I was a bit confused at first because spectral imaging is foundational in astronomy… so that part can’t be the breakthrough. I think what makes it “hyper” is when the spectral resolution is high enough that a separate spectrum is obtained for each pixel of the image.
Hyperspectral imaging (HSI) finds broad applications in various fields due to its substantial optical signatures for the intrinsic identification of physical and chemical characteristics. However, it faces the inherent challenge of balancing spatial, temporal, and spectral resolution due to limited bandwidth. Here we present SpectraTrack, a computational HSI scheme that simultaneously achieves high spatial, temporal, and spectral resolution in the visible-to-near-infrared (VIS-NIR) spectral range. We deeply investigated the spatio-temporal-spectral multiplexing principle inherent in HSI videos. Based on this theoretical foundation, the SpectraTrack system uses two cameras including a line-scan imaging spectrometer for temporal-multiplexed hyperspectral data and an auxiliary RGB camera to capture motion flow. The motion flow guides hyperspectral reconstruction by reintegrating the scanned spectra into a 4D video. The SpectraTrack system can achieve around megapixel HSI at 100 fps with 1200 spectral channels, demonstrating its great application potential from drone-based anti-vibration video-rate HSI to high-throughput non-cooperative anti-spoofing.
Sounds like the advance here is the sheer bandwidth of data being processed: high frame rate, high resolution, and high spectral detail. They suggest “drone-based anti-vibration video-rate HSI” which at first I thought meant military use, but scientific disciplines also use this capability. For example, the wikipedia article mentions using HSI from drones in order to catch disease outbreaks in grape crops before it spreads.
Spatial scanning imaging spectrometers are typically limited in their imaging speed due to the use of mechanical scanning, allowing for HSI mostly in stationary scenes. If there is a relative displacement between the spatial scanning imaging spectrometer and the object in the imaging process, adverse effects such as artifacts and stretching will appear in the acquired images. We found that these effects arise from the blending of scene information from multiple moments within one measurement (Fig. 1C). Each spectrum in the scanned HSI measurement corresponds to a specific acquisition moment. We can reconstruct a complete HSI video of all acquisition moments by establishing the inverse process of the acquisition. Assuming that no variation occurs in the spectral properties of the scene materials or the illumination during the imaging process, this inverse process is to find the spatial positions corresponding to all spectra in the scanned HSI measurement at all moments. Therefore, the reconstruction requires the incorporation of scene motion information, which can be achieved by introducing a common video-rate RGB camera aligned with the scanning-based imaging spectrometer (Fig. 1A, B).
hyperspectral imaging isn’t new, the breakthrough is in combining it with motion video in a way that the information you’re getting with the spectral scan gets assigned correctly to objects within a moving scene, which is really interesting. sounds like it partly depends on new hardware:
The team has overcome multiple bottlenecks and created an on-chip spectral multiplexing sensing architecture, which subverts traditional geometric beam splitting architecture, narrowband measurement mechanism, and physical measurement output mode. It has achieved computational spectral imaging with on-chip broadband differentiation control, completed the difficult process from blank to formed, and opened up new fields of on-chip spectral research.
I was a bit confused at first because spectral imaging is foundational in astronomy… so that part can’t be the breakthrough. I think what makes it “hyper” is when the spectral resolution is high enough that a separate spectrum is obtained for each pixel of the image.
I found the paper published in Nature:
Sounds like the advance here is the sheer bandwidth of data being processed: high frame rate, high resolution, and high spectral detail. They suggest “drone-based anti-vibration video-rate HSI” which at first I thought meant military use, but scientific disciplines also use this capability. For example, the wikipedia article mentions using HSI from drones in order to catch disease outbreaks in grape crops before it spreads.
yeah, this is the really relevant bit:
hyperspectral imaging isn’t new, the breakthrough is in combining it with motion video in a way that the information you’re getting with the spectral scan gets assigned correctly to objects within a moving scene, which is really interesting. sounds like it partly depends on new hardware:
https://english.bit.edu.cn/2024-11/29/c_1049353.htm