Contaminant and foreign fiber detection in cotton using Gaussian mixture model

Abstract

Cotton is a very important material used in producing many fabric types. Contaminants from various sources need to be removed from cotton before fibers can be spun into yarn. Contaminants critically affect the quality of the yarn produced; any foreign material may result in unacceptable yarn or fabric, or even cause damage to the production machines. Automatic detection and removal of foreign fibers and contaminants in cotton is an essential technology for the modern textile industry. Various image processing and computer vision techniques have been proposed for the detection of foreign materials in cotton fibers. We describe a detection method using Gaussian mixture models and thresholding based on pixel probabilities. The proposed method gives promising results.