Low-Cost Dispersive Hyperspectral Sampling Scanner for Agricultural Imaging Spectrometry

Arad, Or; Cheplanov, Loran; Afgin, Yiftah; Reshef, Liad; Brikman, Roman; Elatrash, Saker; Stern, Adrian; Tsror, Leah; Bonfil, David J.; Klapp, Iftach

doi:10.1109/jsen.2023.3282835

Cited by 3 publications

(1 citation statement)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For these compact hyperspectral imagers, sometimes known as do-it-yourself hyperspectral imagers [40][41][42], pushbroom designs are the most popular, with a few using a snapshot design [43]. Diffraction gratings are commonly used [44], although some use prisms as the dispersing mechanism [45]. Some of these can send visualisations to a mobile device using WiFi [35,46], but this is short range by design.…”

Section: Introductionmentioning

confidence: 99%

A Customisable Data Acquisition System for Open-Source Hyperspectral Imaging

Mao,

Betters,

Garske

et al. 2023

Sensors

View full text Add to dashboard Cite

Hyperspectral imagers, or imaging spectrometers, are used in many remote sensing environmental studies in fields such as agriculture, forestry, geology, and hydrology. In recent years, compact hyperspectral imagers were developed using commercial-off-the-shelf components, but there are not yet any off-the-shelf data acquisition systems on the market to deploy them. The lack of a self-contained data acquisition system with navigation sensors is a challenge that needs to be overcome to successfully deploy these sensors on remote platforms such as drones and aircraft. Our work is the first successful attempt to deploy an entirely open-source system that is able to collect hyperspectral and navigation data concurrently for direct georeferencing. In this paper, we describe a low-cost, lightweight, and deployable data acquisition device for the open-source hyperspectral imager (OpenHSI). We utilised commercial-off-the-shelf hardware and open-source software to create a compact data acquisition device that can be easily transported and deployed. The device includes a microcontroller and a custom-designed PCB board to interface with ancillary sensors and a Raspberry Pi 4B/NVIDIA Jetson. We demonstrated our data acquisition system on a Matrice M600 drone at a beach in Sydney, Australia, collecting timestamped hyperspectral, navigation, and orientation data in parallel. Using the navigation and orientation data, the hyperspectral data were georeferenced. While the entire system including the pushbroom hyperspectral imager and housing weighed 735 g, it was designed to be easy to assemble and modify. This low-cost, customisable, deployable data acquisition system provides a cost-effective solution for the remote sensing of hyperspectral data for everyone.

show abstract

Section: Introductionmentioning

confidence: 99%

A Customisable Data Acquisition System for Open-Source Hyperspectral Imaging

Mao,

Betters,

Garske

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

Hyperspectral image dynamic range reconstruction using deep neural network-based denoising methods

Cheplanov,

Avidan,

Bonfil

et al. 2024

Machine Vision and Applications

View full text Add to dashboard Cite

Hyperspectral (HS) measurement is among the most useful tools in agriculture for early disease detection. However, the cost of HS cameras that can perform the desired detection tasks is prohibitive-typically fifty thousand to hundreds of thousands of dollars. In a previous study at the Agricultural Research Organization’s Volcani Institute (Israel), a low-cost, high-performing HS system was developed which included a point spectrometer and optical components. Its main disadvantage was long shooting time for each image. Shooting time strongly depends on the predetermined integration time of the point spectrometer. While essential for performing monitoring tasks in a reasonable time, shortening integration time from a typical value in the range of 200 ms to the 10 ms range results in deterioration of the dynamic range of the captured scene. In this work, we suggest correcting this by learning the transformation from data measured with short integration time to that measured with long integration time. Reduction of the dynamic range and consequent low SNR were successfully overcome using three developed deep neural networks models based on a denoising auto-encoder, DnCNN and LambdaNetworks architectures as a backbone. The best model was based on DnCNN using a combined loss function of $$\ell _{2}$$ ℓ 2 and Kullback–Leibler divergence on images with 20 consecutive channels. The full spectrum of the model achieved a mean PSNR of 30.61 and mean SSIM of 0.9, showing total improvement relatively to the 10 ms measurements’ mean PSNR and mean SSIM values by 60.43% and 94.51%, respectively.

show abstract