Numero del progetto: 26.34.20.04.01_IMAGINE

IMAGINE: analisi delle immagini e intelligenza artificiale per rafforzare i sistemi di produzione agricola

The IMAGINE project (IMAGe analysis and artificial IntelligeNce to Empower agricultural systems) aims to advance agricultural research by integrating artificial intelligence (AI) with high-throughput imaging. Using computer vision, machine learning, and reinforcement learning, it develops domain-specific models for biodiversity assessment, bird deterrence, crop phenotyping, livestock behavior analysis and pest detection. Further, it aims to standardize data acquisition and processing and organize FAIR-compliant repositories for UAV-based imaging to ensure interoperability and reproducibility across Agroscope and beyond. The six interdisciplinary work packages combine multimodal imaging (RGB, LiDAR, thermal), deep learning approaches, and automated trait extraction workflows. By coupling AI innovation with agricultural expertise, IMAGINE delivers protocols, tools and models that enhance data accessibility, accelerate breeding, support sustainable crop protection, and establish novel animal welfare biomarkers. The project unites more than 10 teams and groups, multiple Agroscope sites, and European partners to build a technological ecosystem for data-driven agriculture.

Cognome, Nome Sede
Anken Thomas Tänikon
Buholzer Serge Reckenholz
Chiang Camilo Tänikon
Huber Tobias Reckenholz
Keller Markus Tänikon
Nasser Hassan-Roland Posieux
Santacroce Nicola Posieux
Savary Pascal Posieux
Simmler Michael Tänikon
Stoop Ralph Tänikon
Szerencsits Erich Reckenholz
Winizki Jonas Reckenholz

Nasser H.-R.
Visual AI in agriculture: Field examples and future directions in methods and applications.
In: Swiss AI Days. 24 March, Ed. Swiss AI Center, Fribourg (CH). 2026.

Nasser H.-R., Kasper-Völkl C.
Unlocking the potential of computer vision in precision pig farming: A call for a collaborative data and AI models platform.
Animal Frontiers, In Press, 2026, Articolo vfag003.

Živković V., Nasser H.-R.
Advancing precision livestock farming in pigs through markerless pose estimation: A comparison between DeepLabCut and SLEAP.
Animal Frontiers, In Press, 2026, Articolo vfaf052.