Many methods of defect detecting on plain or patterned textile products show that the application of image processing techniques, identification, learning machine brings good results and has high potential into practical applications. This paper aims at the research, design and development of an automated fabric defect marking system. The system receives the information including the position, the level of seriousity and the type of fabric defects from the automated defect detection system Based on such information, it will automatically make a mark on the fabric surface at the position where the defect is identified. Besides, the system sends the information of winch velocity to the defect detection system in order to guarantee the accuracy of defects localization. This system is designed with high reliability and flexibility that satisfies various installing requirements of textile industry.