/Human-Counting

Project to create a web app that detects and counts humans in images and videos. Currently experimenting with 2 ML models: Pointrend (ResNet backbone) & OpenCV's HOG descriptor

Primary LanguagePython

Human Counting

Project to create a webapp which detects and counts human in images and videos. Currently experimenting with YOLO 2 ML models: Pointrend (ResNet backbone) & OpenCV's HOG descriptor

Install the dependencies by :

pip install -r requirements.txt

Download the yolov3.weights from here and place it in humancounter/ folder

To run the Selenium automation

  1. Please download the respective chromedriver from this link and place it in humancounter/ folder

  2. The image being used for automation is humancounter/upload_test.jpg. Please replace this with your choice of image and change its name to the one mentioned before.

  3. Run python automate_selenium.py

To run this on a Django Server :

python manage.py runserver

To run this on Docker :

  1. Run docker run -d --name myapp ncmohit/capstoneproject -p 8000:8000 myapp