COLLIN optimises post-processing with advanced film inspection system
DI Dr. Friedrich Kastner, CEO & managing partner COLLIN Lab & Pilot Solutions GmbH, highlights how its new film inspection system allows for easier integration and powerful post-processing for optimising quality control even with variable film properties.
Key Highlights:
- Post-processing offers an enormous advantage in the development of quality criteria, even with variable film properties.
- The data collected by inspection systems can be used to analyse and optimise production processes, identifying causes of errors while maintaining a high level of quality.
- COLLIN presented a new film inspection system with a line scan camera at NPE 2024. The new system presented is a camera with a GigE Vision interface.
Film inspection systems play an important role in plastics technology as they contribute to quality assurance and process optimisation. Camera inspection systems can detect surface defects such as scratches, cracks, gels or discolouration. The inspection systems can be used both on the end product and for quality control of materials. The data collected by inspection systems can be used to analyse and optimise production processes. On the one hand, this helps to identify the causes of errors and to continuously improve processes while maintaining a high level of quality.
With over 30 years of experience in the field of line scan cameras, COLLIN Lab & Pilot Solutions uses both, area scan and line scan cameras for its film inspection systems.
Advantages of the GigE Vision Interface
COLLIN presented a new film inspection system with a line scan camera at NPE 2024. While camera systems with camera link interfaces were previously used, the new system presented is a camera with a GigE Vision interface. The new interface protocol offers several advantages over the previously used protocol. For example, standard Ethernet networking hardware can be used instead of special frame grabber cards. These standardised network cards, routers and switches are much more readily available and also more cost-effective. The data transfer rate of up to 1 Gbps is available for many industrial applications and is therefore easier to integrate into existing network infrastructures. Standardised IP protocols can be used with this technology. Additionally, the new standard also offers advantages in terms of scalability and expansion of the system to multiple camera solutions via simple networks. Meanwhile, the data transfer via CAT5e or CAT6 cables can be routed over long distances of up to 100 metres, whereas previously it was limited to approx. 10 metres. Any disadvantages in latency can be compensated for by optimised evaluation algorithms.
Powerful data analysis
The raw data read out by the camera is transferred in a standardised format. This standardised interface enables the data to be exchanged with a wide variety of applications, including web applications. This makes the system very open and independent of the programming language of the HMI client. The data architecture also results in very simple scalability for applications with several cameras. The data from the line scan camera is combined to form an image and this image is linked together with various boxes of properties to form messages. The message boxes are transferred to the HMI and can be further processed there.
These message boxes contain several property levels, which are superimposed on each other and contain all the information about the image and its properties. This makes it possible to analyse the data from the camera online or to store it in parallel as raw data memory. The raw data memory can be analysed again at any time.
The entire image analysis is divided into three sections:
- The first section concerns the camera and light settings, such as aperture, exposure time, etc. Understandably, these values cannot be changed retrospectively and determine the appearance and quality of the image.
- The second section contains the evaluation algorithms for the individual pixels and pixel gradients. Various parameter sets are available here, which can be saved in templates and used to analyse images. Saved image data can be subsequently analysed with different parameter sets. This means that you can specifically influence the image analysis, e.g. to react to changes in the film properties or changed lighting conditions. One example of this is that the appearance of defects in transparent films is different from those in slightly translucent or opaque films.
- The third section contains all the necessary models for classifying the defects. These include defect sizes, aspect ratios, thresholds, brightness, etc. The 14 error criteria and 10 error classes of the previous evaluation software have been retained. These parameters can also be subsequently changed and thus, post-processing can be carried out with saved raw data.
Post-processing therefore offers an enormous advantage in the development of quality criteria, even with variable film properties. One example of this is the quality control of films made from recycled plastics. Despite the possibility of subsequent optimisation, there is no need to forego the rapid informative value of an online measurement with known evaluation criteria.
In summary, COLLIN uses the latest camera and data transmission technology with its new film inspection system and builds this on proven know-how from over 30 years of experience.
The entire evaluation software was developed and programmed in-house at COLLIN. A real highlight is the post-processing function, which enables the subsequent optimisation of evaluation parameters using the raw data stored in the camera. The 4k camera remains the standard application for quality assurance in the technical centre, but the inspection system is also able to process data from 8k and 16k cameras.
Finally, it should also be mentioned that a new user interface was also developed as part of the new development, which was designed in the style of the COLLIN machine controls in a modern and very intuitive way.