Abstract: In this paper, taking the pure cotton pure yarn for example, a method based on digital image processing for yarn fineness visual measurement was proposed. The yarn image captured with video microscope was treated sequentially with 2-D adaptive Wiener filtering, segmentation using 2-D maximum entropy algorithm optimized with genetic algorithm, and morphology operation. Then, the yarn diameter and fineness are calculated from the clear and complete binary image after the above operations. The test data extracted from the yarn images that are taken as carriers can describe the yarn diameter and fineness more intuitively. Furthermore，when the fineness of the sample can not be measured by standard skein method, the sufficient fineness information still can be provide by splitting a certain number of yarns from the sample and carrying out the proposed method The experimental results indicate that the method for measurement the yarn diameter and fineness based on image processing and data analysis is feasible.