MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641052238 A) filed by Hindusthan College Of Engineering And Technology on April 24, 2026, for Optimizing Fisheries With Ai: Enhancing Fish Detection, Efficiency And Sustainability At Sea.
Inventors include M Ravikumar; Aathithyan T; Abi Prabha K; Abinaya P; and Abinaya Sri R.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: ABSTRACT OF THE INVENTION Monitoring fish populations is essential for marine research, fisheries management, and environmental conservation. Traditional fish monitoring techniques usually involve manual obsei'vation and counting, which are time-consuming, laborintensive, and often prone to human error. With the advancement of artificial intelligence and computer vision, automated detection systems have become an effective solution for improving accuracy and efficiency in underwater monitoring. This project presents an ATBased Sonar Fish Detection System that automatically detects and counts fish from sonar images using deep learning techniques, thereby reducing manual effort and improving reliability in fish population analysis. The proposed system uses the YOLOv7 (You Only Look Once Version 7) object detection model, a powerful deep learning algorithm known for its high detection accuracy and fast processing speed. The model is trained to identify fish patterns in sonar images and detect their presence using bounding boxes. When a user uploads an image through the web interface, the system processes the image using the trained YOLOv7 model and generates detection results. The system then calculates the total number of fish detected in the image and displays the result along with the processed image showing the detected fish. This enables users to quickly analyze fish distribution in sonar imagery without performing manual counting. To make the system accessible and user-friendly, a web-based application is developed using the Flask framework. The backend of the system is implemented using Python, while SQLite is used as the database to store user credentials and detection records. The web interface is designed using HTML and Bootstrap to provide an interactive dashboard where users can upload images, view detection results, and access previous detection history. The proposed system offers an efficient and scalable approach for automated fish monitoring and can be applied in marine research, aquaculture management, and environmental studies
Disclaimer: Curated by HT Syndication.
