PAPER WEB IMAGING WITH ADVANCED DEFECT CLASSIFICATION

摘要

Paper web inspection with defect detection and classification helps mills to reduce production disturbances, removerncauses of defects, and deliver products of correct quality to the customers. However, the capabilities of conventionalrninspection systems are limited. The actual severities of spots or holes cannot always be identified precisely enough.rnModern imaging systems open up new possibilities. They utilize CCD-camera technology to create electronic grayrnscale images of the paper. Therefore, the fine details of defects can be discerned, which makes it possible to classifyrnthe defects more precisely than before. This is accomplished by various digital image retrieving and analysisrntechniques. Neural networks can be used for classifying defects which the system has been trained to recognize. Forrnexample, differentiating slime holes from small wire holes at a re-reeler, helps to prevent paper breaks at an off-linerncoater. As a result these modern systems can help the papermaker to enhance mill productivity.