Industry
Sports & Technology
Region
Europe
Duration
5 Months
Overview
Challenges Faced
Massive Video Data Processing
Handling large-scale 4K video streams from multiple camera angles.
Player Identification
Accurate tracking and recognition of players, especially in dynamic settings.
Data Integration
Merging front and back-view footage of players based on real-time timestamps and jersey numbers.
Highlight Generation
Ensuring no repetitive footage or duplicate timestamps across different camera angles.
Solution Provided by Cplus Soft
AI Rekognition Model
Used to recognize players in live video streams and track their movements during matches.
Clipper Model
Once players are identified, the Clipper model is used to extract relevant video clips based on player activity, such as goals, assists, or key moments, creating a 10minute highlight reel per player.
OCR Model
The OCR model detects the player’s jersey number by analyzing shirt patterns. This allows us to capture back-view footage when the front view is insufficient or blocked.
Intelligent Video Stitching
The system ensures no repetitive clips are added and that no same timestamp footage is captured from multiple camera angles, creating a smooth and unique highlight reel.
Impacts after Implementation
Enhanced Fan Engagement
Provides fans with personalized, real-time player highlights.
Efficient Player Analytics
Offers coaches and analysts detailed insights into player activity over the course of a match.
Automation of Highlight Generation
Reduces manual effort in video editing and clip selection.
Improved Viewer Experience
Delivers unique and dynamic content by incorporating both front and back views of players in the highlight reels.