Abstract:
Real time parameter measurement cannot be carried out to dynamic ring parts during automation ring rolling processes so that rolling process parameters cannot be adjusted in time. Considering effects of shaping of ring rolling parts, a visual measurement platform was set up and a machine vision-based non-contact real -time measurement method was put forward. This article improves the subpixel level edge extraction algorithm to extract edge data information of circular rolling pieces. Based on the characteristics of circular rolling pieces, an RG-Hough transform method is proposed to fit the detected edge data information. The conversion relationship between pixel and actual sizes were determined in combination with the camera calibration to gain parameters of ring rolling parts. Measurements of ring parts (OD: 462.12mm; and ID: 315.53mm) were applied to verify the effectiveness of our method. Our measurement error is ±0.25mm and our average speed can be up to 104ms/frame. Our study can provide powerful technical supports for intelligent control of ring rolling pieces.
摘要:
在自动化环轧过程中,无法对动态环件进行实时参数测量,无法及时调整轧制工艺参数。考虑到环件成形的影响,搭建了视觉测量平台,提出了一种基于机器视觉的非接触实时测量方法。本文改进了亚像素级边缘提取算法,以提取圆形滚动件的边缘数据信息。根据圆形轧制件的特点,提出了一种RG-Hough变换方法来拟合检测到的边缘数据信息。结合摄像机标定确定像素与实际尺寸之间的转换关系,以获得环件的参数。环形零件的测量值(OD:462.12mm;ID:315.53mm)用于验证我们方法的有效性。我们的测量误差为±0.25mm,平均速度可高达104ms/帧。本研究可为环件的智能控制提供有力的技术支持。