlation filter it was adaptive binarized.Then the image is processed by the secondary hole localization algorithm.By fine-tuning the hole position close to the real round hole,it solve the position extraction of common distorted holes and provide support for the progress of circular hole position onset measurement technology.The experiment results showed that using the algorithm in this paper,the average value of 4056 circular holes in 1022 furniture panels is 100.05 mm and the standard deviation is 0.041 mm with 5 repetitions.Within the error range of±0.5 mm,the coincidence degree of the ergonomic measurement of the X-coordinate of the hole center is 99.8%,and the coincidence degree of the Y-coordinate is 92.9%.
作者
邓斌攸
池志强
潘云峰
苏真伟
DENG Bin-you;CHI Zhi-qiang;PAN Yun-feng;SU Zhen-wei(Research Center of Machine Vision and Intelligent Detection,Guangdong Polytechnic College,Zhaoqing 526100,Guangdong,China;School of Manufacturing Science and Engineering,Sichuan University,Chengdu 610065,Sichuan,China)
出处
《木材科学与技术》
北大核心
2022年第2期60-64,共5页
Chinese Journal of Wood Science and Technology
基金
广东省普通高校特色创新类项目“基于人工智能技术的板材表面缺陷检测系统研究”(2018KTSCX274)。
关键词
家具板件
孔位
在线检测
机器视觉
furniture panel
hole position
online measurement
machine vision
