Tomasz S. Tkaczyk and Ye Wang were part of the team that developed a new compact imaging spectrometer for remote sensing applications. The instrument can capture more than 30,000 sampling points each containing over 60 wavelengthsCredit: Rice University

Researchers have developed a new compact, fiber-based imaging spectrometer for remote sensing that can capture 30,000 sampling points each containing more than 60 wavelengths. This rich spectral information combined with high spatial resolution provides valuable insight into the chemical makeup of a scene or sample.

"Compact imaging spectrometers such as the one we developed can be used on unmanned aerial vehicles to help increase crop production or inform response after a disaster based on detected pollution," said research team leader Tomasz S. Tkaczyk from Rice University. "In the biomedical field, the system could increase the efficiency of diagnostic tests or help scientists better understand biological processes."

In The Optical Society (OSA) journal Optics Express, the researchers show that their instrument can provide about an order of magnitude more information than has been reported for other fiber-based systems. The device measures only 600 x 150 x 150 millimeters—almost small enough for use aboard unmanned aerial vehicles. The researchers say that the instrument could be made even smaller while capturing more than 100 wavelengths from up to 250,000 points.

The new spectrometer acquires spectral information instantaneously without requiring any scanning. This allows it to image quickly changing objects such as moving targets or changing imaging conditions.

Creating a smaller fiber bundle

Imaging spectrometers can use a fiber bundle to transfer the spectral image to a detector or detector array. Spectrometers incorporating this setup usually feature a bundle of fibers arranged in a square at the input end and a single line at the detector end. This means that the number of fibers, and thus spatial resolution, is limited by the dimensions of the detector.

"To increase the spatial sampling for our new fiber bundle design, we placed the fibers into multiple rows with gaps between each row," said Tkaczyk. "An added benefit of this design is that the size of the gaps can be changed to tune the balance between spatial and spectral sampling to meet specific application requirements."

Commercially available small-diameter fibers suitable for imaging spectroscopy are typically 125-250 microns in diameter, which quickly drives up the size of the fiber bundle. To keep things compact, the researchers used a commercially available multi-core fiber ribbon in which each fiber has a 10-micron core. They also optimized the fibers' numerical apertures to increase light gathering ability.

Collecting spectra from more than 30,000 fibers could require an extensive calibration process. The researchers developed a rapid method that takes less than 5 minutes to calibrate all the spatial samplings of the system and requires acquisition of only a few images.

Capturing spectral images

The research team demonstrated the new spectrometer by using it to image distant scenes and vegetation on the Rice University campus. "These tests demonstrated the system's spectral imaging capability and showed its significant promise for use in environmental and remote sensing applications," said Tkaczyk. "Our new rapid calibration method for imaging spectrometers with high spatial sampling could also be extended to calibrate fiber bundles and other imaging devices."

The researchers are now working on further improvements to the instrument. They want to make the fiber bundle even more compact by using custom fiber ribbons and are also examining ways to increase the system's light throughput.