University of Amsterdam / Amsterdam University of Applied Sciences
Browse

BJTU-UVA: The First Dataset for Automatic Spectral Calibration of Hyperspectral Images

Version 2 2025-01-20, 14:07
Version 1 2025-01-20, 07:19
dataset
posted on 2025-01-20, 14:07 authored by zhuoran du, S. YouS. You, J. LiJ. Li

We are proud to introduce BJTU-UVA, the first dataset designed specifically for the task of automatic spectral calibration of hyperspectral images (HSIs). This dataset addresses the critical challenge of minimizing illumination variability without relying on manual intervention or physical references.

Key Highlights

  • Task Proposal:
    We propose the novel task of automatic spectral calibration, aiming to advance the robustness of hyperspectral imaging in diverse real-world scenarios.
  • Dataset Characteristics:
    • Camera: Specim IQ, featuring a spectral resolution of 3nm across the 400–1000nm range.
    • Recording Method: Each scene is captured twice:
      1. Without reference board: Captures raw scene data.
      2. With white reference board: Records illumination conditions under the same settings.
        This approach ensures asynchronous yet precise pairing of uncalibrated and calibrated HSIs, effectively minimizing illumination variability.
    • Dark Current Correction: Dark current noise, intrinsic to the camera sensor, is carefully recorded and subtracted during post-processing, ensuring high data accuracy.
  • Scene Diversity:
    The dataset encompasses a wide range of urban and natural scenes, captured under various weather conditions, lighting scenarios, and times of day.
  • Benchmarking Standard:
    BJTU-UVA establishes a new standard for spectral calibration by combining real-world scene variability with rigorous illumination recording, offering a robust foundation for testing and advancing spectral calibration techniques.

Citation

@misc{du2024spectral,
title={Automatic Spectral Calibration of Hyperspectral Images: Method, Dataset and Benchmark},
author={Zhuoran Du and Shaodi You and Cheng Cheng and Shikui Wei},
year={2024},
eprint={2412.14925},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={ https://arxiv.org/abs/2412.14925 },
}

History

Retention period

2026-01-20

Usage metrics

    University of Amsterdam / Amsterdam University of Applied Sciences

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC