The Land Parcel Identification System (LPIS) forms the geospatial backbone of the Integrated Administration and Control System supporting implementation of the European Union’s Common Agricultural Policy. The transition to a performance-based CAP delivery model and the ongoing EU enlargement process have increased demands on LPIS accuracy, scalability, and harmonization. This study examines LPIS modernization from a land governance perspective and proposes a regulatory-centric framework integrating stable reference parcel ontology, eligibility logic based on Maximum Eligible Area, and AI-assisted boundary detection within a human-in-the-loop workflow. Using very high-resolution satellite imagery from pilot areas in Albania and Ukraine, automated parcel delineation was benchmarked against manually defined LPIS reference parcels. Results show high boundary localization accuracy but moderate parcel-level agreement, revealing a structural gap between image-driven segmentation and regulatory parcel definitions. The findings highlight how AI can support LPIS creation while remaining embedded within harmonized institutional rules governing agricultural land management.
LUKETIC Natasa;
LEMOINE Guido;
LOUDJANI Philippe;
2026-10-23
TAYLOR & FRANCIS LTD
JRC146262
1747-4248 (online),
https://publications.jrc.ec.europa.eu/repository/handle/JRC146262,
10.1080/1747423X.2026.2719539 (online),
| Name | Country | City | Type |
|---|
This document is only visible at the Commission level.
You are not authorized to publish or distribute it outside the European Commission.
This is a public document. You can share this publication.
Datasets
| ID | Title | Public URL |
|---|
Dataset collections
| ID | Acronym | Title | Public URL |
|---|
Scripts / source codes
| Description | Public URL |
|---|
Additional supporting files
| File name | Description | File type |
|---|