梯度空间下的丝饼表面缺陷检测

doi:10.13475/j.fzxb.20190401708

摘要/Abstract

摘要：

为解决工业生产中人工检测丝饼表面缺陷效率低、漏检率高的问题,提出了一种在梯度空间下根据图像信息熵变化和能量分布的差异来检测丝饼表面缺陷的方法。首先设计一套基于机器视觉的丝饼图像采集装置,用于获取传输过程中的丝饼表面图像;然后将丝饼图像转换到梯度空间域,构建一个信息熵和能量的组合特征用来表征缺陷,选择适当的临界阈值区分丝饼缺陷区域与正常区域;最后对分割出的丝饼缺陷利用形态学处理得到最终的检测结果。实验结果表明,该方法对丝饼表面污渍、压痕、起毛等缺陷具有较好的检测效果,缺陷识别准确率高、速度快,可满足工厂对检测准确性和实时性的要求,实现丝饼表面缺陷的自动化检测。

关键词: 机器视觉, 丝饼, 缺陷检测, 梯度空间, 图像信息熵

Abstract:

A method based on the difference of image information entropy and energy distribution in gradient space was proposed to solve the low efficiency and high missing rate of manual defect detection on the surface of DTY (draw texturing yarn) packages in industrial production. Firstly, the image acquisition device based on machine vision was designed to acquire the surface image of DTY packages during transmission. Then, the image of DTY packages was transformed into gradient space domain, and a combination feature of information entropy and energy was constructed to characterize the defect. Appropriate threshold was selected to distinguish the defect area from the normal area. Finally, the final detection result was obtained by morphological processing. The experimental results show that the method has a good detection effect for the defects such as stain, indentation and hairiness on the surface of DTY packages. The accuracy of defect recognition method is high and the speed is fast, which meets the requirements of the factory for accuracy and real-time detection, and realizes the automatic detection of defects on the surface of the DTY packages.

Key words: machine vision, DTY packages, defect detection, gradient space, image information entropy

中图分类号:

TP391.4

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.

JING Junfeng, ZHANG Junyang, ZHANG Huanhuan, SU Zebin. Defect detection on surface of draw texturing yarn packages in gradient space[J]. Journal of Textile Research, 2020, 41(02): 44-51.

图/表 12

图1

图2

图3

图4

图5

表1

图6

图7

表2

表3

图8

表4

精确评估结果对比"

检测方法	$R D$	$R TD$	精确率/%	召回率/%	调和平均数/%
Gabor滤波法	54 020	38 790	71.81	67.68	69.68
显著性检测法	32 509	27 394	84.27	47.80	61.00
迟滞阈值法	47 377	44 139	93.17	77.02	84.33
自编码器	50 257	46 783	93.09	81.63	86.98
本文方法	56 238	55 362	98.44	96.60	97.51

表4

参考文献 20

[1]	WHITTAKER C M. The dyeing of viscose rayon yarn in cake form[J]. Coloration Technology, 2010,60(5):109-114.
[2]	MACHALABA N, RODIONOV I. Development issues of the Russian chemical-fiber industry[J]. Fibre Chemistry, 2015,47(4):1-9.
[3]	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(1):147-152.
	JING Junfeng, GUO Gen. Yarn packages hairiness detection based on machine vision[J]. Journal of Textile Research, 2019,40(1):147-152.
[4]	汤勃, 孔建益, 伍世虔. 机器视觉表面缺陷检测综述[J]. 中国图象图形学报, 2017,22(12):1640-1663.
	TANG Bo, KONG Jianyi, WU Shiqian. Review of surface defect detection based on machine vision[J]. Journal of Image and Graphics, 2017,22(12):1640-1663.
[5]	MOVAFEGHI A, MOHAMMADZADEH N, YAHAGHI E, et al. Defect detection of industrial radiography images of ammonia pipes by a sparse coding model[J]. Journal of Nondestructive Evaluation, 2018,37(1):1-7.
[6]	JING Junfeng, CHEN Shan, LI Pengfei. Fabric defect detection based on golden image subtraction[J]. Coloration Technology, 2016,133(1):26-39.
[7]	钱晓亮, 张鹤庆, 张焕龙, 等. 基于视觉显著性的太阳能电池片表面缺陷检测[J]. 仪器仪表学报, 2017,38(7):1570-1578.
	QIAN Xiaoliang, ZHANG Heqing, ZHANG Huanlong, et al. Solar cell surface defect detection based on visual saliency[J]. Chinese Journal of Scientific Instrument, 2017,38(7):1570-1578.
[8]	SELVER M, AVŞAR V, ÖZDEMIR H . Textural fabric defect detection using statistical texture transformations and gradient search[J]. Journal of the Textile Institute, 2014,105(9):998-1007.
[9]	张帆, 张团善, 冀永乐, 等. 基于机器视觉的纺纱管颜色分拣算法研究[J]. 西安工程大学学报, 2018,32(5):560-566.
	ZHANG Fan, ZHANG Tuanshan, JI Yongle, et al. Research on color sorting algorithm of spinning tube based on machine vision[J]. Journal of Xi'an Polytechnic University, 2018,32(5):560-566.
[10]	LI Chunlei, GAO Guangshuai, LIU Zhoufeng, et al. Fabric defect detection based on biological vision modeling[J]. IEEE Access, 2018,6(1):27659-27670.
[11]	ZHAO Shiwei, ZHANG Nan, ZHOU Xiaowen, et al. Particle shape effects on fabric of granular random packing[J]. Powder Technology, 2017,310(1):175-186.
[12]	KANG Xuejuan, ZHANG Erhu. A universal defect detection approach for various types of fabrics based on the Elo-rating algorithm of the integral image[J]. Textile Research Journal, 2019,89(21/22):1-28.
[13]	YE Guodong, PAN Chen, HUANG Xiaoling, et al. A chaotic image encryption algorithm based on information entropy[J]. International Journal of Bifurcation and Chaos, 2018,28(1):1850010-1850021.
[14]	包晓敏, 汪亚明, 罗一平, 等. 基于最大多符号信息熵的织物图像匹配[J]. 纺织学报, 2005,26(2):69-71.
	BAO Xiaomin, WANG Yaming, LUO Yiping, et al. Matching of textile image based on the biggest multi-symbol information entropy[J]. Journal of Textile Research, 2005,26(2):69-71.
[15]	WANG Yong, QIAN Guangzhao. Novel approach for InSAR sensors imaging via gradient-based algorithm for the sparse signal reconstruction[J]. IEEE Sensors Journal, 2018,18(6):2385-2394.
[16]	张波, 汤春明. 基于相对总变差模型与自适应形态学的织物瑕疵检测[J]. 纺织学报, 2017,38(5):145-149.
	ZHANG Bo, TANG Chunming. Fabric defect detection based on relative total variation model and adaptive mathematical morphology[J]. Journal of Textile Research, 2017,38(5):145-149.
[17]	HAIRER E, LUBICH C. Energy-diminishing integration of gradient systems[J]. IMA Journal of Numerical Analysis, 2014,34(2):452-461.
[18]	ANWAR S, ABDULLAH M. Micro-crack detection of multicrystalline solar cells featuring an improved anisotropic diffusion filter and image segmentation technique[J]. EURASIP Journal on Image and Video Processing, 2014,2014(1):1-17.
[19]	黄森林, 王耀南, 彭玉, 等. 基于迟滞阈值分割的瓶口缺陷检测方法[J]. 电子测量与仪器学报, 2017,31(8):1289-1296.
	HUANG Senlin, WANG Yaonan, PENG Yu, et al. Bottle mouth defect detection method based on hysteresis thresholding segmentation[J]. Journal of Electronic Measurement and Instrumentation, 2017,31(8):1289-1296.
[20]	张宏伟, 汤文博, 李鹏飞, 等. 基于去噪卷积自编码器的色织衬衫裁片缺陷检测[J]. 纺织高校基础科学学报, 2019,32(2):119-125.
	ZHANG Hongwei, TANG Wenbo, LI Pengfei, et al. Defect detection and location of yarn-dyed shirt piece based on denoising convolutional autoencoder[J]. Basic Sciences Journal of Textile Universities, 2019,32(2):119-125.

相关文章 15

[1]	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.
[2]	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.
[3]	王泽霞, 陈革, 陈振中. 基于改进卷积神经网络的化纤丝饼表面缺陷识别[J]. 纺织学报, 2020, 41(04): 39-44.
[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]	肖志涛, 郭永敏, 耿磊, 吴骏, 张芳, 王雯, 刘彦北. 基于超声相控阵的机织层合复合薄板试件内部缺陷检测方法[J]. 纺织学报, 2019, 40(11): 81-87.
[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]	朱浩, 丁辉, 尚媛园, 邵珠宏. 多纹理分级融合的织物缺陷检测算法[J]. 纺织学报, 2019, 40(06): 117-124.
[10]	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.
[11]	蔡逸超, 周晓, 宋明峰, 牟新刚. 应用多尺度多方向模板卷积的筒子纱缺陷检测[J]. 纺织学报, 2019, 40(04): 152-157.
[12]	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.
[13]	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.
[14]	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.
[15]	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.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

编号	列1	列2	列3	列4	列5	列6	列7	列8	列9	列10
行1	5.240 8	5.202 1	5.286 6	5.246 1	5.260 7	5.277 8	5.283 1	5.286 3	5.118 9	5.299 5
行2	4.232 7	4.325 3	4.313 7	4.264 1	4.230 9	4.304 8	4.257 5	4.285 5	4.334 8	4.234 5
行3	3.903 7	4.032 1	4.195 2	4.350 3	4.231 7	4.348 7	4.259 6	4.268 3	4.277 2	4.273 2
行4	3.815 1	4.064 2	4.294 8	4.250 4	4.288 7	4.172 6	4.285 8	4.370 3	4.392 8	6.483 0
行5	3.954 9	4.179 4	4.352 3	4.338 0	3.914 6	3.983 5	4.130 2	3.946 0	4.250 5	6.522 3
行6	3.976 0	3.924 2	4.198 3	6.731 2	4.336 8	4.217 5	3.981 6	4.145 0	4.283 8	6.926 1
行7	3.866 7	3.840 3	3.921 4	3.886 9	3.962 8	6.301 5	7.156 9	3.719 1	4.280 8	6.536 8
行8	4.022 5	4.075 9	3.806 8	3.914 6	3.943 2	7.855 4	7.228 3	3.274 4	3.925 9	3.949 7
行9	4.200 1	4.261 3	4.223 6	3.928 6	4.103 4	3.977 4	3.970 5	3.967 1	7.301 8	3.842 3
行10	4.290 7	4.306 9	4.280 2	4.237 3	4.323 3	3.861 0	3.794 4	3.997 1	8.343 2	4.321 4

指标	正常无缺陷	异常有缺陷				合计
指标	正常无缺陷	污渍	压痕	起毛	组合缺陷	合计
数量/个	300	500	500	500	200	2 000
正确检测数/个	300	498	487	493	199	1 977
漏检数目/个	0	2	13	7	1	23
检测准确率/%	98.85
检测漏检率/%	1.35
检测时间/ms	87.00

检测方法	检测准确率/%	检测漏检率/%	检测时间/ms
Gabor滤波法	98.00	2.18	1 036
显著性检测法	90.15	9.24	2 153
迟滞阈值法	96.40	4.06	248
自编码器	96.00	4.18	142
本文方法	98.85	1.35	87