SWIR and RGB Sensor Fusion for Industrial Hyperspectral Sorting and Inspection
Spatially aligned SWIR and RGB data in one industrial workflow
By combining RGB and SWIR imaging with controlled illumination and dedicated software, the system enables the acquisition of spatially aligned colour and spectral data for industrial inspection and sorting applications.
SWIR and RGB in one workflow
Industrial sorting and inspection tasks often depend on more than one property of an object.
Combining the BlackIndustry RGB with the BlackIndustry SWIR 1.7 series brings both information layers into one acquisition workflow. The RGB camera captures high-resolution visual information, while SWIR hyperspectral imaging measures spectral characteristics between 950 and 1750 nm that can reveal material-related differences not available from conventional colour imaging alone.
Through software guided spatial alignment of both sensors, corresponding RGB and SWIR information can be assigned to the same position in the material stream and made available for application-specific classification, sorting, inspection and automation.
Why combine SWIR and RGB imaging? A conventional RGB camera and a SWIR hyperspectral camera capture different parts of the electromagnetic spectrum and therefore provide different information about the inspected object.
Used together, they can support applications in which a reliable decision depends on both visible characteristics and spectral material differences.
Because illumination is part of the measurement chain, we provide dedicated illumination for both sensor ranges. The BlackBright FMWIR provides broadband SWIR illumination for spectral acquisition, while the BlackBright RGB LED Bar provides controlled visible illumination with high colour rendering. Together, they support homogeneous, stable and reproducible acquisition conditions for both data streams.
In addition to the cameras and illumination systems, HAIP Solutions provides dedicated software for both parts of the workflow. BlackFusion supports calibration, spatial alignment and data synchronisation between the BlackIndustry RGB and SWIR cameras, while BlackStudio provides the hyperspectral workflow for data acquisition, visualisation, spectral analysis and classifier development. Together, both software tools connect sensor setup and acquisition with application-specific analysis and classification.
Spectral information for material differentiation
The BlackIndustry SWIR 1.7 series captures up to 399 narrow spectral bands between 950 – 1750 nm for each spatial position along the scan line with a high spatial resolution of up to 2592 px.
Depending on the camera model and ROI configuration, acquisition rates of up to 2587 Hz are possible, supporting high-speed measurements while the integrated NVIDIA Jetson GPU enables application-specific preprocessing and classification directly on the camera.
High-resolution information in the visible spectrum
The BlackIndustry RGB is a smart multi-line 4K colour camera that captures the visible appearance of the material stream with high spatial resolution. With its high line rate of 30720 lines/s, it is optimised for high-throughput measurements in combination with the BlackIndustry SWIR 1.7 series.
All BlackIndustry series cameras can be controlled via HAIP software, GenICam or dedicated C++ and Python APIs.
Specifications
Designed for seamless integration
Hyperspectral camera
BlackIndustry SWIR 1.7 series (950 – 1750 nm)
Sensor Type
InGaAs (SenSWIR)
Spatial resolution
640 – 2592 px
Frame rate
up to 2587 Hz in ROI mode
Internal data processing
Integrated GPU
Software integration
BlackStudio
Usage
Material differentiation, chemical analysis
Multi-line colour camera
BlackIndustry RGB
Sensor Type
CMOS
Spatial resolution
4K (3840 px)
Line rate
30720 lines/s
Internal data processing
Integrated GPU
Software integration
BlackFusion
Usage
Visual inspection, colour classification and object localisation
Applications
Spectral and colour information for sorting and inspection
Recycling streams are rarely homogeneous. Plastics, textiles, paper and cardboard, wood, organic fractions, mineral materials and composites can occur together, while sorting decisions may depend on material type, colour, object position or a combination of these characteristics.
SWIR hyperspectral imaging adds material-related information by measuring differences in spectral reflectance and absorption. RGB imaging provides complementary information about colour, shape and visible surface characteristics of individual objects.
By spatially aligning both data streams, the same object can be evaluated from both perspectives. This is particularly valuable for automated sorting, where the system may need to determine not only which material fraction an object belongs to, but also where it is located and whether additional visible characteristics affect the sorting decision. The resulting information can be transferred to downstream sorting systems for targeted separation of the relevant fractions.
Foreign objects in food products are particularly challenging when their colour or shape is similar to the surrounding product.
SWIR hyperspectral imaging adds spectral information that can support the differentiation of suitable foreign materials or composition-related variations. This can be relevant for tasks such as separating product and shell fractions, detecting selected foreign materials or analysing spatial variations associated with moisture or composition. SWIR spectra can contain information correlated with parameters such as moisture, fat, protein or sugar content. For example, in one of our application notes the BlackIndustry SWIR 1.7 Max with 1280 spatial pixels was used to distinguish small plastic fragments that were barely visible among the rice grains.
RGB data remains valuable for visible quality criteria and visual validation. Together, both modalities allow an inspection system to evaluate appearance and spectral behaviour within the same product stream.
Industrial Inspection
Automated sorting and inspection systems may need to determine where an object is located, what it looks like and which material-related properties are relevant to the process decision.
The BlackIndustry RGB provides high-resolution data about object position, geometry, colour and visible surface characteristics. With 4K spatial resolution acquisition and line rates of up to 30720 lines/s, this data can be captured in dynamic machine vision environments. SWIR hyperspectral imaging adds spectral features when materials need to be differentiated according to their characteristics. Through BlackFusion, RGB and SWIR data can be spatially aligned and synchronised so that visual and spectral information correspond to the same object position. The resulting data can be supplied to application-specific computer vision and AI pipelines for tasks such as automated quality inspection and material-based sorting. Integrated GPU processing together with GenICam, C++ and Python interfaces supports integration into existing machine vision and automation architectures.
- For more information, view our application notes.
Combined technology
Discover all products
SWIR Camera
BlackIndustry SWIR 1.7
- 640 px spatial resolution
- 399 spectral bands
- 450 Hz at full frame
- Up to 1330 Hz at ROI
BlackIndustry SWIR 1.7 Max
- 1280 px spatial resolution
- 399 spectral bands
- 197 Hz at full frame
- Up to 1330 Hz at ROI
BlackIndustry SWIR 1.7 Pro Max
- 2592 px spatial resolution
- 399 spectral bands
- Up to 2587 Hz at ROI
- Internal GPU
RGB Camera
BlackIndustry RGB
- High resolution 4K RGB
- Up to 30720 lines/s
- Designed for seamless integration with BlackIndustry SWIR 1.7 series
Illumination
- Let's evaluate the right setup for your application.
What we will present at VISION 2026
From acquisition to certainty
Material purity and material composition are key factors for stable industrial processes, especially in recycling, food inspection and quality control. Undetected foreign materials, incorrect material classification or production deviations can lead to quality losses, process interruptions and high costs.
At VISION 2026, HAIP Solutions will demonstrate an integrated hyperspectral imaging workflow for industrial inspection and sorting applications. Our focus will be on recycling and food inspection, where reliable information about material composition and purity is essential for process stability, quality assurance and the detection of unwanted materials.
Using the BlackIndustry Station as a live demonstration platform, we will show how cameras, controlled illumination and software work together throughout the complete workflow: from data acquisition and visualisation to analysis, classification and actionable results.
The setup will feature the BlackIndustry SWIR 1.7 Max together with the BlackIndustry RGB camera, supported by our BlackBright FMWIR and BlackBright RGB LED illumination systems. Visitors will also be able to see how BlackStudio is used to acquire, explore and classify hyperspectral data.
Our team will be available to discuss individual use cases and practical integration into industrial processes.
Meet us at VISION 2026 in the Paul Horn Halle (Hall 10), booth 10E70.8.
