Zurich Leads on Duplicate Image Replacement in Public Archives — But Other Cities Are Closing the Gap

Zurich's city archive, the Stadtarchiv Zürich on Alfred-Escher-Strasse, completed the first phase of its duplicate image replacement programme in June 2026, clearing more than 14,000 redundant scans from its publicly accessible digital catalogue — a figure the archive confirmed in its mid-year operational report. The clean-up, running since January, used automated perceptual hashing tools cross-referenced with manual curatorial review to flag identical or near-identical image records before replacing low-resolution duplicates with a single high-resolution master file.

The timing matters. Zurich's cultural institutions have spent the past three years accelerating mass-digitisation after a federal push to make Swiss heritage collections fully searchable by 2028. That sprint produced results — and clutter. The Zentralbibliothek Zürich on Zähringerplatz, which holds one of the largest newspaper photograph collections in the German-speaking world, acknowledged publicly in its 2025 annual report that duplicate entries had inflated its publicly listed image inventory by an estimated 18 percent. Weeding that out is not cosmetic housekeeping; duplicates generate broken metadata chains, confuse rights-clearance workflows and cost storage budget that institutions in a city with Zurich's cost structure feel acutely.

How Zurich Compares With Amsterdam and Vienna

The approach differs sharply from what comparable European institutions are doing. Amsterdam's Stadsarchief — the city archive that holds roughly 750,000 publicly accessible photographs — opted two years ago for a crowd-sourced tagging model, inviting registered users to flag suspected duplicates before professional staff review them. The model is slower but has surfaced contextual errors that purely algorithmic tools miss: two images that look identical but were taken minutes apart during the same event, for instance, can carry different documentary value.

Vienna's Wienbibliothek im Rathaus took a different route again, contracting with a specialist digital preservation firm in 2024 to run a one-time bulk deduplication before migrating to a new content management system. The advantage was speed. The risk, which Viennese archivists have discussed at conferences including the European Conference on Digital Preservation held in Edinburgh last October, is that a one-time purge does not build institutional capacity for ongoing maintenance.

Singapore's National Archives ran its own duplicate-reduction programme across its visual collections between 2023 and 2025, applying AI-assisted clustering to a holdings set exceeding two million items. The programme reduced active catalogue entries by approximately 11 percent, according to figures the National Archives of Singapore published in its 2025 corporate report. Crucially, Singapore paired the technical work with updated acquisition guidelines to prevent duplication at the point of ingest — something Zurich's institutions have not yet formalised at city level.

What Zurich's Institutions Are Building Next

The Stadtarchiv's next phase, scheduled to begin in September 2026, will extend the deduplication review to its architectural photography holdings — a collection documenting Zurich's building stock from the Kreis 4 and Kreis 5 neighbourhoods, areas that underwent intensive redevelopment documentation between 1980 and 2010. Those records are among the most heavily duplicated because multiple departments commissioned photography of the same sites during the same periods.

ETH Zürich's computer science department has a working relationship with the Stadtarchiv on image-recognition research, and doctoral work published through the university's e-collection repository has fed directly into the hashing methodology the archive is now using. That proximity to a globally ranked research institution gives Zurich an advantage that cities without an equivalent technical university partnership — such as many mid-sized European capitals — have to buy in from the private sector at considerably higher cost.

For residents and researchers who use Zurich's public digital collections, the practical effect will be felt in the search interface on the Stadtarchiv's online portal, where duplicate results have historically pushed relevant unique images down the results page. The archive has indicated the portal will reflect the cleaned catalogue by the end of the third quarter of 2026. Anyone relying on those collections for property research, journalism, or academic work should cross-check any image records accessed before that date against the post-cleaning catalogue once it goes live.