作者：沈良辰1,2, 郭钢祥2, 孔明1, 郭斌2
作者单位：1. 中国计量大学计量测试工程学院，浙江 杭州 310018;
2. 浙江省计量科学研究院 长度与精密测量研究所，浙江 杭州 310018
关键词：线纹提取;灰度值;grubbs准则;直线线性度;拟合算法
摘要：
针对目前基于图像处理的钢卷尺自动检定方法中存在线纹边缘定位不准确、直线拟合算法精度较差等问题，提出一种基于grubbs准则和直线线性度拟合的线纹提取算法。该算法能去除因边缘定位不准确而引入粗大误差的奇异点，同时还能拟合出线性度较高的直线。首先，分析线纹区域灰度值特征分布特点，定位线纹中心特征点；然后基于该线纹提取算法拟合中心线纹；最后结合定斜率直线拟合方法实现线纹提取。实验表明，该算法在保留目前自动检定方法优点的基础上，将单根刻线检定的示值误差值降低至0.002 mm内，显著提高检定精度。
algorithm for line pattern extraction based on grubbs criterion and linear linearity fitting
shen liangchen1,2, guo gangxiang2, kong ming1, guo bin2
1. college of metrology and measurement engineering, china jiliang university, hangzhou 310018, china;
2. institute of length and precision measurement, zjim, hangzhou 310018, china
abstract: aiming at the problems of inaccurate positioning of line edges and poor accuracy of line fitting algorithms in the current automatic verification methods for steel tape measures based on image processing, a line extraction algorithm based on grubbs criterion and line linearity fitting is proposed. the algorithm can remove singular points that introduce gross errors due to inaccurate edge positioning, and can also fit a straight line with higher linearity. first, analyze the characteristics of the gray value distribution of the line pattern area, and locate the central feature point of the line pattern. then fit the center line pattern based on the line pattern extraction algorithm. finally, combine the fixed slope straight line fitting method to realize the line pattern extraction. experiments show that the algorithm reduces the indication error value of a single engraved line verification to within 0.002 mm on the basis of retaining the advantages of the current automatic verification method, which significantly improves the verification accuracy.
keywords: line pattern extraction;gray value;grubbs criterion;linear linearity;fitting algorithm
2022, 48(8):46-52 收稿日期: 2021-04-07;收到修改稿日期: 2021-07-23
基金项目: 国家市场监督管理总局科技计划项目（2019mk028）；浙江省重点研发计划（2020c01028）；浙江省重点研发计划（2019c03109）
作者简介: 沈良辰（1995-），男，浙江湖州市人，硕士研究生，专业方向为精密仪器及大长度测量
参考文献
[1] 赵保亚, 魏彩乔. 基于边缘检测的钢卷尺图像二值化方法[j]. 现代制造工程, 2012, 4(12): 101-104
[2] 景加慧. 大尺寸钢卷尺目标线纹检定与实现方法[d]. 合肥: 合肥工业大学, 2019.
[3] 苟建松. 新型钢卷尺自动检定装置的研究[d]. 天津: 天津大学, 2016.
[4] menaka r, janarthanan s, deeba k. fpga implementation of low power and high speed image edge detection algorithm[j]. microprocessors and microsystems, 2020, 75: 103053
[5] 戴宗贤, 莫洪波, 周勇, 等. 基于灰度的亚像素插值视觉测量方法[j]. 中国测试, 2019, 45(5): 33-37
[6] 杨继森, 牟智铭, 李路建等. 时栅位移传感器中激励信号引入误差的分析与修正[j]. 中国测试, 2020, 46(8): 80-86
[7] nancy m, miriam f. a. two smart spectrophotometric methods for the simultaneous estimation of simvastatin and ezetimibe in combined dosage form[j]. spectrochimica acta part a: molecular and biomolecular spectroscopy, 2015, 137(5): 685-691
[8] zhou g z, yang f j, wang c z. lifting transform via savitsky-golay filter predictor and application of denoising[j]. journal of coal science & engineering(china), 2006, 4(2): 66-69
[9] 庄未, 张瑞欣, 苏晓, 等. 无人自行车lpv模型在线识别方法研究[j]. 中国测试, 2020, 46(8): 109-115
[10] 张颖, 高灵君. 基于格拉布斯准则和改进粒子滤波算法的水下传感网目标跟踪[j]. 电子与信息学报, 2019, 41(10): 2294-2301
[11] 郭良丹, 谭锐, 林宝森, 等. 提高供热机组的调峰灵活性研究[j]. 中国测试, 2022, 48(7): 16-22
[12] 钟凯强. 中厚壁管相贯线多层多道焊接路径规划与拟合算法研究[d]. 乌鲁木齐: 新疆大学, 2020.
[13] 钢卷尺检定规程: jjg 4—2015 [s]. 北京: 全国几何量长度计量技术委员会, 2015.
