Abstract:
The purpose of this project is to develop an off-line yam defect detection system by a
computerized system based on image processing software. After the yam is produced, it is
wound onto the bobbin package, different yarn bobbins are sampled and sent to the quality
inspection department to test for yam evenness. Yam defect detection is an important index of
quality control in textiles since the unevenness of yams increases the end breakage rate during
post spinning which will ultimately reduce the productivity. In addition, yam unevenness/defects
affect the quality of appearance of textiles.
Methods of yam defects detection include; Human Visualisation, Gravimetric method, and
Electronic capacitive testers (Uster Tester). The Gravimetric method is rarely adopted in general
tests because of large computation, slow and laborious. The Human Visualisation is associated
with drawbacks such as; results are subjective in nature, tiredness and boredom. The Electronic
Capacitive Tester helps get rid of the influence of man-made factors and fast so, it is applied
widely to detect yam defects. However, it has certain limitations such as high cost, testing values
are affected by the testing conditions especially the atmosphere state or humidity.
The development in computer technology using image processing (MATLAB) introduces a cost
effective yarn defect detection system with few components (i.e. computer, USB Web camera,
and Blackboard). The described method in this project represents an effective and accurate
approach to detection of yam defects. In this work, edge detection and scaling techniques are
implemented to examine the structural regularity of yam structures. To improve the efficiency
of the technique and overcome the problem of detection errors, parameters of the detection
process should be maintained constant Basing on the methods and materials employed in this
project, provide a promising stage for the development of an off-line cost effective defect
detection system.
