Paris Pushes to Purge Duplicate Images From Its Urban Archive This Week

Paris city hall confirmed this week that a long-delayed audit of duplicate images across its municipal digital archive has entered an active remediation phase, with technical teams at the Direction des Affaires Culturelles working through an estimated backlog of more than 40,000 redundant image files accumulated since the archive's partial digitisation began in 2018. The cleanup, quietly underway since late June, affects records tied to planning permits, heritage documentation, and the public-facing Bibliothèque historique de la Ville de Paris on rue de Rivoli.

The timing matters. Grand Paris Express construction — the largest urban infrastructure project in Europe, spanning dozens of worksites across the Île-de-France — has generated a surge in photographic documentation submitted to municipal systems by contractors, urban planners, and public bodies. Images of the same tunnel section, the same borehole, the same street-level disruption in Saint-Denis or Aubervilliers have arrived through multiple submission portals, clogging storage and creating version-control headaches for archivists who must match records to planning files.

A Backlog Built Over Years

The duplication problem is not new. A 2023 internal report from the Apur — the Paris urban planning agency headquartered in the 13th arrondissement — flagged that roughly 18 percent of images in certain shared municipal datasets had at least one exact or near-exact duplicate. That figure climbed during the post-Olympics period: the Paris 2024 Games triggered an unprecedented volume of venue documentation, with sites from the Trocadéro esplanade to the Stade de France in Saint-Denis photographed repeatedly by separate public bodies with no unified submission standard. The result was redundancy at scale.

Software capable of automated perceptual hashing — a technique that detects visually near-identical images even when file names and metadata differ — is now being deployed across the city's S3 cloud storage environment, managed under a contract awarded to a French IT services provider in April 2026. Archivist teams at the Bibliothèque historique have been allocated additional hours through the end of July to manually review flagged batches, particularly for pre-digital holdings that were scanned with inconsistent resolution between 2010 and 2016.

What This Means for Public Access

The practical stakes are higher than they might appear. Paris Musées, the federation that runs twelve municipal museums including the Musée Carnavalet on rue de Sévigné, relies on the shared image infrastructure for its open-access digital collection. Duplicate entries create search noise, slow retrieval, and can cause a single artefact to appear catalogued under conflicting metadata — different dates, different photographers, different accession numbers. A Carnavalet spokesperson declined to give precise figures on affected holdings, but the museum's digital collection currently lists more than 300,000 objects publicly accessible online.

The Seine riverside regeneration project, whose documentation requirements were formalised under the 2021 Plan Berges urban contract, has also contributed significantly to the backlog. Aerial surveys, ground-level progress shots, and before-and-after comparison images submitted by the Agence parisienne du climat and private engineering firms overlapped repeatedly in the city's shared folder structure, according to documentation reviewed as part of the audit scope.

City hall has set a target of reducing the duplicate image count by 70 percent before the September 2026 municipal budget review, when the Direction des Systèmes d'Information is expected to present a unified digital asset management proposal. The new system, if approved, would introduce mandatory unique identifiers at the point of upload — a standard already used by the Réunion des musées nationaux at the federal level but not yet adopted by Paris at the city scale.

For residents and researchers who use the public portals, the immediate advice from the Direction des Affaires Culturelles is straightforward: searches run against the Bibliothèque historique's online catalogue may return incomplete or temporarily missing results between now and late August as batches are re-indexed. Anyone requiring urgent access to specific planning or heritage images is being directed to contact the reading room on rue de Rivoli directly, where staff can pull records from the unaffected physical holdings while the digital remediation continues.