MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641125210 A) filed by Rns Institute Of Technology; Anupam Mohapatra; Dr Vipula Singh; Vaibhav Srivastava; Udayashankar Aradhya; Vanitha K S; Ghousia Begum; and Ayush Kumar Gupta on October 21, 2026, for System And Method For Real-Time, Non-Intrusive Detection And Seat-Level Localization Of Unauthorized Image-Capturing Devices In An Enclosed Low-Light Viewing Environment.
Inventors include Dr Vipula Singh; Ghousia Begum; Vanitha K S; Udayashankar Aradhya; Vaibhav Srivastava; Ayush Kumar Gupta; and Anupam Mohapatra.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: Abstract A significant share of a film's theatrical earnings is lost to piracy driven by camcorder and smartphone recording inside movie theatres, a problem that existing countermeasures have failed to keep pace with. This disclosure presents a real-time, non-intrusive detection system that combines the physics of optical retro-reflection with a cascaded, edge-deployed machine-learning pipeline, closing this gap without disrupting the viewing experience. A sensing unit places an infrared (IR) illuminator coaxially with a near-infrared-sensitive camera, so that any lens present within the illuminated field produces a sharp, high-intensity retro-reflective glint even in total darkness — because camera optics return incident light back almost directly along its angle of incidence. An inexpensive image-processing routine continuously screens incoming IR frames for this signature at negligible computational cost, functioning purely as a high-recall trigger that flags any circular, high- intensity region as a glint candidate. Only once a candidate is flagged is a compact, quantised neural-network classifier — executed on the same embedded processing unit — invoked to verify the candidate and classify it as either an unauthorized image-capturing device or a benign non-threat object such as spectacles, jewellery, or a wristwatch. Gating the more expensive verification step behind this inexpensive screening step is what allows the entire pipeline to run continuously on commodity embedded hardware, without either missing genuine threats or generating excessive false alarms. Verified detections are transmitted as compact data payloads to a backend server, where they are persisted in a database as an immutable, time-stamped record and, substantially simultaneously, broadcast over a low-latency, persistent- connection channel to a location-mapped notification interface — closing the loop from optical detection to an actionable, seat-level alert for on-site personnel within a design budget of under two seconds. The disclosed system requires no radio-frequency emission of any kind, is invisible to the human eye, does not alter or disrupt the displayed content, and produces a forensically usable audit trail at the moment of detection.
Disclaimer: Curated by HT Syndication.
