Video Analysis for Sports Game

Video analysis is useful in any sport. Particularly used in one-to-one sessions, identifying team’s or player’s performance. It allows you and the coach to look repeatedly at the whole action in real time with data generated out of it. The analysis generated from video analysis talks more technically about the game. This solution can infer important information from a rugby match video.

The primary application features are following:

• Uniquely identifying the players by ID
• Tracking the ball
• Provide analytics on how long the ball was in air and how long was the ball in which player’s hand.

Some future enhancements:

• Face detection of the players.
• Detecting gestures and postures of the players and predicting the next move.

Technology: Python, Object Detection, Deep Learning, OpenCV, Pillow, Tensorflow Object Detection API, Keras