Research And Realization Of Underwater Image Restoration And Enhancement Algorithm

The propagation process of light in water media is different from that in air.It is subject to the absorption effect of water and the scattering effect is caused by the influence of suspended particles in water.Therefore,most of the collected underwater images have serious color distortion and quality degradation.to this,in recent years,studies of underwater image recovery enhancement algorithms to get much attention,in view of the underwater image color correction,traditional algorithm and the assumption and the improved algorithm can correct deficiencies or correction problem,so for part of the image the restoration and enhancement algorithm that can be applied to various underwater images.The following is the main content of this article.(1)A white balance method based on colorless constancy is proposed.In YCbCr space,the gray areas in the image,namely the originally weak color areas,are first searched,which are similar to "gray CARDS".Before and after the algorithm correction,the color should not change greatly.After finding the pixels in these regions,the difference between each pixel in the image chromaticity channel and the pixels in this region should be judged.The larger the difference,the greater the degree of color deviation.According to this parameter,the correction weight of each pixel is calculated.Finally,the weighted gray world is used to correct the chromatic pixel.This method can correct the color deviation well and protect the weak color area effectively.This method overcomes the problem that some pixels with weak actual chromance are forced to fill in color due to uniform gain calculation in traditional methods,thus resulting in color distortion.(2)A white balance algorithm in Lab color space is proposed.In Lab space,firstly calculates the average color channel,made clear in the case of chromaticity normal nearly neutral grey,its average value is 0,and under the condition of partial color the mean and the fact that there exist deviations in neutral grey,on this basis,the gaussian function on chromaticity with mean parameters structure,and then gaussianfunction and the product of the average as the final color correction parameters,the partial color pixels of chromaticity value correction to the normal range;The algorithm has proper contrast enhancement,wide application range,and can be used for real-time enhancement of video.(3)A new white balance algorithm in HSV color space is put forward.Depending on the desaturation strategy to change the degree of color bias,the gaussian function is first constructed and each pixel is multiplied by the value of the gaussian function to desaturate.In order to protect the bright underwater objects not happen error correction,the algorithm firstly calculates average hue channel,the hue channel to each pixel and the average difference for color difference,color difference is bigger that the pixels with environment color difference,the greater the bright for object,in this to do to protect such pixels,in the end to the color correction of images.(4)A white balance method based on color channel reconstruction rules and improved perfect reflection is proposed.First,according to the distribution of RGB three channels,the independence of each channel is judged by the intersection of histogram,and the channel at a lower gray level is taken as the target of reconstruction.Then,according to the color correlation,the correlation degree between the three channels is determined,and the rules of channel reconstruction are defined.An improved perfect reflection white balance algorithm is proposed to correct the color bias of the reconstructed three channels.Finally,the improved histogram equalization algorithm and sharpening operation are used to improve the global and local contrast of the image.Compared with the traditional perfect reflection method and related improved algorithm,this algorithm overcomes the shortcomings of excessive correction,insufficient correction and perfect reflection method.To sum up,this paper proposes four algorithms based on white balance in response to the shortcomings of existing methods,all of which have achieved good recovery and enhancement effects.