Design And Application Of Underwater Visual Ranging System

With the increasing depletion of land resources in recent years, countries are scrambling todiscover and develop Marine resources. Relative to the use of land resources, ocean developmentmust overcome the problem，such as underwater environment is so complex that underwater work ishard to implement. The way to throw equipment in the offshore could not be determined in a timelymanner and to put in the equipment damage, because the depth of the water environmentinformation is unknown, which is not only adds to the cost of unnecessary waste, and greatly reducethe efficiency of the detection of the offshore work. Since, underwater ranging system mainlyconsists of ultrasonic ranging, laser ranging, electromagnetic wave distance measuring and so on.Surveyors depend on the requirements of the different depth range and accuracy to do the rangingjob by ranging tool used in a specific environment. In recent years, many domestic and foreignscholars to apply computer vision which is an emerging discipline into ranging technology, and hascarried out a systematic study. Underwater visual range as innovative applications in fields rangingmachine vision, in the realization of the measurement precision and the algorithm, compared withthe traditional ranging technology, is gradually revealing the enormous superiority. Especially, asthe mature underwater sonar ranging tools, due to the ranging accuracy just passable short distance,can not meet the protection equipment under water. Unlike terrain scanning system for large scaleunderwater, this topic mainly aims at the visual model and algorithms of fixed-point real-timesounding near sea, to design an underwater vision measuring system which meets the accuracyrequirements.This paper makes statistics and comparison of current ranging technology and products aremainly used to measuring range under water, which also further verify the feasibility in applicationof computer vision measurement of underwater and its superiority in the specific measurementenvironment. According to the principle of measuring the system and implementation process, thestudy of this system mainly consists of acquisition, processing and analysis the two part image.Image acquisition part responsible for image acquisition of submarine surface laser point, imageprocessing and analysis part mainly responsible for target recognition of images, to obtain thenecessary model parameters, and according to the mathematical model of the system, calculatingthe depth information from the ocean floor equipment. Because this system relates to the algorithmand verify a large number, so the use of Microsoft Visual Studio2008development visionmeasurement model.The subject aims to complete the major work: the general design of the system, monocular three-point visual ranging system model and algorithm realization, binocular vision ranging systemmodel and algorithm realization, analysis and system debugging and experiment data.The overall design of the system mainly includes the core algorithm design, system functiondesign, the selection of image acquisition device camera, image acquisition card, laser, and thedescription of image processing algorithms, as well as the development language and developmentenvironment introduction.Monocular three-point vision measurement system design including monocular system model,basic principle of monocular image acquisition device design, monocular image processingalgorithm design and implementation. Binocular three-point visual ranging system design mainlyincludes the binocular vision ranging system model principle, binocular image acquisition devicedesign, binocular image processing algorithm design and implementation.System testing includes camera calibration test, image processing algorithm testing, systemtesting. Among them, the image processing algorithm testing is one of the most complex andimportant part of the test, which is the key to the whole test process. System testing is mainly incharge of the ranging process full implementation, to obtain a series of reliable measurementresults.After testing these during the weeks: the accuracy of the system meets the expectedrequirement in a certain range. Through the verification of the analysis of the test results and theimage processing algorithm, further to improve the visual ranging system theory practice.