Footfall People Counting System

Footfall is a camera-based people counting system designed to run on a Raspberry Pi using a Pi Cam and the openFrameworks framework.

Note: The provided source code is a barebones implementation. It is not a 'plug-and-play' solution and requires customization for every specific installation. Customization is primarily handled by tweaking variables in a configuration file.

After setting up your server (Local or Dedicated), you must create a database and the required data table to store event information.

1. Access phpMyAdmin via http://localhost/phpmyadmin or http://<ServerIP>/phpmyadmin.
2. Create a new database.
3. Select the new database and navigate to the SQL panel.
4. Execute the following statement to create the event data table:

CREATE TABLE IF NOT EXISTS `data` (
  `timestamp` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP,
  `locationID` tinyint(4) NOT NULL,
  `event` tinyint(4) NOT NULL,
  KEY `timestamp` (`timestamp`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1;

To run Footfall applications on Linux, use the provided clickToLaunchApp.sh shell script. This script ensures the application can find its required shared object libraries by setting the LD_LIBRARY_PATH environment variable to point to the local libs directory.

Users are generally only required to have glut installed on their system; all other necessary .so files are included in the libs folder for portability.

./clickToLaunchApp.sh

To customize the system for your specific installation, you must modify the config.json file located inside the bin/data folder of the Footfall App. This file contains all the relevant variables required to tune the system's behavior. For detailed instructions on which variables to change, refer to the Configuration documentation.

bin/data/config.json

If using a remote dedicated server instead of a Raspberry Pi, follow these steps:

1. Database Setup: Create a new database via phpMyAdmin and run the following SQL command to create the data table:

CREATE TABLE IF NOT EXISTS `data` (
  `timestamp` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP,
  `locationID` tinyint(4) NOT NULL,
  `event` tinyint(4) NOT NULL,
  KEY `timestamp` (`timestamp`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1;

2. File Deployment:
   • Create a subfolder named footfall on your server.
   • Use an FTP client (e.g., Transmit or Cyberduck) to upload the contents of the project's public folder into that footfall directory.

To build the Footfall application on Raspberry Pi, you must install specific dependencies using the getrepos.sh script within your openFrameworks addons directory.

1. Prepare the script: Navigate to your openFrameworks addons folder and create a new file named getrepos.sh.
2. Populate the script: Copy the contents of the getrepos.sh file from the Footfall repository and paste them into this new file.
3. Execute: Run the script to pull down the required packages.

Note: This assumes you have already built and compiled openFrameworks (oF) on your Raspberry Pi.

cd ~/openFrameworks/addons
touch getrepos.sh && vim getrepos.sh
# Paste content, save with :wq
./getrepos.sh

If you are accessing the Raspberry Pi via SSH, you can launch the Footfall application and push it to the background using one of the following commands. Setting DISPLAY=:0 ensures the application renders to the local display connected to the Pi.

DISPLAY=:0 make run & disown
# OR
DISPLAY=:0 bin/Footfall & disown

To manage the application lifecycle, create two shell scripts in the /home/pi/ directory and ensure they are executable.

touch startFootfall.sh stopFootfall.sh
chmod 755 startFootfall.sh stopFootfall.sh