HSI-SC-NeRF: NeRF-based hyperspectral 3D reconstruction using a stationary camera for agricultural applications
Abstract
IntroductionAdvances in hyperspectral Q5 imaging (HSI) and 3D reconstruction have enabled accurate characterization of agricultural produce quality and plant phenotypes, both essential for advancing agricultural sustainability and breeding programs. However, integrating these two modalities at scale remains challenging, as conventional approaches involve complex hardware setups incompatible with automated phenotyping systems. Recent advances in neural radiance fields (NeRF) offer computationally efficient 3D reconstruction but typically require moving-camera setups, limiting throughput and reproducibility in standard indoor agricultural environments. To address these challenges, we introduce HSI-SC-NeRF, a stationary-camera multi-channel NeRF framework for high-throughput hyperspectral 3D reconstruction targeting postharvest inspection of agricultural produce.MethodsMulti-view hyperspectral data is captured using a stationary camera while the object rotates within a custom-built Teflon imaging chamber providing diffuse, uniform illumination. Cameraposes are estimated via structure-from-motion using ArUco fiducial markers as persistent features, then transformed from the object frame to the camera frame through simulated pose transformation, enabling standard NeRF training on stationary camera data. A multi-channel NeRF formulation optimizes reconstruction across all hyperspectral bands jointly using a composite spectral loss, supported by a two-stage training protocol that decouples geometric initialization from radiometric refinement to improve spectral fidelity.ResultsOn a representative sample, the reconstruction achieves a close geometric agreement with an independent reference reconstruction; across diverse agricultural samples, it achieves strong spectral fidelity across the visible and near-infrared spectrum, supporting its use in automated agricultural workflows.DiscussionThe full 204-band reconstruction additionally enables exploration of literature-informed wavelength bands, facilitating redeployment with faster multispectral cameras for high-throughput production environments.
Related
- Microbial Biostimulants for Drought Resilience in Plants: Mechanisms, Plant–Microbiome Interactions, and Sustainable Agricultural Applications
- Scale, trust, and the digital divide: a systematic review of AI and ML for agricultural applications
- Multiplex hyperspectral IHC: A 12-plex assay for high-dimensional analysis of the tumor microenvironment
- Data-driven spectroscopic dictionaries and detector-calibrated inference for photon-limited Raman hyperspectral imaging of living cells
- Multi-scale vision transformer model with UAV-based hyperspectral imagery for early identification of pine wilt disease infected trees
BibTeX
@article{ku2026hsiscnerf,
title = {{HSI-SC-NeRF: NeRF-based hyperspectral 3D reconstruction using a stationary camera for agricultural applications}},
author = {Kibon Ku and Talukder Jubery and Adarsh Krishnamurthy and Baskar Ganapathysubramanian},
journal = {Frontiers in Plant Science},
year = {2026},
month = oct,
doi = {10.3389/fpls.2026.1960204},
url = {https://www.frontiersin.org/articles/10.3389/fpls.2026.1960204},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - HSI-SC-NeRF: NeRF-based hyperspectral 3D reconstruction using a stationary camera for agricultural applications
AU - Kibon Ku
AU - Talukder Jubery
AU - Adarsh Krishnamurthy
AU - Baskar Ganapathysubramanian
PY - 2026
DA - 2026/10/30
JO - Frontiers in Plant Science
DO - 10.3389/fpls.2026.1960204
UR - https://www.frontiersin.org/articles/10.3389/fpls.2026.1960204
AB - IntroductionAdvances in hyperspectral Q5 imaging (HSI) and 3D reconstruction have enabled accurate characterization of agricultural produce quality and plant phenotypes, both essential for advancing agricultural sustainability and breeding programs. However, integrating these two modalities at scale remains challenging, as conventional approaches involve complex hardware setups incompatible with automated phenotyping systems. Recent advances in neural radiance fields (NeRF) offer computationally efficient 3D reconstruction but typically require moving-camera setups, limiting throughput and reproducibility in standard indoor agricultural environments. To address these challenges, we introduce HSI-SC-NeRF, a stationary-camera multi-channel NeRF framework for high-throughput hyperspectral 3D reconstruction targeting postharvest inspection of agricultural produce.MethodsMulti-view hyperspectral data is captured using a stationary camera while the object rotates within a custom-built Teflon imaging chamber providing diffuse, uniform illumination. Cameraposes are estimated via structure-from-motion using ArUco fiducial markers as persistent features, then transformed from the object frame to the camera frame through simulated pose transformation, enabling standard NeRF training on stationary camera data. A multi-channel NeRF formulation optimizes reconstruction across all hyperspectral bands jointly using a composite spectral loss, supported by a two-stage training protocol that decouples geometric initialization from radiometric refinement to improve spectral fidelity.ResultsOn a representative sample, the reconstruction achieves a close geometric agreement with an independent reference reconstruction; across diverse agricultural samples, it achieves strong spectral fidelity across the visible and near-infrared spectrum, supporting its use in automated agricultural workflows.DiscussionThe full 204-band reconstruction additionally enables exploration of literature-informed wavelength bands, facilitating redeployment with faster multispectral cameras for high-throughput production environments.
SN - 1664-462X
ER -