- 1Land Economics Group, Institute for Food and Resource Economics, University of Bonn, Bonn, Germany
- 2EcoVision Lab, Department of Mathematical Modeling and Machine Learning, University of Zurich, Zurich, Switzerland
Tea, Camellia sinensis (L.) Kuntze is a globally significant crop, with approximately 6.6 billion cups consumed daily, making it the second most consumed beverage after water. It supports millions of livelihoods and contributes significantly to regional economies, particularly in Africa. Despite its importance, monitoring tea plantations in the continent remains manual as there are no spatially-explicit maps, thereby hindering efficient quantification of forest and biodiversity changes associated with tea cultivation, for instance. Here, we present the first high-resolution map of tea plantations in Africa, developed using computer vision techniques integrated with high-resolution satellite imagery and ground-truth polygons. Our approach achieves unprecedented spatial accuracy in delineating the area under tea cultivation with an overall accuracy of 97%. This milestone lays a foundation for spatially-explicit monitoring of tea plantations, enabling applications such as yield estimation, pest and disease detection, protected area encroachment analysis, carbon stock assessments, biodiversity impacts investigations, and evaluation of climate-driven range shifts, among others.
How to cite: Oluoch, W. A., Drees, L., Wegner, J. D., and Wuepper, D.: Mapping Tea Plantations in Africa with Computer Vision, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11581, https://doi.org/10.5194/egusphere-egu25-11581, 2025.