TehOPeng

CS4243 Computer Vision Project

How to use

Step 1:

Ensure that you have all these files in the same folder

• Stitch.py
• utils.py
• football_left.mp4, football_mid.mp4, football_right.mp4
• Tracking.py
• Render.py
• pitch.jpg

Step 2:

• Create folder pics in the same folder
• Run Stitch.py to generate individual frames image of stitched video

Ensure that you have all these folders and files created in the same folder

• pics/finalxxxx.jpg created by Stitch.py

Step 3:

• Create folder enlarge in the same folder
• Run Tracking.py to get output frames

Ensure that you have all these folders and files created in the same folder

• enlarge/enlargedxxxx.jpg created by Tracking.py

Step 4:

• Run Render.py to get final video output
• football_full.avi is rendered

Explanation:

• Stitch.py (with the help of utils.py) takes in the three football videos and * stitches them into one video and then generates individual video frames. It can take quite awhile, roughly 2-3 hours and can take up quite a big of space, roughly 8GB.

• Tracking.py then reads in the frames generated and output the topdown view and tracking of the players on the field in individual frames

• Render.py finally reads in all the frames from the previous step to recreate the video in a lower resolution