Video Analytics for Understanding Human Behavior

This study aims to use video along with environmental sensors to develop a behavioral model that can be used to predict human action during design and before construction and to analyze pre- and post-COVID use patterns within the built environment.



Cities are designed for people. The success of a city is wholly dependent on the social, environmental, and economic comforts and opportunities they provide. We continue to build cities with theories to how people act and react to the built environment but have little hard data describing the 24-hour cycle of life in urban settings.

This study is developing a camera vision based human identification system that can track people within 2D space while maintaining individual privacy of recorded individuals. This tracking data will then be combined with environmental sensor data to develop a model that predicts human behaviors in spaces before they are built. 


Students will build software to analyze human activities in video and integrate the information with sensor data.

This research is working with a recent NSF RAPID grant on COVID-19 related issues, analyzing density and pre- and post-COVID use patterns within the built environment.

The team is focused on generating new data, grants, and papers along with the software product and analytical model.


The members are expected to have finished one semester of calculus and one programming course.

Meeting Times:

  • Spring 2022: Wednesdays 10:30-11:20 am, Sync Online

  • Fall 2022: Wednesdays 10:30-11:20 am, Sync Online


Barbarash, David, et al. "Artificial Intelligence Systems for Automated Site Analytics and Design Performance Evaluation."  Landscape Research Record, no. 9, March 2020, pp. 192-203.