I am currently working on a computer vision project related to cricket. My goal is to detect the ball, determine its position in each frame, and then plot the trajectory of the ball in a 3D view. However, I’ve encountered an issue with the recorded video, as it has a zooming effect. This makes it difficult to accurately calculate the position of the ball in relation to the cricket pitch. Initially, the entire cricket pitch is detected in the first few frames.
Full Pitch Image

However, as the video zooms in, only half of the frame is captured,

Half Pitch Image

making it impossible to correctly calculate the position of the ball in reference to the full pitch view.

I’m looking for a way to mitigate this zooming effect.