Research On Online Detection Of Fabric Defects And Positioning Of Samples Containing Defects

Modern clothing brand of clothing design, the quality of raw material, processing technology and other requirements of increasingly high, the cloth produced in the machining process of weaving defects is an important factor affecting the quality of clothing materials. Therefore, how to realize the defect in the process of garment processing automatic positioning and clear print automatically weed out, in the process of garment processing of fabric defect detection, location, identification, and with automatic discharging, automatic cutting process, has important significance to improve the automation degree of garment manufacturing. Therefore, according to the requirements of enterprise, we put forward to carry out the automation of defect recognition and clear plate positioning research work.The research object is the cloth which has been initial inspected and did a simple marker in the cloth inspecting machine. the main purpose is to identify and locate the defect marking, matching sample information and defect information in the automatic discharging system, so as to realize the extraction of samples containing defects positioning and sample information, to provide reference information to remove faults samples and samples processing replacement. The main researches of this paper are as follows:Firstly,the design of the whole system. Taking analysis the process of automatic cloth spreading in the enterprise, with machine vision technology, choose the image acquisition hard ware structure, the image acquisition module is installed in the automatic spreading machine, realize the synchronization and automatic cloth inspection process of automatic spreading. Through the simulation experiment, the image acquisition system designed in this paper can realize the accurate, real-time image acquisition, made the preparation for defect marking and location later samples containing defects.Secondly, image processing and the feature extraction of defect’ marking. Through the image defect marking of the collected several tests, using the image processing methods of the median filtering, the maximum entropy threshold segmentation, mathematical morphology, retain the defect marking information; and then using linear scanning image points the feature information extraction of defect location, to realize the recognition and classification of the defect marking.Thirdly, the defect samples in the discharge system positioning, and achieve the defect is positioned in the discharge system. First extracting the discharge pattern of coordinate information, setting the establishment of each sample. Data and then coordinate; classification positioning method by using the method of intersection in set theory and rough finish, to determine the coordinates data collection and sample defect marking the coordinate data set intersection is zero: Zero intersection, on behalf of the samples belong to the normal samples don’t overlap; zero defects, representative samples containing belongs to the last record number; and information containing defects of samples, patch cut samples and provide the basis for subsequent replacement.The study in this paper can achieve defect marking in automatic shop cloth link identification and localization, and automatic row material and automatic cutting process closely, the combination of actual operation of defect detection theory research and clothing production, through a series of experiments to test and verify, here the online detection and location method is feasible and effective to improve efficiency of garment processing.