Background:
Development of Pills (small fuzz balls) collected on the fabric surface not only gives an ugly appearance but also causes premature ware of the fabric. Moreover, some of the properties of synthetic fiber results in greater pilling propensity Fabric resistance to pilling is generally tested using specific machines. A sample of the original fabric is fixed in them and wear is simulated by the action of abrasive materials. The abraded sample is then compared with photographs provided in standards like ASTM, BIS, JIS, AATCC, etc. to determine the degree of pilling. This method has various drawbacks like non-availability of standard photographs for all the fabrics, inaccuracy due to subjective measurement, non availability of seasoned experts, non compatibility of woven and knitted fabric, etc. in order to overcome these drawbacks, an Image Processing based pilling testing system has been developed. This system processes and analyses the image of pilled fabric to determine various pilling parameters as per standards like ASTM, BIS, JIS, etc.

Description:
The procedure for determining pilling parameters is schematically explained in the enclosed Figure In step 2.1 Image of the sample fabric is grabbed using a camera and freezed in RGB image containing all data including the colour information. The RGB image is subsequently converted to Gray Image by removing the colour information. Fast Fourier Transformation (step 2.2) further converts the gray image to a frequency spectrum. Corresponding power spectrum is generated for this frequency spectrum domain (step 2.3) for showing the contributions of each frequency to the frequency spectrum function. Such contributions in the form of peaks related to the fabric design. These peaks are identified (step 2.4) and filtered by replacing them with zeros if found to be greater than a specified threshold (step 2.5) . Replacing the peaks by zeros removes all image information, which belong to weave pattern and designs formed due to colours of warp and weft. The image of the filtered spectrum is then reconstructed using Inverse Fourier transformation (step 2.6) This image is converted to a binary format for further processing using the thresholding technique (step 2.7) Using this binary image, various pilling parameters of the fabric sample such as number of pills, total and mean area of pills and pills per unit area are determined (step 2.8).

Special Features:
Can be used for all types of fabric and weave patterns,Eliminates subjective assessments, More accurate, User friendly, Adopted universal standard testing procedure.

Prospective users
Fabric manufacturers, testing agencies and exporters

Keywords:
Pilling, Fibre, Fabric Image

Type of Technology:
Testing Device with Software

Status of IPR Protection:
Indian Patent Application No 65/DEL/2003 dated 24.1.2003 entitled "Objective Measurement of Pilling using Image Processing.

Status of Development:
Technology licensed for commercialization

Glossary:
ASTM: American Society of Testing Materials; BIS: Bureau of Indian Standards; JIS: Japanese Industrial Standards; AATCC: Association of American Textile Chemists and Colorists; RGB: Red,Green,Blue

Services offered by Technology Transfer:
Know how manual including quality control procedures
Detailed design
Manufacturing drawings
Consultancy services

For further informatiom, Please contact directly the faculty member/inventor under intimation to Managing Director, FITT, sengupta@fitt.iitd.ernet.in.