Development of a machine vision system for yarn bobbin inspection

作者：Celik H Ibrahim^*

来源：Industria Textila, 2016, 67(5): 292-296.

• Machine vision
• fiber mixture
• yarn bobbin inspection
• image processing
• fault detection

When the different raw materials are blended in any step of the yarn production, it causes fiber mixture fault on the yarn bobbin. Since each fiber has different dye affinity characteristics, this fault causes off-quality fabric production. Despite the new automation systems used in yarn manufacturing mills, the fiber mixture fault is still inspected by human eye. This process takes long time and the evaluation is made subjectively. Furthermore, many bobbins with fiber mixture fault may be escaped from the worker notice. In this study, a prototype of vision inspection system was developed to detect the yarn fiber mixture fault by using image processing method. It was aimed that the fiber mixture faults of the yarn bobbins can be detected automatically and the fault evaluation can be made objectively. The fault detection algorithm was based on Wiener filtering, Gaussian filtering and morphological operations. Six yarn bobbins; three of them with fiber mixture fault and other three having no mixture fault as control group were detected successfully and the fault areas were labeled.

• 出版日期2016