Face Recognition using Feature Descriptors and Classifiers
Author(s):
Shehina.T , Mar Baselios christian College of Engineering and Technology,peermade,kuttikanam,idukki; Almaria Joseph, Mar Baselios christian College of Engineering and Technology,peermade,kuttikanam,idukkiKeywords:
Face Recognition, Face Detection, Feature Extraction, Feature Descriptors, Multi Scale Vector, ClassifiersAbstract:
Feature extraction is becoming popular in face recognition method. Face recognition is the interesting and growing area in real time applications. In last decades many of face recognitions methods has been developed. Feature extraction is the one of the emerging technique in the face recognition methods. In this method an attempt to show best faces recognition method. Here used different descriptors combination like LBP and SIFT, LBP and HOG for feature extraction. Using a single descriptor is difficult to address all variations so combining multiple features in common. Find LBP and SIFT features separately from the images and fuse them with a canonical correlation analysis and same procedure also done using LBP and HOG. The SIFT features have some limitations they don’t work well with lighting changes, quite slow, and mathematically complicated and computationally heavy. The combinations of HOG and LBP features make the system robust against some variations like illumination and expressions. Also, face recognition technique used a different classifier to extract the useful information from images to solve the problems. This paper is organized into four sections. Introduction in the first section. The second section describes feature descriptors and the third section describes proposed methods, final sections describes experiments result and conclusion phase.
Other Details:
| Manuscript Id | : | J4RV3I2053 |
| Published in | : | Volume : 3, Issue : 2 |
| Publication Date | : | 01/05/2017 |
| Page(s) | : | 89-92 |
