Rachmadani, Dimas Bagus (2024) Perancangan Dan Implementasi Prototipe Underwater Camera Dengan Teknologi Deteksi Karang Menggunakan Tensorflow Lite Pada Raspberry Pi 4 Untuk Konservasi Dan Pariwisata. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5027201034_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5027201034_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy |
Abstract
Penelitian ini merepresentasikan perancangan dan pengimplementasian prototipe kamera bawah air yang terintegrasi dengan deteksi karang. Kamera ini menggunakan kecerdasan buatan (Artificial Intelligence) sebagai penunjang teknologinya untuk mendeteksi biota laut khusus karang baik secara real-time maupun dengan rekaman video. Prototipe kamera bawah air ini menggunakan Raspberry Pi 4 Model B, modul kamera OV5647 wide angle 130 derajat serta Tenda O3 sebagai access point (AP) dengan menggunakan TensorFlow Lite (TFLite) model MobileNet SSD V2 dan MobileNet SSD V2 FPN Lite untuk kecerdasan buatan dalam mendeteksi karang. MobileNet SSD V2 FPN Lite dan MobileNet SSD V2 pada Raspberry Pi menunjukkan peforma deteksi mendapatkan 3 dan 4 frame per second (FPS) masing-masing dengan mean average precision (mAP) rata-rata sebesar 14.81% dan 13.29%. Meskipun terdapat kendala pada akses real-time yang terbatas hingga kedalaman ±70cm, fitur rekaman video dengan deteksi karang menawarkan solusi inovatif untuk mendukung konservasi terumbu karang. Dengan demikian, prototipe ini diharapkan dapat memberikan kontribusi berharga dalam upaya konservasi terumbu karang dan pengembangan pariwisata bawah air yang berkelanjutan.
=================================================================================================================================
This research presents the design and implementation of an underwater camera prototype integrated with coral detection technology. This camera utilizes artificial intelligence (AI) as a technological support to detect marine biota, specifically coral, both in real-time and through video recordings. The underwater camera prototype employs Raspberry Pi 4 Model B, OV5647 wide-angle 130-degree camera module, and Tenda O3 as an access point (AP), utilizing TensorFlow Lite (TFLite) models MobileNet SSD V2 and MobileNet SSD V2 FPN Lite for artificial intelligence in coral detection. MobileNet SSD V2 FPN Lite and MobileNet SSD V2 on Raspberry Pi demonstrate detection performance of 3 and 4 frames per second (FPS), respectively, with mean average precision (mAP) averaging 14.81% and 13.29%. Despite limitations in real-time access up to a depth of approximately ±70cm, the video recording feature with coral detection offers an innovative solution to support coral reef conservation efforts. Thus, this prototype is expected to make a valuable contribution to coral reef conservation efforts and the development of sustainable underwater tourism.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Deteksi Karang, Internet of Things, Kamera Bawah Air, Kecerdasan Buatan, MobileNet SSD V2, MobileNet SSD V2 FPNLite, Raspberry Pi 4 Model B, TensorFlow Lite, Artificial Intelligence, Coral Detection, Real-time, Underwater Camera |
| Subjects: | T Technology > T Technology (General) > T59.7 Human-machine systems. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5103.2 Wireless communication systems. Two way wireless communication |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis |
| Depositing User: | Dimas Bagus Rachmadani |
| Date Deposited: | 09 Feb 2024 08:11 |
| Last Modified: | 09 Feb 2024 08:11 |
| URI: | http://repository.its.ac.id/id/eprint/106500 |
Actions (login required)
 |
View Item |