- Open Source Computer Vision based personal project to detect Human Faces and different objects coming in front of the webcam for a specific time frame.
- Programming Language: Python
- Libraries: OpenCV, Bokeh, Pandas
- For Human Face detection I've used HaarCascade Classifier XML file created by OpenCV.
-
Before you execute the project, please cover your webcam and remove the cover once your webcam starts and you can see some video recording windows on your desktop.
-
You will see four different windows on your screen.
-
Color Frame: Which detects the different objects
-
Current Frame: Which detects the human face
-
Delta Frame*: Which compares the current situation the with the initial situation
-
Threshold frame* - for identification of an object
-
*not important for end-user
-
Try Like this...
-
At any moment if you want to stop, please press
q
-
A motion graph will be generated and automatically your default browser will pop-up where you can see it like this...
How to run on your machine!! (Don't worry if you don't have Python installed, I've already taken care of it!)
- Please download or clone this repo from here.
- Extract the folder and reach to the root of the project, by using the command line.
- Write command
venv/bin/python plotting.py
- Virtual Environment for Windows user will be available soon.