Berlin Takes a Methodical Approach to Duplicate Image Replacement — But Is It Moving Fast Enough?

Berlin's network of public museums, libraries and city agencies is midway through a multi-year effort to identify and replace duplicate and low-resolution images across its digital collections — a technical project that sounds mundane until you consider what's at stake: public access to decades of digitised heritage, infrastructure for a growing creative economy, and the credibility of government open-data portals that researchers and developers rely on daily.

The push gained urgency after the Europeana aggregation platform — which pools cultural data from institutions across the European Union — tightened its metadata quality standards in early 2025, flagging duplicate image records as a category-one compliance issue. Berlin institutions that supply content to Europeana, including the Staatliche Museen zu Berlin and the Zentral- und Landesbibliothek Berlin on Blücherplatz, were among those notified that duplicate or mislinked image assets needed resolution before the next major export cycle.

What the Work Actually Involves

Duplicate image replacement is not simply deleting copies. Archivists must verify that a replacement file carries identical or superior resolution, that rights metadata transfers correctly, and that every external link pointing to the old file — from Wikipedia articles to university course pages — is redirected or updated. At the ZLB alone, staff have described the catalogue as containing well over 200,000 digitised image assets, accumulated across years of scanning projects with inconsistent naming conventions. The library's digital infrastructure team has been running a deduplication pass since the third quarter of 2024, using open-source perceptual hashing tools to flag near-identical files before human review.

The Stadtmuseum Berlin, whose collection spans sites including the Märkisches Museum on Köllnischer Park, is taking a parallel approach. Rather than a single sweep, it has embedded duplicate checking into its routine acquisition workflow — meaning new uploads are automatically compared against existing holdings before they enter the live catalogue. The goal is to stop the backlog from growing while the older material is cleared.

Berlin's approach is notably more decentralised than what Amsterdam has done. The Gemeente Amsterdam began a city-wide unified digital asset management rollout in 2023, consolidating collections from its municipal archive, city library and public art fund into a single repository with built-in deduplication. That project, budgeted at roughly €4.2 million over three years according to municipal budget documents, means Amsterdam's institutions work from one canonical image ID system. Berlin's landscape of individually governed institutions — each with its own IT contracts and collection management software — makes that kind of consolidation politically difficult under the current SPD-led Senate coalition, which has prioritised housing and transport spending in its capital budget.

How Berlin Compares Globally

Seoul's National Museum of Korea completed a comparable deduplication exercise across its online collection portal in late 2024, processing roughly 380,000 records with a dedicated data team of six people over eighteen months. The key difference was centralised mandate: South Korea's Ministry of Culture directed the project from the top down, with a fixed completion deadline of December 31, 2024. Berlin has no equivalent directive from the Senate Department for Culture.

In London, the Victoria and Albert Museum published its approach to image-record deduplication as part of its Collections Data Strategy in 2023, making its methodology openly available for peer institutions. The Staatliche Museen zu Berlin has reportedly reviewed that framework, though it has not yet published a comparable strategy document of its own.

For the roughly 1,400 Berlin-based tech startups and creative agencies that license images from public collections — a sector concentrated in Mitte, Prenzlauer Berg and Kreuzberg — the practical consequence of unresolved duplicates is broken API calls, inconsistent licensing terms on files that share content but not metadata, and unreliable search results on portals like the Deutsche Digitale Bibliothek. Developers building cultural applications have raised the issue in forums run by the CityLAB Berlin, the public innovation lab on Platz der Luftbrücke in Tempelhof.

The Senate's digital affairs office has indicated that a cross-institution working group on collections data quality is scheduled to report back by the end of the third quarter of 2026. Whether that produces binding standards or another round of recommendations will determine whether Berlin closes the gap on Amsterdam and Seoul — or keeps managing the problem institution by institution, one duplicate at a time.