Approaches to the task of searching for anomalies on the conveyor line using neural networks
When transporting the fabric along the conveyor, it is necessary to check the quality of the fabric. The authors have developed approaches to the problem of finding anomalies on the conveyor line. The developed approach uses neural networks. Tissue defect detection is a quality control process that aims to identify defects and determine their location. An important task is also to detect the location of defective areas. This process allows for the precise identification of defective areas and avoidance of them entering the finished product, which is of great importance for textile manufacturers. The ability to accurately pinpoint defect points to support a fabric quality control process is the primary goal of an automated patterned fabric defect detection and classification system. This should be achieved at the expense of good processing speed, less computational complexity, and less computation time. Thus, the designed systems require reliable and efficient algorithms for detecting defects. Although various types of defects have been mentioned in the literature, only a few of them have been mentioned in the case of the transportation of textiles along the conveyor. Therefore, the purpose of this article is to present personal experience of applying various approaches to detecting defects on the conveyor using technical vision and machine learning technologies.
Artificial intelligence and machine learning; Computer vision; Deep learning; Defect detection; Event-triggered control in conveyor line