Made using python and importing opencv library in it using haar casscade algorithm that allows us to detect images in the grayscale.
A Real-time smile and eye detection app\ Detection happens in a single frame with two loops running side by side
It allows us to open our webcam and detect our eyes and smile in the real time and multiple smiles and eyes can be detected for that a loop is implemented to detect them separately.
Added a waitkey also to control the clicks of the webcam and make it to the real-time.
Learning from this:
Haar features and algorithms
1.how the haar cascade algorithm works in real-time upon grayscaled images
2.why it works better on grayscaled images than taking colored frames instead.
3.simple lines of code can do magic just putting the right things at right places
The result from this:
1.we can detect faces using either from our gallery and importing them or we can use webcam support to get the real time frame.
2.we can also detect real-time video faces from its frontal face if being classified into
3.multiple real-time faces can be detected and also with regular changing of dimensions
Challenges faced
1.the most important challenge is to train the data and it's time-consuming so to build a simple prototype taking OpenCV trained data is beneficial as it saves lots of time.
2.haar algorithm how it works is again one of the most important challenges as it has to be quite accurate to detect the face in real-time
3.importing OpenCV required installation of multiple packages and different versions of python have different versions of that library.