Vision system
based on AI in the food industry

Through the use of high-resolution cameras, machine learning (ML)-based algorithms and image processing technology, vision systems offer much more than traditional visual inspection methods. Their use not only increases productivity, but also allows the elimination of defective products at early stages of the production process, thus reducing costs and wasting raw materials. In the food industry in particular, where product quality is not only a matter of economics but also consumer safety, vision systems play a key role.

One of the most advanced examples of the application of these technologies is a vision system designed for the inspection of protein casings in food production. Its detailed capabilities, benefits and implementation effects are outlined below.

Key functionalities of the vision system

The implemented vision system is an example of advanced technology that combines business continuity, flexibility of customisation and integration with other production systems. As a result, it provides a comprehensive solution for quality monitoring in the protein casing production process. Here are its key features:

1. Continuous imaging of protein casings
   The system is based on an RGB ruler camera, working in conjunction with a ruler illuminator, which enables continuous imaging of production processes. As a result, every element of the casing is precisely recorded and the image data can be analysed immediately.
2. Real-time flaw detection
   The use of algorithms based on greyscale image intensity analysis and neural networks allows the rapid and precise detection of defects such as creases, inclusions, pinholes or other structural anomalies. Real-time analysis increases efficiency and allows for immediate reactions.
3. Real-time flaw detection
   The use of algorithms based on greyscale image intensity analysis and neural networks allows the rapid and precise detection of defects such as creases, inclusions, pinholes or other structural anomalies. Real-time analysis increases efficiency and allows for immediate reactions.
4. Integration with SCADA systems
   Communication with external SCADA systems allows defect and process data to be transmitted in real time. This enables operators and control systems to make quick decisions and the data can be used to optimise processes.
5. Flexible adaptation to user requirements
   The system allows the user to adapt the defect detection parameters to the specific requirements of the process. This approach allows the system to be better tuned to specific production conditions, increasing the effectiveness of fault identification.
6. Real-time image display
   This function allows operators to monitor the process in real time, which not only increases quality control, but also allows potential problems to be identified quickly.
7. Possibility of adding new casing variants
   The system allows the user to define new casing types, which are automatically recognised and handled. This gives the company the flexibility to adapt to changing market requirements.

Results and achievements of the system

The Scanway vision system used has significantly increased the quality of production, allowing precise detection of many types of defects such as:

• High-contrast black inclusions – detected from 3 mm in size,
• White spots, dams and creases – detected from 5 mm in size.

Despite its precise detection capabilities, the system operates on a greyscale, meaning that it does not identify colour blemishes such as yellow spots. Verification of the ability to detect other colour anomalies is planned for the next stages of implementation.

Why are Scanway vision systems changing the food industry?

The introduction of advanced vision systems into the food industry brings benefits on many levels. First and foremost, it increases the quality of final products, which influences customer satisfaction and builds trust in the brand. Reducing the number of defective products early in the process reduces waste and reduces operating costs. In addition, the ability to integrate with other production systems, such as SCADA, makes vision systems part of comprehensive production automation.

The system implemented in the production of protein casings is an excellent example of how artificial intelligence and advanced algorithms can revolutionise quality control processes, increasing efficiency, minimising waste and enhancing consumer safety.

This implementation demonstrates that modern vision technology is not just an innovative addition, but a key tool in achieving business and quality goals in the food industry.