Ahmad Fadhlur Rohman Rohman, Henny Dwi Bhakti

Perancangan Deteksi Wajah pada Aplikasi Berbasis React Native Menggunakan Metode Haar Cascade

Introduction

Perancangan deteksi wajah pada aplikasi berbasis react native menggunakan metode haar cascade. Rancang sistem deteksi wajah untuk aplikasi React Native (Bangbeli) dengan metode Haar Cascade & OpenCV. Tingkatkan keamanan autentikasi aplikasi digital Anda dan cegah kejahatan siber.

Abstract

Peningkatan keamanan aplikasi digital di era sekarang penting dilakukan, karena ancaman kejahatan siber yang semakin meningkat. Sistem keamanan yang tepat, seperti sistem autentikasi, dapat mencegah terjadinya kejahatan siber pada suatu aplikasi. Bangbeli sebagai salah satu startup yang bergerak di aplikasi pembayaran digital, penting untuk meningkatkan sistem kemananan pada aplikasinya. Penelitian ini bertujuan untuk merancang sistem deteksi wajah dan halaman verifikasi wajah untuk aplikasi Bangbeli yang berbasis framework React Native. Metode yang digunakan dalam mengimplementasikan sistem deteksi wajah adalah metode Haar Cascade dan menggunakan library OpenCV. Metode Haar Cascade memiliki kemampuan mendeteksi suatu objek dengan cepat, seperti objek berupa wajah. OpenCV merupakan library sumber terbuka yang banyak digunakan untuk pengolahan citra digital, serta dapat diimplementasikan ke berbagai macam bahasa pemrograman. Hasil pada penelitian ini, didapatkan halaman verifikasi berhasil dibuat dan sistem deteksi wajah dapat berjalan dengan menggunakan metode Haar Cascade yang dikembangkan pada aplikasi android berbasis framework React Native.

Review

This paper addresses a highly pertinent issue in today's digital landscape: enhancing application security against increasing cybercrime threats. Titled "Perancangan Deteksi Wajah pada Aplikasi Berbasis React Native Menggunakan Metode Haar Cascade," the research focuses on designing a face detection and verification system for the Bangbeli digital payment application, developed on the React Native framework. The authors clearly articulate the necessity of robust authentication systems for startups handling sensitive user data. The proposed approach, utilizing the well-established Haar Cascade method with the OpenCV library, offers a practical solution for fast object detection, which is often a critical requirement for mobile applications. Overall, the paper presents a timely and relevant contribution to securing mobile applications. The methodology employed is straightforward and suitable for an initial design and implementation phase. The choice of Haar Cascade, known for its computational efficiency, paired with OpenCV, a powerful and versatile image processing library, is a logical and proven combination for face detection tasks, especially where processing speed is a factor. The implementation within a React Native application for Bangbeli grounds the research in a real-world context, underscoring its practical applicability. The abstract states that the verification page was successfully created and the face detection system became operational on an Android application using the described method. This indicates a successful proof-of-concept for integrating the specified technology stack to achieve the desired security feature. While the paper successfully demonstrates the design and basic functionality of a face detection system, there are areas that, while not explicitly within the scope of this "design" paper, would significantly strengthen future work or a more comprehensive evaluation. The abstract primarily focuses on the system's operability rather than a detailed performance analysis. Future research could benefit from quantifying the system's accuracy, precision, recall, and robustness under varying real-world conditions (e.g., different lighting, angles, occlusions) and comparing its performance against more contemporary, perhaps deep learning-based, face detection algorithms that often offer higher accuracy, albeit potentially at a higher computational cost. Furthermore, given the context of a digital payment application, incorporating anti-spoofing measures (liveness detection) would be a a crucial next step to prevent malicious circumvention of the security system. Nonetheless, this paper provides a solid foundation for enhancing security in React Native-based mobile applications.

Full Text

Comments

You need to be logged in to post a comment.