IMPLEMENTASI COMPUTER VISION MENGGUNAKAN DRONE UNTUK DETEKSI DAN PEMETAAN SAMPAH SECARA REAL - TIME

    Farhan Naufal Nurdiansyah, - and MUNAWIR, - and MUHAMMAD TAUFIK DWI PUTRA, - (2025) IMPLEMENTASI COMPUTER VISION MENGGUNAKAN DRONE UNTUK DETEKSI DAN PEMETAAN SAMPAH SECARA REAL - TIME. S1 thesis, Universitas Pendidikan Indonesia.

    Abstract

    Pencemaran lingkungan akibat sampah menjadi salah satu permasalahan serius yang berdampak pada kesehatan masyarakat dan kelestarian ekosistem. Laporan dari Kementerian Lingkungan Hidup dan Kehutanan Indonesia menunjukkan bahwa pada tahun 2024 lebih dari 40% sampah tidak terkelola dengan baik. Seiring berkembangnya teknologi, pemanfaatan drone dan computer vision berpotensi menjadi solusi inovatif dalam pengelolaan sampah pada daerah yang sulit dijangkau. Penelitian ini bertujuan untuk mengembangkan sistem deteksi dan pemetaan sampah secara real-time berbasis drone yang terintegrasi dengan teknologi computer vision. Penelitian ini mengimplementasikan algoritma deteksi objek YOLOv8n menggunakan drone DJI Mini 3, yang dikendalikan melalui Mobile SDK (MSDK) versi 5.1.3. Sistem ini terdiri dari integrasi antara aplikasi mobile, drone, dan antarmuka web untuk visualisasi hasil deteksi dalam bentuk peta interaktif yang dilengkapi dengan fitur heatmap. Pengujian dilakukan secara menyeluruh, meliputi pengujian aplikasi mobile, akurasi model YOLOv8n, dan visualisasi pada website. Evaluasi akurasi deteksi dilakukan berdasarkan dua parameter utama, yaitu intensitas pencahayaan dan variasi ketinggian drone saat pemindaian. Hasil pengujian menunjukkan bahwa sistem memberikan performa deteksi terbaik pada ketinggian 5 hingga 6 meter di kondisi pencahayaan siang hari, dengan akurasi mencapai 80%. Namun, pada ketinggian di atas 7 meter, akurasi deteksi menurun secara signifikan, terutama dalam kondisi saat mendung atau malam hari. Oleh karena itu, untuk memastikan efektivitas deteksi sampah secara real-time, sistem direkomendasikan beroperasi pada ketinggian rendah dan dalam kondisi pencahayaan siang hari. ------------ Environmental pollution caused by waste has become a serious issue that impacts both public health and ecosystem sustainability. A report from the Indonesian Ministry of Environment and Forestry indicates that in 2024, more than 40% of waste remains unmanaged. With the advancement of technology, the utilization of drones and computer vision offers the potential to serve as an innovative solution for waste management, particularly in areas that are difficult to access. This research aims to develop a real-time waste detection and mapping system using drones integrated with computer vision technology. The study implements the YOLOv8n object detection algorithm on a DJI Mini 3 drone, controlled through the Mobile SDK (MSDK) version 5.1.3. The system consists of an integration between a mobile application, the drone, and a web interface for visualizing detection results in the form of an interactive map equipped with a heatmap feature. Comprehensive testing was carried out, covering the mobile application, the accuracy of the YOLOv8n model, and visualization on the website. The accuracy evaluation was conducted based on two main parameters: lighting intensity and variations in drone altitude during scanning. The experimental results show that the system achieves the best detection performance at an altitude of 5 to 6 meters under daylight conditions, reaching an accuracy of up to 80%. However, at altitudes above 7 meters, detection accuracy decreases significantly, especially under cloudy or nighttime conditions. Therefore, to ensure effective real-time waste detection, the system is recommended to operate at lower altitudes and during daytime lighting conditions.

    Baca Full Text klik disini
    [thumbnail of S_TEKOM_2102555_Title.pdf] Text
    S_TEKOM_2102555_Title.pdf
    Download (1MB)
    [thumbnail of S_TEKOM_2102555_Chapter1.pdf] Text
    S_TEKOM_2102555_Chapter1.pdf
    Download (318kB)
    [thumbnail of S_TEKOM_2102555_Chapter2.pdf] Text
    S_TEKOM_2102555_Chapter2.pdf
    Restricted to Staf Perpustakaan
    Download (644kB)
    [thumbnail of S_TEKOM_2102555_Chapter3.pdf] Text
    S_TEKOM_2102555_Chapter3.pdf
    Download (718kB)
    [thumbnail of S_TEKOM_2102555_Chapter4.pdf] Text
    S_TEKOM_2102555_Chapter4.pdf
    Restricted to Staf Perpustakaan
    Download (1MB)
    [thumbnail of S_TEKOM_2102555_Chapter5.pdf] Text
    S_TEKOM_2102555_Chapter5.pdf
    Download (211kB)
    [thumbnail of S_TEKOM_2102555_Appendix.pdf] Text
    S_TEKOM_2102555_Appendix.pdf
    Restricted to Staf Perpustakaan
    Download (11MB)
    Official URL: https://repository.upi.edu/
    Item Type: Thesis (S1)
    Additional Information: https://scholar.google.com/citations?hl=en&user=YeITW5sAAAAJ SINTA ID : 6745899 SINTA ID : 6745726
    Uncontrolled Keywords: Waste Detection, Computer Vision, Mobile application, YOLOv8n, DJI Mini 3 Drone, Deteksi Sampah
    Subjects: L Education > L Education (General)
    Q Science > QA Mathematics
    Q Science > QA Mathematics > QA75 Electronic computers. Computer science
    Q Science > QA Mathematics > QA76 Computer software
    T Technology > T Technology (General)
    Divisions: UPI Kampus cibiru > S1 Teknik Komputer
    Depositing User: Farhan Naufal Nurdiansyah
    Date Deposited: 04 Sep 2025 08:15
    Last Modified: 04 Sep 2025 08:25
    URI: http://repository.upi.edu/id/eprint/137118

    Actions (login required)

    View Item View Item