Istanbul's War on Duplicate Images: The Numbers Exposing a City's Broken Digital Archive

Istanbul's municipal digital archive holds more than 4.2 million photographs, maps and scanned documents across its 39 district municipalities — and a 2026 internal audit commissioned by the Istanbul Metropolitan Municipality (IBB) found that roughly 18 percent of all image files are exact or near-exact duplicates. That translates to an estimated 756,000 redundant files consuming server infrastructure, slowing data retrieval, and, in some cases, generating conflicting records on the same heritage building or earthquake-risk structure.

The timing is not accidental. Since the February 2023 Kahramanmaraş earthquakes killed more than 50,000 people across southern Turkey, IBB and the Turkish Disaster and Emergency Management Authority (AFAD) have been racing to digitise structural survey data for Istanbul's estimated 90,000 high-risk buildings. Duplicate image files in that context are not just an administrative nuisance — they can mean two different damage assessments sitting under the same address with no clear flag indicating which record is authoritative.

What the Data Actually Shows

The IBB audit, covering the January 2024 to March 2026 period, classified duplicates into three categories: exact binary copies, visually identical images with different metadata, and near-duplicate images taken within 30 seconds of each other at the same GPS coordinate. Exact copies accounted for 41 percent of the duplicate total. Metadata-variant duplicates — the trickiest category to catch algorithmically — made up another 35 percent. The remaining 24 percent were near-duplicates, often generated when field inspectors uploaded a photo burst rather than a single selected frame.

The Fatih district archive, which covers the historic peninsula and includes records on roughly 8,400 registered heritage structures catalogued with the General Directorate of Cultural Heritage and Museums, showed the highest duplication rate of any single district at 23 percent. Beyoğlu came second at 21 percent. Both districts have been subject to repeated re-surveys over the past decade — first for earthquake preparedness mapping, then for tourism infrastructure audits tied to the Galata and Sultanahmet heritage zones — which explains the layered accumulation of redundant files.

Storage costs matter here. The IBB's cloud and on-premise server bill for its document management system ran to approximately 47 million Turkish lira in 2025, according to the municipality's published budget summary. Even assuming that data storage represents only a fraction of that total, administrators estimate that eliminating confirmed duplicate image files could reduce the active storage footprint by around 12 terabytes — freeing capacity for the ongoing AFAD-linked structural surveys without a new procurement cycle.

The Deduplication Program Now Underway

IBB's Department of Geographic Information Systems, based at the municipality's Saraçhane headquarters in Fatih, began a phased deduplication programme in April 2026 using a combination of perceptual hashing algorithms and manual review queues. The first phase targets the earthquake-risk building dataset exclusively — around 340,000 image files linked to structural assessments in Avcılar, Bağcılar and Küçükçekmece, three western districts that geological surveys consistently rank among Istanbul's highest seismic-risk zones.

Phase two, scheduled to begin in October 2026, will extend to heritage and planning records, including the contested Bosphorus shoreline development files that have generated repeated legal challenges before the Istanbul Administrative Court. Officials want a clean, deduplicated image record before any further court submissions, given that conflicting photographs of the same site on the same date have previously complicated evidentiary packages.

For residents and property owners, the practical upshot is this: if you have submitted building survey photographs to your district municipality in the past three years, it is worth confirming with your local belediye office — particularly in Fatih, Beyoğlu or the three western risk districts — that your records appear as a single, consolidated file rather than fragmented duplicates. The IBB has opened an online portal at the e-belediye platform where owners can check whether their property's archive entry carries a duplication flag. District offices in Bağcılar on Yurt Caddesi and in Avcılar near Cihangir Mahallesi are also accepting in-person queries. The audit team expects to publish a full public-facing dataset on duplicate resolution rates by September 2026.