Impact of Face Partitioning On Face Recognition Performance
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2016, Vol 18, Issue 4
Abstract
Abstract: Face partitioning algorithm is presented in this paper. Face is first divided into face parts namely head, eye pair, left eye, right eye, mouth and nose. Instead of giving the entire face as input in testing and training phases of face recognition algorithm, the face parts are given individually. Eigen features of all the face parts are extracted separately and given to the individual classifiers. Finally the classifier outputs are given to the decision making algorithm. This accepts all the face parts and generates a face based on the algorithm. ORL data base is used for evaluating the performance of this new technique. Results are separately calculated with and without face partitioning technique. Results show that face recognition rate is increased by using the combination of face partitioning technique and PCA. The new algorithm is also verified on 8 different data sets. There is an improvement of 15% face recognition rate using the new algorithm on ORL database.
Authors and Affiliations
Harihara Santosh Dadi, Krishna Mohan P. G.
Keywords
Related Articles
- EP ID EP159495
- DOI -
- Views 84
- Downloads 0
IOT Based Coal Mine Safety Monitoring and Control Automation
In this paper we are organizing an IoT (Internet of Things) screen, a safety efforts for excavators which is most basic in underground mining domains. In this undertaking, the system is build using particular sensors sor...
Risk Management in Agile Model
Abstract: In traditional waterfall-model risks were usually managed by using project risk management frameworks. Nowadays agile methods have started replacing the traditional models. Agile development is based on s...
An Enhanced Scheme for Hiding Text in Wave Files
Abstract: Steganography is a term applied to any number of processes that embed an object into another object in order to deceive any observer or adversary. An embedding algorithm for hiding messages into wave files or a...
Image Noise Removal by Dual Threshold Median Filter for RVIN
Abstract: Removal of random valued impulse noise in digital images with edge preservation is one of thechallenging tasks in digital image processing. For removal of impulse noise as well as preserve edge proposeda...
Pattern Analytical Module for EDOS Attacker Recognition
Abstract: Cloud computing has provided a platform to its users where they are charged on the basis of usage of the cloud resources; this is known as “pay-as-you-use”. Today, Cloud computing is the most hyped techno...