Abstract:

Aimed at the defects of yarn hairiness measurement by image process in the prior image methods, a new algorithm was proposed for analyzing hairiness and measuring hairiness length by tracking yarn hairiness skeleton. First, using 10 pixels as the step value, the dividing line of hairiness skeleton by the curves of yarn edge were plotted, and the hairiness starting points were obtained. And then, the upper 5-neighborhood points or lower 5-neighborhood points were judged in the extension direction of yarn hairiness, and the new hairiness path points were obtained. The all of path points were judged until no hairiness path points exist, all of which were recorded in sequence to generate a hairiness path. and the solutions of multi-hairiness path points and crossed hairiness were peoposed. Finally, the pixel of two hairiness path points was calculated according to the distance between two points, and the hairiness measurement length was obtained. The results of tracking measurement and fixed segment measurement of long hairiness show that the measurement length can be increased by 24.3%∽666% by the new method of yarn hairiness measurement using the hairiness tracking, and the measurement results are more accurate.

Key words: yarn hairiness, hairiness skeleton, hairiness length, tracking algorithm