Delhi's Duplicate Image Problem: How the Capital Stacks Up Against Mumbai, London and Seoul

Delhi's official digital infrastructure is riddled with duplicate images — the same photograph of a road, monument or public facility appearing hundreds of times across government portals, civic databases and heritage archives — and the agencies responsible for cleaning them up are, by most accounts, only beginning to grasp the scale of the problem.

The issue has sharpened in 2026 because three separate Delhi government initiatives are actively digitising records simultaneously: the Delhi Urban Shelter Improvement Board is cataloguing slum resettlement sites, the Archaeological Survey of India is uploading its photographic archive of Old Delhi monuments, and the Delhi Metro Rail Corporation is building a visual asset library for Phase 4 stations along the Janakpuri West–RK Ashram corridor. Each of these projects ingests thousands of images per month, and none of them currently shares a deduplication protocol with the others.

Duplicate images are not merely a storage nuisance. When civic portals serving residents in neighbourhoods like Shaheen Bagh or Rohini pull the wrong photograph — say, an outdated image of a road that has since been relaid — the mismatch erodes trust in official data. Urban planners and journalists alike have flagged instances where Delhi's open data portal, data.delhi.gov.in, lists visually identical images under different file names and different geotags, effectively poisoning search results for anyone trying to verify on-the-ground conditions.

What Other Cities Are Doing

London's Ordnance Survey introduced a mandatory deduplication check for all locally sourced geographic imagery uploaded to its GeoPlace platform in January 2025. The system uses perceptual hashing — an algorithmic fingerprint for visual content — to flag near-identical images before they enter the live database. Seoul's Smart City Division, operating out of its Digital Foundation Centre in Mapo-gu, went further: it mandated that all city-owned images pass through an AI-assisted similarity filter before publication, a move the city said reduced redundant file storage by roughly 34 percent in its first six months. Mumbai, the most comparable Indian city, launched a deduplication drive across its BMC portal in March 2026, targeting roughly 1.2 million legacy photographs accumulated since the portal's 2017 launch.

Delhi has no equivalent citywide programme. The Delhi government's IT department has acknowledged the problem in internal communications shared with stakeholders — this reporter has reviewed a summary document circulated to the Delhi e-Governance Society in April 2026 — but a formal remediation plan has not been made public. The Delhi e-Governance Society, headquartered in IP Estate near ITO, oversees the city's digital infrastructure and is the body best positioned to coordinate a cross-agency response.

The Old Delhi Complication

The problem is especially acute in heritage documentation. The Aga Khan Trust for Culture, which has been involved in restoration work in the Nizamuddin Basti area of South Delhi, maintains its own photographic archive of before-and-after conservation work. That archive overlaps partially with ASI's own records, and neither institution has a formal data-sharing agreement that would allow automated cross-referencing. The result is that the same image of, say, a restored haveli façade near Chandni Chowk may exist in four separate databases with four different metadata tags — none of them talking to one another.

Technology experts tracking smart city rollouts in Asia point to perceptual hashing and vector-embedding tools, both now freely available through open-source libraries, as the most practical first step for cities without large IT budgets. Mumbai's BMC contracted a Pune-based firm for its deduplication exercise at a reported cost of ₹47 lakh for the initial audit phase. Delhi, with a larger and more fragmented image corpus, would likely face higher costs — but without a formal scoping exercise, no reliable estimate exists.

The Delhi e-Governance Society's next scheduled stakeholder meeting is in August 2026. Civic tech advocates say that meeting is a practical opportunity for the society to table a unified deduplication framework that all three major digitisation projects — DUSIB, ASI, and DMRC — could adopt before their archives grow any larger. Without it, each agency will keep building its own silo, and Delhi's public image databases will keep growing messier, one duplicate photograph at a time.