Pengenalan wajah merupakan bagian dari pattern recognition. Pengenalan wajah yang telah banyak diterapkan dimasyarakat, diaplikasikan pada sistem keamaan penerbangan, kepolisian dan pelayanan kependudukan. Penelitian ini menggunakan citra sebanyak 400 citra wajah, proses pengenalan wajah pada sistem ini dilakukan 4 tahap. Tahap pertama adalah data citra diproses dengan menjadikan citra wajah menjadi citra diagonal. Tahap kedua citra wajah diproses ekstraksi fitur menggunakan metode Principle Component Analysis (PCA). Tahap ketiga ekstraksi fitur citra wajah menggunakan metode Linear Discriminant Analysis (LDA). Tahap keempat hasil nilai matrik bobot dari masing - masing proses ekstraksi fitur. Tahap selanjutnya tahap pengklasifikasian dengan menggunakan jarak kemiripannya dengan menggunakan metode Manhattan Distance dan Naive Bayes Classifier, dalam skenario ujicoba sistem dari 4 skenario menggunakan Manhattan Distance menghasilkan rata - rata akurasi 93.75%. Sedangkan menggunakan Naive Bayes Classifier menghasilan rata - rata 88.15% Hasil analisa tersebut menyimpulkan bahwa input citra diagonal metode Diagonal LDA Linear Discriminant Analysis bisa diterapkan dalam pengenalan wajah. Kata Kunci: Pengenalan wajah, Diagonal LDA Linear Discriminant Analysis, Manhattan Distance, Naive Bayes Classifier.

Face recognition is part of the pattern recognition. Face recognition has been widely applied in the community, was applied to the flight's security system, police and service population. This study uses image database ORL 400 face images, face recognition process is performed on the system 4 process. The first stage is to make the processed image data into the image of the face image diagonal. The second stage of face images are processed using the method of feature extraction Principle Component Analysis (PCA). The third stage of face image feature extraction using the method of Linear Discriminant Analysis (LDA). The fourth stage results of each weight matrix value - each feature extraction process. The next stage of the classification stage by using a similarity distance using the Manhattan Distance and Naive Bayes classifier, the system test scenario of 4 scenarios using Manhattan Distance yield average 93.75%. While Naive Bayes classifier using produce average 88.15% Average results of this analysis concluded that input image diagonal method Diagonal Linear Discriminant Analysis LDA can be applied in face recognition. Keywords : Face recognition, LDA Diagonal Linear Discriminant Analysis, Manhattan Distance, Naive Bayes Classifier.

Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal Baca Jurnal