Research On Binocular Vision Based Locating System For Underwater Vehicle

The Ocean is the origin of life on the earth, which has vast underwater space and enormous underwater resources. The exploration of the ocean and the utilization of submarine resources will greatly benefit the progress of human society. The research about the technology of Underwater Robot is the main component of the undersea technology research. Moreover, the Underwater Robot with machine vision on it will do a great help to the undersea inspection and operation. Therefore, the research of underwater machine vision is of great significance, especially, as the basis of which, the research of vision-based underwater locating system.In this thesis, the author made researches of vision-based underwater locating system on the underwater hardware design, underwater calibration method, underwater image processing algorithm and underwater locating algorithm. The locating experiment of the vision-based underwater locating system showed the accuracy of system.This thesis is divided into 6 chapters.ChapterⅠpresented the background of the underwater vision-based locating system. The underwater robot system and machine vision systems are introduced at the same time. And the chapter also introduced the development of vision-based manipulator of robots, both at home and abroad. At last of the chapter, the aim and the main content of the research was presented.ChapterⅡpresented the overall hardware configuration and system control algorithm of the vision-based underwater manipulator. New improved methods of the vision-based locating system about underwater system hardware design, underwater calibration algorithm, and underwater locating algorithm were proposed, according to the configuration and system control algorithm of the underwater robot manipulator system.ChapterⅢpresented the hardware architecture research of the vision-based underwater locating system. The method of components selection of the system hardware was introduced. And a sealing structure for the cameras was designed to meet the need of the underwater application of the system. An experiment of the sealing structure in the undersea environment simulation vessel was made to test the ability of the waterproofing of it.ChapterⅥpresents the research on the modeling of the underwater vision system and the calibration experiments. Aiming at the special perspective combination of water, glass, lens and air, a new calibration method was introduced to fix the distortion of the camera image. The results of experiment of the new calibration method proved the efficiency of it.ChapterⅤpresented a new locating algorithm called front-parallel correction locating algorithm of the vision-based underwater locating system, compared with the merits and drawbacks of old traditional sight-intersect correction locating algorithm. Experiment of the front-parallel correction locating algorithm was made to test the accuracy of it.ChapterⅣsummarized the full thesis, and made prospects of the following works of the vision-based underwater robot manipulator.