基于总变差的织物疵点分割方法

doi:10.13475/j.fzxb.20201200807

摘要/Abstract

摘要：

为解决当前机织物疵点检测方法精度不足的问题,提出了基于总变差模型的织物疵点分割方法,并着重分割经纬向尺寸小且异常不显著的疵点。首先应用奇异值分解低秩重建的方法将织物纹理背景去除,获取疵点异常图;然后通过构建总变差模型对疵点异常图进行最优化求解处理,得到不同约束下的疵点增强图;最后通过常规分割算法实现疵点的准确分割。实验结果表明:经总变差模型处理后的疵点异常图,其疵点与背景的可分割性得到显著提升。通过讨论总变差模型的参数对分割结果的影响,进一步验证了基于总变差的织物疵点分割方法的有效性和稳定性。

关键词: 织物疵点, 纹理低秩重建, 总变差模型, 疵点分割, 图像识别

Abstract:

Aiming at the problem of insufficient accuracy of the current woven fabric defect detection methods, this paper proposes a fabric defect segmentation method based on the total variation model, focusing on solving unobvious defects along warp and weft directions. The singular value decomposition low-rank was used to reconstruct the textile texture image by removing the texture background of the fabric so as to obtain the defect abnormal map. Following that, by constructing a total variation model, the defect anomaly map was optimized to obtain the enhanced abnormal map under different constraints. Finally, the accurate segmentation of defects was achieved through adopting the conventional segmentation algorithms. The experimental results show that the separability of the defects and the background of the defect abnormal image processed by the proposed total variation model has been significantly improved. The influence of the parameters of the model on the segmentation results was discussed to further verify the effectiveness and stability of the method.

Key words: fabric defect, texture low-rank reconstruction, total variation model, defect segmentation, image recognition

中图分类号:

TS101.91

刘国维, 潘如如, 高卫东, 周建. 基于总变差的织物疵点分割方法[J]. 纺织学报, 2021, 42(11): 64-70.

LIU Guowei, PAN Ruru, GAO Weidong, ZHOU Jian. Fabric defects segmentation using total variation[J]. Journal of Textile Research, 2021, 42(11): 64-70.

图/表 11

图1

图2

图3

图4

表1

图5

图6

图7

图8

表2

图9

参考文献 15

[1]	ZHOU Jian, WANG Jun. Unsupervised fabric defect segmentation using local patch approximation[J]. Journal of The Textile Institute, 2016, 107(6):1-10. doi: 10.1080/00405000.2014.999478
[2]	LATIF-AMET A, ERTÜZÜN A, ERÇIL A. An efficient method for texture defect detection: sub-band domain co-occurrence matrices[J]. Image and Vision Computing, 2000, 18(6):543-553. doi: 10.1016/S0262-8856(99)00062-1
[3]	CHAN C H, PANG G K H. Fabric defect detection by Fourier analysis[J]. IEEE Trans on Ind Appl, 2000, 36(5):1267-1276. doi: 10.1109/28.871274
[4]	李仁忠, 张缓缓, 景军锋, 等. 基于EM算法的高斯混合型的织物疵点检测研究[J]. 计算机工程与应用, 2014, 50(10):184-187.
	LI Renzhong, ZHANG Huanhuan, JING Junfeng, et al. Fabric defect detection based on Gaussian mixture models of EM algorithm[J]. Computer Engineering and Applications, 2014, 50(10):184-187.
[5]	潘如如, 高卫东, 钱欣欣, 等. 基于互相关的印花织物疵点检测[J]. 纺织学报, 2010, 31(12):134-138.
	PAN Ruru, GAO Weidong, QIAN Xinxin, et al. Defect detection of printed fabrics using normalized cross correlation[J]. Journal of Textile Research, 2010, 31(12):134-138.
[6]	JING Junfeng, MA Hao, ZHANG Huanhuan. Automatic fabric defect detection using a deep convolutional neural network[J]. Coloration Technology, 2019, 135(3):2-4.
[7]	厉征鑫, 周建, 潘如如, 等. 应用单演小波分析的织物疵点检测[J]. 纺织学报, 2016, 37(9):59-64.
	LI Zhengxin, ZHOU Jian, PAN Ruru, et al. Fabric defect detection using monogenic wavelet analysis[J]. Journal of Textile Research, 2016, 37(9):59-64.
[8]	屈博, 卢朝阳, 李静, 等. 一种改进的多通道Gabor滤波器布匹瑕疵检测方法[J]. 纺织学报, 2009, 30(12):37-40.
	QU Bo, LU Chaoyang, LI Jing, et al. An improved multichannel Gabor filter algorithm for fabric defect detection[J]. Journal of Textile Research, 2009, 30(12):37-40. doi: 10.1177/004051756003000104
[9]	祝双武, 郝重阳. 基于纹理周期性分析的织物疵点检测方法[J]. 计算机工程与应用, 2012, 48(21):163-166.
	ZHU Shuangwu, HAO Chongyang. Fabric defect detection approach based on texture periodicity analysis[J]. Computer Engineering and Applications, 2012, 48(21):163-166.
[10]	ANTONIN Chambolle. An algorithm for total variation minimization and applications[J]. Journal of Mathematical Imaging and Vision, 2004, 20(1/2):89-97. doi: 10.1023/B:JMIV.0000011321.19549.88
[11]	BRESSON X, CHAN T F. Fast dual minimization of the vectorial total variation norm and applications to color image processing[J]. Inverse Problems and Imaging, 2008(2):455-484.
[12]	CHATTERJEE. An introduction to the proper orthogonal decomposition[J]. Current Science, 2000, 78:808-817.
[13]	ZHOU Jian, WANG Jun. A real-time computer vision-based platform for fabric inspection: part 1: algorithm[J]. Journal of The Textile Institute, 2015, 106(12):1282-1292. doi: 10.1080/00405000.2014.996333
[14]	刘红毅. 结构保持的图像去噪方法研究[D]. 南京:南京理工大学, 2011:10-13.
	LIU Hongyi. Structure preserved image denoising algorithms[D]. Nanjing: Nanjing University of Science and Technology, 2011:10-13.
[15]	李燕民, 季维勋, 李维国. 一种新的图像去噪方法[J]. 微计算机应用, 2010, 31(3):17-21.
	LI Yanmin, JI Weixun, LI Weiguo. A new method for image denoising[J]. Microcomputer Applications, 2010, 31(3):17-21.

相关文章 15

[1]	吕文涛, 林琪琪, 钟佳莹, 王成群, 徐伟强. 面向织物疵点检测的图像处理技术研究进展[J]. 纺织学报, 2021, 42(11): 197-206.
[2]	朱磊, 任梦凡, 潘杨, 李博涛. 基于相似性定位和超像素分割的织物疵点检测[J]. 纺织学报, 2020, 41(10): 58-66.
[3]	李涛, 杜磊, 黄振华, 蒋玉萍, 邹奉元. 服装款式图识别与样板转换技术研究进展[J]. 纺织学报, 2020, 41(08): 145-151.
[4]	杨恩君, 廖义辉, 刘安东, 俞立. 基于低秩分解的织物疵点检测[J]. 纺织学报, 2020, 41(05): 72-78.
[5]	张缓缓, 马金秀, 景军锋, 李鹏飞. 基于改进的加权中值滤波与K-means聚类的织物缺陷检测[J]. 纺织学报, 2019, 40(12): 50-56.
[6]	许倩, 陈敏之. 基于深度学习的服装丝缕平衡性评价系统[J]. 纺织学报, 2019, 40(10): 191-195.
[7]	杜帅, 李岳阳, 王孟涛, 罗海驰, 蒋高明. 基于改进局部自适应对比法的织物疵点检测[J]. 纺织学报, 2019, 40(02): 38-44.
[8]	胡克满, 罗少龙, 胡海燕. 应用Canny算子的织物疵点检测改进算法[J]. 纺织学报, 2019, 40(01): 153-158.
[9]	徐启永胡峰王传桐吴雨川. 改进频率调谐显著算法在疵点图像分割中的应用[J]. 纺织学报, 2018, 39(05): 125-131.
[10]	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(05): 145-149.
[11]	尉苗苗李岳阳蒋高明丛洪莲. 应用最优Gabor滤波器的经编织物疵点检测[J]. 纺织学报, 2016, 37(11): 48-54.
[12]	石美红张正郭仙草陈永当. 基于显著纹理特征的织物疵点检测方法[J]. 纺织学报, 2016, 37(10): 42-049.
[13]	厉征鑫周建潘如如刘建立高卫东. 应用单演小波分析的织物疵点检测[J]. 纺织学报, 2016, 37(09): 59-64.
[14]	景军锋范晓婷李鹏飞张蕾张宏伟. 应用Gaussian回代交替方向图像分解算法的色织物疵点检测[J]. 纺织学报, 2016, 37(06): 136-141.
[15]	黎聪闫学娜曾祥忠梁猛张莹. 应用一维傅里叶变换的剖幅区自动识别与定位[J]. 纺织学报, 2016, 37(01): 147-151.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

样本编号	密度/(根·(10 cm)^-1)		疵点	组织
样本编号	经向	纬向	疵点	组织
1^#	402	254	沉纱	斜纹
2^#	348	276	双纬	平纹
3^#	292	224	缺纬	斜纹
4^#	346	274	双纬	斜纹
5^#	428	296	纬缩	平纹
6^#	428	296	稀纬	平纹
7^#	538	340	紧经	平纹
8^#	438	304	松经	斜纹

样本编号	λ值	F值
1^#	0.106	0.96
2^#	0.064	0.98
3^#	0.064	1.00
4^#	0.106	0.93
5^#	0.064	0.84
6^#	0.085	0.92
7^#	0.085	0.81
8^#	0.085	0.74