See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/339749654

Fault Detection in Fabric Using Classification and Automated approach

Article · June 2015

CITATIONS READS
0 24

2 authors:

Saravanan P. s. Balaji
Dhanraj baid jain college(Autonomous), Chennai DHANRAJ BAID JAIN COLLEGE
24 PUBLICATIONS   18 CITATIONS    2 PUBLICATIONS   2 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Analysis of Various Contrast Improvement Techniques for Dehazing an Image View project

Wireless sensor networks View project

All content following this page was uploaded by Saravanan P. on 06 March 2020.

The user has requested enhancement of the downloaded file.

 International Journal of Trend in Research and Development, Volume 2(3), ISSN: 2394-9333
www.ijtrd.com

Fault Detection in Fabric Using Classification and Automated

Approach
S.Balaji 1,P.Saravanan 2
1,2
Assistant Professor,
P.G.Dept of Research and Computer Science,
Dhanraj Baid Jain College(Autonomous), Tamil Nadu, India

Abstract: Garments with defects normally affects J.lewis dorrity etl.study indicated the survey of
the quality of the products. The garments with major textile defects as well as the associated
defects in the market needs marginalized discounts tangible and intangible costs. Jagruti mahure & y.c.
for a saleable one. In order to ensure quality and kulkarni paper mainly describes the various
defect free end product, there is a need for fault problems faced while the processing of the system.
detection mechanism in the fabric production It also describes on what ground do one needs to
stage. Even though manual fault detection is the take a particular step forward for the killing of the
traditional approach, automated approach towards anomaly occurred in that particular situation.
the fault detection is considered in this paper. This description is done about fault detection,
paper proposes the neural network with thresolding positioning and classification of the faults that
for the fault detection in fabrics. exists or can occur in the weaving machine during
weaving by using the principle of image
Keywords: Quality, Defect detection, Neural processing, an automatic fabric evaluation system,
network, Thresolding. which enable computerized defect detection
analysis of weaved fabrics. Wastage reduction
I.INTRODUCTION through accurate and early stage detection of
defects in fabrics is an important aspect of quality
Textile industry has been gaining its improvement is indicated in the work of
importance in Industrial sector for the years. arivazhagan and ganesan(2006). Kang et
Inspecting the quality of the fabrics is the main al.analyzed fabric samples from the images
part of industrial manufacturing. The process of obtained from transmission and reflection of light
reducing defects in the fabrics improves the quality to determine its interlacing pattern.
of the fabrics. The main aim of inspecting the
fabrics is identifying and locating defects. Quality III. DRIVING FORCES BEHIND DEFECTS
improvement aspect is concerned with accurate IN FABRICS
and early stage detection of defects in fabrics. The
most common defects in fabrics are hole, scratch, The national economy has the significant
stretch, fly, yarn, dirty spot, slub, cracked point contribution of textile industry in the form of
and color bleeding. The impact of these defects exports. Therefore defects has the serious impact
can affect the production process massively. on quality of the fabrics. Different stages and
processes of fabrics may be avenue for defects in
II. LITERATURE SUPPORT fabrics. The defects may occur during one or more
stages like spinning, weaving, knitting, printing,
M.rallo.,etl. observed textured samples dyeing, stitiching etc. Missing ends is one of the
analysis have been done with digital image defects in loom static cloth which may consume
processing techniques. Machine vision automated upto 50% of total defects in fabric.
inspection system for textile defects has been in the Knots with long tail ends, oozy, entanglement of
research industry for longtime was highlighted in adjoining ends in the regions between heads can
the work of B.G. Batchelor and P. F. Whelan. cause the defect jala. Unbalanced shuttle,damaged
or broken picking,wrong timing of shedding can

IJTRD | May-Jun 2015

Available [email protected] 138
 International Journal of Trend in Research and Development, Volume 2(3), ISSN: 2394-9333
www.ijtrd.com

cause break to a shuttle trap. Hole in the cloth may pixel in the gray-scale image contains only the
be due to shuttle smash,sharp edge on cloth intensity information.
roller.Due to poor oiling,lack of care on the part of Third phase concerned with removal of noise free
operatives and cleaning practicies there is a and disturbances free image.The subsequent
possibility for stains the fabrics. process of conversion to Binary format of the
input.
IV. METHODOLOGY The distribution of data is represented by a
histogram.The histogram helpful for decisions
Fabric texture refers to the feel of the regarding the classification of the faults. The image
fabric. It is smooth, rough, soft, velvety, silky, thresholding is a type of image segmentation.
lustrous, and so on. The different textures of the The conversion process of images from grayscale
fabric depend upon the types of weaves used. to binary isolates objects.The images with high
Textures are given to all types of levels of constrast is most effective.The
fabrics,cotton,silk, wool, leather etc.,In textile, thresholding is a simple and effective process of
different types of faults are available partitioning foreground and background.
i.e.hole,scratch,stretch,fly yarn,dirty spot,slub, Matlab is used for the implementation of the
cracked point, colour bleeding etc; if not detected system.Its efficient behavior with various types of
properly these faults can affect the production images and the ease of operations leads to effective
process massively. output.
The work proposed in this paper is the
identification of defect in the fabric.The location CONCLUSION
and the type of the defect is indicated if there is a
defect in the fabric.The input for the system is the As the accuracy of the human inspection
fabric which may be faulty one or defect free.The of faults in fabric diminishes with routine jobs and
output of the histogram helps to conclude the lack of quality manpower.Erractic inspection and
defect rate in the fabrics. slow process may result with manual
The various Phases in processing the fabric are as inspection.The approach followed in this paper
follows yield better results than the manual textile quality
1.Image acquisition control.It detects most common faults such as
2.Conversion of color image to gray image hole,cut or tear,gout and broken picks.
3.Removal of disturbances and Noise The problem over visual inspection (human) like
4.Noise free image is to be converted to Binary tiring,fatiguing and tedious process in attention and
image observation to detect the fault occurrence.
5.Drawing conclusion from Histogram output
6.Thresholding process REFERENCES
In the above said phases,the first phase concerned
with getting the image of the fabric with the help 1. M. Ralló, M. S. Millán, J. Escofet,
of digital camera or charged couple device camera “Wavelet based techniques for textile
or CMOS camera.The image of size 512 x 512 inspection”, Opt.Eng. 26(2), 838-844
pixels.Two dimensional camera can be used.The (2003).
orthogonal placement of camera at a distance of 2. B. G. Batchelor and P. F. Whelan,
50m from the textile produces quality image of the “Selected Papers on Industrial Machine
fabric. Vision Systems,” SPIE Milestone Series,
The acquired image has to retain luminance 1994.
information with the elimination of hue and 3. T. S. Newman and A.K. Jain, “A Survey
saturation.For this purpose the confirmation of Automated Visual Inspection,”
process of color to gray image is a part of second Computer Vision and Image
phase. Gary scale image is of black-and-white Understanding, vol. 61, 1995, pp.231–262
nature.The variation from black to white is being 4. J.Lewis Dorrity, G. Vachtsevanos, Warren
referred as weak intensity to strong intensity.The Jasper, “Real-time fabric defect detection
and control in weaving processes”,

IJTRD | May-Jun 2015

Available [email protected] 139
 International Journal of Trend in Research and Development, Volume 2(3), ISSN: 2394-9333
www.ijtrd.com

National Textile Center Annual Report:

August 1995,Project No.G94-2.
5. Jagruti mahure & y.c. kulkarni,"fabric
faults processing: perfections and
imperfections",international journal of
computer networking,wireless and mobile
communications (ijcnwmc),vol. 4, issue
2,apr 2014,pp.101-106.
6. Arivazhagan S., Ganesan L. and Bama S.
(2006),Fault segmentation in fabric images
using Gabor wavelet transform‟,
International Journal of Machine Vision
and Applications,Vol. 16, No. 6, pp. 356-
363.
7. Kang T.J. et al. “Automatic Recognition of
Fabric Weave Patterns by Digital Image
Analysis”, Textile Res. J. 69(2), 77-83
(1999).
8. Tolba A.S. and Abu-Rezeq A.N.: A self-
organizing feature map for automated
visual inspection of textile products,
Computers in Industry,Vol. 32, 1997, pp.
319-333.
9. Ojala T., Pietikainen M., and Harwood D.:
A comparative study of texture measures
with classification based on featured
distribution,Pattern Recognition, Vol. 29,
Issue 1,January, 1996, pp. 51-59.
10. Tsai D.-M. and Huang T. Y.: Automated
surface inspection for statistical textures,
Image and Vision Computing, Vol. 21,
2003, pp. 307-323.
11. Escofet J., Millan M. S., and Rallo M.:
Modeling of woven fabric structures based
on Fourier image analysis, Applied
Optics,Vol. 40, No.34, December, 2001,
pp. 6170-6176.
12. Ralló M. l., Millán M. S. and Escofet J.:
Reference less segmentation of flaws in
woven fabrics, Applied Optics, Vol. 46,
No . 27,September, 2007, pp. 6688-6699.
13. Sakaguchi A., Kim H., Matsumoto Y.-I.
and Toriumi K.: Woven Fabric Quality
Evaluation Using Image Analysis, Textile
Research Journal, Vol. 70,
November,2000, pp. 950-956.
14. Tsai I.-S. and Hu M.-C.: Automatic
Inspection of Fabric Defects Using an
Artificial Neural Network Technique,
Textile Research Journal,Vol. 66, July,
1996, pp. 474- 482.

IJTRD | May-Jun 2015

Available [email protected] 140

View publication stats