Coders Packet

Face Recognition using OpenCV

By Jahnavi Ravi

Face recognition is a software recognizing an individual from a digital image from an image or video source. In this project, OpenCV library is used for recognizing faces.

Face Recognition System is capable of identifying and verifying an individual from an image or video source.

In this "Face Recognition using OpenCV project", the dataset is created by the user itself. 

HAAR Cascade Algorithm is used in this project. HAAR Cascade Algorithm is a Machine Learning algorithm which is used to identify objects in images or videos. HAAR Cascade Classifier can be accessed from the OpenCV library. The webcam is used for live capture of an individual till a few seconds and around 60 images of an individual is generated from the live capture which is used as the dataset. The size of the images created in the dataset is set to be 130*100 pixels. We assign a name for each of the image dataset that is being created. We can create images of as many persons as we want.

The model that is used to recognize faces in the project is the "FisherFaceRecognizer" model. FisherFaceRecognizer is an OpenCV built in face recognizer model. FisherFaces algorithm extracts principal components that separates one individual from the other and recognizes them. The FisherFaceRecognizer model is used on the camera stream and recognizes which individual is in front of the camera and displays the individual's name continuously.