MAHAKUMBH MELA 2025 (PRAYAGRAJ): FIELD PHOTOGRAPHS OF RITUAL LIFE, AKHARAS, INFRASTRUCTURE AND PUBLIC SERVICES
Description
This dataset contains 144 field photographs taken by the authors at the Maha Kumbh Mela 2025, Prayagraj, India, one of the largest religious gatherings in the world, during January and February 2025 (the Mela ran from 13 January to 26 February 2025). The photographs document the Mela as a temporary city. They cover religious life (akhara camps and processions, royal baths, Naga and women sannyasis, sannyas initiation at the Sangam, the Kinnar Akhada and its rituals, and kalpvas practice such as shayya-daan, daily worship and cooking on dung-cake hearths); infrastructure (pontoon bridges under construction and in use, tent cities and luxury camps); public services (the Mela Authority office, media centre, computerised lost- and-found centre, hospitals and first aid, fire and emergency services, railway arrangements, temporary bank branches and tourist information); welfare and labour (sanitation workers and the Vidya Kumbh school for their children); culture and commerce (Kala Kumbh and Kala Gram exhibitions, state pavilions, folk performances, vendors and bhandaras); and Prayagraj city during the Mela. Each image has a stable identifier (KUMBH_0104.jpg to KUMBH_0247.jpg). The numbering continues from the companion dataset of banners and hoardings (doi:10.17632/xmf482njn2.1), so identifiers are unique across both. An accompanying catalogue (XLSX and CSV) gives one row per image: event, collection details, photographers and rights, scene category, visible features, a crowd annotation (none / sparse / moderate / dense), technical details, processing applied and sharing status, together with the photographers' original Hindi captions. The original camera files are no longer available. The images come from a compilation the authors prepared after fieldwork, which had reduced the resolution (median width about 880 px) and removed camera metadata. Exact dates are therefore known only for five images that carry a GPS Map Camera stamp (8–9 February 2025). To protect privacy, the faces of real people were blurred in 75 images (217 regions); faces printed on posters and hoardings were left unchanged. Eleven images are marked "Needs review" because they show children (faces blurred) or the ritual nudity of Naga ascetics. The data can support research on religion and pilgrimage, mass-gathering management and public services, temporary urbanism, the anthropology of Hindu ascetic orders and the Kinnar Akhada, and visual culture. Companion dataset ----------------- This dataset complements: Maral, Priyaranjan; Kumar, Sumit (2026). "Mahakumbh Mela 2025 (Prayagraj) Field Photographs of Banners, Hoardings and Public Messaging". Mendeley Data, V1. doi: 10.17632/xmf482njn2.1 That dataset contains images KUMBH_0001–KUMBH_0103. This dataset continues the numbering (KUMBH_0104–KUMBH_0247), so every identifier is unique across both.
Files
Steps to reproduce
1. Fieldwork: the authors photographed religious camps and processions, ritual practice, infrastructure, public services, exhibitions and street life in the Maha Kumbh Mela area and Prayagraj city during January and February 2025. 2. Compilation: soon after fieldwork the authors assembled the photographs, with Hindi captions, into a thematic compilation (PowerPoint, 42 slides). The camera files were later deleted, so this compilation is the only surviving copy. 3. Extraction and selection: all 147 embedded images were extracted unchanged. Two slide-template background textures and one lower-resolution duplicate were removed, leaving 144 photographs. 4. Identifiers: files were renamed KUMBH_NNNN.jpg (0104–0247) in the order of the compilation. The slide number and original Hindi caption of each image are listed in the catalogue. 5. Orientation: one image stored sideways and mirrored (KUMBH_0237) was restored to upright. 6. Privacy: faces large enough to identify were blurred by hand (pixelation plus Gaussian blur over manually placed regions) in 75 images. All files were saved at JPEG quality 95. No cropping, colour adjustment or other editing. 7. Cataloguing: each image was described using the fields defined in the "Definitions & Method" sheet of the catalogue. Scene categories follow the authors' captions. Crowd annotation is a visual estimate of the people in the frame (None 0; Sparse 1–5; Moderate 6–20; Dense >20). Image quality was rated from a sharpness score (variance of the Laplacian). Date, time and GPS were read from the GPS Map Camera stamp where one is printed in the image. 8. Verification: catalogue entries were checked against the images by the authors.
Institutions
- National Forensic Sciences UniversityGujarat, Gandhinagar
- University of AllahabadUttar Pradesh, Prayagraj