London Leads on Duplicate Image Crackdown — But Other World Cities Are Closing the Gap

London's planning and property data teams are quietly running one of the largest duplicate-image-removal operations of any major European city, scrubbing thousands of repeated or misattributed photographs from public-facing databases that underpin everything from housing applications to heritage listings. The scale of the effort — touching records held by the Greater London Authority, individual borough planning departments, and the Land Registry's digital archive — reflects how badly the problem had accumulated over two decades of ad hoc digitisation.

The practical consequences are not abstract. When a planning officer in Lewisham or a homebuyer's solicitor in Hackney pulls up a property file and finds the same stock photograph attached to three separate addresses, or a Victorian terrace illustrated with a picture of a different street entirely, decisions slow down. Appeals get lodged. In a housing system already grinding under pressure from the Starmer government's 1.5-million-homes target, that friction costs time and money that neither councils nor applicants can easily spare.

What London Is Actually Doing

Southwark Council began a structured deduplication programme in January 2026, working with its internal GIS team to cross-reference images stored in its planning portal against a hash-based matching system. The borough holds records for more than 48,000 live planning applications, and early internal reviews suggested that roughly one in twelve contained at least one image file that was either a duplicate of another record or had been mislabelled during batch uploads carried out between 2008 and 2014. The council has not yet published final figures from the audit.

The Museum of London's digital collections team at its new Smithfield site has run a parallel exercise across its photographic archive, which contains more than 260,000 images. Staff there have used open-source perceptual hashing tools to flag near-identical files — a method that catches not just exact copies but also slightly cropped or brightness-adjusted versions of the same original. The approach is now being shared informally with the National Archives in Kew, which manages millions of scanned government documents, some carrying repeated imagery inserted during bulk digitisation contracts awarded in the early 2010s.

Transport for London's asset management division has also been working through its station and infrastructure image libraries, a project that began after a 2024 internal review found duplicate photographs were being served to engineers consulting maintenance records for the Northern and Jubilee lines. No safety incidents have been attributed to the duplication, but the review recommended standardising file-naming conventions across all 272 stations by the end of 2026.

How London Compares With Amsterdam, New York and Singapore

Amsterdam's Stadsarchief — the city archive — completed a comparable deduplication exercise in 2023, covering roughly 800,000 digitised photographs of the canal district and surrounding neighbourhoods. The Dutch team used a combination of AI-assisted clustering and manual verification, and the project cost approximately €340,000 over eighteen months. The result was a cleaned, open-access dataset that has since been integrated into the city's planning portal.

New York City's Department of City Planning began a similar audit of its ZOLA land-use mapping system in late 2025, focusing on images attached to zoning filings in Brooklyn and the Bronx. Singapore's Urban Redevelopment Authority completed a full image-registry overhaul in 2022, partly driven by its Smart Nation initiative, and now operates a centralised image-management layer shared across all public agencies — a model that London boroughs, which still maintain largely separate systems, have not replicated.

The gap between London and Singapore in particular illustrates a structural problem: London's 33 borough councils each maintain their own planning portals, making a single deduplication standard difficult to mandate. The GLA has no direct power to compel boroughs to adopt uniform image-management protocols, though the government's Planning and Infrastructure Bill, currently before Parliament, includes provisions that could give central bodies more leverage over data standards.

For residents and professionals navigating London's planning system now, the practical advice is straightforward: when submitting planning applications, label every image file with the full address, date taken, and a unique reference number before upload. Several borough planning departments, including those in Camden and Tower Hamlets, have updated their pre-application guidance to request this as standard. Checking the Planning Portal's own submission guidelines — last updated in March 2026 — is the most reliable starting point.