Design for Application of Grey Relational Analysis to On-Line Cloth Defect Inspecting System
| Autor: | Su Te-Li, 蘇德利 |
|---|---|
| Rok vydání: | 2002 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 90 The purpose of this thesis is to develop a set of on-line cloth-defect inspecting system, so as to seek the feature values of the cloth defect image through the use of wavelet transformations and gray-level co-concurrence matrix. As a follow-up, the cloth defects will be classified by applying the grey relational grade analysis of the grey theory. The main task is to identify the classification research for the spider weave, holes, oil stains, broken warps and broken wefts that are commonly seen. During the grey relational grade analysis, the relativity of factors to be analyzed and researched can be located from the random factor serial through data processing in the selected feature indicators. By this way, the one having the highest relativity will be the judged defect. In the meantime, the fuzzy theory will be used to design a fuzzy controller in maintaining consistent tension on the cloth surface so as to elevate the identification rate of cloth defects and enhance the cloth inspection efficiency. In this way, the system may be provided with higher performance efficiency. In the software design, the Visual C++ and MFC (Microsoft Foundation Class) will be used as the software developing tool while the Microsoft Windows operating system will be used for the platform development. By using the load cell in conjunction with A/D, D/A signal conversion card and two units of frequency changers, the cloth tension control module will directly adjust the feeding and rolling rollers so as to control the cloth tension and its conveying speed. The test result indicated a very satisfactory 92.5% of total identification rate when using 40 sheets of defective samples for the on-line inspection. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|