Derived in the 1980s from the development of NASA’s Airborne Imaging Spectrometer (AIS) in 1983 followed by the more advanced Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) in 1987, hyperspectral imaging has gained popularity in a variety of industries and research applications due to its non-invasive and non-destructive ability to identify and quantify material. The AVIRIS became one of the most important hyperspectral instruments of this era. AVIRIS was capable of capturing images across 224 spectral bands, spanning from 400nm to 2500nm, covering the visible, near-infrared (NIR) and short-wave infrared (SWIR) spectrum, and was successfully applied for environmental monitoring and mineral exploration, among many other applications. It marked a significant leap forward in hyperspectral imaging capabilities.
Despite NASA preferring the earlier term “imaging spectroscopy” over “hyperspectral imaging”, a 2021 search in Web of Science revealed that the use of the latter term has become more prevalent in the scientific and non-scientific vernacular.
What is Hyperspectral Imaging?
Hyperspectral imaging (HSI) is an advanced imaging technique that captures and processes information across a broad range of the light spectrum. Unlike traditional imaging, which typically captures images in three bands (red, green and blue), HSI captures data in dozens or even hundreds of spectral bands, spanning visible light, NIR, SWIR and even ultraviolet (UV). This allows for the detection and identification of materials, objects and processes based on their unique spectral signatures, much beyond what our human eye can directly see within only the visible spectrum range.
What Are the Advantages of Hyperspectral Imaging?
Material Identification and Differentiation
Unlike traditional imaging, HSI captures hundreds of narrow spectral bands. This provides a detailed spectral signature for each pixel, allowing precise identification of materials, chemicals or objects based on their unique spectral properties.
It can distinguish between materials that may look the same in a conventional image but have different spectral characteristics, making it ideal for applications such as mineral exploration, precision agriculture and medical diagnostics.
Non-Destructive and Non-Invasive
As a remote sensing technique, HSI can be used to analyse objects or scenes without physically altering or damaging them. This makes it particularly useful for sensitive applications like art conservation, medical imaging and environmental monitoring.
It allows for real-time, non-invasive analysis of dynamic processes, such as monitoring plant health or tracking environmental changes over time.
High Sensitivity and Precision
HSI is highly sensitive to small variations in material composition or condition. It can detect subtle changes in an object's spectral signature, revealing shifts in chemical composition, contamination or degradation.
This technology can also identify and quantify multiple materials or substances within a single pixel, making it highly useful for applications requiring multicomponent analysis, such as pollution detection, food quality control and chemical monitoring.
What is Hyperspectral Imaging Used For?
As demonstrated above, HSI is extremely versatile and can be applied across a wide range of industries:
- Agriculture: Monitors crop health, detects diseases and assesses soil composition.
- Environmental Monitoring: Maps vegetation, analyses water quality and detects pollution.
- Medical Imaging: Applied in diagnostics, such as detecting tumours or other tissue anomalies.
- Remote Sensing: Utilised by satellites and airborne systems including drones to monitor land use, track deforestation and study geological formations.
- Art Conservation: Analyses paintings and artefacts to identify pigments, detect underdrawings and assess deterioration.
- Forensics: Assists in crime scene investigations by detecting residues or trace materials.
- Military and Defence: Used in surveillance and target identification, as different materials and objects reflect unique spectral signatures.
- Food & Drink: Ensures quality control, detects contaminants and analyses freshness.
- Smart manufacturing: Quality grading of materials and components as well as quality control of the products even in additive manufacturing.
- Pharmaceutical: Analyses pharmaceutical compositions to ensure product consistency and quality.
Is the National Subsea Centre Involved in Any Hyperspectral Imaging Projects?
The NSC’s Transparent Ocean team has accomplished several HSI projects and continues to work on finding solutions for industry partners.
The team utilised the NSC’s dedicated Hyperspectral Imaging (HSI) Lab to help Carpenter Additive, the market leader in the additive manufacturing, development and processing of high-quality metal powders and solutions and services for the metal Additive Manufacturing (AM) industry, characterise different metal powders. The company can now more accurately control material for improved product quality and reduced waste. Discover more about the project here.
As part of a collaborative project with The Scotch Whisky Research Institute (SWRI), the team used HSI technology with advanced AI algorithms to accurately quantify the levels of phenolic compounds in peat. Discover more about the motivation behind the project and the external parties reaping the benefits of improved quality control, process optimisation and environmental sustainability here.
Funded by the Royal Society of Edinburgh, the team collaborated with National Yang Ming Chiao Tung University in Taiwan to apply HSI in ocean colour remote sensing, wetland mapping and coastal sensing.
In 2023, the team achieved recognition for the collaborative paper, ‘CBANet: An End-to-End Cross-Band 2-D Attention Network for Hyperspectral Change Detection in Remote Sensing’ in the internationally prestigious journal, ‘IEEE Transactions on Geoscience and Remote Sensing’. An overview of the paper can be found here.
Hyperspectral imaging is expected to play a crucial role in future technologies, including autonomous vehicles, smart agriculture, environmental sustainability and advanced medical diagnostics. The continuous improvement in sensor technology, data processing and artificial intelligence will likely drive further adoption and innovation in this field.
To discover more about how our team is solving real-world problems and the other impactful research projects currently being undertaken, view our dedicated Transparent Ocean webpage.
