<![CDATA[The technology is developing rapidly with various sophisticated features in every branch of science. One of the developing branches of science is the payment system. The face of people being has a specific personal identification that has its own characteristics. Therefore, faces can be used as objects as personal identification marks to increase the level of security. The working principle of Face Recognition is to detect faces first before carrying out the recognition to be compared with the database. The method used for Face Recognition is facial landmark. Facial landmark is a method of localizing the prominent points on the face. Before the image is processed, the image must be detected whether there is a face. The face detection method uses haar cascade. The results of facial landmarks will be classified based on training data with k-Nearest Neighbor. The results of these studies resulted in facial recognition accuracy levels of 76.47%, 64.71%, and 47.06% and computation times of 0.5016, 0.1322, and 0.1298 seconds with values of k = 5, 7 and 9. In the integrated system between facial recognition and electronic payment systems with the value of k which has the highest accuracy, namely 5, has a system integration level of 71.43%. In the field of technology, there are many things that need to be resolved to make human activities easier. Therefore, this thesis will discuss an electronic payment system based on Face recognition in trading with the facial landmark method.]]>
Copyrights © 2021
