Unsupervised crop anomaly detection at the parcel-level using optical and SAR images: application to wheat and rapeseed crops

Florian Mouret (1, 2) , Mohanad Albughdadi (2) , Sylvie Duthoit , Denis Kouamé (1) , Hervé Poilvé (3) , Guillaume Rieu (2) , Jean-Yves Tourneret (1)

1 IRIT-MINDS - CoMputational imagINg anD viSion
2 TerraNIS
3 Airbus Defence and Space [Toulouse]

This paper proposes a generic approach for crop anomaly detection at the parcel-level based on unsupervised point anomaly detection techniques. The input data is derived from synthetic aperture radar (SAR) and optical images acquired using Sentinel-1 and Sentinel-2 satellites. The proposed strategy consists of four sequential steps: acquisition and preprocessing of optical and SAR images, extraction of optical and SAR indicators, computation of zonal statistics at the parcel-level and point anomaly detection. This paper analyzes different factors that can affect the results of anomaly detection such as the considered features and the anomaly detection algorithm used. The proposed procedure is validated on two crop types in Beauce (France), namely, rapeseed and wheat crops. Two different parcel delineation databases are considered to validate the robustness of the strategy to changes in parcel boundaries.

Mots clés: Crop anomaly detection; Wheat crops; Sentinel-2 images; Unsupervised crop monitoring; Local Outlier Factor; Isolation Forest; One-Class SVM; Parcel-level analysis; Sentinel-1; Rapeseed crops

Domaines: Machine Learning [stat.ML]

Florian Mouret, Mohanad Albughdadi, Sylvie Duthoit, Denis Kouamé, Hervé Poilvé, et al.. Unsupervised crop anomaly detection at the parcel-level using optical and SAR images: application to wheat and rapeseed crops. 2020. ⟨hal-02546260v1⟩