The existing paper is concerned with the improvement of the traditional cotton yarn appearance grading system with the help of computerized MATLAB image processing and fuzzy logic system. The complete system of this research comprises of yarn winder blackboard, digital camera for taking sample image, computer for processing and analyzing the sample quality by using MATLAB graphical user toolbox to facilitate the investment of algorithm during image processing and analyzing of yarn quality. Since in textile industry, mostly the cotton yarn appearance grading evaluation is principally based on labor-intensive assessment. This research used to solve the inherent limitations of the human visual inspection. The MATLAB image processing in graphical user toolbox integrates the image acquisition, digital image extraction, and yarn quality classification based on the sample yarn count and the ASTM (D2255) standard of yarn image of number of pixels. Totally this paper proposes the new method for cotton yarn appearance grading system by using matlab image processing methods and fuzzy logic systems to modify and improve the existing manual appearance grading system.