Sukarno Bahat Nauli, Agung Priambodo, Bosar Panjaitan, Faizal Zuli, Turkhamun Adi Kurniawan, Anita Ratnasari, Melani Indah Sari Manik, Dian Ayu Palapa Putri, Istiqomah Sumadikarta, Fachrel Muhammad

APLIKASI MOBILE MENDETEKSI WARNA BERBASIS ARTIFICIAL INTELLIGENCE BAGI PENDERITA BUTA WARNA

Introduction

Aplikasi mobile mendeteksi warna berbasis artificial intelligence bagi penderita buta warna. Aplikasi mobile deteksi warna berbasis AI untuk penderita buta warna. Memakai K-Means Clustering, memberikan info warna real-time, nilai RGB, dan kuis interaktif.

Abstract

Aplikasi deteksi warna berbasis Android ini dirancang untuk membantu penyandang tunanetra dan penderita buta warna dalam mengenali warna secara lebih mudah, cepat, dan akurat. Aplikasi memanfaatkan kamera perangkat untuk menangkap citra objek, kemudian mengolahnya menggunakan algoritma K-Means Clustering guna menentukan warna dominan. Hasil proses deteksi ditampilkan dalam bentuk nama warna, nilai RGB, tingkat keyakinan (confidence), serta deskripsi warna yang informatif sehingga dapat dipahami oleh pengguna. Selain fitur deteksi warna, aplikasi ini juga dilengkapi dengan kuis interaktif yang bertujuan untuk mengidentifikasi kemungkinan jenis buta warna yang dialami pengguna berdasarkan pola jawaban yang diberikan. Seluruh data pengguna, riwayat hasil deteksi warna, serta hasil kuis disimpan secara otomatis menggunakan Firebase Firestore sebagai basis data berbasis cloud. Metode pengembangan perangkat lunak yang digunakan dalam penelitian ini adalah Waterfall, yang meliputi tahap analisis kebutuhan, perancangan sistem, implementasi, pengujian, dan pemeliharaan. Pengujian aplikasi dilakukan menggunakan metode black-box testing untuk memastikan setiap fungsi berjalan sesuai dengan spesifikasi yang telah ditetapkan. Hasil pengujian menunjukkan bahwa aplikasi mampu bekerja dengan baik, memberikan informasi warna secara real-time, serta memiliki tingkat akurasi yang memadai dalam membantu pengguna mengenali warna.

Review

The paper presents a commendable effort in developing an Android-based mobile application to assist individuals with color blindness in identifying colors. The application's core functionality, which utilizes a device's camera and K-Means Clustering for dominant color detection, directly addresses a significant practical need. The resulting output, including color names, RGB values, confidence levels, and descriptive text, is designed to be user-friendly and informative, making it a valuable tool for improving daily life for the target demographic. The real-time processing capability is a crucial feature that enhances its utility and responsiveness. Beyond its primary function, the application incorporates several thoughtful features. The interactive quiz designed to help identify potential types of color blindness adds an important diagnostic dimension, further empowering users. The use of Firebase Firestore for secure cloud-based data storage for user profiles, detection history, and quiz results is a practical decision that ensures data persistence and a richer user experience. The adoption of the Waterfall development model and black-box testing for validation indicates a structured approach to development and quality assurance, confirming that the application functions as intended and provides adequate accuracy in color recognition. While the application demonstrates clear utility and a solid foundational design, there are opportunities for further scientific depth and clarification. The title's emphasis on "Artificial Intelligence" could be more thoroughly justified within the abstract by detailing how K-Means Clustering (a classical machine learning algorithm) contributes to this AI claim, or by outlining plans to integrate more advanced AI techniques. Furthermore, while "tingkat akurasi yang memadai" is reported, providing specific quantitative metrics (e.g., precision, recall, F1-score, or accuracy against a standardized color palette under various lighting conditions) would significantly enhance the paper's empirical rigor. Detailing the methodology behind calculating the "tingkat keyakinan" would also add valuable technical insight, and clarifying the specific features or accessibility considerations for "tunanetra" (visually impaired) beyond "buta warna" would refine the understanding of its broad applicability.

Full Text

Comments

You need to be logged in to post a comment.