SISTEM DETEKSI ROAD ACCIDENT BERBASIS WEB MENGGUNAKAN MODEL DEEP LEARNING YOLOV11 DAN FRAMEWORK FLASK

    Muhammad Imron Maulana, - and Arjuni Budi Pantjawati, - (2025) SISTEM DETEKSI ROAD ACCIDENT BERBASIS WEB MENGGUNAKAN MODEL DEEP LEARNING YOLOV11 DAN FRAMEWORK FLASK. S1 thesis, Universitas Pendidikan Indonesia.

    Abstract

    Penelitian ini bertujuan mengembangkan sistem deteksi kecelakaan lalu lintas dan tindakan kriminal bersenjata menggunakan model deep learning YOLOv11 yang terintegrasi dengan platform web berbasis Flask. Sistem dirancang untuk memberikan notifikasi real-time guna meningkatkan respons unit tanggap darurat. Metode penelitian menggunakan Research and Development (R&D) dengan pendekatan Waterfall yang meliputi pengumpulan dataset hybrid, pra-pemrosesan, pelatihan model YOLOv11n, serta pengembangan dan integrasi sistem web. Dataset terdiri dari 2.301 gambar hasil augmentasi yang dibagi menjadi 85% pelatihan, 10% validasi, dan 5% pengujian. Hasil evaluasi model menunjukkan kinerja tinggi dengan precision 97,1%, recall 98,2%, mAP50 98,9%, dan mAP50-95 87,9%. Pada pengujian sistem berbasis web, video lokal mencapai akurasi 91,11%, kecepatan deteksi rata-rata 61,14 ms, frame rate 8,66 FPS, dan latensi 63,4 ms. Pada IP CCTV real-time, akurasi mencapai 64,44%, kecepatan deteksi 65,72 ms, frame rate 6,57 FPS, dan latensi total 119,1 ms. Sistem alert berhasil memicu notifikasi suara dan menyimpan bukti kejadian dengan akurasi 100%. Kesimpulan penelitian menegaskan efektivitas sistem dalam mendeteksi kejadian darurat secara cepat, meskipun perlu pengembangan dataset lebih variatif dan optimasi infrastruktur komputasi untuk meningkatkan kinerja real-time. This research aims to develop a road accident detection system using the YOLOv11 deep learning model integrated with a Flask-based web platform. The system is designed to provide real-time notifications to improve emergency response unit responsiveness. The research methodology employs Research and Development (R&D) with a Waterfall approach, encompassing hybrid dataset collection, pre-processing, YOLOv11n model training, and web system development and integration. The dataset consists of 2,301 augmented images divided into 85% for training, 10% for validation, and 5% for testing. Model evaluation results demonstrate high performance with 97.1% precision, 98.2% recall, 98.9% mAP50, and 87.9% mAP50-95. In web-based system testing, local video achieved 91.11% accuracy, average detection speed of 61.14 ms, frame rate of 8.66 FPS, and latency of 63.4 ms. On real-time IP CCTV, accuracy reached 64.44%, detection speed of 65.72 ms, frame rate of 6.57 FPS, and total latency of 119.1 ms. The alert system successfully triggered audio notifications and saved incident evidence with 100% accuracy. The research conclusion affirms the system's effectiveness in rapidly detecting emergency incidents, although more diverse dataset development and computational infrastructure optimization are needed to enhance real-time performance.

    Baca Full Text klik disini
    [thumbnail of S_TE_2108071_Title.pdf] Text
    S_TE_2108071_Title.pdf
    Download (607kB)
    [thumbnail of S_TE_2108071_Chapter1.pdf] Text
    S_TE_2108071_Chapter1.pdf
    Download (279kB)
    [thumbnail of S_TE_2108071_Chapter2.pdf] Text
    S_TE_2108071_Chapter2.pdf
    Restricted to Staf Perpustakaan
    Download (815kB)
    [thumbnail of S_TE_2108071_Chapter3.pdf] Text
    S_TE_2108071_Chapter3.pdf
    Download (537kB)
    [thumbnail of S_TE_2108071_Chapter4.pdf] Text
    S_TE_2108071_Chapter4.pdf
    Restricted to Staf Perpustakaan
    Download (1MB)
    [thumbnail of S_TE_2108071_Chapter5.pdf] Text
    S_TE_2108071_Chapter5.pdf
    Download (235kB)
    [thumbnail of S_TE_2108071_Appendix.pdf] Text
    S_TE_2108071_Appendix.pdf
    Restricted to Staf Perpustakaan
    Download (3MB)
    Official URL: https://repository.upi.edu/
    Item Type: Thesis (S1)
    Additional Information: https://scholar.google.com/citations?view_op=new_profile&hl=en ID Sinta Dosen Pembimbing Arjuni Budi Pantjawati: 5994602
    Uncontrolled Keywords: Deteksi road accident, Deep learning, YOLOv11, Real-time monitoring, Framework Flask Road accident detection, Deep learning, YOLOv11, Real-time monitoring, Flask Framework
    Subjects: L Education > L Education (General)
    L Education > LB Theory and practice of education
    T Technology > T Technology (General)
    Divisions: Fakultas Pendidikan Teknik dan Industri > Jurusan Pendidikan Teknik Elektro > Program Studi Pendidikan Teknik Elektro
    Depositing User: Muhammad Imron Maulana
    Date Deposited: 03 Nov 2025 03:27
    Last Modified: 03 Nov 2025 03:27
    URI: http://repository.upi.edu/id/eprint/144853

    Actions (login required)

    View Item View Item