I am looking to create a proof of concept of counting people in large spaces using the AWS Rekognition video analysis platform. I have uploaded a 30 second clip from a shopping mall. I would like the system to tell me:
1: How many people went up the escalator?
2: How many people went down the escalator?
3: How many people entered the 'PAUL' store?
4: How many people exited the 'Karen Millen' store?
This would then populate a dashboard with the relevant counts.
The model will be tested using more footage from the same camera.