Delhi's Duplicate Image Problem: How the Capital Compares to Cities Racing to Clean Up Digital Records

Delhi's municipal and heritage agencies are sitting on hundreds of thousands of duplicated, mislabelled and outdated photographs embedded across government portals, property records and urban planning databases — a backlog that civic technologists say is slowing down everything from land title verification in Shahdara to heritage conservation assessments in Chandni Chowk.

The problem has sharpened because Delhi is mid-way through several large digitisation drives. The Delhi Development Authority is pushing property records online as part of a broader e-governance push, while the Archaeological Survey of India maintains visual documentation of protected monuments concentrated heavily in Mehrauli and Nizamuddin. When source photographs are duplicated across systems, it creates conflicting records that courts, planners and citizens then have to untangle manually.

What Other Cities Have Done

Other large cities with comparable digitisation timelines have moved faster. Singapore's Urban Redevelopment Authority completed a systematic deduplication of its GeoSpace portal — which holds planning imagery and land parcel records — by 2023, using hash-based matching to flag redundant files automatically before they entered the master database. The city-state's approach cut storage costs and, more importantly, reduced the rate of duplicate planning objections filed by residents who were looking at different versions of the same site photograph.

London's Ordnance Survey, which supplies imagery to the Greater London Authority, has operated a formal duplicate-detection pipeline since 2019. By contrast, Mumbai's municipal corporation — often the closest comparison to Delhi given the scale of its informal settlement documentation work — has acknowledged backlogs in its own GIS imagery layers, particularly around Dharavi, where redevelopment planning depends on accurate, non-redundant visual records. Mumbai and Delhi share a common structural problem: imagery enters government systems through multiple uncoordinated channels, from ward offices and contractor submissions to satellite data feeds, with no single gateway to enforce uniqueness at the point of upload.

Delhi's situation is compounded by the political architecture of the city. The DDA reports to the central government rather than to the elected Delhi government led by Chief Minister Arvind Kejriwal. The result is that two parallel digitisation efforts — one run through the AAP administration's e-district portal and one through DDA's own systems — can hold conflicting imagery of the same property without any automatic reconciliation step between them.

Where the Gaps Show Up on the Ground

The practical friction is visible at the Delhi Heritage Conservation Committee's documentation work in areas like Lal Kuan and the lanes off Ballimaran in Old Delhi. Field officers working on heritage listings have reported submitting photographic evidence of structures that already exist in central databases under different file names, different coordinates and different condition assessments — producing records that appear to document separate buildings but in fact show the same haveli photographed on different dates.

The National Informatics Centre, which provides technology backbone for multiple Delhi government portals, has been developing deduplication tools as part of its broader Digital India infrastructure work. The centre's systems are already deployed in some state-level land record portals elsewhere in India, where pilot programmes have reportedly reduced duplicate image entries by flagging files with matching perceptual hash values before ingestion. Delhi has not yet integrated this tooling across all relevant agencies, according to publicly available procurement documents reviewed by The Daily Delhi.

Globally, the benchmark is moving. Barcelona's smart city programme completed an audit of its municipal image repositories in 2024, reducing a database of roughly 2.4 million civic photographs by an estimated 18 percent after deduplication. Nairobi's county government, working with UN-Habitat support, has been building duplicate-detection into its informal settlement mapping from scratch — an approach urban planners call cheaper than retrofitting older systems.

For Delhi residents, the immediate consequence is administrative delay. Property mutation applications that reference photographs held in conflicting records can take weeks longer to process than applications with clean documentation. The DDA's online property portal, which went through a significant upgrade in early 2025, still does not display the source date or file origin of images attached to individual plot records — meaning applicants have no way to verify whether the image on their file is current or a years-old duplicate.

The most straightforward fix, according to the technical literature, is a mandatory perceptual-hash check at the upload stage of every government image portal in the city — a change that would cost relatively little to implement but requires coordination across agencies that currently operate independently. Without that coordination, Delhi's duplicate image problem will keep compounding with every new record added to an already cluttered system.