Cuvis AI
CUVIS AI extends CUVIS with real-world applications, showcasing the potential of hyperspectral imaging combined with machine learning.
Explore the Possibilities
CUVIS AI brings hyperspectral imaging to life with practical applications across diverse industries. From ensuring product quality in food processing to monitoring vegetation health from the air, these real-world use cases highlight the transformative power of hyperspectral imaging combined with machine learning. Dive into these examples to see how CUVIS AI addresses complex challenges with precision and efficiency.
Foreign Object Detection
Use Case: Ensuring Product Quality and Safety
This example demonstrates the effectiveness of hyperspectral imaging for quality control in food production. Using the ULTRIS XMR hyperspectral camera, foreign plastic fragments are identified among Oreo cookies in food processing. These fragments, which are nearly indistinguishable in RGB True Color, are detected with precision through hyperspectral data.
What the Video Shows
The video highlights a real-time application of foreign object detection. Combining RGB True Color imaging with a hyperspectral overlay, the analysis highlights plastic fragments in red. Leveraging the Spectral Angle Mapper (SAM), even subtle differences in material composition are detected, ensuring a higher standard of safety and quality in production environments.
Application Potential
This approach is invaluable for food production, enabling the detection of contaminants that are invisible to traditional imaging. It can also be applied in pharmaceuticals, packaging industries, or any sector requiring strict quality control measures.
Detecting hidden markers
Use Case: Friend or foe detection using non-visible markers
This example showcases the power of hyperspectral imaging for live tracking applications. Using the ULTRIS XMR with a 100 mm f/2.8 lens, we detected infrared ink applied to a cap worn by one of our team members. While invisible to standard cameras and the human eye, the ink, which absorbs at 793 nm and emits at 840 nm, is easily identified and highlighted in red through hyperspectral analysis.
What the Video Shows
The video demonstrates real-time tracking of the cap’s spectral signature. Using a Spectral Angle Mapper (SAM), the analysis isolates the inked areas and overlays the results onto the live feed. This application is ideal for scenarios requiring the detection and monitoring of unique spectral signatures.
Applications Potential
This use case is highly relevant for security applications, such as verifying authenticity in documents or tracking items marked with invisible identifiers. It can also be applied in logistics, anti-counterfeiting, and surveillance technologies requiring precise, non-visible markers.
Monitoring Blood Flow with Hyperspectral Imaging
Use Case: Real-Time Perfusion Analysis
This example demonstrates how hyperspectral imaging can visualize blood flow and oxygen saturation in real-time. Using a spectral index derived from bands at 566 nm and 750 nm, the video shows the effects of restricting and restoring blood flow in a finger. The ULTRIS XMR hyperspectral camera captures this intricate process, revealing information that standard imaging cannot.
What the Video Shows
The video begins with a hand viewed in RGB True Color, transitioning to a false-color representation based on a hyperspectral index similar to NDVI but tuned for oxygen-related bands. The finger is tied off with a rubber band, and areas behind the restriction turn blue, indicating reduced oxygen saturation. After the band is removed, the blood rushes back, and the values shift to red, showing a rapid increase in perfusion. The colormap (HSV) represents oxygen levels, with high values in red and low in blue. This capability highlights the unmatched precision of hyperspectral imaging for monitoring narrow spectral bands.
Application Potential
This approach is particularly valuable for medical applications, such as real-time monitoring during surgeries. The ability to detect subtle changes in oxygen saturation and blood flow makes it a powerful tool for ensuring patient safety in critical scenarios.
UAV-Based Hyperspectral Mapping
Use Case: Advanced Vegetation Analysis from Above
This example highlights the capabilities of the ULTRIS X20 Plus for aerial remote sensing. Captured from approximately 40 meters above ground, the pansharpened images provide detailed insights into vegetation health and land conditions. By combining high-resolution data with multispectral analysis, this application demonstrates how hyperspectral imaging enhances UAV-based environmental monitoring.
What the Video Shows
The video features a sequence of images captured during a UAV flight:
- RGB imagery provides a true-color representation of the scene.
- CIR (Color Infrared) reveals vegetation health through infrared reflectance
- NDVI (Normalized Difference Vegetation Index), derived from spectral data, highlights plant vitality, with healthy vegetation displayed in vibrant hues and stressed areas appearing subdued.
This progression showcases how hyperspectral imaging extends beyond standard visual representations, offering unparalleled detail for environmental analysis.
Applications Potential
The Ultris X20 Plus’ pansharpened imagery is particularly valuable for precision agriculture, forestry, and land management, providing actionable insights into crop health, soil conditions, and ecosystem monitoring.
Authentication of Banknotes with Hyperspectral Imaging
Use Case: Detecting Counterfeit Currency
This example demonstrates the precision of hyperspectral imaging in distinguishing between authentic and counterfeit banknotes. Using the ULTRIS XMR, oversized counterfeit €50 notes are analyzed alongside genuine ones. While visually identical in RGB True Color, hyperspectral analysis reveals clear differences, ensuring reliable authentication.
What the Video Shows
The video begins with an RGB representation, where both genuine and counterfeit notes appear identical. When the Spectral Angle Mapper (SAM) is applied, the notes are classified: genuine €50 notes are highlighted in green, and counterfeit ones in red. This overlay clearly shows the power of hyperspectral imaging for authentication tasks that go beyond human perception or standard imaging technologies.
Application Potential
This technique is crucial for banking and security industries, enabling rapid and accurate counterfeit detection. Beyond currency authentication, it can be applied to document verification, brand protection, and anti-counterfeiting efforts across various sectors.
Analyzing Vineyards with Early Hyperspectral Imaging
Use Case: Early Vegetation Analysis for Precision Agriculture
This example highlights one of Cubert’s earliest hyperspectral imaging achievements. Captured in 2016 over vineyards in Italy, the data demonstrates how hyperspectral technology can reveal critical insights into plant health. Using the Firefleye S185, Cubert’s first HSI camera, the imagery combines a 50x50x128 spectral sensor with a high-resolution 1 Megapixel panchromatic sensor for pansharpened analysis.
What the Video Shows
The video presents aerial images of the vineyard in four visualizations:
- RGB: A true-color representation of the scene.
- CIR (Color Infrared): Highlighting vegetation health using infrared reflectance.
- CAI (Chlorophyll Absorption Integral): A spectral index derived from hyperspectral data, indicating chlorophyll absorption levels.
- Spectral Response: Reflectance data from the central pixel (indicated by a white box) displayed alongside the imagery.
The video consists of sequential frames captured during the UAV flight, illustrating the power of hyperspectral imaging for vegetation monitoring.
Applications Potential
This approach remains a cornerstone for precision agriculture, helping vineyard managers monitor plant health, optimize irrigation, and manage nutrient levels. The analysis of chlorophyll absorption supports decisions that maximize crop yield and quality. Beyond vineyards, the technique can be applied to forestry, orchard management, and broader environmental monitoring efforts.
For more information on the Chlorophyll Absorption Index, read further at IntechOpen.
Discover CUVIS AI on GitHub
Explore the full potential of hyperspectral imaging with CUVIS AI. Check out our open-source platform on GitHub and see how you can leverage machine learning to create powerful applications. If you have any questions or need guidance, feel free to reach out to us – we’re here to help!
