Estimation of Yarn Hairiness using Image Processing

Yarn hairiness is one of the significant quality parameter for producing quality fabrics .Fibers protruding out from the main body of the yarn is called hairiness. The upshot of yarn hairiness on the textile operations such as spinning, knitting, etc, led to the measurement of hairiness. This paper includes the analysis of cotton yarn [17], Lenin yarn and bamboo yarn. The hairiness analysis without charging leads to the problems of less accuracy in the hair count and the area of measurement. In the proposed system the videos of charged yarn is taken as input and frames are extracted. The yarn core extraction is performed using graph cut segmentation. Then by masking, the yarn hairs are extracted. The labeling of the yarn hair is done for the analysis of hair nature and for the parameter measurements like hair count, types of hairs and the length of hair. Through this analysis the yarn quality is estimated .Finally the other types of yarns can also be analyzed.