REAL TIME FACE RECOGNITION SYSTEM FOR CHALLENGING ENVIRONMENTS
ABSTARCT :
Real-World face recognition in unconstrained scenarios is still a major challenge for biometrics. The reasons are manifold. Among them is the fact that gallery (stored) enrolled face images are usually captured in controlled settings, using a predefined arrangement of subjects and capture devices, whereas the probes (test images) are captured in quite different settings. The latter usually present looser restrictions, which significantly increase intra, class variability, so that pose and illumination, as well as expression and occlusions, may become disturbing factors.
EXISTING SYSTEM :
The Principal Component Analysis (PCA) is one of the most successful techniques that have been used in image recognition and compression. PCA is a statistical method under the broad title of factor analysis. The purpose of PCA is to reduce the large dimensionality of the data space (observed variables) to the smaller intrinsic dimensionality of feature space (independent variables), which are needed to describe the data economically. This is the case when there is a strong correlation between observed variables. The jobs which PCA can do are prediction, redundancy removal, feature extraction, data compression, etc.
DISADVANTAGE :
Like all technology however, biometrics also comes with some disadvantages. One disadvantage of biometrics is cost. Different biometric technologies need the use of different devices that have a range of costs. Also the use of these biometric devices may cause delay in people’s day.
PROPOSED SYSTEM :
The function performs photometric normalization of the image X using the non-local means algorithm. The algorithm constructs a smoothed image based on a weighted sum of similar patches comprising the image. The smoothed image is then used to estimate the reflectance which should be illumination invariant.
ADVANTAGE :
In practice, for this set of experiments, the proposed normalization procedure is not exploited, and images are normalized using the technique this implies that only its local correlation module is used. For LFW, we only considered those identities with at least eight images, five of which were used for training (for those techniques like PCA which require it) and three for testing.
|