Inline process control with chemical imaging

Inline process control with chemical imaging

HSI technology needs powerful hardware, fast software and the expertise to put it all together into a reliable overall system. We offer all this so that you can monitor the chemical homogeneity of your products inline.

Hyperspectral imaging (HSI) combines the advantages of optical spectroscopy and spatially resolved imaging

RGB colour cameras are well suited for checking external quality characteristics such as size, shape and colour. However, the plastics industry needs to check other quality parameters such as chemical composition and the distribution of additives. Spectroscopic methods can help with this. They provide information about the chemical composition and physical properties of a product. HSI can show what is invisible to the human eye.

Hyperspectral imaging includes numerous spectral channels covering different wavelengths, from ultraviolet to long-wave infrared. This technology uses the reflectivity of a material to measure and evaluate certain chemical properties with spatial resolution, also known as chemical imaging.

product image: HSI NIR camera: RedEye from Inno-Spec

RedEye: HSI NIR camera from Inno-Spec. One of several cameras we can use and select to suit each customer project.

Use of hyperspectral imaging offers potential for new applications

Hyperspectral systems provide spectral data per object pixel and not just monochrome or colour values. Depending on the wavelength range and spectroscopic processing, this enables highly accurate colour coordinates, chemical material properties and layer thickness information. The information output is more complex, but also offers greater variety and selectivity for applications.

Objects that look the same may have different light spectra under broadband illumination due to their chemical properties. Hyperspectral systems have the unique ability to distinguish between these objects, which is not possible with other imaging solutions.

Sorting and recycling plastics is an important application for hyperspectral systems. HSI can automatically separate plastic parts such as polyethylene and polypropylene. With the help of colour sorting and the differentiation of the molecular composition, the materials can be distinguished. This improves the result quality of the sorting process.

Challenges and barriers

Currently, hyperspectral imaging is not a widely used field. This is due to the high cost of hyperspectral technology, which is a major barrier to entry. In addition, the technique is complex and requires expertise in spectroscopy. Despite some promising applications, it remains an exotic discipline in imaging.

Hyperspectral camera systems produce a large amount of data that must be processed correctly. Advanced software algorithms are required to extract spectral information from the noise. For speedy processing, the processors must have sufficient power.

One of the challenges is hardware, as hyperspectral imaging is not compatible with LED lighting commonly used in imaging. Halogen lamps that emit a broad wavelength spectrum are needed. Suitable illuminators are still required.

The need for powerful hyperspectral software, reliable spectral data and experience are reasons why this new technology is spreading rather slowly

Our expertise – customised inline systems with real-time data processing

Our company specialises in developing industrial system solutions that are tailored to specific needs and can be seamlessly integrated into existing plants. Working with our partners, we solve the difficulties and provide the necessary hardware, powerful software and expertise to create effective and reliable automated optical inspection systems.

Our hyperspectral camera systems provide precise imaging and data acquisition in real time. Together with our partners, we offer complete systems and can draw on the latest lighting technologies and cameras from various manufacturers. Our technology partner’s software platform is designed for real-time data processing. Among other things, we can develop chemometric data models for your application. Data access is possible via various interfaces such as C++, .NET and Python if a stand-alone solution is not desired. The software can be integrated into higher-level systems to obtain even more powerful functions.

HSI systems are complex and require considerable investment. Contact us so that we can clarify together whether HSI technology is suitable for your task.