Face Detection ABSTRACT: Dimensionality reduction methods have been successfully employed for face recognition.

Among the various dimensionality reduction algorithms, linear discriminate analysis (LDA) is one of the popular supervised dimensionality reduction methods, and many LDA-based face recognition algorithms systems have been reported in the last decade. However, the LDA-based face recognition systems suffer from the scalability problem. To overcome this limitation, an incremental approach is a natural solution. The main difficulty in developing the incremental LDA (ILDA) is to handle the inverse of the within-class scatter matrix. In this paper, based on the generalized singular value decomposition LDA (LDA/GSVD), we develop a new ILDA algorithm called GSVD-ILDA. Different from the existing techniques in which the new projection matrix is found in a restricted subspace, the proposed GSVD-ILDA determines the projection matrix in full space. Extensive experiments are performed to compare the proposed GSVD-ILDA with the LDA/GSVD as well as the existing ILDA methods using the face recognition technology face database and the Carnegies Mellon University Pose, Illumination, and Expression face database. Experimental results show that the proposed GSVD-ILDA algorithm gives the same performance as the LDA/GSVD with much smaller computational complexity. The experimental results also show that the proposed GSVD-ILDA gives better classification performance than the other recently proposed ILDA algorithms. INTRODUCTION: Face recognition has been an active area in the field of pattern recognition and computer vision, owing to its wide range of applications in commerce and law enforcement. To solve the high-dimensionality problem in face images, dimensionality reduction methods have been employed in the last decade. One of the well-known dimension reduction algorithms is linear discriminant analysis (LDA). Incremental learning has already been proposed in the computer vision community for decades. Most of the recent works are designed for incremental principal

component analysis. PCA (principle components analysis) is an optimal signal reconstruction method in the sense

of minimum-mean-squared error, but class label information is not considered. However, there are only few works for incremental learning algorithm for LDA, which may be due to the fact that it is very hard, if not impossible, to handle the inverse of the within-class matrix. Moreover, it is difficult to design an incremental solution of the ED on the product of two matrices. This paper loosely divides these methods into two categories. If the projection matrix is determined in full space, it is called conventional LDA algorithm, and if the projection matrix is determined in subspace, it is called subspace LDA. Theoretically, the projection matrix determined in full space will give better discriminate power than that from subspace. EXISTING SYSTEM: In the Existing system of the face detection the LDA-based face recognition systems suffer from the scalability problem. To overcome this limitation, an incremental approach is a natural solution. The main difficulty in developing the incremental LDA (ILDA) is to handle the inverse of the within-class scatter matrix. PROPOSED SYSTEM: In the Proposed system the face detection solves the scalability problem in the recognition systems by LDA method. The experimental results also show that the proposed GSVD-ILDA gives better classification performance than the other recently proposed ILDA algorithms.

SOFTWARE AND HARDWARE REQUIREMENTS: Hardware Specification Processor

Speed

: : : :

Pentium III 1.5 GHZ 5 GB 256MB

Hard disk RAM Software Specification Operating System Development Environment Platform Languages

: Window XP : Microsoft VS 2008 : .NET : VB.Net