Istanbul's Duplicate Image Crisis: The Key Decisions That Will Define the City's Digital Heritage

Istanbul's municipal digital archive has a problem hiding in plain sight. Tens of thousands of duplicate images — photographs of the Bosphorus shoreline, construction-site documentation from the 2023 earthquake-retrofit programme, and street-level records from Beyoğlu to Üsküdar — have accumulated across servers managed by the Istanbul Metropolitan Municipality (İBB) and its affiliated heritage bodies, creating a storage and governance headache that archivists say can no longer be ignored.

The timing is not accidental. The İBB launched its citywide Digital Istanbul initiative in early 2024, promising a unified, publicly searchable repository of urban photographs and planning documents. Eighteen months in, the project has ingested material from at least seven legacy databases — including records from the Istanbul Archaeological Museums on Osman Hamdi Bey Yokuşu and from the KUDEB heritage inspection unit responsible for overseeing protected structures across the historic peninsula. The merger exposed the duplication at scale.

Why the Choices Made Now Actually Matter

Duplicate images are not merely a storage nuisance. In a city where earthquake preparedness shapes almost every major planning decision since the February 2023 Kahramanmaraş disaster, accurate and non-redundant photographic records of building facades, foundation exposures, and structural interventions carry legal weight. Insurers, municipal engineers assessing properties under the Urban Transformation Law (Kentsel Dönüşüm Kanunu, Law No. 6306), and neighbourhood headmen (muhtars) in districts like Bağcılar and Zeytinburnu rely on these records when disputes arise over which version of an image reflects actual site conditions at a given date.

The İBB's technology directorate is now facing a fork in the road. Option one is an automated deduplication sweep using perceptual hashing algorithms — fast, cheap, but prone to deleting near-identical images that differ in legally meaningful ways, such as a photograph taken before and after emergency scaffolding was erected on a Taksim-area building. Option two is human-curated review, slower and more expensive, with archivists working through collections neighbourhood by neighbourhood. A third path — hybrid, AI-assisted triage followed by expert sign-off — is under active internal discussion but would require procurement of specialist software and potentially delay the Digital Istanbul public launch beyond its current target window.

The financial stakes are real. Municipal cloud storage in Turkey has been running at elevated cost since the lira's prolonged depreciation pushed technology procurement prices sharply higher. Industry estimates for enterprise-grade cloud storage in Turkey have hovered around 8 to 12 Turkish lira per gigabyte per month depending on provider and contract terms, and the İBB archive is understood to run into hundreds of terabytes — figures that make the deduplication question a budget issue as much as an archival one.

The Institutions and Neighbourhoods Watching Closely

Two organisations are particularly invested in the outcome. The Istanbul branch of TMMOB, the Union of Chambers of Turkish Engineers and Architects, has been pushing for open, verified access to building-condition photographs as part of its earthquake-risk transparency campaign. Meanwhile, the Sultanahmet-based Foundation for the Preservation of Turkish Monuments and Environment (TÜRKÇEV) has its own photographic collections documenting Ottoman-era structures — collections that partially overlap with İBB holdings and have never been formally cross-referenced.

The Kadıköy municipality, which operates with some administrative independence on the Asian shore, completed its own smaller-scale deduplication exercise for neighbourhood planning records in late 2025, finishing the project in roughly four months. That experience is now being studied by İBB technicians as a potential model, though the scale difference is enormous: Kadıköy processed roughly 40,000 images; the metropolitan archive holds an estimated 1.2 million files flagged as potentially duplicated.

The next concrete decision point is September 2026, when the İBB technology committee is expected to formally adopt a deduplication methodology ahead of the Digital Istanbul public beta. Residents, researchers, and community groups in earthquake-vulnerable neighbourhoods like Avcılar and Fatih have a direct interest in pressing for the more rigorous, human-verified approach — even if it takes longer. The photographs that get deleted in the next round of server cleanup may never be reconstructed. That makes the procurement meeting in September considerably more consequential than it sounds.