A Novel Face Recognition Approach Based on Genetic Algorithm Optimization

In the field of image processing and recognition, discrete cosine transform (DCT) and principal component analysis (PCA) are two widely used techniques. In this paper we present a face recognition approach based on them. Feature selection (FS) is a global optimization problem in machine learning, which reduces the number of features, removes irrelevant, noisy and redundant data, and results in acceptable recognition accuracy. It is the most important step that affects the performance of a face recognition system. Genetic Algorithms (GA), one of the most recent techniques in the field of feature selection, are a type of evolutionary algorithms that can be used also to solve this issue. The application of a GA in the resolution of a problem requires the coding of the potential solutions to this problem in finite bit chains in order to constitute the chromosomes coming from a population formed by candidate points. The aim is to find a selective function allowing good discrimination between chromosomes and to define the genetic operators that will be used. In this sense, this approach seeks to develop a system of face recognition using Genetic Algorithm and a DCT-PCA combination for feature selection and dimensionality reduction, to be applied to an archive of images of human faces. The proposed approach is applied on various Face Databases. Experimental results demonstrate the effectiveness of this approach compared to state of the art in face recognition.