Implementasi Metode Local Binary Patterns Histogram (LBPH) Untuk Verifikasi Kemiripan Swafoto dengan KTP Berbasis Android

Abstract

With the large use of financial technology such as startup payments, lending, financial planning (personal finance), retail investment, financing (crowd funding), electronic money (e-wallet) and others, OJK (Financial Services Authority) requires every provider company to implement the Know-Your-Customer principle, one of which is to find out the identity and background of a user. To fulfill this principle, several electronic money provider companies require users to take selfies with identity cards, either in the form of identity card or driving license. Selfies are taken using the camera of the user's cellphone. The system will verify whether the selfie is similar to the photo on the identity card. The similarity in question is whether the selfie and the photo on the identity card are the same person. The method used in this study is the Local Binary Patterns Histogram (LBPH). The program will be built on Android mobile. Image training, namely 10 selfie images and identity card as testing images will go through a pre-processing then the results will be processed by Local Binary Patterns Histogram. After testing 33 data, it was concluded that the proposed method has the ability to verify selfie with identity card photos with an accuracy of 84.8%.