SPECTRAL IMAGING SENSOR TECHNOLOGY
Headwall's Hyperspec® family of imaging sensors includes commercial-off-the-shelf (COTS) sensors (Nano-Hyperspec VNIR) as well as applications-specific imagers (the high-resolution chlorophyll fluorescence and the co-aligned VNIR+SWIR). Sensors are available for airborne and satellite deployment, and for advanced machine vision applications where ruggedized enclosures assure uninterrupted performance in harsh inspection environments.
For small sample analysis, Headwall's Scanning Kits provide a motorized moving stage, a gantry to attach the imaging sensor, proper illumination for the wavebands of use (QTH, for example), and full software to capture the hyperspectral data cube. Scanning kits are available in standard (one sensor attachment) and large-format (up to two sensors can be attached). In addition, Headwall's UV-VIS scanning kit includes a fully-darkened enclosure.
DATA STORAGE AND PROCESSING
Hyperspectral data cubes are large, often gigabytes in size. Sufficient storage and fast computational speeds are needed. Headwall's Hyperspectral Data Processing Units (HDPU's) are available that provide 1TB of storage and the latest CPU/GPU processing technology. The Standard HDPU is sufficient for lab and ground use. The Compact HDPU is suitable for these same applications as well as certain airborne ones. Finally, Headwall's Hypercore™ is a small and lightweight UAV-suitable 'data fusion' hub. This product provides 480GB of internal solid-state storage as well as connection points for other instruments such as the GPS/IMU, LiDAR, RGB, and more.
Because Headwall embeds its own diffractive optics technology in each Hyperspec sensor, image accuracy is assured. High signal-to-noise, high spectral and spatial resolution, low stray light, and aberration-corrected imaging are additional advantages. The use of holographic gratings and concentric mirrors means that every Hyperspec sensor is small, robust and light. This makes them perfect for UAV and cubesat deployment. Additionally, there are no moving parts that can lead to the unnecessary weight and complexity seen in other designs.
Hyperspectral imaging and multispectral imaging differ with respect to their abilities to accurately characterize and classify materials within the field of view. The more spectral bands, the easier it is to discern between similar-looking materials.
The five typical spectral ranges for Hyperspec sensors are noted below. Thanks to Headwall's control over the diffractive optics in each instrument, grating efficiency can be 'tuned' and maximized to deliver high resolution performance in the waveband of interest.