摘要
针对外螺纹零件关键参数测量效率低、工作强度大、精度低,且难以实现在线自动检测的问题,提出一种基于视觉的非接触外螺纹关键参数测量方法。首先,针对螺纹图像中不存在明显直线的情况,设计一种基于Hough变换的特征圆提取螺纹图像倾角校正算法,利用特征圆圆心构造螺纹虚拟直线以实现螺纹倾角计算并旋转图像;进一步,在螺纹的感兴趣区域(ROI)区域内进行一系列图像处理得到螺纹轮廓点,采用最小二乘算法拟合螺纹上、下轮廓点直线,实现螺纹的大径与小径高精度测量。现场测试表明所提方法能对螺纹图像倾斜进行有效校正,大径与小径尺寸测量精确达到0.0085 mm,验证提出方法可行性和有效性。
Aiming at the problems of low efficiency,high working intensity,low measurement accuracy and difficult to realize online automatic detection of the key parameters of external threaded,a vision-based measurement method for non-contact key parameters of external threaded was proposed.First,in the case that there was no obvious straight line in the thread image,a feature circle based on Hough transformation was designed to extract the angle correction algorithm of the thread image.Furthermore,a series of image processing was carried out in the ROI area of the thread to obtain the thread contour points.The least square algorithm was adopted to fit the upper and lower points of the thread to achieve the high-precision measurement of the large diameter and small diameter of the thread.Field tests show that the proposed method can effectively correct the tilt of thread image,and the measurement accuracy of large diameter and small diameter is up to 0.0085 mm,which verifies the feasibility and effectiveness of the proposed method.
作者
刘阳
刘超
LIU Yang;LIU Chao(College of Software,Chongqing Institute of Engineering,Chongqing 400056,China;Guizhou Aerospace Electronics Co.,Ltd.,Guiyang 550009,China)
出处
《组合机床与自动化加工技术》
北大核心
2021年第8期105-108,共4页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家重点研发专项(2020YFB1710500)。
关键词
螺纹
倾斜校正
霍夫变换
最小二乘法
视觉测量
screw thread
skew correction
Hough transformation
least square method
vision measurement
