3D Face Recognition Using Pca, Knn, And Svm

Kwee, Chun Yoong (2015) 3D Face Recognition Using Pca, Knn, And Svm. [Final Year Project Report] (Unpublished)

PDF (Please get the password by email to repository@unimas.my , or call ext: 3914 / 3942 / 3933) Kwee Chun (fulltext).pdf Restricted to Registered users only Download (29MB)

Abstract

Recent years, with the rapid development of face recognition technology in various applications such as biometric personal recognition, human-to-machine interaction as well as access control, face recognition has become one of the most interesting and significant research topic in the field of computer vision. There are many face recognition techniques and approaches has been proposed and suggested by many researchers in the past few decades. Existing methods such as PCA and LDA has been used as dimensionality reduction that compute a face feature vector that has the largest associated variance. KNN and SVM has been used as a classifier to classify the faces. However plenty of these methods have tested mainly with 2D face data. In fact, most 2D face recognition result a slightly less satisfied recognition rate, and this is mostly due to the 2D dataset itself. Factors such as illumination, expression, scale, and pose have weaken the recognition rate. In this project, the general ideas and structures of the recognition is studied. A 3D face recognition by using PCA - KNN and PCA= SVM is proposed and the result between these methods are compared. 3D face recognition is less likely to be affected by factors such as illumination, henceforth the use of 3D face dataset in this project together with the proposed method has shown a better recognition rate. Meanwhile, by using PCA - LSVM an average of recognition rate 83.33% is observed.

Actions (For repository members only: login required)

View Item