Perancangan Sistem Transmisi Data Pada Kamera Nightvision Pendeteksi Reflektor Life Jacket Berbasis Mikroprosesor Raspberry Pi 4

Laksana, Unggul Tri (2026) Perancangan Sistem Transmisi Data Pada Kamera Nightvision Pendeteksi Reflektor Life Jacket Berbasis Mikroprosesor Raspberry Pi 4. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5019221109-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Keselamatan korban kecelakaan kapal menjadi tantangan besar terutama pada malam hari ketika jarak pandang sangat terbatas. Data BASARNAS periode Januari sampai April 2025 mencatat 100 korban hilang dan 113 korban meninggal dunia, sehingga dibutuhkan alat yang mampu mendeteksi posisi korban pengguna life jacket secara akurat dari jarak jauh. Penelitian ini merancang sistem deteksi reflektor life jacket menggunakan kamera nightvision yang diproses oleh mikroprosesor Raspberry Pi 4, dipadukan dengan sensor LiDAR untuk mengukur jarak dan sudut objek, serta modem 4G LTE dan VPN Tailscale untuk mengirim data secara real time. Pengolahan citra memanfaatkan metode Top Hat Transform agar reflektor tetap dikenali dalam kondisi gelap, dan hasilnya ditampilkan pada dashboard berbasis web (Flask) yang memuat live streaming kamera, jarak dan arah objek, peta GPS, serta data telemetri sistem. Pengujian dilakukan dengan membandingkan sistem nirkabel dan sistem berkabel pada berbagai jarak dan jumlah objek. Hasilnya, F1 Score rata rata untuk satu objek mencapai 96,7 persen pada sistem nirkabel dan 100 persen pada sistem berkabel, sementara untuk dua objek masing masing 77,5 persen dan 79,5 persen, sehingga selisih keduanya hanya berkisar 2 sampai 3 persen. Delay deteksi tercatat sebesar 0,895 detik untuk satu objek dan 1,048 detik untuk dua objek, lebih cepat dibandingkan penelitian terdahulu. Sistem nirkabel juga terbukti mampu mengirim data hingga jarak 27 km tanpa kehilangan paket data, meski akurasi deteksi masih menurun pada jarak dekat dan saat objek terdeteksi lebih dari satu, serta ditemukan gangguan kamera yang bersumber dari kestabilan daya perangkat keras. Dengan hasil ini, sistem transmisi data nirkabel yang dirancang dapat memperluas jangkauan pemantauan tanpa memberikan dampak signifikan terhadap akurasi deteksi, sehingga berpotensi mendukung efisiensi operasi SAR di laut.
=====================================================================================================================================
Survivor safety after ship accidents remains a serious concern, especially at night when visibility is very limited. BASARNAS data from January to April 2025 recorded 100 missing victims and 113 deaths, showing the need for a tool that can accurately detect the position of survivors wearing life jackets from a distance. This study designed a life jacket reflector detection system using a nightvision camera processed by a Raspberry Pi 4 microprocessor, combined with a LiDAR sensor to measure object distance and angle, along with a 4G LTE modem and Tailscale VPN to transmit data in real time. Image processing uses the Top Hat Transform method so the reflector remains recognizable in dark conditions, with results shown on a Flask based web dashboard displaying live camera streaming, object distance and direction, a GPS map, and system telemetry. Testing compared the wireless and wired systems across various distances and object counts. The average F1 Score for single object detection reached 96.7 percent for the wireless system and 100 percent for the wired system, while for two objects the scores were 77.5 percent and 79.5 percent respectively, keeping the difference between both systems within 2 to 3 percent. Detection delay was recorded at 0.895 seconds for one object and 1.048 seconds for two objects, which is faster than the results reported in previous studies. The wireless system also proved capable of transmitting data up to 27 km without losing any data packets. However, detection accuracy still declined at close range and when more than one object was detected. Additionally, a camera malfunction was traced to hardware power instability. These findings show that the designed wireless transmission system can extend monitoring range without significantly reducing detection accuracy, offering potential support for more efficient maritime SAR operations.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Deteksi Life Jacket, Mikroprosesor, Night Vision, Sistem Telemetri, Life Jacket Detection, Microprocessor, Night Vision, Telemetry System
Subjects:	T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing V Naval Science > VK > VK200 Merchant marine--Safety measures V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering > VM365 Remote submersibles. Autonomous vehicles. V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering > VM453 Oceanographic research ships.
Divisions:	Faculty of Marine Technology (MARTECH) > Marine Engineering > 36202-(S1) Undergraduate Thesis
Depositing User:	Unggul Tri Laksana
Date Deposited:	04 Aug 2026 07:12
Last Modified:	04 Aug 2026 07:12
URI:	http://repository.its.ac.id/id/eprint/143109

Actions (login required)

View Item