Human-OpenCV

OpenCV Human Detection Based on COCO Dataset.
The project counts the number of humans of each side of the frame (Left & Right), which could be used for input emulation based on positions of the humans ; similar to a voting system of which input should be emulated.

Features

Existing Features

• Object Detection - Detect Objects based on COCO Dataset, Not only for humans.
• Object Region Counter - Calculates and Estimates the number of obj in a part of the frame. (aka, Region of interest (ROI))

Features Left to Implement

• Implement CUDA - Harness GPU powers to compute (Smoother, Higher FPS) [Not Supported On All Systems]

Dependency Installations

1. Download Project

   • Download the project
   • Drag all files into a folder

2. Install Virtual Environment (venv) [Optional]

   • Open CMD
   • Type into CMD: >> pip install virtualenv

3. Setup Virtual Environment (venv)

   • Open CMD
   • Navigate to folder where files are extracted
   • Type into CMD: >> python -m venv (name)
   • Replace (name) with your desired Virtual Environment Name
   • Activation: Type into CMD: >> .(name)\Scripts\activate

4. Install OpenCV

   • Ensure you have activated your Virtual Environment (prev step) [optional]
   • Type into Terminal/CMD: >> pip install opencv-python
   • Done! 🏁