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.
- 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))
- Implement CUDA - Harness GPU powers to compute (Smoother, Higher FPS) [Not Supported On All Systems]
-
- Download the project
- Drag all files into a folder
-
- Open CMD
- Type into CMD: >> pip install virtualenv
-
- 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
-
- Ensure you have activated your Virtual Environment (prev step) [optional]
- Type into Terminal/CMD: >> pip install opencv-python
- Done! 🏁