Deteksi Objek Manusia pada Citra menggunakan Single Shot Detecctor (SSD) Berbasis Edge Ccomputing

Penulis

  • Muhammad Iqbal Universitas Tanjungpura, Pontianak
  • Dwi Marisa Midyanti Universitas Tanjungpura, Pontianak
  • Syamsul Bahri Universitas Tanjungpura, Pontianak

DOI:

https://doi.org/10.25126/jtiik.938446

Kata Kunci:

Edge Computing, ESP32-CAM, Keamanan, MQTT, Single Shot Detector

Abstrak

Keamanan merupakan salah satu aspek penting di kehidupan manusia. Kemajuan teknologi yang ada dapat dimanfaatkan untuk meningkatkan keamanan, khususnya untuk kasus pencurian. Sistem kamera pengawas seperti CCTV telah terbukti dalam meningkatkan keamanan. Tetapi, CCTV mengharuskan pengawas untuk memantau layar CCTV secara manual 24/7 untuk melihat adanya pergerakan manusia. Pada penelitian ini, dibangun sistem deteksi secara otomatis menggunakan sensor passive infrared (PIR) HC-SR501 yang dikontrol oleh ESP32-CAM. Algoritma single shot detector (SSD) digunakan untuk mendeteksi objek pada citra dan edge computing digunakan untuk mengurangi latency pada proses transimisi data sehingga dapat memberikan informasi secara real-time. Proses transmisi data citra menggunakan protokol MQTT. Setiap ESP32-CAM akan menjadi publisher dan edge akan menjadi subscriber. Data citra yang digunakan berformat jpg dan memiliki resolusi 800x600 pixels. Sebanyak 1050 data digunakan untuk membangun model algoritma SSD. Seluruh data dibagi menjadi 3 bagian yaitu 735 data latih, 210 data evaluasi, dan 105 data uji. Data uji pada penelitian ini terdiri dari 3 jenis data, di antaranya 35 data uji siang hari, 35 data uji sore hari, dan 35 data uji malam hari. Algoritma SSD pada penelitian ini menghasilkan ketepatan deteksi 87.51% mAP pada data uji siang hari, 81.39% mAP pada data uji sore hari, dan 76.82% mAP pada data uji malam hari. Proses dari saat sensor HC-SR501 mendeteksi gerakan hingga informasi sampai ke user membutuhkan rata-rata waktu 2,843 detik.

 

Abstract

 

Security is one of the most important aspects of human life. Technological advances can be utilized to improve security, especially in the case of theft. Surveillance camera systems such as CCTV have been proven to improve security. However, CCTV requires the person to monitor the CCTV screen manually 24/7 to see any human movement. In this research, an automatic detection system was built using HC-SR501 passive infrared (PIR) sensor controlled by ESP32-CAM. The single shot detector (SSD) algorithm is used to detect an object on the image and edge computing is used to reduce the latency of the transmission process so that it can achieve real-time information. The transmission process of image data uses the MQTT protocol. Each ESP32-CAM will be the publisher and the edge will be the subscriber. The image data used are in jpg format and have a resolution of 800x600 pixels. A total of 1050 data were used to build the SSD algorithm model. All data is divided into 3 parts, namely 735 training data, 210 evaluation data, and 105 test data. The test data consists of 3 types of images, including 35 daytime images, 35 afternoon images, and 35 nighttime images. The SSD algorithm in this research resulted in 87.51% mAP on daytime images, 81.39% mAP on afternoon images, and 76.82% mAP on nighttime images. The process from the moment HC-SR501 sensor detects movement until the information reaches the user takes an average of 2.843 seconds.

Downloads

Download data is not yet available.

Biografi Penulis

  • Dwi Marisa Midyanti, Universitas Tanjungpura, Pontianak

    Rekayasa Sistem Komputer

  • Syamsul Bahri, Universitas Tanjungpura, Pontianak

    Rekayasa Sistem Komputer

Referensi

ALBAR, B., AMBARITA, A. AND IBRAHIM, A., 2019. Sistem Keamanan Ruangan Laboratorium Politeknik Sains dan Teknologi Wiratama Maluku Utara Menggunakan Sensor PIR (Passive Infra Red) dengan Metode Pengembangan Prototyping Berbasis Mikrokontroller ATmega328. Jurnal Ilmiah ILKOMINFO - Ilmu Komputer & Informatika, 2(2). https://doi.org/10.47324/ilkominfo.v2i2.34.

ALI, H.H., NAIF, J.R. AND HUMOOD, W.R., 2023. A New Smart Home Intruder Detection System Based on Deep Learning. Al-Mustansiriyah Journal of Science, 34(2), pp.60–69. https://doi.org/10.23851/mjs.v34i2.1267.

BHATTI, M.T., KHAN, M.G., ASLAM, M. AND FIAZ, M.J., 2021. Weapon Detection in Real-Time CCTV Videos Using Deep Learning. IEEE Access, 9, pp.34366–34382. https://doi.org/10.1109/ACCESS.2021.3059170.

CRUZ, M., MAFRA, S., TEIXEIRA, E. AND FIGUEIREDO, F., 2022. Smart Strawberry Farming Using Edge Computing and IoT. Sensors, 22(15), p.5866. https://doi.org/10.3390/s22155866.

DÜNTSCH, I. AND GEDIGA, G., 2019. Confusion Matrices and Rough Set Data Analysis. Journal of Physics: Conference Series, 1229(1), p.012055. https://doi.org/10.1088/1742-6596/1229/1/012055.

KHAN, I.U., QURESHI, I.M., AZIZ, M.A., CHEEMA, T.A. AND SHAH, S.B.H., 2020. Smart IoT Control-Based Nature Inspired Energy Efficient Routing Protocol for Flying Ad Hoc Network (FANET). IEEE Access, 8, pp.56371–56378. https://doi.org/10.1109/ACCESS.2020.2981531.

LIU, W., ANGUELOV, D., ERHAN, D., SZEGEDY, C., REED, S., FU, C.-Y. AND BERG, A.C., 2015. SSD: Single Shot MultiBox Detector. https://doi.org/10.1007/978-3-319-46448-0_2.

MARCHISIO, A., HANIF, M.A., KHALID, F., PLASTIRAS, G., KYRKOU, C., THEOCHARIDES, T. AND SHAFIQUE, M., 2019. Deep Learning for Edge Computing: Current Trends, Cross-Layer Optimizations, and Open Research Challenges. In: 2019 IEEE Computer Society Annual Symposium on VLSI (ISVLSI). IEEE. pp.553–559. https://doi.org/10.1109/ISVLSI.2019.00105.

MISHRA, B., MISHRA, B. AND KERTESZ, A., 2021. Stress-Testing MQTT Brokers: A Comparative Analysis of Performance Measurements. Energies, 14(18), p.5817. https://doi.org/10.3390/en14185817.

NUGRAHA, K.A. AND SEBASTIAN, D., 2021. Designing Consultation Chatbot Using Telegram API and Webhook-based NodeJS Applications. In: Proceedings of the 7th International Conference on Education and Technology (ICET 2021). [online] Atlantis Press. pp.119–122. https://doi.org/10.2991/assehr.k.211126.047.

PADILLA, R., NETTO, S.L. AND DA SILVA, E.A.B., 2020. A Survey on Performance Metrics for Object-Detection Algorithms. In: 2020 International Conference on Systems, Signals and Image Processing (IWSSIP). IEEE. pp.237–242. https://doi.org/10.1109/IWSSIP48289.2020.9145130.

PANDYA, S., GHAYVAT, H., KOTECHA, K., AWAIS, M., AKBARZADEH, S., GOPE, P., MUKHOPADHYAY, S. AND CHEN, W., 2018. Smart Home Anti-Theft System: A Novel Approach for Near Real-Time Monitoring and Smart Home Security for Wellness Protocol. Applied System Innovation, 1(4), p.42. https://doi.org/10.3390/asi1040042.

SANDLER, M., HOWARD, A., ZHU, M., ZHMOGINOV, A. AND CHEN, L.-C., 2018. MobileNetV2: Inverted Residuals and Linear Bottlenecks.

TOYIB, R., BUSTAMI, I., ABDULLAH, D. AND ONSARDI, O., 2019. Penggunaan Sensor Passive Infrared Receiver (PIR) Untuk Mendeteksi Gerak Berbasis Short Message Service Gateway. Pseudocode, 6(2), pp.114–124. https://doi.org/10.33369/pseudocode.6.2.114-124.

WANG, B., 2022. A Parallel Implementation of Computing Mean Average Precision.

WU, X., SAHOO, D. AND HOI, S.C.H., 2020. Recent advances in deep learning for object detection. Neurocomputing, 396, pp.39–64. https://doi.org/10.1016/j.neucom.2020.01.085.

ZHOU, Z., CHEN, X., LI, E., ZENG, L., LUO, K. AND ZHANG, J., 2019. Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing. Proceedings of the IEEE, 107(8), pp.1738–1762. https://doi.org/10.1109/JPROC.2019.2918951.

Diterbitkan

31-07-2024

Terbitan

Bagian

Ilmu Komputer

Cara Mengutip

Deteksi Objek Manusia pada Citra menggunakan Single Shot Detecctor (SSD) Berbasis Edge Ccomputing. (2024). Jurnal Teknologi Informasi Dan Ilmu Komputer, 11(3), 547-556. https://doi.org/10.25126/jtiik.938446