Research On Technologies Of Underwater Camera Calibration

As the land resources drying up，More and more, we need marine resources. In theprocess of human exploration，images and videos are two important ways for us to gaininformation underwater. However, because of the complexity of the Marineenvironment, the quality of the underwater photographs seriously decline, in which theblurry is one of the most serious problems underwater images show. At the same time,the space measurement of the underwater positions and3d reconstruction technologyare widely used, the camera calibration technique is the most important link in theprocess of these applications. However, the camera is in the air environment, refractionoccurs inevitably which makes originally used for calibration technology in the aircannot be directly used for underwater, so solving the problem of underwater cameracalibration has become particularly important.Paper makes a scrutiny into the propagation characteristics of light underwater andthen analyses the reason for what underwater images are blurry. Mathematical model ofthe degradation process is formulated. Technology of image restoration is used forrecovery of underwater blurred image with the model and a satisfactory result isobtained. The refraction of light in the process of propagating underwater is analyzed indetail and the influence of refraction to the imaging model is also analyzed in this paper.The feasibility of the imaging model in the air is also discussed and the congruentrelationship between space points and points image plane and the mathematics ofrefraction in the whole process of propagation are used to establish the nonlinear modelof imaging system. At last, Tsai calibration algorithm used in the air is improved tocomplete the underwater camera parameters calibration and nonlinear imaging model isvalidated by experimental results.Paper models the process of the formation of blurry images underwater and thismodel has the characteristics of simple and comprehensive. Experiments have beenperformed to validate the correctness of the model. In the practical application, thismodel is valuable for deblurring of underwater images. In addition, paper analyses theunderwater imaging system. A nonlinear model is established and makes a goodbedding for further application and an important foundation for the3d reconstruction.