search

SforAiDl / Playground

A python library consisting of pipelines for visual analysis of different sports using Computer Vision and Deep Learning.
MIT License
17 stars 17 forks source link

Football player tracking #91

Open ashwinvaswani opened 4 years ago

ashwinvaswani commented 4 years ago

When the players are detected in the first frame, instead of running detections on all frames thereafter, we need to track the players. See the link for example https://drive.google.com/file/d/1FwGkApZwX03mCpvA1BD1aRT1nVa6QMnZ/view?usp=sharing

ashwinvaswani commented 4 years ago

https://towardsdatascience.com/detecting-soccer-palyers-and-ball-retinantet-2ab5f997ab2

ashwinvaswani commented 4 years ago

Can also look at: https://www.youtube.com/watch?v=Tjx8BGoeZtI Instead of us manually selecting the box, in our case, we will get the initial boxes from the detector so smooth.

khizirsiddiqui commented 4 years ago

Can I work on this issue if no one is already?

ashwinvaswani commented 4 years ago

Sure. Assigning it to you. You can co-ordinate with @rajanarasimhan and @ShreyPandit as well as they're working on the detector and you might need that as a starting point for the tracker.

khizirsiddiqui commented 4 years ago

Sure! I will talk to them.

ashwinvaswani commented 4 years ago

@khizirsiddiqui Detector is ready and working. You can start with tracking. With tracking, another goal is to get the passes mapped on the top view of the football field for better insights into training patterns.