Proceedings of 2011 International Conference on Signal Processing, Communication, Computing and Networking Technologies (ICSCCN 2011)

A Novel Approach to Fabric Defect Detection Using

Digital Image Processing

S.Priya

Asst. Professor Dept. of Computer Sc. & Engg:

T. Ashok kumar

Asst. Professor Dr. Varghese Paul

Dean, Dept. of CS/IT Research, TocH Institute of Science& Technology, Arakkunnam,

Ernakulam Dist, Kerala, India vp@tistcochin.edu.in

Govt. Model Engineering College, Ernakulam , Kerala, India priyavarkala@yahoo.com,

priya@mec.ac.in

Dept. of Electronics & Com. Engg: College of Engineering, Cherthala

Alappuzha, Kerala, India ashokkumart@yahoo.com

Abstract-This paper presents a novel approach to the fast

detection and extraction of fabric defects from the images of

textile fabric. Automated visual inspection systems are much

needed in the textile industry, especially when the quality control

of products in textile industry is a significant problem. In the

manual fault detection systems with trained inspectors, very less

percentage of the defects are being detected while a real time

automatic system can increase this to a maximum number .Thus,

automated visual inspection systems play a great role in assessing

the quality of textile fabrics. For the detection of fabric defects,

we first decompose the image into its bit planes. The lower order

bit planes are found to carry important information of the

location and shape of defects. Then we find the exact location by

means of mathematical morphology.

The algorithm has been tested on a subset of TILDA I image

database with various visual qualities. Robustness with respect to

the changes of the parameters of the algorithm has been

evaluated.

Keywords-Fabric defects, bitplane decomposition, mathematical morphology ,dilation, erosion, opening, closing

I. INTRODUCTION

One of the industry fields where automated visual inspection systems are highly needed is the textile industry. Other than classifying a certain appearance of the fabric, registration of the exact location of the defects and determining their type are also important in several applications. The advantage for the manufacturer here is to get a warning when a certain amount of defect or imperfection occurs during the production of the fabric so that precautionary measures can be taken before the product hits the market. It has been observed [ 1] that price of textile fabric is reduced by 45% to 65% due to defects. Manual defect detection in a fabric quality control system with a width of 1.60-2.0 m and which moves with an average speed of 10 m per minute is a difficult task to be performed by inspectors. The work of an observer is very tedious and time consuming. They have to detect small details that can be located in a wide area that is moving through their visual field. The identification rate is only about 70% [3]. Moreover, the effectiveness of visual inspection decreases quickly with fatigue. Digital image processing techniques have been increasingly applied to textured sample analysis over the past

978-1-61284-653-8/11/$26.00 ©2011 IEEE

several years. Wastage reduction through accurate and early stage detection of defects in fabrics is also an important aspect of quality improvement.

The high cost ,along with other disadvantages of human visual inspection has led to the development of on-line machine vision systems that are capable of performing inspection tasks automatically. The problem of textile web inspection is particularly complex, since there is a large variety of fabrics of different structures, compositions, colors, and other properties. Human inspection systems are still preferred for a number of tasks because an alternate solution, reliable and versatile enough, is not available yet. Therefore, the search in the field is still wide open.

Textile quality control involves, among other tasks, the detection of defects that cause a distortion of the basic structure of the material, which commonly shows a high degree of periodicity. Several techniques of image analysis have been proposed for this purpose. Inspection of 100% of fabric is necessary first to determine the quality and second to detect any disturbance in the weaving process to prevent defects from reoccurring. Performance and assessment is never constant and effectiveness decreases quickly with fatigue. Owing to the very slow speed of human visual inspection compared to production rate, automatic inspection is more important than ever. Many researchers have worked on the automation of inspection systems.

II. BACKGROUND

Fabric defect detection using digital image processing has received considerable attention during the past two decades and numerous approaches have been proposed in the literature. Wang et a1. [2] reported that 90% of the defects in a plain fabric could be detected simply by thresholding. Table 1 summarizes a comparison between human visual inspection and automated inspection [4]. Zhang et al. [7] have introduced two approaches to detect defects: gray-level statistical and morphological methods. Lanes [8] has defmed a number of convolution masks to detect the defect. These methods, which

228

Proceedings of 2011 International Conference on Signal Processing, Communication, Computing and Networking Technologies (ICSCCN 2011)

depend on intensity change on the fabric image, can only

capture significant defects such as knot, web, and slub.

TABLE I

Visual inspection versus automated

Inspection Type inspection

Visual Automated

Fabric Types 100% 70%

Defect Detection 70% 80%+ Reproducibility 50% 90%+ Objective Defect

50% 100% Judgment Statistics Ability 0% 95%+ Inspection Speed 30 mlmin 120 mlmin Response Type 50% 80% Information Content 50% 90%+ Information

20% 90%+ Exchange

In view of the high degree of periodicity for textile fabrics,

Fourier transform based approaches were developed for defect

detection by some researchers. Wood [9] has used Fourier and

associate transform to characterize carpet patterns, while in

[10], the approach used is Histogram equalization followed by

FFT and central spatial frequency spectrum analysis.

Unfortunately, most of the algorithms used today for fabric

defect localization or detection are computationally intensive

and are less accurate, particularly in the presence of a number

of patterns and print. The proposed algorithm is simple and

more efficient for implementation. There is no mathematical complexity as in other methods and hence there is a

significant improvement in computational time also. This

algorithm has been evaluated on a subset of the TILDA I

image database with various visual qualities .

III. BITPLANE DECOMPOSITION

Separating a digital image into its bit planes is useful for

analyzing the relative importance played by each bit of the image. Instead of highlighting gray level images, highlighting

the contribution made to total image appearance by specific

bits is examined here[I4]. In an 8 bit gray level image, each

pixel in an image is represented by 8 bits. The image is

composed of 8, I-bit planes ranging from bit plane 0 (LSB)to

bit plane 7 (MSB). In terms of 8-bits, plane 0 contains all lowest order bits in the bytes comprising the pixels in the

image and plane 7 contains all higher order bits. Thus bitplane

decomposition of an 8 bit image yields eight binary images.

The bit-plane representation of an 8 bit image is shown in

Fig. I

For the images, bit-plane 0 and bit-plane I are carrying vital

data corresponding to the position and shape of the defects.

Bit-planes 0 and I of the chosen image are shown in Fig. 3 and

Fig.4 respectively. As bit-plane images are binary images,

they are highly suited for morphological image processing.

978-1-61284-653-8/11/$26.00 ©2011 IEEE

In general, the higher order bit planes contain a majority of

visually significant data while the lower order ones contribute to more subtle details in an image . On examining the eight bit planes of the image , the lower order bit planes are found

to carry significant information regarding the location and

shape of the fabric defect in the image.

g·bilpixel

Bit Plane 7 -- 4----------./ (MSB)

Bit Plane 0 (LSB) -�-'-------.../

Fig. 1.. Bitplane decomposition

Fig. 2. Original Image

6 5 4 3 2

o

::�:'!,':::::::::=::::�:�:�:::�:!:!&::::�::!t:� ::�::::::::::::::=I::::;''''I1:II.'1:;�!'�':: ,·_ .. ·· ..... · .. ,· .... ······ ........ m .. , ...... ........ ' .... , .......... ,....... I.

1'I .... .:::=a:a::=:.U=1rm:&: "

�"'=::''D.'Ift'�r.I:n.'I:t!1 __ ••• .. .... , .... '.,.iil ...... ....... ... .. =, ..... ,1 .• ,., •• , •• :::::�!:��:!!:':l:�==::lr.:��n�:� • .• ,&..... ... . . .. . . . J , •• , .... 1111 .... .. •• --� .' ....... , •. ., ..... ; ......... '.,.'.,.,.·.1.· ..• J. a,e . ............................. ......... , ••••• 1 ....... " ... .,.,.; ..... , ... ,.,.,., ... ,.,.,.,.' ., ... ,.,.,.,. " ."'.",.,11".,., •.• , •••• ··,···,11.'.'.'.'.'.'."" •.••.•.• , ..... '.1.,.,.' •.• ' .' •.• , •.• , ••.•.• ,. ' •.•..

........... , ....... , ..... , ......... ' ., ............. .. • ,111.' ....... , .,.;. , ••• ,.,.' •••••• ,., •••.• ,., •.•.••• I .• '. '. , ••• , ..... ",., •.• ,., ••• 11'.'.'1'.'. I.'. "'." ., ..... '.,., •.• '.,.J ••• '.,., •.•.•.• 1. I.e_.l.'." _I' , .......... '.'.11 ............ ' ..... ,.·."" ,.,.", ... 1, ... ;1 .• , ., •.•.• ,.,.:., •. 1,11".,11' ••••••••••• •••••.• •••••• ,., ........ , ..... , ......... I. ,.la, ... '81' � •.•• e ......... , •• a •••• I ••• ,.,.:atl'.ilf.J ••.•. 1

Fig. 3. Bitplane 0

229

Proceedings of 2011 International Conference on Signal Processing, Communication, Computing and Networking Technologies (ICSCCN 2011)

Fig. 4. Bitplane 1

IV. MATHEMATICAL MORPHOLOGY

Mathematical morphology deals with non-linear processes which can be applied to an image to remove details smaller

than a certain reference shape called the structuring element.

The most widely used morphological operations used in image

processing are dilation, erosion, opening and closing. Binary

images are best suited for performing morphological

operations. The images obtained after bit plane decomposition are binary images, which are thus suitable for performing

morphological operations[15]. Note that bit planes 0 and 1

contain the most significant information regarding the location

and shape of the fabric defect .

Dilation is an operation in which the binary image is

expanded from its original shape. The amount of expansion is

controlled by the structuring element. The dilation process is

similar to convolution, in which the structuring element is

reflected and shifted from left to right and then from top to

bottom. In this process, any overlapping pixels under the

centre position of the structuring element are assigned with 1

or black values. If X is the reference image and B is the

structuring element, the dilation ofXby B is represented as

x ffi B = {ZI[(B)znX] !; X}

Where B is the image B rotated about the origin. When an

image X is dilated by a structuring element B, the outcome

element Z would be that there will be at least one element in B

that intersects with an element in X

Erosion is a thinning operator that shrinks an image. The

amount by which shrinking takes place is determined by the

structuring element. Here, if there is a complete overlapping

with the structuring element, the pixel is set white or O. The erosion of X by B is given as

X 8 B = {Z I [(B)z] !; X}

In erosion, the outcome element Z is considered only when the

structuring element is a subset or equal to the binary image X

Opening operation is done by first performing erosion,

followed by dilation. Opening smoothens the inside of object

978-1-61284-653-8/11/$26.00 ©2011 IEEE

contours, breaks narrow strips and eliminate thin portions of

the image. It is mathematically represented as

X 0 B = (X 8 B) ffi B

Closing operation does the opposite of opening. It is dilation

followed by erosion. Closing fills small gaps and holes in a

single pixel object. The closing process is represented by

X • B = (X ffi B) 8 B

Closing operation protects coarse structures, closes small gaps

and rounds off concave comers.

Morphological operations are widely used in the detection of

boundaries in a binary image. For an image X , the following

can be applied to obtain a boundary image

Y = X - (X 8 B)

Y = (X ffi B) - X

or Y = (X ffi B) - (X 8 B)

where, the operator 'ffi' denotes dilation , '8' denotes

erosion and '- ' indicates the set theoretical subtraction.

Most binary morphological operations have natural

extensions to gray scale processing. Some, like morphological

reconstruction, have applications that are unique to gray scale

images , such as peak filtering.

V. THE METHODOLOGY

We first convert the RGB image obtained from the camera

to an equivalent grayscale image using the Matlab function

'rgb2gray'. Now bit plane slicing is performed on the gray

scale image to decompose it into its bit planes. The lower

order bit planes are preserved for further processing and the

higher order ones are discarded. Figure 3 and Figure 4 shows

bit plane 0 and bit plane 1 respectively for a representative

image. Bit plane 0 is now opened by a disc shaped

structuring element to obtain a clear white region

corresponding to the defect such as shown in Figure 5.

Fig. 5. Opened Bitplane 0

230

Proceedings of 2011 International Conference on Signal Processing, Communication, Computing and Networking Technologies (ICSCCN 2011)

Fig. 6. Dilated Outline

A suitably processed bit plane I can be used to compensate

for errors, if any, in the previous operation.

The matlab function 'bwperim' is used to obtain the outline of the defect. This outline is subsequently dilated by a carefully selected structuring element to obtain an image as

shown in Figure 6.

Now the fabric defect can be localized by superimposing

the outline on the original grayscale image as shown in

Figure 7.

Fig. 7. Defect Marked Image

The algorithm has been implemented by using Matlab version

7.9 (Release 2009 b) and is found to be reasonably fast and

accurate than the existing computationally intensive methods.

The results are promising even when it is applied to localize

defects on images with varying lighting or exposure levels .

The algorithm has been tested and compared with the

978-1-61284-653-8/11/$26.00 ©2011 IEEE

commonly used methods and the results show that the method

proposed here can not only detect defects but also provide

more acceptable location of the defect .

VI. RESULTS AND DISCUSSION

Modern industrial manufacturing of materials such as fabric,

paper, wood, leather, etc. requires a large number of

inspection tasks concerning the visual appearance of the

material surface (texture, color, fault detection, etc.). The

problem of textile web inspection is particularly complex,

since there is a large variety of fabrics of different structures,

compositions, colors, and other properties. Textile quality

control involves, among other tasks, the detection of defects

that cause a distortion of the basic structure of the material,

which commonly shows a high degree of periodicity. Several

techniques of image analysis have been proposed for this

purpose. If the fabric sample exhibits an overall distortion, Fourier-domain-based techniques allow one to obtain

successful results to detect shrinking, abrasion and skewness.

If the fabric sample has local defects or faults such as holes,

stains, broken threads, etc., then it is a good practice to apply

methods based on wavelet transforms, used as multiresolution

spectral filters, that localize and analyze features in both the spatial and the frequency domain. Some other tasks

concerning pattern recognition, weave-repeat identification

and classification can also be performed using techniques

based on Fourier analysis. In general, only uniformly colored

materials are usually considered because, the existence of hierarchical structures given by bands, squares, circles, or

drawings of varied colors along with the basic woven structure

makes the inspection more difficult.

We, human beings, have a unique capability to easily fmd

imperfections in spatial structures. This visual mechanism works even when we do not know what the ideal pattern is and

what the possible types of defects are. Just looking at a

relatively regular structure containing an imperfection, we can

usually tell what is wrong there. But human inspector based

defect detection is subject to errors for many reasons. There

are large influences of human errors and subjectivity on the results of inspection. Presence of other factors such as noise,

non-uniform illumination and variety of defect types in textiles

make the defect detection a challenging problem. Due to these

reasons, an expert system for the automatic detection of such

anomalies has inspired much research in this direction.

The proposed algorithm is simple and more efficient for

automation. There is no mathematical complexity as in other

methods and hence there is a significant improvement in

computational time also. Moreover, this method does not just

detect the defect but the shape and size of the defect also. It extracts most defect pixels accurately. This algorithm has been

evaluated on a subset of the TILDA I image database with

various visual qualities. The algorithm is found to be superior

to the existing ones in terms of computational speed and

accuracy. Several images with different types of defects were

231

Proceedings of 2011 In ternational Conference on Signal Processing, Communication, Computing and Networking Technologies (ICSCCN 2011)

also tested using the algorithm. The false alarm rate is found to

be less than 5% even in low contrast images with multiple

defects. Our algorithm has an average accuracy of90.8%. The main attraction of the proposed method is its simplicity , accuracy and computational time. This algorithm demonstrates

its strong ability to differentiate defects from other regions in

the image. The method works pretty well even when the

input image is a low-contrast one. The experimental results

demonstrate that the proposed algorithm is fast and robust.

VII. CONCLUSION

A novel fabric defect detection and localization algorithm is

proposed in this paper based on bitplane decomposition and

mathematical morphology. The algorithm is superior to the

existing algorithms in terms of computational time and

accuracy.

ACKNOWLEDGMENT

The authors are highly indebted to Ms. Cynthia Findlay of

TILDA TECHNO-VISION Project. TILDA is a Textile­

Texture Database which was developed within the framework

of the working group Texture Analysis of the

DFG's (Deutsche Forschungsgemeinschaft) major research

programme "Automatic Visual Inspection of Technical

Objects". The images were kindly provided by the TILDA

program partners.

REFERENCES

[I] K. Srinivasan, P. H. Dastoor, P. Radhakrishnaiah, and S. Jayaraman,

"FDAS: A knowledge-based framework for analysis of defects in woven

texti Ie structures" ,J. Textile inst., pt. 1, vol. 83, no. 3, pp. 431-448, 1992

[2] J. Wang, R.A. Campbell, and R.J. Harwood, "Automated inspection of

978-1-61284-653-8/11/$26 .00 ©2011 IEEE

carpets" , in Proc. SP IE, vol. 2345, 1995, pp. 180-191.

[3] T. E Mursalin, F. Z Eishita and AR Islam, "Fabric defect inspection using Neural network and Microcontroller," Journal of Theoretical and Applied Information Technology, 2008,pp 560-570.

[4] R. Meier, "Uster Fabriscan, The Intelligent Fabric Inspection," [Online document], cited 20, Apr. 2005,

http://www.kotonline.com/engl ish �ages/ana _ ba sl iklar/uster.asp.

[5] A Serdaroglu, A Ertuzun, and A Ercil, "Defect detection in textile fabric images using wavelet transforms and independent component analysis ", Journal Pattern Recog-nition and Image Analysis 16 (2006), no. 1, 61-64.

[6] A Kumar and G.K.H. Pang, "Defect detection in textured materials

using Gabor filters", IEEE Trans. Ind. Appl., vol. 38, no.2, pp. 425-440,

2002.

[7] X. F. Zhang and R. R. Bresee, "Fabric defect detection and classification using image analysis," Textile Res. J., vol. 65, no. I, pp. 1-9, 1995.

[8] J. S. Lane and S. C. Moure, "Textile fabric inspection system," U.S.

Patent 5 774 177, 1998.

[9] E. J. Wood, "Applying Fourier and associated transforms to pattern charecterization in textiles," Textile Res. J., vol. 60, pp. 212-220, 1990.

[10] C. Chan and G. K. H. Pang, "Fabric defect detection by Fourier analysis", IEEE Trans. on Ind. Appl, vol 36, no.5,ppI267-1276 Oct 2000

[Il] S. Jayaraman., S. Esakkirajan and T. Veerakumar Digital image Processing, Tata Mc Graw Hill, New Delhi, 2009,ch 10.

[12] http://en.wikipedia.org/wiki/Mathematical morphology

[13] J . Serra, Image analysis and mathematical morphology, vol. 2, Academic Press, New York, 1988

[14] T. Ashok kumar ,So Priya and M.G. Mini"Optic disc localization in ocular fundus images(Accepted for publication)," Proc. of iCVCi International Conference., India 2011

[15] T. Ashok kumar ,So Priya and M.G. Mini"Fast Optic disc localization and Extraction in retinal fundus images Using Bitplane Decomposition and Mathematical Morphology," CffT International Journal., India,March 2011

232