Face Recognition Application with the Complete Kernel Fisher Discriminant (CKFD) Method
Johanes Terang Kita Perangin Angin, Johan, Sukiman, Sugianto Sugianto, B. Ricson Simarmata, Suharjito Suharjito · 2020
Facial recognition applications are currently being developed because they can be applied in various problem areas such as criminal recognition, security systems, attendance, or human-computer interaction. However, to develop a face recognition-computing model is quite difficult, because the human face presents something complex, so to develop an ideal computing model for human face recognition is quite difficult. One of the best facial recognition methods is the Complete Kernel Fisher Discriminant (CKFD) method. The Complete Kernel Fisher Discriminant (CKFD) method has two advantages when compared to the previous Kernel Fisher Discriminant (KFD). First, the implementation of this algorithm can be divided into two phases, namely Kernel Principal Component Analysis (KPCA) plus Fisher Linear Discriminant analysis (FLD) so that the results are more transparent and simpler. Second, CKFD can make two categories of discriminate information so that the results are stronger. This research aims to produce an application with a face recognition algorithm with CKFD Method that is quite accurate. This research will produce a face recognition application with the CKFD method. The existence of this application is expected to help various parties who need a face recognition process. The CKFD method used can provide excellent facial recognition results when compared to other methods. The best result was given by using limit value with a False Acceptance Rate(FAR) value equal to 3.3 and a False Rejected Rate (FRR) value equal to 36.7.