deco

Award-Winning School Technology: Sports AI Advances with Real-Time Soccer Tactical Analysis System.

2024.01.18

deco

Director Hwai Jung Hsu from the Artificial Intelligence Research Center at Feng Chia University led a team to participate in the first "2023 Sports Data Ideathon," organized by the Administration for Digital Industries. Their proposal for the "Real-Time Tactical Analysis System for Soccer Matches" stood out in the finals and received an outstanding achievement award, with a total prize of NT$500,000.

Director Hwai Chung Hsu leading center staff members Zhen Wei Wang and students Zu An Ping, Yu Hsuan Chen and Yu Han Wu from the Department of Information Engineering and Computer Science. Their proposal for the "Real-Time Tactical Analysis System for Soccer Matches" utilizes footage captured by cameras and employs artificial intelligence computer for analysis. This system calculates data such as possession rate, number of passes, pass success rate, player running distance, and speed variations, enabling the analysis of tactics and event trajectories and generating tactical boards. This analysis system is characterized by high efficiency and user-friendliness, effectively saving time and costs required for manual interpretation and securing the "Outstanding" award in the competition.

The first "Sports Data Ideathon" invited the general public to propose demands for sports data applications. Teams from enterprises or campuses brainstormed and provided verifiable solutions to address these demands. Through collecting problems and team problem-solving, the ideathon aimed to discover more diverse applications of sports data or creative innovative services.

人工智慧研究中心許懷中主任(左二)帶領學生團隊以「足球賽事即時戰術分析系統」,榮獲「2023運動數據點子松」特優佳績。

人工智慧研究中心許懷中主任(左二)帶領學生團隊以「足球賽事即時戰術分析系統」,榮獲「2023運動數據點子松」特優佳績。

逢甲人工智慧研究中心許懷中主任擔任指導老師,帶領中心職員王振唯與資訊工程學系平祖安、陳淯萱、吳伃涵同學,以「足球賽事即時戰術分析系統」提案,透過攝影機拍攝之畫面,以人工智慧電腦視覺模組進行解析,可計算出持球率、傳球次數、傳球成功率、球員跑動距離與速度變化等數據,進而分析戰術及事件軌跡,產出戰術板。此分析系統具有高效率及高易用性的特性,有效地節省人工判讀所需的時間成本,在競賽中勇奪「特優」獎項。

 

第一屆「運動數據點子松」廣邀社會大眾提出運動數據應用需求,由企業組織或校園團隊針對需求進行發想,提出可驗證的解決方案;以蒐集問題、團隊解題的方式,期望透過點子松發掘更多元的運動數據應用或創意創新服務。

「2023運動數據點子松」決選獲獎名單。

「2023運動數據點子松」決選獲獎名單。