作者：王蔚¹, 王晓凯¹, 刘慧锋²

作者单位：1. 山西大学物理电子工程学院，山西 太原 030006;
2. 山西迪迈沃科光电工业有限公司，山西 太原 030006

关键词：光学玻璃;线结构光;Hough变换;最小二乘法

摘要：

针对目前国内光学玻璃切割工艺造成的料损及质量离散度较大的问题，以光学玻璃定重切割工艺为研究基础，提出一种基于光切法的质量实时累加式视觉定重方法。根据光切法原理建立定重测量系统，并由3个线型激光器及2台CCD相机测量光学玻璃条料左、右部分的横截面轮廓，再利用Hough变换和最小二乘法提取高精度的激光轮廓中心线直线方程，从而根据直线相交得到轮廓特征点，即可知当前的横截面面积，结合光学玻璃条料的进给量求取每一帧截面之间的体积，根据光学玻璃经验环境下的密度累计得出质量。在智能定重切割实验中，光学玻璃块的质量相对标准差在1%内，实际切割耗损小于0.1 g，表明该装置能高效、精确地进行自动计重切割。

Intelligent cutting system for limited weight of optical glass based on light sectioning method
WANG Wei¹, WANG Xiaokai¹, LIU Huifeng²
1. College of Physics and Electronic Engineering, Shanxi University, Taiyuan 030006, China;
2. Shanxi Teamwork Photoelectric Industries Co., Ltd., Taiyuan 030006, China
Abstract: In order to solve the problem of material loss and weight dispersion caused by optical glass cutting process in China, a visual method of limited weight was presented by real-time accumulated weight based on light sectioning method in the cutting technology. The limited weight measurement system was established based on light sectioning method. Firstly, the cross-section profile of the left and right half of the optical glass strip were measured by three linear lasers and two CCD cameras. Secondly, the high-precision centre line linear equation was extracted combining Hough transform with least square method, and the cross section area can be known by obtaining the contour feature points based on the intersection of straight lines. Thirdly, the volume between each frame section can be calculated by cross section area and the feed rate of optical glass strip. Finally, the limited weight was obtained by cumulative volume and density under the experience environment. In the intelligent limited cutting experiment, the  relative standard deviation of mass is within 1%, and the actual cutting loss is less than 0.1 g, which indicates that the system can complete intelligent cut for limited weight efficiently and accurately.
Keywords: optical glass;line structured light;Hough transform;least square method
2020, 46(11):83-89  收稿日期: 2019-12-31;收到修改稿日期: 2020-02-04
基金项目: 山西省重点研发计划项目（201803D121102）
作者简介: 王蔚(1990-),男,山西文水县人,硕士研究生,专业方向为机器视觉
参考文献
[1] HERRERA-GRANADOS G, MORITA N, HIDAI H, et al. Experimental stress analysis of glass cutting using a non-rigid cutting mechanism[J]. Manufacturing Letters, 2015, 6: 14-18
[2] MATSUMURA T, ARISTIMUNO P, GANDARIAS E, et al. Cutting process in glass peripheral milling[J]. Journal of Materials Processing Technology, 2013, 213(9): 1523-1531
[3] BAO G, JIA M, XUN Y, et al. Cracked egg recognition base on machine vision[J]. Computers and Electronics in Agriculture, 2019, 158: 159-166
[4] GAO Y, QIAO T, ZHANG H, et al. A contactless measuring speed system of belt conveyor based on machine vision and machine learning[J]. Measurement, 2019, 139: 127-133
[5] IRERI D, BELAL E, OKINDA C, et al. A computer vision system for defect discrimination and grading in tomatoes using machine learning and image processing[J]. Artificial Intelligence in Agriculture, 2019, 2: 28-37
[6] JIAN C X, GAO J, AO Y. Automatic surface defect detection for mobile phone screen glass based on machine vision[J]. Applied Soft Computing, 2017, 52: 348-358
[7] 彭超. 透明玻璃平板表面缺陷在线检测系统的研究[D]. 杭州: 浙江工业大学, 2016.
[8] 杨尉. 结构光光条中心提取方法研究[D]. 哈尔滨: 哈尔滨工业大学, 2009.
[9] 梅峻华, 赖磊捷. 结构光测量系统光条中心提取算法研究[J]. 电子测量技术, 2018, 41(21): 8-13
[10] 陈至坤, 邸跃, 曾凯, 等. 基于线结构光的角钢截面几何参数三维测量[J]. 中国测试, 2019, 45(2): 128-133
[11] 王昊. 基于结构光投影的钢轨廓形动态测量方法研究及应用[D]. 北京: 中国铁道科学研究院, 2018.
[12] 曲岳. 基于FPGA的线结构光中心线提取关键技术的研究[D]. 长春: 中国科学院长春光学精密机械与物理研究所, 2017.
[13] 王世平. 线结构光测量技术在接触轨几何参数检测中的应用研究[D]. 成都: 西南交通大学, 2015.
[14] 周凯黎. 基于图像轮廓检测的带约束草图自动生成[D]. 杭州: 浙江工业大学, 2014.
[15] 赵洪亮. 基于Hough变换改进的直线检测与定位[J]. 激光与红外, 2009, 39(9): 1009-1011
[16] 韦飞云. 钢板尺寸视觉测量方法研究[D]. 沈阳: 沈阳理工大学, 2016.
[17] 汪博, 桂江生, 周建平, 等. 基于最小二乘法的早期玉米作物行检测研究[J]. 浙江理工大学学报(自然科学版), 2015, 33(4): 547-551