Underwater distance ranging implemented through a stereo vision system

• Main Authors: Cabarle, Luis Eduardo, Sybingco, Edwin, Dadios, Elmer P.
• 格式: text
• 出版: Animo Repository 2016
• 主題: Robot vision; Underwater vision; Electrical and Electronics; Robotics; Systems and Communications
• 在線閱讀: https://animorepository.dlsu.edu.ph/faculty_research/4282
• 機構: De La Salle University

Vision systems are becoming popular as a method of gathering information and stimuli from a robot's environment. The goal of this paper is to implement an underwater distance ranging system using a stereo vision system. Target tracking is based on the color of the targets. The ranging algorithm was designed so that it can range targets up to 2 meters away with a minimum accuracy of 90%. An algorithm for ranging above ground and underwater was implemented and the two were compared. Algorithms were setup using samples from an aquarium, pool and beach. It was found that the algorithm may be setup more accurately using samples from a wider range of distances. Different colors of lighting was used to illuminate a target to see which wavelength will best lights up the underwater environment. It was found that wavelengths above the visible range could be ruled out. Camera performance was measured under varying conditions of salinity. It was found that the camera used performed worse for salt water as the recorded videos became blurred with increase in salinity. © 2015 IEEE.