MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641108206 A) filed by Srm University-Ap on September 09, 2026, for Embedded System And Method For Offline Multi-Class Agricultural Pest Detection.
Inventors include Harsha Siva Akash Saladi; Devi Sree Honne; Chakka Madhu Bhushan; Fazal Shaik; Mohammad Abdussami; Firoj Gazi; and Muzakkir Hussain.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention provides an offline multi-class agricultural pest detection system (100) and method comprising an embedded microcontroller unit (200) integrated with an INT8-quantized lightweight convolutional neural network (CNN) model (202) and has a communication rail (204) for communication and a power rail (206); a CMOS camera module (300) configured to capture real-time digital images of pest targets; an interface unit (400) configured to locally output diagnostic pest classification results; and a power supply unit (500) coupled to supply electrical power through the power rail (206) to the embedded microcontroller unit (200), the CMOS camera module (300) and the interface unit (400). The system (100) and method executes on-device pest detection locally without external processing hardware or cloud connectivity and performs real-time pest classification into eight categories using captured image, without relying on cloud connectivity or external processing hardware. Figure of Abstract –
Disclaimer: Curated by HT Syndication.
