A UAV-Based VNIR Hyperspectral Benchmark Dataset for Landmine and UXO Detection

TL;DR

This work addresses the lack of publicly available drone-based hyperspectral data for landmine detection by introducing a UAV VNIR hyperspectral benchmark dataset collected over a controlled field with 143 inert targets. The dataset uses a Headwall Nano-Hyperspec mounted on a multi-sensor drone, acquiring 270 bands from 398–1002 nm at ~20.6 m, and is radiometrically calibrated with reflectance retrieved via the Empirical Line Method, using ground-truth spectra and precise GCP/AeroPoint georeferencing to produce a 3123×6631×272 hyperspectral cube. Validation against reference spectra yields RMSE 0.5–4.5 (400–1000 nm) and SAM 1°–12° (400–1000 nm), improving to RMSE < 1.0 and SAM 1°–6° in 400–900 nm, demonstrating high spectral fidelity. The dataset, released with raw radiance cubes and reference spectra, enables reproducible research and serves as a multi-sensor benchmark when combined with EMI data from the same field, promoting robust spectral analysis and fusion methods for landmine/UXO detection. A key contribution is the explicit calibration and processing workflow, including $R = aL + b$ for reflectance retrieval, and the emphasis on open data to advance humanitarian demining technologies.

Abstract

This paper introduces a novel benchmark dataset of Visible and Near-Infrared (VNIR) hyperspectral imagery acquired via an unmanned aerial vehicle (UAV) platform for landmine and unexploded ordnance (UXO) detection research. The dataset was collected over a controlled test field seeded with 143 realistic surrogate landmine and UXO targets, including surface, partially buried, and fully buried configurations. Data acquisition was performed using a Headwall Nano-Hyperspec sensor mounted on a multi-sensor drone platform, flown at an altitude of approximately 20.6 m, capturing 270 contiguous spectral bands spanning 398-1002 nm. Radiometric calibration, orthorectification, and mosaicking were performed followed by reflectance retrieval using a two-point Empirical Line Method (ELM), with reference spectra acquired using an SVC spectroradiometer. Cross-validation against six reference objects yielded RMSE values below 1.0 and SAM values between 1 and 6 degrees in the 400-900 nm range, demonstrating high spectral fidelity. The dataset is released alongside raw radiance cubes, GCP/AeroPoint data, and reference spectra to support reproducible research. This contribution fills a critical gap in open-access UAV-based hyperspectral data for landmine detection and offers a multi-sensor benchmark when combined with previously published drone-based electromagnetic induction (EMI) data from the same test field.

Figures (4)

• Figure 1: Pre-burial test field with target locations on bare soil in 2023. Each marker indicates a precisely geolocated target..
• Figure 2: Condensed overview of all target types deployed in the test field pre-burial in 2023 Baur2023.
• Figure 3: RGB composite of the final ELM-retrieved and georeferenced VNIR hyperspectral dataset.
• Figure 4: Comparison of reference and ELM-retrieved spectra for various in-scene materials. The label ending with ‘SVC’ refers to reference spectra measured by the SVC spectrometer, while the corresponding label with ‘Image’ indicates the ELM-retrieved spectra extracted from the hyperspectral image.