The Research Of Industrial Vision System Of Fabric Defects Automatic Detecting

In fabric industry, the quality of control and detection is very important. Detection of defects is the most important part. Even to nowadays, in most fabric factories, operators beside the product line watch the running fabric and decide whether defects occur or not. This artificial method has much shortcoming, such as omitting defects, confusing defects, and the worst is bad efficiency. So in order to realize automation of detecting ,recent years many experts focus on this research topic with the fast developments of industrial vision (IV), image processing and Microsoft computer technology. Computer real-time analyzes and processes image data grabbed by a camera and image card. This thesis designs a method of automatic detecting fabric defects and emphasizes the algorithm based on wavelet analysis. Then software realization of the vision system is specially discussed.First, introduce the statement of fabric defects detecting and achievement foreign and home experts have acquired in this field. Software and hardware design of this fabric defects automatic detecting industrial vision system are expatiated.Second, gives brief of wavelet algorithm and analyzes the feasibility of detecting defects through wavelet analysis. Then, tell the details of how to apply this method on defects detecting, and the whole process consists of wavelet creation and optimization, wavelet analysis and picking-up parameters to judge effects occurrence or not.Third, design and realize the software of IV system by Oriented-Object method. Also build an image database by SQL Server2000. Fabric images and some other information including defects are stored in database. Operators can query and edit these images anytime.At the end of the thesis, discuss the development direction of this IV system. The detecting method should be optimized and the system function should be extended.The fabric defect automatic defecting IV system is still on the first step. Though some achievement have gotten, also occurs much shortcoming and need more research in order to satisfy the fabric industry.