Hitungin

Hitungin is a camera based people counting system using OpenCV Framework. This project is aimed to create a robust counting system to generate the mapping crowd in some building.

System Used

• Ubuntu 18.04
• OpenCV3 Python
• Flask Framework

Installation

Hitungin consists of 2 dependecies framework :

• OpenCV3 Python Install the OpenCV Python following this tutorial.
• Flask Install flask using pip with this command after entering your python virtual environment.

  $ pip install flask

After installing the dependecies you can start to use the project

First you must clone the project to your system

$ git clone https://github.com/oracl4/hitungin.git

After cloning the project you can start running the project using runScript.sh in hitungin workspace

hitungin$ ./runScript.sh

Development

Counter Framework

• You can change the video source and destination for the project with changing the input tags on src/counter/runScript.sh like this

  python src/counter/people_counter.py
  --prototxt src/counter/mobilenet_ssd/MobileNetSSD_deploy.prototxt \
  --model src/counter/mobilenet_ssd/MobileNetSSD_deploy.caffemodel \
  --input input/nameoftheinputfile.mp4 \
  --output output/nameoftheoutputfile.avi

• The default configuration is using webcam and no output file

Streamer Configuration

• You can also change the source of video stream in flask app with another input file that "renamed" into input.mp4 that placed on input folder
• The default configuration is also using webcam

After Running > Website Integration

• Change the source ip and port on website HTML with the local ip address and port in flask
• Change the database PHP configuration in people_counter.py and uncomment it

Reference

• https://www.pyimagesearch.com/2018/08/13/opencv-people-counter/
• https://blog.miguelgrinberg.com/post/video-streaming-with-flask