Research Article
Estimating Object Distances Using Vision On Underwater Robots
@INPROCEEDINGS{10.4108/eai.21-9-2023.2342960, author={Syaiful Amri and M. Nur Faizi and Khairudin Syah}, title={Estimating Object Distances Using Vision On Underwater Robots}, proceedings={Proceedings of the 11th International Applied Business and Engineering Conference, ABEC 2023, September 21st, 2023, Bengkalis, Riau, Indonesia}, publisher={EAI}, proceedings_a={ABEC}, year={2024}, month={2}, keywords={underwater robot vision color filtering}, doi={10.4108/eai.21-9-2023.2342960} }- Syaiful Amri
M. Nur Faizi
Khairudin Syah
Year: 2024
Estimating Object Distances Using Vision On Underwater Robots
ABEC
EAI
DOI: 10.4108/eai.21-9-2023.2342960
Abstract
Underwater robot technology is one of the robot technologies that is very necessary for carrying out underwater activities. Arduino as the main controller is connected to the control joystick via cable so that the robot can be controlled from above the water surface. Apart from controlling movement maneuvers, a watertight mechanical design for the ROV robot is also very important. In testing the robot's movement while in water, the robot is expected to be able to maneuver according to commands and can help with underwater monitoring activities. The focus of this research is how underwater robots can detect and measure the distance of objects in front of them when maneuvering in the water. The object in this research is an orange ball, while the water in which the robot maneuvers is clear, white and clean water without interference from other objects. The process of detecting and measuring the distance of an object is a stage so that the robot can run autonomously later. The way to get the distance of an object is that the frame obtained from the camera's video capture is then converted into an HSV image which is segmented using the color filtering method and the pixel area of the object whose distance is measured can be obtained. The color classification method for segmenting spherical objects on underwater robots produces an average distance estimation error with a measurement range of 20 cm to 300 cm of 5.63% with the color classification method Color Filtering.