MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641051225 A) filed by Jeevaraj S; Alagu Manikandan A; Sam Varghese George; and Sumathi S on April 22, 2026, for Ai Powered Volleyball Match Tactical Analysis.
Inventor includes Sam Varghese George.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: Through the use of Al technologies, we have built an intelligent platform that provides detailed tactical analysis of volleyball matches. Our Al-powered video analysis and tactical visualization solution combines computer vision, deep learning, and structured gameplay modeling to deliver a comprehensive, decision-support match analysis. The system processes live and/or recorded real-time volleyball footage and produces tactical insights. The system's core detection pipeline is based on the use of multiple YOLO-based models for player, referee, and ball identification in each frame, and multi-object tracking algorithms (in this case, ByteTrack) to ensure that all players receive a consistent identity and are tracked throughout the match. The system will also recognize and utilize visible boundaries and the net and provide a consistent set of coordinates for each player action and the respective trajectory of the ball. The system will automatically classify players into either opposing team categories, and will also utilize this classification to identify interactions between players during game play and to recognize key volleyball event types (e.g. serve, pass, set, spike, block). In addition to these capabilities, the gameplay modelling engine will additionally record the actions of the players, thereby displaying the entire flow of the rally on one screen and enabling context-sensitive event classification. Not just looking at particular games, but instead looking at how players interact with one another throughout the course of a game, as well as evaluating those performances using different metrics (such as the distance a player moves during the course of a game, how many times they jump for a spike, how many spikes they make, and how long they’ve possessed the ball). Then, using heat maps, propping up attack zones, and finding patterns within the distance players move throughout the court, tactical input can be created for easier interpretation of team tactics and player positioning. All structured data (statistics for players, player trajectory information and event logs) can be exported as CSV and PDF reports and are stored in the database. A dashboard based on Streamlit is used to create an intuitive interface to allow individuals to upload their match video, view an interactive timeline of the match, review match footage that has been annotated, and view analytical summaries. Also, highlight reels and visual overlays are created to facilitate review of the match after the game has concluded. By synthesizing three areas: Video Understanding, Spatial Analytics, and Al-Driven Modeling - the invention takes raw volleyball video and converts it to actionable insights that can be used for coach-led data-driven improvement of a player's and/or team’s performance, as well as making decisions regarding tactics used at all levels of volleyball.
Disclaimer: Curated by HT Syndication.
