Face Recogintion System Using Backpropagation Artifical Neural Network and Principal Component Analysis

Abstract

 

Abstrak

 

Pengenalan wajah dapat dilakukan dengan menggunakan metode Backpropagation Artificial Neural Network (ANN) dan Principal Component Analysis (PCA). ANN dibuat menyerupai sistem syaraf manusia. Dengan beberapa parameter pada Backpropagation, dapat diketahui karakteristik Backpropagation sehingga dapat memperkecil error dan epoch, serta memperbesar Recognition Rate. Hasil percobaan menunjukkan hubungan antara parameter eigenvalaue, parameter alpha, dan koefisien momentum terhadap Recognition Rate yang diperoleh.

 

Kata kunci: ANN, Backpropagation, JST, Recognition Rate, Face Recognition, PCA.

 

 

Abstract

 

Face recognition can be performed using Backpropagation Artificial Neural Network (ANN) and Principal Component Analysis (PCA). ANN is made to resemble human neural system. Through several parameters on backpropagation, backpropagation characteristics could be known so that errors and epoch would be minimized and Recognition Rate would be enlarged. The experimental results show the relationship between the parameter of eigenvalues, the parameter of alpha, and the momentum coefficien, with the obtained recognition rate.

 

Keywords: ANN, Backpropagation, JST, Recognition Rate, Face Recognition, PCA.