Evolution on Football Players Tracking

[massimo1108]

I've several question about your videos:

1. the model request to have the same perspective. It means any time extra manual work is needed to fine-tune the identification
2. How difficult should be tracking on CSV speed or acceleration for players and ball? Would it be tricky defining the start and stop of a sudden change in speed? Only peak speed or number of runs above its average is possible.
3. Distance between players could be a fascinating perspective to be tracked, probably more accessible than speed. What do you think?

M

[SkalskiP]

Hi, @massimo1108 👋🏻!

the model request to have the same perspective. It means any time extra manual work is needed to fine-tune the identification

Could you elaborate? I don't understand that question.

How difficult should be tracking on CSV speed or acceleration for players and ball? Would it be tricky defining the start and stop of a sudden change in speed? Only peak speed or number of runs above its average is possible.

The first big problem you must solve is converting the camera perspective into a bird's eye view. Without that, you can't calculate any speed.

Distance between players could be a fascinating perspective to be tracked, probably more accessible than speed. What do you think?

Same thing as above. We first need to be able to get BEV.