Abstract
Near-earth hyperspectral big data present both huge opportunities and challenges for spurring developments in agriculture and high-throughput plant phenotyping and breeding. In this article, we present data-driven approaches to address the calibration challenges for utilizing near-earth hyperspectral data for agriculture. A data-driven, fully automated calibration workflow that includes a suite of robust algorithms for radiometric calibration, bidirectional reflectance distribution function (BRDF) correction and reflectance normalization, soil and shadow masking, and image quality assessments was developed. An empirical method that utilizes predetermined models between camera photon counts (digital numbers) and downwelling irradiance measurements for each spectral band was established to perform radiometric calibration. A kernel-driven semiempirical BRDF correction method based on the Ross Thick-Li Sparse (RTLS) model was used to normalize the data for both changes in solar elevation and sensor view angle differences attributed to pixel location within the field of view. Following rigorous radiometric and BRDF corrections, novel rule-based methods were developed to conduct automatic soil removal; and a newly proposed approach was used for image quality assessment; additionally, shadow masking and plot-level feature extraction were carried out. Our results show that the automated calibration, processing, storage, and analysis pipeline developed in this work can effectively handle massive amounts of hyperspectral data and address the urgent challenges related to the production of sustainable bioenergy and food crops, targeting methods to accelerate plant breeding for improving yield and biomass traits.
Highlights
R EMOTE sensing imagery in the visible, near-infrared (VNIR) and shortwave infrared (SWIR) (400–2500 μm) domains are affected by noise and uncertainty attributed to the atmosphere, solar and sensor viewing geometries, terrain, environmental context, and mechanical degradation of sensor optics over time
Band-wise strong linear relationships characterized with high R2 between hyperspectral sensor digital numbers (DNs) and downwelling sensor irradiance values were observed for both VNIR (Fig. 7) and SWIR (Fig. 8) sensors, and band-wise linear function that plays transformation role between DNs and irradiance values flxdwnλ was built for each band of both VNIR (Fig. 7) and SWIR (Fig. 8) sensors, and Climate and Forecast (CF)(flxdwnλ) was computed and provided to the following equation that was mentioned in the methodology section, to derive reflectance values of VNIR and SWIR hyperspectral imagery: rflimg =
It is worth noting that, the data points used to compute the CFλ were collected diurnally, as well as at multiple seasons, to account for low, median, and high solar geometry. Both VNIR and SWIR downwelling irradiance sensors have higher spectral resolution, spectral resampling was applied to VNIR and SWIR downwelling spectral flux data to match with a spectral resolution of VNIR and SWIR hyperspectral sensors, respectively
Summary
R EMOTE sensing imagery in the visible, near-infrared (VNIR) and shortwave infrared (SWIR) (400–2500 μm) domains are affected by noise and uncertainty attributed to the atmosphere (e.g., effects due to molecular scattering and absorption), solar and sensor viewing geometries, terrain, environmental context, and mechanical degradation of sensor optics over time. Massive amounts of structured and unstructured datasets from various sensors (multispectral, hyperspectral, thermal, SAR) and platforms (e.g., drones, polar or solar orbiting, and geostationary satellites) have accumulated over decades, and the remote sensing community has been drowning in big data. These diverse datasets are crucial for various applications and informed decision-making. They may carry nonuniform spectral information that can be different both in magnitude and shape for the same target partly due to the differences in spectral sensitivity, signalto-noise ratio (SNR), and ground sampling distance (GSD) of these sensors but most importantly disturbances resulted from atmospheric scattering and absorption and anisotropy associated with terrain heterogeneity. Methods developed for one site or data source may not be
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Transactions on Geoscience and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
