Feature extraction of scale pattern in animal textile fibers

Abstract

Animal fibers have variety in their longitudinal sections. The main difference of the longitudinal section of these fibers is the geometry and dimension of scale patterns, which with attention to one of the physical fiber detection, one can inspect longitudinal section of fiber images, which determine the fiber characteristics. The method of the determination is image processing which extracting the closed contours and then the chain code algorithm inspected characteristic of fibers. Regularity or irregularity of animal scale patterns can be recognized by feature extraction. As result it can be accompanied with others such as scale length, scale area, etc. for classification of fibers.