Post date: Nov 20, 2015 4:47:6 PM
In order to validate the data that the team is receiving from Placemeter's algorithm, the group has been manually counting pedestrians at each of the measurement points. This procedure involves the team physically counting the bidirectional flow of pedestrians for an hour at a time using mechanical clickers. These counts are then put into a spreadsheet maintained by the team in order to compare to the Placemeter provided data. Currently, 2 of the installed cameras have reported totals within 5% of the the physical counts. Two more have been able to return data within 10%. These cameras have all been located at residences with particularly strong wifi, supporting the team's conclusion that the more stable the connection, the more accurate the data will be.
The team has been looking into the locations of existing camera networks such as security cameras operated by stores, the city of Venice, and the public transportation network. Being able to have data from a wider range of streets, boat stops, public squares, and bridges would enable a more thorough study of pedestrian movement throughout the city. Ideally this would increase the ability of the city to monitor growing congestion due to increased tourism and make more informed decisions about handling high pedestrian volume.
Here is a link to a website which shows every camera in the city of Venice:
http://tramaci.org/anopticon/
A picture of the sites homepage: