AITEX FABRIC IMAGE DATABASE

by Javier Miró

The textile fabric database consists of 245 images of 7 different fabrics. There are 140 defect-free images, 20 for each type of  fabric. With different types of defects, there are 105 images.

Images have a size of 4096×256 pixels. Defective images have been denominated as follows: nnnn_ddd_ff.png, where nnnn  is the image number, ddd is the defect code, and ff is the fabric code.

There is a mask of defect, denominated as: nnnn_ddd_ff_mask.png, where white pixels represent the defect area of the defective image.

Defect free images have been denominated as follows:

nnnn_000_ff.png, where defect code has been replaced by 0000 code.

Defect Codes used in the database:

Broken end

2

Broken yarn

6

Broken pick

10

Weft curling

16

Fuzzyball

19

Cut selvage

22

Crease

23

Warp ball

25

Knots

27

Contamination

29

Nep

30

Weft crack

36

Defect free images 15 and 08 of fabrics types 00 and 03 (only an area of 256×256 is shown)

 

0015_000_00.png

0008_000_03.png

Defect 19 on fabric 02 and its mask (only an area of 256×256 is shown)

  

0030_019_02

0030_019_02_mask

Links: (this links will be available after the publication of paper)

Please, if you use this images and the rest of information in your research, we kindly ask that you reference our paper:

AFID: a  public fabric image database for defect detection.

Javier Silvestre-Blanes, Teresa Albero-Albero, Ignacio Miralles, Rubén Pérez-Llorens, Jorge Moreno

AUTEX Research Journal, No. 4, 2019

https://content.sciendo.com/view/journals/aut/ahead-of-print/article-10.2478-aut-2019-0035.xml