The conservation and sustainable management of Italian small ruminant breeds are crucial for preserving livestock biodiversity. Italy counts over 100 sheep and goat breeds, often raised in extensive farming systems and mountainous and marginal areas, many of which hold high cultural and environmental value. Their presence supports rural economies and contributes to the identity and landscape of regions with strong tourist appeal. However, many local breeds are at risk due to limited distribution and declining populations. One key challenge in their management is the correct registration of an animal to a breed, traditionally based on expert morphological evaluation and assessment of standard adherence. Artificial intelligence (AI), particularly deep learning techniques, offers innovative solutions to support this sector. A promising application, explored in few studies with encouraging results, is the development of AI-powered image-based classification tools for breed identification. Extending this approach to a broader scale could have significant practical implications. With photographic data collected under diverse conditions and image augmentation techniques, it becomes possible to build a robust and efficient system adaptable to the morphological variability of local breeds. Also, georeferencing images might be especially valuable, as many populations are strongly localized in specific regions; spatial data could thus enhance breed recognition, even in cases of high phenotypic variability or morphological similarity between breeds. A possible outcome of this approach could be a user-friendly mobile application to assist breeders, technicians, and associations in breed identification and registration, streamlining procedures and minimizing animal handling. Additionally, such a tool could serve an educational purpose, providing users—including tourists and local communities—with information about each breed’s history, risk status, and typical products. By raising public awareness and fostering a connection between people and livestock biodiversity, this AI-based solution may contribute to the conservation and valorization of Italy’s small ruminant heritage, while supporting the resilience of rural areas.
AI Meets Tradition: Enhancing Italian Small Ruminant Biodiversity through Breed Identification / A. Bionda, P. Crepaldi - In: Book of Abstracts of the 1st EAAP Conference on Artificial Intelligence 4 Animal Science[s.l] : EAAP, 2025. - pp. 46-46 (( Intervento presentato al 1. convegno EAAP Conference on Artificial Intelligence 4 Animal Science tenutosi a Zurich nel 2025.
AI Meets Tradition: Enhancing Italian Small Ruminant Biodiversity through Breed Identification
A. Bionda;P. Crepaldi
2025
Abstract
The conservation and sustainable management of Italian small ruminant breeds are crucial for preserving livestock biodiversity. Italy counts over 100 sheep and goat breeds, often raised in extensive farming systems and mountainous and marginal areas, many of which hold high cultural and environmental value. Their presence supports rural economies and contributes to the identity and landscape of regions with strong tourist appeal. However, many local breeds are at risk due to limited distribution and declining populations. One key challenge in their management is the correct registration of an animal to a breed, traditionally based on expert morphological evaluation and assessment of standard adherence. Artificial intelligence (AI), particularly deep learning techniques, offers innovative solutions to support this sector. A promising application, explored in few studies with encouraging results, is the development of AI-powered image-based classification tools for breed identification. Extending this approach to a broader scale could have significant practical implications. With photographic data collected under diverse conditions and image augmentation techniques, it becomes possible to build a robust and efficient system adaptable to the morphological variability of local breeds. Also, georeferencing images might be especially valuable, as many populations are strongly localized in specific regions; spatial data could thus enhance breed recognition, even in cases of high phenotypic variability or morphological similarity between breeds. A possible outcome of this approach could be a user-friendly mobile application to assist breeders, technicians, and associations in breed identification and registration, streamlining procedures and minimizing animal handling. Additionally, such a tool could serve an educational purpose, providing users—including tourists and local communities—with information about each breed’s history, risk status, and typical products. By raising public awareness and fostering a connection between people and livestock biodiversity, this AI-based solution may contribute to the conservation and valorization of Italy’s small ruminant heritage, while supporting the resilience of rural areas.| File | Dimensione | Formato | |
|---|---|---|---|
|
AI4AS_2025_AI Meets Tradition Enhancing Italian Small Ruminant Biodiversity through Breed Identification.pdf
accesso aperto
Tipologia:
Publisher's version/PDF
Licenza:
Creative commons
Dimensione
203.26 kB
Formato
Adobe PDF
|
203.26 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.




