Abstract

The automatic segmentation of flaws in woven fabrics is achieved by applying Fourier analysis to the image of the sample under inspection, without considering any reference image. No prior information about the fabric structure or the defect is required. The algorithm is based on the structural feature extraction of the weave repeat from the Fourier transform of the sample image. These features are used to define a set of multiresolution bandpass filters, adapted to the fabric structure, that operate in the Fourier domain. Inverse Fourier transformation, binarization, and merging of the information obtained at different scales lead to the output image that contains flaws segmented from the fabric background. The whole process is fully automatic and can be implemented either optically or electronically. Experimental results are presented and discussed for a variety of fabrics and defects.

© 2007 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |

  1. W. J. Jasper and H. Potlapalli, "Image analysis of mispicks in woven fabric," Text. Res. J. 65, 683-692 (1995).
    [CrossRef]
  2. Y. F. Zhang and R. R. Bresee, "Fabric defect detection and classification using image analysis," Text. Res. J. 65, 1-9 (1995).
    [CrossRef]
  3. C. Ciamberlini, F. Francini, G. Longobardi, P. Sansoni, and B. Tiribilli, "Defect detection in textured materials by optical filtering with structured detectors and self-adaptable masks," Opt. Eng. 35, 838-844 (1996).
    [CrossRef]
  4. W. J. Jasper, S. J. Garnier, and H. Potlapalli, "Texture characterization and defect detection using adaptive wavelets," Opt. Eng. 35, 3140-3149 (1996).
    [CrossRef]
  5. L. M. Hoffer, F. Francini, B. Tiribilli, and G. Longobardi, "Neural networks for the optical recognition of defects in cloth," Opt. Eng. 35, 3183-3190 (1996).
    [CrossRef]
  6. M. S. Millán and J. Escofet, "Fourier-domain-based angular correlation for quasi-periodic pattern recognition. Applications to web inspection," Appl. Opt. 35, 6253-6260 (1996).
    [CrossRef] [PubMed]
  7. J. Escofet, R. Navarro, M. S. Millán, and J. Pladellorens, "Detection of local defects in textile webs using Gabor filters," Opt. Eng. 37, 2297-2307 (1998).
    [CrossRef]
  8. D. M. Tsai and C. Y. Hsieh, "Automated surface inspection for directional textures," Image Vision Comput. 18, 49-62 (1999).
    [CrossRef]
  9. M. C. Hu and I. S. Tsai, "Fabric inspection based on best wavelet packet bases," Text. Res. J. 70, 662-670 (2000).
    [CrossRef]
  10. A. Conci and C. B. Proença, "A computer vision approach for textile inspection," Text. Res. J. 70, 347-350 (2000).
    [CrossRef]
  11. T. J. Kang, S. H. Choi, S. M. Kim, and K. W. Oh, "Automatic structure analysis and objective evaluation of woven fabric using image analysis," Text. Res. J. 71, 261-270 (2001).
    [CrossRef]
  12. D. Chetverikov and A. Hanbury, "Finding defects in texture using regularity and local orientation," Pattern Recogn. 35, 2165-2180 (2002).
    [CrossRef]
  13. A. Bodnarova, M. Bennamoun, and S. Latham, "Optimal Gabor filters for textile flaw detection," Pattern Recogn. 35, 2973-2991 (2002).
    [CrossRef]
  14. A. Kumar, "Neural network based detection of local textile defects," Pattern Recogn. 36, 1645-1659 (2003).
    [CrossRef]
  15. H. Y. T. Ngan and G. K. H. Pang, "Novel method for patterned fabric inspection using Bollinger bands," Opt. Eng. 45, 087202 (2006).
    [CrossRef]
  16. D. R. Rohrmus, "Invariant and adaptive geometrical texture features for defect detection and classification," Pattern Recogn. 38, 1546-1556 (2005).
    [CrossRef]
  17. D. M. Tsai and T. Y. Huang, "Automated surface inspection for statistical textures," Image Vision Comput. 21, 307-323 (2003).
    [CrossRef]
  18. J. Escofet, M. S. Millán, and M. Ralló, "Modeling of woven fabric structures based on Fourier image analysis," Appl. Opt. 40, 6170-6176 (2001).
    [CrossRef]
  19. M. Ralló, J. Escofet, and M. S. Millán, "Weave repeat identification by Fourier analysis of fabric images," Appl. Opt. 42, 3361-3372 (2003).
    [CrossRef] [PubMed]
  20. R. Navarro and A. Tabernero, "Gaussian wavelet transform: two alternative fast implementations for images," Multidimens. Syst. Signal Process 2, 421-436 (1991).
    [CrossRef]
  21. D. Dutton, M. P. Givens, and R. E. Hopkins, "Some demonstration experiments in optics using a gas laser," Am. J. Phys. 32, 355-361 (1964).
    [CrossRef]
  22. J. Goodman, Introduction to Fourier Optics, 2nd ed. (McGraw-Hill, 1996).

2006 (1)

H. Y. T. Ngan and G. K. H. Pang, "Novel method for patterned fabric inspection using Bollinger bands," Opt. Eng. 45, 087202 (2006).
[CrossRef]

2005 (1)

D. R. Rohrmus, "Invariant and adaptive geometrical texture features for defect detection and classification," Pattern Recogn. 38, 1546-1556 (2005).
[CrossRef]

2003 (3)

D. M. Tsai and T. Y. Huang, "Automated surface inspection for statistical textures," Image Vision Comput. 21, 307-323 (2003).
[CrossRef]

A. Kumar, "Neural network based detection of local textile defects," Pattern Recogn. 36, 1645-1659 (2003).
[CrossRef]

M. Ralló, J. Escofet, and M. S. Millán, "Weave repeat identification by Fourier analysis of fabric images," Appl. Opt. 42, 3361-3372 (2003).
[CrossRef] [PubMed]

2002 (2)

D. Chetverikov and A. Hanbury, "Finding defects in texture using regularity and local orientation," Pattern Recogn. 35, 2165-2180 (2002).
[CrossRef]

A. Bodnarova, M. Bennamoun, and S. Latham, "Optimal Gabor filters for textile flaw detection," Pattern Recogn. 35, 2973-2991 (2002).
[CrossRef]

2001 (2)

J. Escofet, M. S. Millán, and M. Ralló, "Modeling of woven fabric structures based on Fourier image analysis," Appl. Opt. 40, 6170-6176 (2001).
[CrossRef]

T. J. Kang, S. H. Choi, S. M. Kim, and K. W. Oh, "Automatic structure analysis and objective evaluation of woven fabric using image analysis," Text. Res. J. 71, 261-270 (2001).
[CrossRef]

2000 (2)

M. C. Hu and I. S. Tsai, "Fabric inspection based on best wavelet packet bases," Text. Res. J. 70, 662-670 (2000).
[CrossRef]

A. Conci and C. B. Proença, "A computer vision approach for textile inspection," Text. Res. J. 70, 347-350 (2000).
[CrossRef]

1999 (1)

D. M. Tsai and C. Y. Hsieh, "Automated surface inspection for directional textures," Image Vision Comput. 18, 49-62 (1999).
[CrossRef]

1998 (1)

J. Escofet, R. Navarro, M. S. Millán, and J. Pladellorens, "Detection of local defects in textile webs using Gabor filters," Opt. Eng. 37, 2297-2307 (1998).
[CrossRef]

1996 (4)

C. Ciamberlini, F. Francini, G. Longobardi, P. Sansoni, and B. Tiribilli, "Defect detection in textured materials by optical filtering with structured detectors and self-adaptable masks," Opt. Eng. 35, 838-844 (1996).
[CrossRef]

W. J. Jasper, S. J. Garnier, and H. Potlapalli, "Texture characterization and defect detection using adaptive wavelets," Opt. Eng. 35, 3140-3149 (1996).
[CrossRef]

L. M. Hoffer, F. Francini, B. Tiribilli, and G. Longobardi, "Neural networks for the optical recognition of defects in cloth," Opt. Eng. 35, 3183-3190 (1996).
[CrossRef]

M. S. Millán and J. Escofet, "Fourier-domain-based angular correlation for quasi-periodic pattern recognition. Applications to web inspection," Appl. Opt. 35, 6253-6260 (1996).
[CrossRef] [PubMed]

1995 (2)

W. J. Jasper and H. Potlapalli, "Image analysis of mispicks in woven fabric," Text. Res. J. 65, 683-692 (1995).
[CrossRef]

Y. F. Zhang and R. R. Bresee, "Fabric defect detection and classification using image analysis," Text. Res. J. 65, 1-9 (1995).
[CrossRef]

1991 (1)

R. Navarro and A. Tabernero, "Gaussian wavelet transform: two alternative fast implementations for images," Multidimens. Syst. Signal Process 2, 421-436 (1991).
[CrossRef]

1964 (1)

D. Dutton, M. P. Givens, and R. E. Hopkins, "Some demonstration experiments in optics using a gas laser," Am. J. Phys. 32, 355-361 (1964).
[CrossRef]

Am. J. Phys. (1)

D. Dutton, M. P. Givens, and R. E. Hopkins, "Some demonstration experiments in optics using a gas laser," Am. J. Phys. 32, 355-361 (1964).
[CrossRef]

Appl. Opt. (3)

Image Vision Comput. (2)

D. M. Tsai and C. Y. Hsieh, "Automated surface inspection for directional textures," Image Vision Comput. 18, 49-62 (1999).
[CrossRef]

D. M. Tsai and T. Y. Huang, "Automated surface inspection for statistical textures," Image Vision Comput. 21, 307-323 (2003).
[CrossRef]

Multidimens. Syst. Signal Process (1)

R. Navarro and A. Tabernero, "Gaussian wavelet transform: two alternative fast implementations for images," Multidimens. Syst. Signal Process 2, 421-436 (1991).
[CrossRef]

Opt. Eng. (5)

H. Y. T. Ngan and G. K. H. Pang, "Novel method for patterned fabric inspection using Bollinger bands," Opt. Eng. 45, 087202 (2006).
[CrossRef]

J. Escofet, R. Navarro, M. S. Millán, and J. Pladellorens, "Detection of local defects in textile webs using Gabor filters," Opt. Eng. 37, 2297-2307 (1998).
[CrossRef]

C. Ciamberlini, F. Francini, G. Longobardi, P. Sansoni, and B. Tiribilli, "Defect detection in textured materials by optical filtering with structured detectors and self-adaptable masks," Opt. Eng. 35, 838-844 (1996).
[CrossRef]

W. J. Jasper, S. J. Garnier, and H. Potlapalli, "Texture characterization and defect detection using adaptive wavelets," Opt. Eng. 35, 3140-3149 (1996).
[CrossRef]

L. M. Hoffer, F. Francini, B. Tiribilli, and G. Longobardi, "Neural networks for the optical recognition of defects in cloth," Opt. Eng. 35, 3183-3190 (1996).
[CrossRef]

Pattern Recogn. (4)

D. Chetverikov and A. Hanbury, "Finding defects in texture using regularity and local orientation," Pattern Recogn. 35, 2165-2180 (2002).
[CrossRef]

A. Bodnarova, M. Bennamoun, and S. Latham, "Optimal Gabor filters for textile flaw detection," Pattern Recogn. 35, 2973-2991 (2002).
[CrossRef]

A. Kumar, "Neural network based detection of local textile defects," Pattern Recogn. 36, 1645-1659 (2003).
[CrossRef]

D. R. Rohrmus, "Invariant and adaptive geometrical texture features for defect detection and classification," Pattern Recogn. 38, 1546-1556 (2005).
[CrossRef]

Text. Res. J. (5)

W. J. Jasper and H. Potlapalli, "Image analysis of mispicks in woven fabric," Text. Res. J. 65, 683-692 (1995).
[CrossRef]

Y. F. Zhang and R. R. Bresee, "Fabric defect detection and classification using image analysis," Text. Res. J. 65, 1-9 (1995).
[CrossRef]

M. C. Hu and I. S. Tsai, "Fabric inspection based on best wavelet packet bases," Text. Res. J. 70, 662-670 (2000).
[CrossRef]

A. Conci and C. B. Proença, "A computer vision approach for textile inspection," Text. Res. J. 70, 347-350 (2000).
[CrossRef]

T. J. Kang, S. H. Choi, S. M. Kim, and K. W. Oh, "Automatic structure analysis and objective evaluation of woven fabric using image analysis," Text. Res. J. 71, 261-270 (2001).
[CrossRef]

Other (1)

J. Goodman, Introduction to Fourier Optics, 2nd ed. (McGraw-Hill, 1996).

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (14)

Fig. 1
Fig. 1

Model of a flawed woven fabric in the spatial domain.

Fig. 2
Fig. 2

Model of the flawed woven fabric of Fig. 1 in the spatial frequency domain (magnitude). Energy is not normalized in this figure. Instead, each image in the spatial frequency domain is independently presented using logarithmic gray scale. In this way, the relationship between its general appearance here and their content in the spatial domain (Fig. 1) becomes easier to notice.

Fig. 3
Fig. 3

Magnified central part of F ^ ( u , v ) (magnitude) presented using logarithmic scale (Fig. 2). The vector basis { s , t } generates the peak positions of H m ( u , v ) . The vector basis { s , t } generates the peak positions of H ( u , v ) represented by the intersection points of the dotted grid.

Fig. 4
Fig. 4

(a) Basic shape selected to build the spatial filters in the frequency domain. (b) Design of the set of multiscale concentric spatial filters, with the smallest one just contained in the rectangle ABCD. Background: F ^ ( u , v ) is shown in magnitude and using logarithmic gray scale.

Fig. 5
Fig. 5

Product of each spatial filter of the set by F ^ ( u , v ) (shown in magnitude and with logarithmic gray scale).

Fig. 6
Fig. 6

(Reverse contrast) Spatially filtered images obtained by applying the inverse Fourier transform to the filtered spectrum elements of Fig. 5.

Fig. 7
Fig. 7

Left-truncated zero-mean normal distribution f σ ( x ) used to model the histogram of the spatially filtered images of a defect-free fabric sample. As an example, the defect-free fabric of Fig. 12(a) has been considered for this figure. The histogram corresponds to the image filtered by the smallest spatial filter of the set.

Fig. 8
Fig. 8

(a)–(d) Binary filtered images obtained from Fig. 6. (e) Output image obtained by applying the or-logical operator to (a)–(d).

Fig. 9
Fig. 9

Scheme of the processor for an optical implementation:λ, wavelength; WP, half-wave plate; {SLM1, SLM2}, spatial light modulators; {L1, L2}, lenses; CCD sensor; charge coupled device sensor.

Fig. 10
Fig. 10

(a) Flawed sample of twill fabric with warp and weft threads of different colors; (b) magnitude of the Fourier transform of (a) in logarithmic scale.

Fig. 11
Fig. 11

(a) Set of four bandpass filters adapted to the woven fabric under inspection. From left to right:low bandpass to high bandpass masks. (b) (Reverse contrast) Filtered images in the four channels. (c) Binary versions obtained from the histogram analysis of images in (b). (d) Output image obtained by merging the images of (c) with the or-logical operator.

Fig. 12
Fig. 12

Examples of twill fabric sample image (top) and output image (bottom):(a) a defect-free sample; (b) … (e) different sorts of flaws.

Fig. 13
Fig. 13

Top:Examples of flawed fabrics of different structure, thread colors, and sorts of defects. Bottom: Output images with defects segmented.

Fig. 14
Fig. 14

(a) Output image obtained from the input fabric sample image of Fig. 10(a) when the set of filters are limited by ellipses E i ( u , v ) that are mirror images (i.e., the reflection about the horizontal) of the ellipses E i ( u , v ) of the filters originally designed by the algorithm [Fig. 11(a)]. (b) Output image obtained when the set of filters designed by the algorithm [Fig. 11(a)] are scaled by a 0.7 factor.

Equations (5)

Equations on this page are rendered with MathJax. Learn more.

f ^ ( x , y ) = f ( x , y ) + d ( x , y ) .
f ^ ( x , y ) = f ( x , y ) + d ( x , y ) = b m ( x , y ) h m ( x , y ) + d ( x , y ) = b ( x , y ) h ( x , y ) + d ( x , y ) .
F ^ ( u , v ) = F ( u , v ) + D ( u , v ) = B m ( u , v ) H m ( u , v ) + D ( u , v ) = B ( u , v ) H ( u , v ) + D ( u , v ) ,
g σ ( x ) = 1 2 π σ e 1 2 ( x σ ) 2 , < x < ,
f σ ( x ) = { 0 x < 0 2 g σ ( x ) x 0 .

Metrics