Project Overview:

• Detect basketball using HSV value, masking techniques, and contours.
• Track ball movement by storing initial bounce coordinates.
• Identify ball bounces when reaching a specified distance from the initial coordinates.
• Using frame number to track fastest and slowest bounce

Specific to the video:

• This project is specific to the video or the ball movement.
• The logic and distance is tuned to the temporal characteristics of the ball movement.

Main Logic:

1. Detect the basketball using HSV value, masking, and contours.
2. Track ball coordinates and store initial bounce coordinates.
3. Identify a bounce when the ball reaches a distance REACH_DISTANCE from the initial coordinates.
4. Count bounces when the ball moves a distance greater than BOUNCE_DISTANCE + 10.
5. Fastest and slowest bounce are identify using the frame number

Important Parameters:

• REACH_DISTANCE: Distance from initial coordinates considered as reaching the ground.
• BOUNCE_DISTANCE: Distance moved after which a bounce is counted.

NOTE This project only use cv2 and the analysis perform can significantly improve by adding deep learning techniques.

Watch the couting of jogging (bouncing) of ball using opencv here