Investigating surface properties of fibers and yarns by image processing and statistical analysis techniques

Natural fibers are usually described in terms of lengths and diameters or fineness, respectively. These parameters can be determined by optical examination and image-processing technologies. Methods to find and evaluate bifurcations, however, would not only be helpful for developers and users of automatized fiber examination equipment but also interesting in examining the influence of side branches in fibers on the mechanical properties of yarns or fiber-reinforced plastic composites. Similarly, evaluating the hairiness of textile fabrics is much more complicated than measuring this value in a yarn. This chapter thus aims at defining and describing a new statistical method of processing fiber, yarn, or fabric pictures after pretreatment with simple image-processing. The method measures the diffusive fractal dimension, or its inverse value known as the Hurst exponent, of the image to classify fibers, yarns, or fabrics. The chapter gives an overview of the technology and depicts its advantages and limits.