WebInspector AI

WebInspector® systems for web inspection not only ensure production efficiency and product quality, but also efficiency of the entire downstream converting process – from the winder to converting to complaint handling. The unique image processing system based on neural networks reliably detects and documents all defects in the paper web.

YOUR BENEFITS

• Reliable defect segmentation using neural networks – resistant to process variations (lighting, grade, etc.).
• Distinct classification of typical defect classes using AI models (e.g. streaks, wrinkles, shreds, halos, …).
• Fastest start-up due to pre-trained AI model.
• Line scan or matrix cameras – the right technology for every application.
• Tracking the root cause of defects with TotalVision.

WebInspector® systems combine defect detection and quality analysis in one system. The unique image processing system based on neural networks reliably detects and documents all defects in the paper web. One of the key advantages of neural networks is their ability to analyze large amounts of image data in a very short time and thus detect defects or deviations in real time. In order to continuously improve results, neural networks can also be trained using machine learning methods. Their use in image analysis for paper production brings decisive advantages compared to conventional methods:

• Increased accuracy.
• Faster and more efficient defect detection and segmentation.
• Improved adaptability to new data patterns.

Unlike conventional systems, WebInspector® systems
can also process HDR (high dynamic range) images. With these, larger amounts of information are stored in a single image. The larger contrast range of HDR images improves the visibility of details in difficult lighting conditions and favors defect segmentation. In this way, defects in the paper web can be detected at an early stage and the efficiency in data analysis can be massively increased.

In addition to defect detection, WebInspector® systems
can also be used for quality analysis of the produced
product:

Online DirtInspection
Online measurement of dirt in paper. Meets the requirements of TAPPI T437 “Dirt in Paper and Paperboard“ and T213 “Dirt in Pulp“.

Online formation measurement
Formation is one of the most important indicators of paper quality and homogeneity. With online formation measurement, the structure of the paper can be measured during production. Furthermore, floc size, floc
orientation and distribution of the flocs provide information about the quality of the dewatering in the wire section.