FabVis-Preprocess

Setup

Clone the repository

git clone https://github.com/IntellisenseLab/FabVis-RD-Preprocess.git

Create a folder to hold the base dataset

mkdir dataset

Optionally download existing data set from here and extract.

Dataset labelling

For installing up dataset labelling software run,

pip3 install labelImg

For starting dataset labelling software run,

cd dataset
labelImg

Dataset processing

For starting the dataset preprocessing process run,

python3 preprocess.py

For checking the dataset distribution after preprocessing run,

python3 checkStatistics.py

For checking the dataset for unannotated images run,

python3 checkDataset.py

Statistics

Annotated Defects

Defect	Count
Slubs	44
Barre	7
Thick Yarn	0
Foreign Yarn	3
Missing Line	64
Holes	68
Knots	0
Misknit	8
Dye Spot	12
Crease line/Crush Mark	423
Stains/Dirty	380
Stop marks	0
Snagging	34
Laddering	26

Dataset Distribution

Unannotated Defects

Defect	Image Number
Abrasion	1
Dead Cotton	88, 89, 90