The example demonstrates how to use VQPy to detect and send alerts when individuals loiter in specified areas beyond set time thresholds. Real-time loitering detection is crucial for smart city safety, preventing potential risks such as suicides on infrastructure, retail store burglaries, and parking lot assaults.
You can follow the instructions here to prepare your environment for VQPy.
You can use python new_frontend.py
to run the example. Below are the arguments you need to specify.
--path
: your own video dataset path.--save_folder
: the folder that you preferred to save the query result.--polygon
: the polygon coordinates of the loitering area.--time_warning
: the time threshold for warning.--time_alert
: the time threshold for alert.