The Underwater Image Restoration Method Based On Polarization Information

Underwater imaging technology has been widely used in underwater target discovery,marine environment monitoring.As the light's transmission in the water could be affected by water 's absorption and scattering effect,leading to underwater target imaging is blurred,visual range is limited.In view of the current strong scattering underwater environment and serious image degradation problem,this thesis puts forward two kinds of underwater image polarization recovery method,one kind is to set up simplified underwater reconstruction model,and use the way of actively estimated global optimal reconstruction parameters to restore image;The other is to integrate the forward scattering into the underwater model and use the method of oblique edge to estimate the point spread function(PSF),to inhibit the forward scattering effect on the image detail resolution.Through constructing laser underwater imaging system and carrying out the underwater target imaging experiments,the proposed method are verified and underwater image quality is improved substantially.This main research content is as follows:(1)In view of existing questions rely on an area to calculate the parameters in the background without target,artificially estimated parameters are poor,and recovery methods cannot meet the demand of practical application in underwater recovery method,underwater target polarization recovery method based on global parameters estimation is proposed.Take underwater target imaging with active polarization light source,and establish a new type of underwater simplified reconstruction model,take the initiative to estimate the underwater background light degree of polarization information and infinity underwater background light intensity information,automatically recover the underwater target from turbid medium,the establishment of a new simplified reconstruction model and reconstruction parameters of global solution improve the efficiency of image restoration and robustness of image results.(2)Due to the complexity of the forward scattering imaging mechanism,and the majority existing methods ignore its effects,in view of the forward scattering problem,putting forward oblique edge method to estimate the point spread function(PSF)for underwater target imaging in this thesis.Firstly,estimating the point spread function PSF,then take convolution method between directly transfer underwater target optical and the point spread function,obtaining the forward scattering information,finally through conversion imaging model in the frequency domain space to suppress the forward scattering factor,thus improve the details of the underwater image resolution and image quality.(3)The laser underwater imaging system was designed and built,meanwhile the underwater target polarization imaging experiments was carried out.The other several different underwater recovery method was taken to restore underwater image for subjectivity and objectivity comparison analysis.The experimental results show that the proposed method can obtain clear goals in turbid water environment,not only has better visual effect compared with other methods,and processing mode can meet the practical engineering application of full automatic estimation of model parameter in demand.