基于机器视觉的单组分染液浓度快速检测方法

doi:10.13475/j.fzxb.20200502207

摘要/Abstract

摘要：

针对印染行业的单组分染液浓度的检测需求,提出一种基于机器视觉的对单组分染液浓度进行快速检测的方法。该方法采用颜色亮度可调的背光光源和彩色工业相机搭建机器视觉检测平台,将质量浓度范围在0~0.4 g/L的单组分染液样品通过玻璃皿置于背光光源上进行图像采集和处理,获取染液颜色特征值(R、G、B值),拟合不同光源条件下的R、G、B值随染液浓度变化的曲线,以朗伯比尔定律为依据建立R、G、B值与染液浓度的关系模型,实现染液浓度的预测。结果表明,对于DRA-3R染料溶液,光源为蓝光(亮度等级为50)条件下建立的最小值模型预测精度最高,平均相对误差绝对值为3.35%,并且该方法检测速度快、成本低,为染液浓度在线检测的工业化应用提供了一定的研究基础。

关键词: 机器视觉, 彩色背光光源, 染液浓度检测, 染液, 颜色特征值

Abstract:

A method based on machine vision was proposed to detect the concentration of dye liquor with a single-component for the printing and dyeing industry.With this method, a machine vision detection platform was built with a color adjustable backlight source and an industrial color camera. The single-component dye liquor sample with a concentration range of 0-0.4 g/L was placed on a backlight light source through a glass dish for image acquisition and processing, in order to obtain the color characteristic values of the dye liquor, and to get the relationship between the RGB value and concentration of dye liquor under different light source conditions.Based on Lambert-Beer law, the relational model between the RGB value and dye liquor concentration was built to predict the concentration of dye in the dye liquor.It is verified by experimental fact that for the DRA-3R dye liquor, the minimum value model established under the condition of blue light intensity 50 has the highest prediction accuracy. The absolute value of the average relative error of the results is 3.35%. Experiments show that the method offers fast detection speed and low cost, which provides a certain research base for industrial application of on-line detection of dye liquor concentration.

Key words: machine vision, color backlight source, detection of dye liquor concentration, dye liquor, RGB value

中图分类号:

TS190.4

田宇航, 王绍宗, 张文昌, 张倩. 基于机器视觉的单组分染液浓度快速检测方法[J]. 纺织学报, 2021, 42(03): 115-121.

TIAN Yuhang, WANG Shaozong, ZHANG Wenchang, ZHANG Qian. Rapid detection method of single-component dye liquor concentration based on machine vision[J]. Journal of Textile Research, 2021, 42(03): 115-121.

图/表 9

图1

图2

表1

图3

图4

表2

表3

线性回归方程及决定系数"

蓝光亮度等级	因变量	回归方程	决定系数r²
50	lg(R₀/R)	$y = 4.27157 x - 0.02536$	0.991 2
20	lg(G₀/G)	$y = 3.55831 x - 0.0204$	0.997 5
15	lg(G₀/G)	$y = 3.40354 x + 0.02737$	0.993 3
10	lg(G₀/G)	$y = 4.52617 x - 0.22988$	0.996 6

表3

表4

表5

参考文献 15

[1]	叶早萍. 染浴在线监控的染色工艺[J]. 印染, 2008,34(9):47-49.
	YE Zaoping. Dyeing process of on-line surveillance in dye bath[J]. China Dyeing & Finishing, 2008,34(9):47-49.
[2]	房文杰. 染液中染料组分的浓度分析[D]. 上海: 东华大学, 2011: 16-18.
	FANG Wenjie. Concentration analysis of dyestuffs in dye solution[D]. Shanghai: Donghua University, 2011: 16-18.
[3]	MORISHITA SI, YOKOTA H, HIMENO R, et al. Estimation method of fluorescent dye's concentration from multiple fluorescence and the estimation method of fluorescent intensity from multiple fluorescence:US20060200318[P]. 2006-09-07.
[4]	杨军浩. 国外活性染料高效液相色谱分析的新进展[J]. 染料与染色, 2009,46(5):56-60.
	YANG Junhao. Technological progresses in HPLC analysis of reactive dyes[J]. Dyestuffs and Coloration, 2009,46(5):56-60.
[5]	向忠, 王宇航, 吴金波, 等. 过氧化氢检测方法研究进展[J]. 纺织学报, 2020,41(10):197-204.
	XIANG Zhong, WANG Yuhang, WU Jinbo, et al. Research progress in detection of hydrogen peroxide concentration[J]. Journal of Textile Research, 2020,41(10):197-204.
[6]	KATRC R, LUTFIYE Dahil. Spectrophotometric determination and computational study of fluoride in hexavalent chromium electroplating bath[J]. Journal of Fluorine Chemistry, 2020. DOI: 10.1016/j.jfluchem.2020.109462. pmid: 26213416
[7]	张建新, 张银露, 胡旭东. 光谱优化处理结合多层次支持向量机的混合染液浓度检测方法[J]. 纺织学报, 2017,38(7):90-94.
	ZHANG Jianxin, ZHANG Yinlu, HU Xudong. Method of mixed dye liquor concentration detection based on spectral optimization and multi-level support vector machine[J]. Journal of Textile Research, 2017,38(7):90-94.
[8]	张建新, 徐文杰, 许珺. 双波长分光光度法染液吸光度在线检测装置[J]. 纺织学报, 2014,35(9):149-154.
	ZHANG Jianxin, XU Wenjie, XU Jun. Development of on-line detection device of dyebath absorbance based on dual wavelength spectrophotometry[J]. Journal of Textile Research, 2014,35(9):149-154.
[9]	房文杰, 屠天民, 张骞, 等. 染液浓度的在线监测初探[J]. 印染, 2011,37(7):12-15, 33.
	FANG Wenjie, TU Tianmin, ZHANG Qian, et al. On-line monitoring of dye concentration[J]. China Dyeing & Finishing, 2011,37(7):12-15, 33.
[10]	FERUS-COMELO M, 张奔. 溶液中分散染料状态的光谱分析[J]. 国际纺织导报, 2008(6):44-46.
	FERUS-COMELO M, ZHANG Ben. Spectroscopic analysis of the state of disperse dyes in solution[J]. Melliand China, 2008(6):44-46.
[11]	NISHIDATE I, MAEDA T, NIIZEKI K, et al. Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method[J]. Sensors, 2013,13(6):7902-7915. pmid: 23783740
[12]	王克如, 李少昆, 王崇桃, 等. 用机器视觉技术获取棉花叶片叶绿素浓度[J]. 作物学报, 2006,32(1):34-40.
	WANG Keru, LI Shaokun, WANG Chongtao, et al. Acquired chlorophyll concentration of cotton leaves with technology of machine vision[J]. Acta Agronomica Sinica, 2006,32(1):34-40.
[13]	聂玉广. 基于RGB三基色数据的色度测量及不确定度评定研究[D]. 长春: 长春理工大学, 2014: 2-5.
	NIE Yuguang. The uncertainty evaluation of color measurement based on RGB trichromatic data[D]. Changchun: Changchun University of Science and Technology, 2014: 2-5.
[14]	汪晨. 紫外-可见光谱技术在连缸染色浓度分析中的应用研究[D]. 杭州: 中国计量学院, 2013: 9-10.
	WANG Chen. Research on dye concentration analysis of successive vats dyeing based on UV-Vis spectroscopy[D]. Hangzhou: China Jiliang University, 2013: 9-10.
[15]	杨海燕, 贾贵儒. 基于数字色度学的有色透明溶液浓度快速检测方法[J]. 中国农业大学学报, 2006,11(3):47-50.
	YANG Haiyan, JIA Guiru. Study on rapid concentration detection of colored transparentsolution based on digital chromatology[J]. Journal of China Agricultural University, 2006,11(3):47-50.

相关文章 15

[1]	冯文倩, 李新荣, 杨帅. 人体轮廓机器视觉检测算法的研究进展[J]. 纺织学报, 2021, 42(03): 190-196.
[2]	朱世根, 杨宏贤, 白云峰, 丁浩, 朱巧莲. 长条状细薄带钩零件变形自动检测系统[J]. 纺织学报, 2020, 41(10): 158-163.
[3]	张建新, 李琦. 基于机器视觉的筒子纱密度在线检测系统[J]. 纺织学报, 2020, 41(06): 141-146.
[4]	路浩, 陈原. 基于机器视觉的碳纤维预浸料表面缺陷检测方法[J]. 纺织学报, 2020, 41(04): 51-57.
[5]	王文胜, 李天剑, 冉宇辰, 卢影, 黄民. 筒子纱纱笼纱杆的定位检测方法[J]. 纺织学报, 2020, 41(03): 160-167.
[6]	金守峰, 林强强, 马秋瑞, 张浩. 基于BP神经网络的织物表面绒毛质量的检测方法[J]. 纺织学报, 2020, 41(02): 69-76.
[7]	景军锋, 张君扬, 张缓缓, 苏泽斌. 梯度空间下的丝饼表面缺陷检测[J]. 纺织学报, 2020, 41(02): 44-51.
[8]	孙卫红, 阮棉奖, 邵铁锋, 梁曼. 基于机器视觉的生丝抱合性能检测方法[J]. 纺织学报, 2019, 40(08): 164-168.
[9]	景军锋, 张星星. 基于机器视觉的玻璃纤维管纱毛羽检测[J]. 纺织学报, 2019, 40(05): 157-162.
[10]	徐洋, 朱治潮, 盛晓伟, 余智祺, 孙以泽. 基于机器视觉的鞋面特征点自动识别改进方法[J]. 纺织学报, 2019, 40(03): 168-174.
[11]	景军锋, 郭根. 基于机器视觉的丝饼毛羽检测[J]. 纺织学报, 2019, 40(01): 147-152.
[12]	牟新刚蔡逸超周晓陈国良. 基于机器视觉的筒子纱缺陷在线检测系统[J]. 纺织学报, 2018, 39(01): 139-145.
[13]	张建新张银露胡旭东. 光谱优化处理结合多层次支持向量机的混合染液浓度检测方法[J]. 纺织学报, 2017, 38(07): 90-94.
[14]	杨松林马帅丁朝鹏范红丽薛欢欢. 应用机器视觉的织物表面绒毛率测试系统[J]. 纺织学报, 2017, 38(06): 118-123.
[15]	蒋双歌王蕾王静刘建立高卫东. 织物折皱回复的各向异性表征[J]. 纺织学报, 2016, 37(06): 42-47.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

染料型号	染液质量浓度/(g·L^-1)	R值	G值	B值
DRA-3B	0.4	205.77	27.6	8.27
DRA-3R	0.4	230.97	118.75	30.13
DRA-3G	0.2	4.26	22.86	49.43

蓝光亮度等级	因变量	回归方程	决定系数r²
50	R	y=78.551 03×10^-4^.^{038 84}^x	0.997 30
	G	y=410.039 33×10^-3^.^{122 55}^x	0.989 66
	B	y=2 121.081 35×10^-4^.^{575 25}^x	0.976 26
20	R	y=27.567 31×10^-4^.^{204 28}^x	0.996 97
	G	y=163.335 7×10^-3^.^{374 1}^x	0.998 79
	B	y=690.187 ×10^-4^.^{410 42}^x	0.992 08
15	R	y=20.579 4×10^-4^.^{569 75}^x	0.997 63
	G	y=122.772 88×10^-3^.^{624 28}^x	0.998 91
	B	y=444.944 75×10^-4^.^{304 12}^x	0.996 11
10	R	y=13.768 9×10^-4^.^{314 87}^x	0.994 49
	G	y=84.228 05×10^-3^.^{363 2}^x	0.996 65
	B	y=13.768 9×10^-4^.^{043 69}^x	0.993 99

蓝光亮度等级	染液质量浓度/(g·L^-1)		相对误差/%	平均相对误差绝对值/%
蓝光亮度等级	实际值	预测值	相对误差/%	平均相对误差绝对值/%
50	0.05	0.050 4	0.75	3.83
	0.10	0.051 4	6.61
	0.25	0.106 6	-3.58
	0.30	0.241 1	-5.25
	0.35	0.284 2	-2.28
	0.40	0.342 0	-4.53
20	0.05	0.050 6	1.29	3.80
	0.10	0.103 4	3.45
	0.25	0.241 1	-3.57
	0.30	0.283 1	-5.64
	0.35	0.337 8	-3.49
	0.40	0.378 5	-5.36
15	0.05	0.037 1	-25.84	7.33
	0.10	0.096 3	-3.69
	0.25	0.242 1	-3.15
	0.30	0.284 0	-5.34
	0.35	0.346 5	-1.00
	0.40	0.380 2	-4.95
10	0.05	0.051 2	2.31	3.45
	0.10	0.103 2	3.23
	0.25	0.244 4	-2.26
	0.30	0.281 5	-6.16
	0.35	0.349 5	-0.14
	0.40	0.373 6	-6.60

蓝光亮度等级	染料质量浓度/(g·L^-1)		相对误差/%	平均相对误差绝对值/%
蓝光亮度等级	实际值	预测值	相对误差/%	平均相对误差绝对值/%
50	0.05	0.047 4	-5.20	3.35
	0.10	0.104 1	4.10
	0.25	0.245 3	-1.88
	0.30	0.290 3	-3.23
	0.35	0.342 8	-2.06
	0.40	0.385 6	-3.60
20	0.05	0.047 4	-5.20	3.71
	0.10	0.108 4	8.40
	0.25	0.244 8	-2.08
	0.30	0.290 8	-3.07
	0.35	0.347 1	-0.83
	0.40	0.389 2	2.70
15	0.05	0.041 6	-16.80	6.39
	0.10	0.095 7	-4.30
	0.25	0.238 0	-4.80
	0.30	0.280 4	-6.53
	0.35	0.344 3	-1.63
	0.40	0.382 9	-4.95
10	0.05	0.045 7	-8.60	3.98
	0.10	0.102 0	2.00
	0.25	0.258 3	3.32
	0.30	0.301 2	0.40
	0.35	0.376 8	7.66
	0.40	0.407 5	1.87