Research On Image Restoration Technology Of Underwater-to-air Imaging Based On Structured Light

Underwater-to-air imaging technology is an effective way to realize sea and air detection and early warning.As a key technology of virtual periscope imaging system,it plays an important role in the military and civilian fields.At present,there are the following problems in the research of underwater-to-air imaging technology:(1)The complexity of the underwater environment.The absorption and scattering of light in water will affect the quality of optical imaging,resulting in image color degradation and relatively blurred texture of underwater airborne detection;(2)The random fluctuation of the water surface and the complex refraction of the light change the propagation direction of the light,and finally the position of the light reaching the imaging surface is disordered,causing serious distortion of underwater air detection images,resulting in the inability to identify the true shape of the air target.Hence,it is meaning to carry through the study on the distorted image restoration technology of underwater aerial imaging.This paper starts from the problem of airborne imaging of virtual periscope through Snell window,and focuses on the research of underwater-to-air imaging technology and distorted image restoration technology,aiming to realize the restoration and correction of random distorted images of underwater airborne detection.The main work is as follows:(1)The characteristics of underwater-to-air imaging are analyzed and studied.The optical properties of water and its effect on optical imaging and the image degradation model are described in detail,and the Snell window for underwater-to-air imaging observation is systematically analyzed,thereby clarifying the image distortion and degradation caused by underwater aerial imaging.various factors.(2)According to the designed underwater space imaging model,a locally weighted surface fitting image restoration algorithm based on structured light is proposed.Firstly,the feature information of the distorted structured light image is extracted by the corner detection algorithm,and the spatial transformation model of the distorted structured light image is obtained by combining the local weighted surface fitting calculation.Then the corresponding relationship between active and passive imaging is analyzed and deduced,so as to restore and correct the distorted target image.Finally,according to the objective evaluation method of image quality,the restoration effect of the algorithm is analyzed and compared,and the effectiveness of the algorithm is verified.(3)A Helmholtz-Hodge decomposition image restoration algorithm based on structured light is proposed.Firstly,by extracting the features of the structured light with wavefront information,a two-dimensional warped vector field is obtained,and the Helmholtz-Hodge decomposition of the warped vector field is realized by the finite difference method and interpolation algorithm to obtain continuous and smooth three-component information,and then combined with the image The distorted image is restored by registration,and the algorithm is optimized and improved.Finally,through the calculation and analysis of the restoration effect and running time of the algorithm,while satisfying the high-quality restoration of the distorted image,it also overcomes the difficulty of the restoration efficiency of the traditional iterative registration algorithm,and greatly improves the restoration speed.(4)Underwater air imaging experiment and result analysis.By building the underwater air-to-air imaging experimental platform,and combining the two image restoration algorithms proposed in this paper,the underwater air-to-air imaging image restoration software is designed to realize the instantaneous restoration and correction of the distorted image under the water wave disturbance.The experimental results show that the two restoration algorithms proposed in this paper have a good correction and restoration effect on the distorted image of underwater air imaging,which verifies the feasibility and effectiveness of underwater air imaging experiment and restoration algorithm.