Automatic Recognition of African Bust using Modified Principal Component Analysis (MPCA)

Abstract

This study identified and analysed the pattern recognition features of African bust. It also developed and evaluated a Modified Principal Component Analysis (MPCA) for recognizing those features. This was with a view to providing information on the developed MPCA for a robust approach to recognition of African bust.The developed MPCA used varying number of eigenvectors in creating the bust space. The characteristics of the bust in terms of facial dimension, types of marks, structure of facial components such as the eye, mouth, chin etc were analysed for identification. The bust images were resized for proper reshaping and cropped to adjust their backgrounds using the Microsoft Office Picture Manager. The system code was developed and run on the Matrix Laboratory software (MatLab7.0).The use of varying values of eigenvectors has proven positive result as far as the system evaluation was concerned. For instance, a sensitivity test carried out revealed that thirteen out of seventeen bust’s images were recognized by selecting only vectors of highest eigenvalues while all the test images were recognized with the inclusion of some vectors of low energy level. That is, the modification made to the Conventional PCA (i.e. Eigenface Algorithm) gave rise an increment of about twenty five percent (25%) as far as recognizing the test images was concerned.The study concluded that the Modification made to the conventional PCA has shown very good performance as far as the parameters involved were concerned. The performance of the MPCA was justified by the identification of all the test images, that is, the MPCA proved more efficient than the Conventional PCA technique especially for the recognition of features of the African busts.

Keywords: Eigenvectors, Bust recognition, Modified Principal Component Analysis Technique (MPCA), African Bust.