How Barcelona Is Tackling the Duplicate Image Problem — and Where It Stands Against Amsterdam, Lisbon and Tokyo

Barcelona's municipal digital office confirmed this spring that a systematic audit of the city's public image repositories — covering everything from the Ajuntament's tourism portal to the open-data library maintained by the Institut Municipal d'Informàtica — had identified thousands of duplicated photographs circulating across official channels. The problem is not trivial. Duplicated images inflate apparent visual coverage of certain areas while starving others of representation, and they create legal headaches when licensing records are inconsistent across copies.

The timing matters. Mayor Jaume Collboni has pushed hard to reframe Barcelona's international image away from overcrowded postcard clichés — the same dozen shots of La Sagrada Família and Las Ramblas recycled endlessly — toward a broader portrait of the city's 73 neighbourhoods. Duplicate image proliferation directly undermines that goal. When the same file appears under 14 different metadata entries on a public portal, algorithms serving those images to travel platforms keep reinforcing the same visual monoculture the city says it wants to move past.

What Barcelona Is Actually Doing

The Ajuntament awarded a contract in March 2026 to a Poblenou-based digital services firm to deploy perceptual hashing tools across its three main image repositories. Perceptual hashing identifies near-identical images even when file names, formats or compression levels differ — a more reliable method than simple filename comparison. The audit scope covers roughly 340,000 assets held across the Barcelona Turisme image bank, the open data portal at opendata-ajuntament.barcelona.cat, and the city's press office library. A first-phase report is expected before the end of the third quarter.

The district of Gràcia has been used as a test case. Its visual archive was comparatively small — around 8,200 images — making it a manageable starting point. According to the contract documentation published in the city's official procurement register, the deduplication pass on the Gràcia subset alone removed or merged 1,140 redundant files, roughly 14 percent of the total. Officials have not yet released comparable figures for denser districts like Eixample or Sant Martí.

Barcelona Activa, the city's economic development agency, has separately flagged duplicate image contamination as a concern for startups using public datasets to train machine-learning models. Low-quality or duplicated training data skews model outputs, and several computer vision firms working out of the 22@ innovation district in Poblenou have raised the issue directly with the agency over the past 18 months.

How Other Cities Compare

Amsterdam's Gemeente has been running a comparable program since 2024 under its Digitaal Erfgoed (Digital Heritage) framework, with a reported 1.2 million assets reviewed across municipal and heritage institution databases. The Dutch capital found a duplication rate of around 11 percent — slightly lower than Barcelona's preliminary Gràcia figure, though the datasets are not directly comparable in scope or curation history.

Lisbon presents a cautionary contrast. The Câmara Municipal de Lisboa launched a similar audit in late 2023 but suspended it after six months when the vendor relationship broke down over data sovereignty disputes. The city's tourism image bank, which feeds directly into the Visit Lisboa platform, still has no confirmed deduplication timeline as of this writing.

Tokyo's metropolitan government took a different route. Rather than retroactively cleaning existing archives, the Tokyo Metropolitan Government Bureau of Industrial and Labor Affairs set strict ingestion rules starting in January 2025 — every image submitted to public repositories must pass an automated hash-check at the point of upload. New duplicates are blocked before they enter the system. It is a prevention-first model rather than a cure-first one, and several European cities, including Barcelona, have studied it.

For residents and journalists who rely on the city's open image data — researchers at the Universitat Pompeu Fabra have used the municipal archive for urban studies projects, for instance — the practical advice is to treat any downloads made before September 2026 as potentially containing duplicates and to run independent verification before publishing or citing visual data. The city's digital office says a cleaned, versioned dataset should be publicly available by the fourth quarter of this year, though procurement documents do not specify a hard delivery date.