Delhi's Duplicate Image Crisis: What Happens Next and the Key Decisions Ahead

Delhi's municipal digital infrastructure has a problem it can no longer ignore. Across the South Delhi Municipal Corporation's online property records portal, the Delhi Tourism website, and the heritage documentation database maintained by the Archaeological Survey of India's Delhi circle, duplicate and mislabelled images have accumulated in the tens of thousands — the result of years of piecemeal digitisation drives conducted without a unified tagging protocol. Officials who oversee the civic records have now been given until September 30, 2026 to produce a remediation plan, according to internal timelines circulating within the corporation.

The issue matters now because two major projects converge on this problem simultaneously. Delhi Metro Rail Corporation is building out Phase 4 corridors, including the Janakpuri West to RK Ashram Marg stretch, and the tender documentation for station design reviews relies on georeferenced image libraries that are riddled with misfiled photographs. Separately, the Delhi government's Yamuna Rejuvenation Authority has been preparing a public-facing progress dashboard — its credibility depends on verified before-and-after imagery from specific ghats between Wazirabad and Okhla. Duplicate images appearing under incorrect location tags undermine both efforts.

Where the Problem Is Sharpest

Ground zero for the duplication mess is the SDMC's central server farm in Sarita Vihar, which stores roughly 4.2 million civic photographs uploaded between 2017 and 2025. Auditors brought in earlier this year flagged that approximately 18 percent of those files share a hash value with at least one other file in the same database — meaning nearly 756,000 images are exact or near-exact duplicates clogging retrieval systems. A parallel audit of the Delhi Urban Art Commission's image library, which documents building permission cases across neighbourhoods from Lajpat Nagar to Rohini Sector 3, found mislabelling rates above 22 percent in photographs uploaded before 2021.

The Archaeological Survey of India's Delhi circle faces a different but related challenge. Its photographic archive for protected monuments — covering sites from the Qutb Minar complex in Mehrauli to Humayun's Tomb in Nizamuddin East — contains survey photographs taken across multiple decades. When those older images were digitised and uploaded into a shared government cloud system in 2023, file-naming conventions from three separate predecessor systems were merged without reconciliation. The result: the same doorway arch at Tughlaqabad Fort appears under at least six different catalogue entries, each associated with a different purported date of survey.

The Decisions That Cannot Wait

Three choices now sit on the table for Delhi's digital governance officers, and each carries real cost and political weight.

First, the Corporation must decide whether to adopt an AI-assisted deduplication tool — the National Informatics Centre has offered a pilot framework that has already been tested in Pune's municipal records — or commission a manual review. Manual review of even a fraction of the 4.2 million files at standard contractor rates would run into several crore rupees and take well past the 2027 election cycle to complete. The AI pilot, by contrast, completed a 500,000-image test batch in Pune in under three weeks, though its accuracy rate on heritage photographs with subtle contextual differences has not been independently verified for Delhi's archive.

Second, the Yamuna Authority dashboard launch — provisionally scheduled for October 15 — must either be delayed or proceed with a caveat banner acknowledging data quality limitations. Both options carry political risk for the AAP administration, which has tied its credibility to measurable Yamuna cleanup progress ahead of assembly elections.

Third, DMRC's Phase 4 project design teams need a confirmed handover date for clean georeferenced imagery around the proposed Maujpur-Majlis Park corridor stations. Every week of uncertainty pushes detailed design reviews further back.

The National Informatics Centre's Delhi unit is expected to present a formal recommendation to the Lieutenant Governor's office before the end of July. What happens after that depends largely on whether the Corporation, the Delhi government, and the central agencies can agree on a single image-management standard — something they have failed to do in three previous attempts since 2019. The September 30 deadline is firm on paper. Whether the institutional will exists to meet it is the question that will define Delhi's digital civic record for the next decade.