TY - GEN
T1 - Design of Remotely Operated Vehicle Prototype for Ship Biofouling Inspection on Berth
AU - Azhary, Sutan
AU - Purwanto, Dedi Budi
AU - Nurhadi, Hendro
AU - Pramujati, Bambang
AU - Effendi, Mohammad Khoirul
AU - Widjaja, Sjarief
AU - Wahjudi, Arif
AU - Arief, Irfan Syarief
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Biofouling in ships is generally treated using antifouling paint on the hull and cleaning on the dock using high-pressure water. With the current development of Remotely Operated Vehicle (ROV) technology, biofouling inspection can be done while the ship is loading and unloading without having to dry dock. Therefore, this study aimed to design an ROV prototype, and the system required to perform biofouling inspection on a ship's haul when berthing. The inspection capability was designed by implementing underwater image enhancement and object detection methods using machine learning. The obtained prototype design was determined with the dimension of 42.9cm long, 18.7cm wide and 11.7cm high. To improve the underwater image, the FUNIE-GAN model resulted in the average Peak Signal-to-Noise Ratio (PSNR) value of 28.6dB and average Structural Similarity Index Metrics (SSIM) value of 0.798. On average, the vehicle's ability to detect biofouling objects later increased by 155%.
AB - Biofouling in ships is generally treated using antifouling paint on the hull and cleaning on the dock using high-pressure water. With the current development of Remotely Operated Vehicle (ROV) technology, biofouling inspection can be done while the ship is loading and unloading without having to dry dock. Therefore, this study aimed to design an ROV prototype, and the system required to perform biofouling inspection on a ship's haul when berthing. The inspection capability was designed by implementing underwater image enhancement and object detection methods using machine learning. The obtained prototype design was determined with the dimension of 42.9cm long, 18.7cm wide and 11.7cm high. To improve the underwater image, the FUNIE-GAN model resulted in the average Peak Signal-to-Noise Ratio (PSNR) value of 28.6dB and average Structural Similarity Index Metrics (SSIM) value of 0.798. On average, the vehicle's ability to detect biofouling objects later increased by 155%.
KW - Biofouling
KW - Machine Learning
KW - Object Detection
KW - Remotely Operated Vehicle
KW - Underwater Image Enhancement
UR - http://www.scopus.com/inward/record.url?scp=85135008834&partnerID=8YFLogxK
U2 - 10.1109/ICAMIMIA54022.2021.9807777
DO - 10.1109/ICAMIMIA54022.2021.9807777
M3 - Conference contribution
AN - SCOPUS:85135008834
T3 - 2021 International Conference on Advanced Mechatronics, Intelligent Manufacture and Industrial Automation, ICAMIMIA 2021 - Proceeding
SP - 223
EP - 228
BT - 2021 International Conference on Advanced Mechatronics, Intelligent Manufacture and Industrial Automation, ICAMIMIA 2021 - Proceeding
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2021 International Conference on Advanced Mechatronics, Intelligent Manufacture and Industrial Automation, ICAMIMIA 2021
Y2 - 8 December 2021 through 9 December 2021
ER -