Intellectual Property India Publishes Patent Application for Embedded Intelligent Microscopy System For Real-Time Identification And Enumeration Of Microscopic Marine Organisms

MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641126370 A) filed by Srm Institute Of Science And Technology, Ramapuram Campus; and Easwari Engineering College on October 23, 2026, for Embedded Intelligent Microscopy System For Real-Time Identification And Enumeration Of Microscopic Marine Organisms.

Inventors include S. S. Madhavan; Vijaya Karthick Raja U M; Divyesh Hari G; K. Rakshithasri; M. N. Raksha; G. K. Akashgautham; and Dr. S. Nancy Lima Christy.

The application for the patent was published on October 30, 2026, under issue no. 44/2026.

Abstract: Abstract: An embedded intelligent microscopy system is disclosed for the automated real-time identification and enumeration of microscopic marine organisms including phytoplankton and zooplankton. The system integrates a USB digital microscope with a Raspberry Pi 5 single-board computer augmented by a Coral USB Tensor Processing Unit and executes a sequential two-stage deep learning inference pipeline comprising a YOLOv11-Nano object detection model for organism localization and a MobileNetV3-Large convolutional neural network for species-level classification. An Image Quality Assessment module upstream of the inference pipeline rejects blurred or improperly illuminated frames using Laplacian variance and histogram-based metrics, ensuring that computational resources are applied exclusively to analytically valid inputs. Organism counts and population density estimates are computed per frame and accumulated across monitoring sessions, with all results persistently stored in a local SQLite database and exportable in Darwin Core and OBIS-compliant structured formats. A locally served Flask web dashboard provides real-time annotated visualization, species count monitoring, and system performance metrics without requiring internet connectivity. The complete system operates within a power consumption envelope of ten to twelve watts and can be assembled at a hardware cost below twenty thousand Indian rupees, enabling scalable, field-deployable, and affordable marine biodiversity monitoring across coastal stations, research vessels, and resource-constrained laboratory environments. The architecture supports future extension via active learning feedback loops and federated model improvement across distributed device networks.

Disclaimer: Curated by HT Syndication.