Automatic Determination of Slub Yarn Geometrical Parameters Based on an Amended Similarity-based Clustering Method

Abstract

An automatic method for characterizing the parameters of ring slub yarn is presented. Firstly, size parameters of slub yarn were measured by using a capacitance-type sensor, and data was expressed as voltage signals through a data acquisition card (DAQ) installed in a PC. Then the voltage signals of the slub yarn were transferred to a two-dimensional (2D) image. Consequently, the repetition pattern of the slub yarn was determined by analyzing the 2D image using cluster analyses with an amended similarity-based clustering method. The samples were produced using a practical slub yarn production system. Results were validated by comparison of measurement parameters with set parameters.