Android Quiz Application Based on Face Recognition

Abstract

Background: Technology in the field of education is permanently developing, growing and this raise will repeatedly offer new and unusual advances. The significant objective of this project is to encourage the understudies in participating in learning and enhancing their insight abilities. Methods: An Android application provides a new technique of developing a test or quiz using smart devices. This project implements a mobile quiz application based on face recognition as an authentication process to ascertain students' identity. The authentication process was implemented in two steps face detection using Mobile Vision APIs and face recognition using a Speed Up Robust Features (SURF) algorithm. Image classification and retrieval process is applied using SURF algorithm to extract features vector, then compare those vector with those of all stored images at server database and the matching process is applied based on RANSAC algorithm. A Wi-Fi ad hoc network in this project is established using jmdns java library to enable accessing the application by students. For training purposes, a data set containing 10 persons is added with 5 images per person. Results: A quiz environment has been arranged, in class with seven examiners each one separately accessed the quiz application with randomly chosen questions by the server. The achieved recognition rate was 85%, with a total average computation time 8.816 s per user login. Conclusion: This quiz application decreases manual intervention and brings adaptability to users with ease of use.