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

Delhi's municipal digital infrastructure is sitting on a growing problem: tens of thousands of duplicate photographs embedded in public records, from Yamuna Riverfront project files to the Delhi Metro Rail Corporation's Phase 4 construction documentation. The issue is no longer a minor administrative nuisance. It is slowing project clearances, inflating storage costs and, in at least two documented cases flagged by civic groups, causing mis-identification errors in heritage building assessments near Shahjahanabad in Old Delhi.

The timing matters. Delhi is in the middle of its most ambitious digitisation push in a decade. The Archaeological Survey of India's Delhi Circle has been cataloguing structures along Chandni Chowk and the Walled City precincts as part of a broader smart-city compliance drive. Simultaneously, the Delhi Development Authority is processing thousands of site images weekly for Phase 4 Metro corridor approvals, covering stretches from Janakpuri West to RK Ashram Marg. When duplicate images slip through, the downstream effects can range from redundant environmental-impact citations to, more seriously, administrative files that reference the wrong building entirely.

What Other Cities Are Doing

London's Ordnance Survey began deploying automated duplicate-detection algorithms across its national geographic database in 2023, integrating perceptual hashing tools that flag near-identical images before they enter the master archive. Seoul's Smart City Division, operating under the Seoul Digital Foundation, rolled out a similar pipeline for its urban data lake in late 2024, reducing image-record duplication by a figure the Foundation publicly reported at roughly 34 percent within the first six months. Mumbai's Brihanmumbai Municipal Corporation has been piloting an AI-assisted asset image registry since early 2025, specifically targeting construction inspection photographs uploaded from ward offices, where duplication rates had reportedly reached significant levels in legacy records.

Delhi does not yet have a unified equivalent. The DDA, the Municipal Corporation of Delhi and the Delhi Metro Rail Corporation each maintain separate image repositories with different metadata standards. There is no single deduplication protocol shared across these three bodies, according to publicly available procurement and IT policy documents on the DDA website. That fragmentation puts Delhi closer to where Mumbai was before its 2025 pilot than where Seoul is now.

Costs and Local Pressure Points

Storage redundancy is one measurable pressure point. A 2025 Central Public Works Department circular on government cloud procurement noted that image data constitutes a disproportionate share of total storage use across Union Territory administrations, though it did not publish Delhi-specific figures. Independent estimates from civic technology researchers affiliated with Delhi-based organisations such as the Centre for Communication Governance at NLU Delhi have pointed to image deduplication as one of the lower-cost, higher-impact efficiency interventions available to city governments — requiring relatively modest initial tooling investment compared to, say, full database migration.

The political backdrop adds pressure. The AAP administration in Delhi and the BJP-led central government have both claimed credit, at separate points, for elements of the capital's digital governance agenda, making it harder for either to openly acknowledge systemic gaps. The Yamuna cleanup project — which has generated substantial photographic monitoring data over multiple years — is one area where image-record integrity directly intersects with political accountability. Duplicate or mis-filed photographs in monitoring archives can complicate independent audits of whether embankment work has actually progressed.

For residents and small business owners near heritage zones like Lal Darwaza or the Turkman Gate area, the practical consequence is slower responses to building permit applications when documentation reviews get tangled in mis-matched image records. Heritage conservation bodies and resident welfare associations in these neighbourhoods have repeatedly cited documentation inconsistencies in submissions to the Delhi Urban Heritage Foundation.

The next concrete step, if Delhi follows the Seoul or London model, would be commissioning a cross-agency audit of existing image repositories before the end of the current financial year — March 2027. Without that baseline count, any deduplication tool procurement is essentially guesswork. Mumbai's BMC learned that lesson the expensive way before resetting its approach in 2025. Delhi still has time to move faster, but that window is narrowing as Phase 4 construction documentation volumes continue to grow every week.