الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 116 |  |  |
| | | from | 116 | | |
Abstract
The use of artificial intelligence has widespread in textile industry over the last decades. Many studies have been dedicated to investigating the application of artificial intelligence techniques and their advantages in the various fields of the industry. One of these fields is quality assessment for the manufactured textile products, starting from raw materials to the final products. Yarn quality influences the subsequent products quality. Therefore, testing of yarn features is an essential requirement that can be tedious, time consuming, and demands trained labor and expensive appliance. Employing artificial intelligence technologies can lead to more objective systems and higher product specifications that meet the demands of the manufacturing and end user. This research is an attempt to model several yarn properties using a low-cost system that employs machine vision and learning. Work was carried out on an Egyptian spinning mills where samples were collected and tested. Sample images were captured and treated using image processing technique. Subsequently Three neural network models were developed to estimate different yarns’ properties: for cotton ring spun yarns, blended ring spun yarns and compact yarns. Each model contains three modules for the assessment of: first, yarn tenacity and elongation. Second and third modules evaluate yarn CVm% and Imperfections respectively By utilizing image filtering and a multi-layer artificial neural network, better results were acquired for the various yarn parameters. Therefore, yarn properties modelling can be carried out successfully. It was also proved that the system can be applied for several yarn types and different tested properties.