MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641018694 A) filed by Dr. R. Murugadoss, Chennai, Tamil Nadu, on Feb. 19, for 'yolo-based ai-driven threat detection and real-time alert system for intelligent surveillance.'
Inventor(s) include Ms. Chandramugi P. S; Ms. Roobini Supraja R. K; and Ms. Shimimah Sherin C.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "Something new shows up - a smart camera setup using artificial intelligence to spot metal dangers such as guns or sharp tools while video is playing. Most old-style monitoring leans heavily on people watching screens nonstop or reviewing clips once something happens, which often means reacting too late. This one fix that gap by linking quick visual recognition software with instant alerts during ongoing footage capture. Video footage flows nonstop from regular security cameras into the system, where every image gets cleaned up first - adjusting for blur, shadows, or graininess by a preprocessor. Right after, a streamlined neural network scans each scene once, having learned beforehand what guns or knives look like, marking possible dangers inside outlined boxes along with how sure it is about them. Since everything happens in one go, there's no lag from extra steps. When certainty crosses a set level, the software flags that item as risky through its judgment-based component. Right after spotting a threat, the screen shows clear alerts right inside the video feed, so people watching get aware fast. Because it helps teams react quickly, extra tools might display things like estimated age or mood clues alongsidethe footage - yet these details never change how dangers are classified. Watching happens nonstop without delays, needing fewer people to watch, working inside banks, city structures, secure zones, also smart observation tech. Speed jumps up, responses become steady plus quicker, it fits right into cameras already installed, no overhaul needed. Built-in trust grows because alerts come faster, tied directly to what the gear sees, avoids waiting. Works behind the scenes, joins current setups smoothly, cuts lag between detection and reaction. Keeps pace with live moments, adapts without extra hardware, stays alert quietly. Fits where security matters most, runs without constant checks, reduces slips through gaps."
Disclaimer: Curated by HT Syndication.
