BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.earthmonitor.org//soil-health-now-2026//RJCTF8
BEGIN:VTIMEZONE
TZID:Europe/Athens
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:EET
TZOFFSETFROM:+0300
TZOFFSETTO:+0200
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:EEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0300
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-soil-health-now-2026-YECZSZ@pretalx.earthmonitor.org
DTSTART;TZID=Europe/Athens:20261207T170000
DTEND;TZID=Europe/Athens:20261207T180000
DESCRIPTION:This work presents a UAV-based hyperspectral imaging pipeline f
 or high-resolution Soil Organic Carbon (SOC) mapping at the field scale. T
 he system combines hyperspectral remote sensing\, ground-truth soil sampli
 ng\, and machine learning to generate detailed SOC distribution maps that 
 can support precision agriculture and soil monitoring applications.\nThe p
 ipeline begins with data acquisition using a hyperspectral camera mounted 
 on an unmanned aerial vehicle (UAV). In this study\, a real-time hyperspec
 tral imager based on a large Fabry-Pérot Interferometer (FPI) filter was 
 used. Unlike conventional fixed-band hyperspectral systems\, the tunable F
 PI design developed by VTT allows the user to programmatically select the 
 spectral bands and wavelength combinations most suitable for a specific ap
 plication\, enabling more flexible and efficient data acquisition. The sen
 sor operates in the 650–1200 nm wavelength range using an InGaAs detecto
 r with 1280 × 1024 pixel resolution and captures full 2D spatial informat
 ion for every spectral channel while minimizing acquisition delays. \nThe 
 UAV captures hyperspectral imagery at sub-meter spatial resolution\, allow
 ing detailed observation of soil variability across agricultural fields. C
 ompared to conventional multispectral imagery\, hyperspectral data provide
 s significantly richer spectral information\, enabling the detection of su
 btle soil characteristics and absorption features related to organic carbo
 n content.\nTo create reference data for model development\, soil samples 
 were collected from multiple locations within each field and analyzed in t
 he laboratory to determine SOC content. The laboratory measurements were t
 hen spatially linked with the hyperspectral imagery to form a training and
  testing dataset consisting of spectral signatures and corresponding SOC v
 alues. Image preprocessing included radiometric correction\, geometric ali
 gnment\, and bare-soil identification to reduce the influence of vegetatio
 n and non-soil materials on the spectral signal.\nThree machine learning a
 lgorithms were evaluated for SOC prediction\, including Convolutional Neur
 al Networks (CNN)\, Random Forest\, and Gradient Boosting (XGBoost) models
 . Model performance was assessed using 5-fold cross-validation with metric
 s such as Root Mean Squared Error (RMSE) and coefficient of determination 
 (R²). Among the evaluated approaches\, XGBoost achieved the best predicti
 ve performance and was selected for final SOC map generation.\nThe trained
  model was subsequently applied to the full hyperspectral dataset to produ
 ce continuous high-resolution SOC maps for multiple agricultural fields. T
 he resulting maps reveal fine-scale spatial variability in soil carbon con
 tent that would be difficult to capture using traditional sampling approac
 hes alone.
DTSTAMP:20260825T175950Z
LOCATION:Basement (Foyer)
SUMMARY:UAV-Based Hyperspectral SOC Mapping Using a Tunable Fabry-Pérot Im
 ager and Machine Learning - Wannes De Man\, Haris Ampas\, Tuna Coppens\, N
 ick Berkvens
URL:https://pretalx.earthmonitor.org/soil-health-now-2026/talk/YECZSZ/
END:VEVENT
END:VCALENDAR