How Face recognition System Works - OpenCv Python смотреть онлайн

Face recognition is the process of detecting and identifying human being in-front of camera in the process of authentication. Actually, it remains a challenge on computer vision and image processing, endeavoring face features such as in-variance, aging, illumination and partial occlusion are going to be addressed.
Techniques of Face detection and recognition
Techniques which are being used are Viz, Eigen face, Artificial neural network, Support vector machine, Principle component analysis. Apart from these techniques face recognition problems are face verification and face identification.
Face recognition databases

The following are the testing databases which can be used to train data are AT & T, AR, FERET, LFW and YTF though LFW is mostly used public database which is used in training of data. Face recognition database vary with controllable photos to controllable videos where YTM is used for photos and YTM is used for videos.

 Components of Faces recognition system modules
o Pre-processing.
o Feature Selection.

o Classification.

 Methods and Techniques of Face Recognition.
A. Eigen Face
This algorithm is used to recognize features in the image, it is based on principal component analysis. It is suitable for front face.
B. Artificial Neural Network
It is more effective for face recognition since the emergence of artificial intelligence for evaluation LFW dataset have been used for cropped image. Though it need more time to train it.
C. Support Vector Machine
This is supervised machine learning; it can be used to identify feature of face after features extraction.
D. Gabor Wavelet
Is the technique used for face tracking and position estimating in face recognition, it works well in extraction edge features of face and shape of face.
E. Hidden Markov Model
It is reinforcement learning, where the system undergoes Markov process to detect hidden state, it can be used to recognize face expression.

Application of Face Recognition it has been used in access control, surveillance, time and attendance and security. Conclusion and Future Direction most popular methods with great result are neural network, PCA and SVM and CNN.
#facerecognition #raspberrypi #python #haarcascade #facedetection