PENGEMBANGAN SISTEM DETEKSI FORENSIK CITRA DIGITAL MENGGUNAKAN ERROR LEVEL ANALYSIS DAN CONVOLUTIONAL NEURAL NETWORK

    Muthmainah, R. Irzia Fitri (2025) PENGEMBANGAN SISTEM DETEKSI FORENSIK CITRA DIGITAL MENGGUNAKAN ERROR LEVEL ANALYSIS DAN CONVOLUTIONAL NEURAL NETWORK. S1 thesis, Universitas Pendidikan Indonesia.

    Abstract

    Penyalahgunaan teknologi yang semakin mudah khususnya dalam menyunting gambar dengan berbagai aplikasi menyebabkan semakin besar potensi manipulasi gambar untuk tujuan pencemaran nama baik, penyebaran informasi tidak benar dan kasus asusila. Manipulasi citra digital menjadi salah satu tantangan besar dalam keamanan informasi, terutama dalam memastikan keaslian bukti visual pada ranah forensik digital. Jenis manipulasi gambar yang umum terjadi meliputi copy-move forgery dan image splicing. Dalam investigasi, keakuratan proses identifikasi bukti visual menjadi krusial agar dapat dipertanggungjawabkan di ranah hukum. Oleh karena itu, mengacu pada standar ISO/IEC 27037 pada tahap identifikasi bukti citra digital untuk memastikan data sejak awal proses pemeriksaan. Untuk menjawab tantangan tersebut, penelitian ini mengembangkan sistem berbasis website untuk mendeteksi manipulasi citra menggunakan metode Error Level Analysis (ELA) dan Convolutional Neural Network (CNN) sebagai alat bantu analisis citra digital pada investigasi. Proses ELA digunakan sebagai tahap pra-pemrosesan untuk menonjolkan perbedaan tingkat kompresi pada citra yang kemudian diklasifikasikan oleh model CNN. Hasil pengujian menunjukkan bahwa sistem mencapai akurasi 95%. Hal ini membuktikan bahwa kombinasi ELA dan CNN efektif dalam mengidentifikasi manipulasi citra digital sehingga dapat menjadi alat bantu dalam analisis forensik digital. ------- The increasingly easy misuse of technology, particularly in editing images using various applications, has amplified the potential for image manipulation aimed at defamation, the spread of false information, and immoral cases. Digital image manipulation has become one of the major challenges in information security, especially in ensuring the authenticity of visual evidence in the field of digital forensics. Common types of image manipulation include copy-move forgery and image splicing. In investigations, the accuracy of the visual evidence identification process is crucial to ensure its accountability in legal proceedings. Therefore, this study refers to the ISO/IEC 27037 standard in the identification stage of digital image evidence to ensure data integrity from the very beginning of the examination process. To address these challenges, this research develops a web-based system to detect image manipulation using Error Level Analysis (ELA) and Convolutional Neural Network (CNN) as tools for digital image analysis in investigations. The ELA process is utilized as a pre-processing stage to highlight differences in compression levels within the images, which are then classified by the CNN model. Testing results show that the system achieved an accuracy of 95%. This demonstrates that the combination of ELA and CNN is effective in identifying digital image manipulation, making it a useful tool for digital forensic analysis.

    Baca Full Text klik disini
    [thumbnail of S_TEKOM_2100470_Title.pdf] Text
    S_TEKOM_2100470_Title.pdf
    Download (737kB)
    [thumbnail of S_TEKOM_2100470_Chapter1.pdf] Text
    S_TEKOM_2100470_Chapter1.pdf
    Download (464kB)
    [thumbnail of S_TEKOM_2100470_Chapter2.pdf] Text
    S_TEKOM_2100470_Chapter2.pdf
    Restricted to Staf Perpustakaan
    Download (670kB)
    [thumbnail of S_TEKOM_2100470_Chapter3.pdf] Text
    S_TEKOM_2100470_Chapter3.pdf
    Download (675kB)
    [thumbnail of S_TEKOM_2100470_Chapter4.pdf] Text
    S_TEKOM_2100470_Chapter4.pdf
    Restricted to Staf Perpustakaan
    Download (890kB)
    [thumbnail of S_TEKOM_2100470_Chapter5.pdf] Text
    S_TEKOM_2100470_Chapter5.pdf
    Download (241kB)
    [thumbnail of S_TEKOM_2100470_Appendix.pdf] Text
    S_TEKOM_2100470_Appendix.pdf
    Restricted to Staf Perpustakaan
    Download (4MB)
    Official URL: http://repository.upi.edu
    Item Type: Thesis (S1)
    Additional Information: https://scholar.google.com/citations?hl=en&user=-YAaqX0AAAAJ SINTA ID: 6745751 SINTA ID: 6680849
    Uncontrolled Keywords: Forensik Digital, Error Level Analysis, Convolutional Neural Network, Copy-move forgery, Image Splicing.
    Subjects: L Education > L Education (General)
    L Education > LB Theory and practice of education
    L Education > LB Theory and practice of education > LB1501 Primary Education
    Divisions: UPI Kampus cibiru > S1 Teknik Komputer
    Depositing User: R IRZIA FITRI MUTHMAINAH
    Date Deposited: 08 Sep 2025 02:42
    Last Modified: 08 Sep 2025 02:42
    URI: http://repository.upi.edu/id/eprint/137678

    Actions (login required)

    View Item View Item