NEURAL NETWORK & GENETIC ALGORITHM BASED FACE RECOGNITION SYSTEM
Journal Title: International Journal of Research in Computer and Communication Technology - Year 2012, Vol 1, Issue 5
Abstract
Facial image analysis plays a significant role for human computer interaction. Automatic analysis of human face is still a challenging and emerging problem with many applications. In this paper, we proposed a neural network and genetic algorithm based face recognition system to recognize and verify the person from a digital image. This paper shows the advantages of using Neural network along with Genetic Algorithm for face recognition system. The neural network has the ability to adapt to unknown situations & it trains itself by learning through datasets and is fault tolerant while the Genetic algorithm is an optimization technique used in computing to find the exact or approximate solutions. It handles large & poorly understood search spaces easily and handles noisy functions well. The proposed system consists of three basic steps which automatically detect the human face image using BPNN, the feature extraction based on a fusion approach of different techniques like PCA,LDA,LBA, KDDA and face recognition will be performed based on BPNN and GA. The dimensionality of face image will be reduced by the fusion extraction block and the recognition will be done by the combined BPNN and GA for efficient and robust face recognition.
Authors and Affiliations
Lalit P. Bhaiya, Vivek Pali
Keywords
Related Articles
- EP ID EP27466
- DOI -
- Views 386
- Downloads 10
Security Enhancement in Mobile Ad hoc Networks using Non-zero Non-cooperative Game Theory
Mobile Ad hoc Network (MANET) is a multi hop infrastructure less wireless networks which offers unrestricted mobility. As wireless ad-hoc networks lack a centralized infrastructure, they are exposed to a lot of attac...
Digital Video Frame Rate Up-Conversion Based on Modified Block Matching Technique
This paper presents a novel digital video frame rate up-conversion technique based on modified block matching algorithm. The primary target is to achieve an observable improvement in the test video playback quality (in...
Enhancement of the image by using Histogram Modification and High-pass Filtering Mask
Generally the major problem is low contrast image analysis in medical Field. Low contrast digital images reduce the ability of observer in analyzing the image. This low contrast images are obtained from low radiated...
Inter Transactional Association Rule Mining using Boolean Matrix
One of the extensions of the data mining is Temporal data mining, which mines or discovers knowledge and patterns from temporal information which includes time attribute analysis. From the various types of temporal...
A Cost Effective Design of Reversible Single Precision Floating Point Multiplier
The emerging computing technologies like quantum computing, optical computing, low power computing etc. make use of reversible logic. Also applications like image processing and signal processing make use of floating...