Heatmaps for Consumer Behaviour Tracking by using Deep Learning based Human Detection

About this project

Autors:

Ivan, Changoluisa Ahmad, Alrashed

This project has been created as part of the Human-Robot Interaction course of the IFROS Master in the University of Zagreb.

---

User's Guide

---

To test the system, some input videos are included in the folder /test_videos.

Testing the People Heatmap

The first argument is the path of the input video and the second is the desired name for the resulting video that will be generated after the test.

    python ./people_heatmap.py test_videos/urban.mp4 urban_people_heatmap

After this, a result similar as the one below will be shown.

Testing the Customer Heatmap

Similarly as for the People Heatmap, the first argument correspond to the input video file and the second argument is the desired name for the resulting heatmap in video format.

    python ./customer_heatmap.py test_videos/urban.mp4 urban_customers_heatmap

After this, a result similar as the one below will be shown.