Analysis of PCA Method in Image Recognition with MATALAB

ZHAO Ping

(1.Qilu University of Technology，Ji＇nain，250353;2.Zaozhuang University，Zaozhuang Shandong 277160)

1.INTRODUCTION

Information and Communication Technologies (ICT)are increasingly entering in all aspects of our life，opening a world of unprecedented scenario in which people can interact with electronic devices，embedded in environments that are sensitive and responsive to the presence of users. Image analysis is a process of discovering，identifying，and understanding patterns that are relevant to the performance of an image － based task.［1，2］Face recognition has recently received more and more extensive attention. It plays an important role in many applicable areas，such as human － machine interaction，authentication and surveillance.［3，4］However，the wide－ range variations of human face，due to pose，illumination，and expression，result in a highly complex distribution and deteriorate the recognition performance. In addition，the problem of machine recognition of human faces continues to attract researchers from multi － disciplines，such as image processing，pattern recognition，neural networks，computer vision，computer graphics，and psychology. As for the identification problems，the input to the system is an unknown face，and the system reports back thedetermined identity from a database of known individuals，whereas in verification problems，the system needs to confirm or reject the claimed identity of the input face. The solution to the problem involves segmentation of faces (face detection)from cluttered scenes，feature extraction from the face regions，recognition or verification. Robustand reliable face representation is crucial for the effective performance of face recognition system yet still a challenging problem.［5，6］Feature extraction is realized by means of some linear or nonlinear transform of the data with subsequent features election for reducing the dimensionality of facial image so that the extracted feature can be as representative as possible.

2. SITIMULATING WITH MATALAB

PCA is a useful statistical technique that has found application in diverse fields such as face recognition and image compression，and is a common technique for finding patterns in data of high dimension. PCA is a powerful tool for analyzing data［7］. Another straightforward image projection technique，called two －dimensional principal component analysis (2DPCA)，is developed for image feature extraction. As opposed to conventional PCA，2D PCA is based on 2D matrices rather than 1D vector. In the general PCA /2D PCA methods，eigenvectors are calculated from training images that include all the poses or classes. But for classification，a large number of hand poses for a large number of users need large number of training datasets from which eigenvectors generation is tedious and may not be feasible for a personal computer.［8］

This face pattern is classified using the eigenface method， whether it belongs to known person or unknown person. The eigenvectors are calculated from the known persons＇face images for each face class and k －number of eigenvectors corresponding to the highest eigenvalues is chosen to form principal components for each class. The Euclidean distance is determined between the weight vectors generated from the training images and the weight vectors generated from the detected face by projecting them onto the eigenspaces. If minimum Euclidian distance is lower than the predefined threshold value，thecorresponding person is identified;otherwise，the result is an unknown person.In this experiment，the threshold values are defined from the given equation. We use the image from standard “ORL Database of Faces”.Through matlab programming，we realized the training，testing，adding noise，compressing，and rebuilding process. The results are as below:

Fig 1.The original image Fig2.The noise is added Fig 3. The rebuilt image

The original test image ORL078. bmp and the other images used for training are all from the ORL database. The noise is added by the function in Matlab with density of 0.05. The rebuilt image has a root mean square error of 30. 59 with the original image and looks very nice，which means PCA method has a better de －noise effect.

3. CONCLUSION

In this paper，the face recognition systems，PCA algorithms are examined. The feature projection vectors obtained through the PCA methods with matlab language and these vectors are applied to test image. PCA face recognition systems use Euclidean Distance based classifier. Eigenvalues is also a very important aspect of the faceand the result shows that the first largest Eigenvalues keeps the large amount of principal component information;the secondlargest keeps the second largest principal component information. The result of the experiment shows that PCA method is effective in compressing and rebuilding，and has an outstanding de －noise effect.

［1］Beumer. G. M.，Tao，Q.，Bazen，A. M.，and Veldhuis. R.N.J.，A landmark paper in face recognition，2006 G.M.Beumer，Q. Tao，A.M. Bazen，R.N.J. Veldhuis，A Landmark Paper in Face Recognition，.［J］.Seventh IEEE International Conference on Automatic Face and Gesture Recognition (FG＇06)，2006，73 －78.

［2］Rafel，C. Gonzalez，Richard，and Woods. E.，Digital Image Processing. ［M］.Pearson Education，2001.

［3］Hui Kong，Xuchun Li，Lei Wang，EamKhwangTeoh，Jian－Gang Wang，Ronda Venkateswarlu. Generalized 2D Principal Component Analysis.［J］.IEEE International Joint Conference on NeuralNetwork (IJCNN)，2005:23 －26.

［4］Serge Lang，Linear Algebra，Addison－Wesley，Second Edition，［M］.Newyork:Springer，1985.

［5］S. Z. Li and A. K. Jain，(Eds.)，Handbook of FaceRecognition. ［M］.Newyork:Springer，2005.

［6］Liwei Wang，Xiao Wang，Ming Chang，and JufuFeng. Is Two－Dimensional PCA a New Technique. ［J］.ActaAutomatica2005，31(5):782 －787.

［7］Matthew，A. Turk and Alex，P. Pentland，Facerecognition Using Eigenfaces，［J］.Proc CVPR 2013，2(2):586 －591.

［8］J. Yang，D. Zhang，A. F. Frangi，and J. Yang，Two－dimensional PCA:A new approach toappearance－basedfacerepresentationandrecognition，［J］.IEEE Transactions on PatternAnalysis and Machine Intelligence，2014，26(1):131 －137.