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

Delhi's municipal and civic databases are carrying tens of thousands of duplicate photographic records — redundant images of roads, heritage sites, and infrastructure that are slowing down everything from pothole repairs in Lajpat Nagar to property dispute resolution in Old Delhi's Chandni Chowk corridors. The problem, long acknowledged inside the Delhi Development Authority, has quietly become a practical crisis as Phase 4 of the Delhi Metro expansion generates new digital asset requirements at scale.

Duplicate image records may sound like a bureaucratic footnote, but the operational cost is real. When field teams in Shahdara or Rohini upload inspection photographs and those images exist in multiple versions across competing government servers, verification slows, storage costs climb, and the risk of acting on outdated visual evidence rises. With the Smart Cities Mission pushing Delhi's civic bodies toward AI-assisted infrastructure monitoring since its 2015 launch, the integrity of underlying image data has become a foundation-level problem.

What Delhi Is Actually Doing

The Municipal Corporation of Delhi has been piloting a de-duplication protocol across its grievance portal since late 2024, integrating hash-based image matching — a technique that assigns each photograph a unique numerical fingerprint — into its 311-equivalent citizen complaint system. The effort covers roughly 14 civic zones, including high-volume areas like Karol Bagh and Dwarka. The DDA has separately been auditing its land-record imagery archive under a digitisation push that began in fiscal year 2023-24, though that process has faced delays tied to server procurement.

The Yamuna Riverfront Authority's monitoring programme, which uses drone imagery to track encroachment and cleanup compliance between the Wazirabad Barrage and the Okhla Bird Sanctuary stretch, has reportedly grappled with the same duplication challenge. Multiple drone runs over the same coordinates generate near-identical frames, and without automated filtering, analysts reviewing compliance data wade through redundant footage. The authority has not publicly disclosed how much storage capacity this has consumed, but civic technology observers familiar with similar programmes place the overhead at 20 to 40 percent of total image storage in unoptimised municipal systems.

How Other Major Cities Are Handling It

London's Transport for London body began mandating SHA-256 hash verification on all CCTV and infrastructure images uploaded to its asset management system in January 2025, following an internal audit that found duplicate records were delaying maintenance scheduling on the Elizabeth line. Seoul's Smart City Division, operating out of the Digital Innovation Bureau at City Hall, deployed a machine-learning deduplication layer across its urban data lake in 2023, reducing redundant visual assets by approximately 31 percent within the first six months, according to a public report the bureau released that year.

São Paulo's Secretaria Municipal de Urbanismo faced a comparable crisis when digitising favela boundary maps between 2022 and 2024 — duplicated aerial survey images caused boundary disputes to linger months longer than necessary. The city contracted a Brazilian tech firm to run perceptual hashing across 1.2 million archived images, a project that cost roughly R$4.7 million and took 14 months to complete.

Delhi's challenge is compounded by institutional fragmentation. Unlike Seoul, where a single Digital Innovation Bureau holds authority over city-wide data standards, Delhi's image assets sit across the DDA, MCD, Delhi Police, the Metro Rail Corporation, and multiple state departments — none of which share a common data standard as of mid-2026. The Lieutenant Governor's office and the AAP-led state government have historically competed over jurisdiction in precisely these kinds of administrative technology decisions, which observers say has slowed the adoption of a unified civic data policy.

For citizens and contractors working with public records — whether filing heritage conservation applications for properties near Nizamuddin Basti or submitting road-damage reports via the MCD app — the practical advice is straightforward: always timestamp and GPS-tag submitted photographs before upload, since systems that do attempt deduplication rely on metadata to distinguish genuinely new images from re-submitted ones. Those navigating property disputes involving DDA records should request confirmation from the agency that the image version on file carries a unique document identifier, a step that adds a few days to the process but avoids the risk of a case being assessed against a duplicated, outdated photograph.