Aerial drone photograph of homes nestled in the vegetation, facing the ocean, on the coast of Martinique
DRONE · AI · GIS

Geospatial intelligence serving territories

A geospatial intelligence consultancy. We support local authorities in understanding and monitoring their territory through aerial drone imagery, artificial intelligence and human expertise.

Context and challenges

Territories are constantly changing

Aerial drone view of a town in Martinique whose homes climb the hillsides, facing the sea

Over time, territories change: new buildings, extensions, swimming pools, developments, demolitions. A building can change long after it was built, and these changes are not always reflected in the administrative information available.

For local authorities, staying up to date is a challenge, with tight budgets, limited resources and time-consuming field checks. Across territories that can be large, these checks tie up staff for long periods, and some properties remain hard to see from public space.

Administrative information can then drift away from reality, with an impact on the reliability of tax bases and on the authority's resources. Having reliable, up-to-date knowledge of the territory has become essential to identify, understand and monitor these changes.

01

Territorial knowledge that needs updating

Having an up-to-date inventory of buildings, extensions and developments.

02

Significant administrative and tax stakes

Making the databases used by departments more reliable and securing the authority's revenue.

03

Limited resources and time-consuming methods

Reducing the time spent on manual research and systematic field checks.

Our approach

Reliable knowledge of the territory

We identify the changes that have occurred in your territory (new buildings, extensions, demolitions, swimming pools, marine jetties). Each one is precisely located, checked by an operator, then delivered turnkey, ready for your departments to use.

Drone orthophotograph of a residential neighbourhood in Martinique: AI-based detection of buildings and swimming pools, with additions shown in green
Automatic detection of buildings and swimming pools on a drone orthophotograph (additions in green).

Observe

Gain precise, up-to-date knowledge of the territory.

Detect

Identify and locate changes between survey campaigns.

Analyse

Turn observations into usable information.

GEO DATA ANALYSE identifies, analyses and documents territorial changes.The administrative, regulatory or tax classification of the situations observed is the exclusive responsibility of the competent authorities.
Benefits

The benefits for your local authority

Up-to-date, objective and usable territorial information that saves your departments time and secures your databases.

Tax challenge

Making tax bases more reliable

Identifying situations that need checking can contribute to the reliability of tax bases and, where a regularisation applies, to better use of resources by the authority.

This is the goal of « Foncier innovant », a project of the French tax authority (DGFiP), in the service of tax fairness. See the national case study

  • Time savings

    Fewer field checks, focused on the situations that need them: your staff step in where it matters.

  • Prioritisation

    Situations to check are located and ranked by importance, to guide your departments' action.

  • Updatable data

    More recent territorial information, updated with each new campaign.

  • Fairness

    The same level of information for every area and every owner, including properties that are hard to see from public space.

  • Reusable data

    Maps and GIS data that also serve urban planning, development and land management.

  • Documentation

    Objective, traceable geographic evidence: dated image, precise location, report.

Monitoring over time

Observing and detecting changes

GEO DATA ANALYSE works in two stages: an initial audit that identifies existing features and notes any cadastral discrepancies, followed by successive campaigns designed to detect and document changes in the territory.

Drone orthophotograph, T0 campaign: initial state of the territory before change detection
Drone orthophotograph, T1 campaign: building changes detected compared with the T0 campaign
T0 · Initial auditT1 · Change monitoring
AdditionModificationRemovalUnchangedExisting swimming pool

Drag the slider to compare the two campaigns. Illustrative example based on drone imagery and simulated data.

T0 · Initial audit

Initial analysis of the territory

  • Identification and location of the objects present in the area of interest
  • Cross-checking with available cadastral data
T1 / T2… · Change monitoring

Comparing campaigns

  • Comparing the data with the previous campaign
  • Detecting changes: additions, modifications and removals
Monitoring over timeEach campaign builds the area's history and makes it possible to track precisely how the territory changes over time.
T0Initial audit
T1Comparison with T0
T2Comparison with T1
T3Subsequent campaigns
Our detection capabilities Buildings Swimming pools Marine jetties

Specific requests: contact us

Our methodology

A complete processing chain

To turn imagery into usable territorial information.

1Acquisition

High-resolution aerial imagery by drone.

A drone flies over the area of interest and captures georeferenced aerial images, where each point of the image corresponds to a real position on the territory.

2Processing

Orthophotography and data preparation.

The images captured during the flight are stitched together to form an orthophotograph, an overview of the area, which is then prepared for analysis.

3Analysis

Automated analysis of objects and changes.

Automated analysis tools scan the orthophotograph to spot the features being sought: buildings, swimming pools, marine jetties, or any other specific request. During monitoring campaigns, they also flag what has changed: addition, modification or removal.

4Quality control

Checking and validation of results by our teams.

An automated result can contain errors. Our teams therefore check every detected item before delivering it, to qualify the results.

5Data cross-checking

Cross-checking with available cadastral and geographic data.

The features observed are matched against the available cadastral and geographic data, such as parcels. Any discrepancies between what exists on the ground and what is on record can then be identified, and each observation linked to the parcel concerned.

6Delivery of deliverables

Production of data and observation reports.

Results are delivered as an orthophotograph, a change map and a PDF observation report.

Deliverables

Deliverables your departments can use directly

High-resolution drone orthophotograph of a residential neighbourhood in Martinique

High-resolution orthophotograph

An up-to-date view of the territory.

Map of changes detected between two drone campaigns: new buildings, extensions and swimming pools located

Change map

Location of the changes detected.

Map of detected buildings and swimming pools in GIS software, with its cadastral layers

Structured GIS data

Information that can be integrated into your geographic tools.

Observation reports: initial audit T0 and T0 / T1 monitoring, with orthophotographs, cadastral map and summary

Observation report

Observation sheets linked to the parcels concerned.

COMING SOON

A web platform to access your maps, data and reports, online, at any time.

Case studies

Examples of observation reports

Whether it is the initial audit (T0) or change monitoring (T1, T2…), each sheet presents the observations made, their cadastral location and a summary designed to help the competent departments make use of them.

T0 initial audit report: building and swimming pool identified on cadastral parcel R 831 from a drone orthophotograph

What we observe

On the T0 orthophotograph, a building (179.67 m²) and a swimming pool (5.88 m²) are identified on parcel R 831.

What this brings

Cross-checking with the cadastral map reveals a potential discrepancy and helps target the situations that need verification.

What happens next

The information collected forms a dated analytical baseline for detecting and documenting changes in subsequent campaigns.

Illustrative examples based on demonstration data, modified or adapted for testing purposes. The observations shown are not official findings.

Each observation is checked by our teams, then linked to the parcel concerned before delivery.
A proven technology

Performance assessed against quality indicators*

Our analysis chain relies in particular on AI-based geospatial extraction technologies that meet high accuracy requirements.

≥ 95%

Precision

At least 95% of detected objects are correctly identified.

≥ 95%

Recall

At least 95% of the objects actually present are found.

≥ 95%

Geometric accuracy (IoU)

An overlap of at least 95% with the reference geometry.

100%

Valid geometries

Geometries that meet the defined validity criteria.

* Reference performance assessed against defined quality criteria and analysis conditions.

Additional human review to check and refine the results, strengthening their accuracy and reliability.

Aerial view of the Caravelle peninsula in Martinique, with homes nestled in the vegetation

A powerful combination

Artificial intelligence, human expertise and GIS analysis.

Case study

A national-scale case study

Launched in 2021, the « Foncier innovant » project allows the DGFiP, France's tax authority, to use aerial imagery with artificial intelligence, in the service of tax fairness. Initially trialled on swimming pools in nine pilot départements in 2022, it delivered concrete results.

20,000+

new swimming pools taxed¹

Detected by AI during the pilot phase.

≈ €10M

in additional revenue¹

€4.1M of recurring annual property tax and €5.7M of back-taxes.

94%+

of owners confirmed¹

That their swimming pool was taxable.

≈ €40M

in revenue expected in 2023¹

National projection issued by the DGFiP in 2022.

A broader approach with GEO DATA ANALYSE

Our expertise extends observation to buildings, swimming pools and other developments, with validation of the results before delivery.

¹ DGFiP, L'intelligence artificielle au service de la lutte contre la fraude : bilan de l'expérimentation « Foncier innovant » (Artificial intelligence in the fight against fraud: review of the « Foncier innovant » trial), August 2022.

Florian PRIAM, founder of GEO DATA ANALYSE, a geospatial intelligence consultancy in Martinique
The founder

Get to know us

Florian PRIAM has built his career between buildings, data and professional drones. Find out what led him to create GEO DATA ANALYSE.

Discover his background
Contact us

Let's discuss your authority's needs

Every territory has its own challenges. Let's explore together the possibilities of a territorial observation programme tailored to your priorities.

Write to us

The information you enter is used only to reply to your request. It is neither passed on nor used for any other purpose. To find out more and exercise your rights: privacy policy.