Plataforma de análisis de imágenes satelitales para el descubrimiento de recursos hídricos mediante la aplicación de técnicas basadas en inteligencia artificial
Bibliographic record
Abstract
[ES] España es el segundo país de Europa con más piscinas. Sin embargo, la literatura jurídica estima que el 20% de las piscinas no están declaradas de forma legal o son irregulares. La Administración cuenta con un cuerpo de personas que analizan mediante procedimientos manuales, imágenes de satélite o de drones para detectar estructuras ilegales o irregulares. Este método es costoso en términos de esfuerzo, implicación de recursos humanos y tiempo, además de ser un método basado en la subjetividad de la persona que lo lleva a cabo. La propuesta de este trabajo de investigación pretende diseñar una plataforma basada en sistemas multiagente que incluya técnicas de visión artificial y que permita la detección automática de estructuras ilegales, pudiendo destacar, por ejemplo, la detección de balsas irregulares. Para la consecución exitosa de este trabajo, se emplearán herramientas de información geográfica (SIG) basadas en ortofotografía, combinadas con técnicas avanzadas de visión artificial basadas en redes convolucionales para la detección de objetos. Además, el uso de una arquitectura multiagente permitirá que el sistema diseñado sea modular, con la posibilidad de que las diferentes partes del sistema trabajen conjuntamente, equilibrando la carga de trabajo. El sistema propuesto ha sido validado mediante pruebas en diferentes ciudades de España. El sistema ha mostrado resultados prometedores en la realización de esta tarea, con una tasa de acuerdo superior al 97%. [EN] Spain stands as the second-ranked European nation in terms of the abundance of swimming pools. However, it has come to light in legal circles that a substantial 20% of these aquatic facilities either evade declaration or exist in an irregular manner. To tackle this issue, the governing bodies employ a team of individuals who manually scrutinize satellite and drone imagery. Their objective is to pinpoint structures that run afoul of legality or convention. This approach demands significant expenditure of both labor and time, compounded by the inherent subjectivity associated with human interpretation. This proposal sets forth the ambition to craft a platform capable of autonomously identifying aberrant pools. This endeavor draws upon geographical information systems (GIS) grounded in orthophotography, coupled with cutting-edge machine learning methodologies for precise object detection. Moreover, a multi-agent architecture comes into play, introducing modularity into the system's framework. This modular design facilitates the collaborative functioning of distinct system components, enabling the equitable distribution of workloads. The efficacy of the proposed system has been established through rigorous testing across various municipalities in Spain. Encouragingly, the system has yielded promising outcomes in its execution of this task, boasting an impressive F1-Score of 97.1%
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".