Research On Underwater Image Quality Improvement Algorithm

Underwater image quality improvement algorithm is a major challenge in underwater visual field.Underwater images have different color cast due to different attenuation conditions,such as bluish,greenish and yellowish.In addition,due to floating particles and special illumination,underwater images will have problems such as the lack of detail and unnecessary noise.The existing algorithm has limitations for the underwater images of different degrees of color cast,and therefore,the researchers urgently need to introduce an effective way to solve the above problems.This thesis proposes a simple and effective underwater image enhancement method with color correction methods and detail enhancement methods.It can restore the underwater image of different degrees of color matrix to restore the natural appearance,and the method can restore the detail features of the water image and increase the contrast of the output image.The main results and innovations are as follows:1.By analyzing the correlation between underwater image and land images,we introduce an improved strategy to handle underwater images of different color cast,which adjusts the histograms of different color cast images by setting different reference values.Even though its simplicity is both in theory and implementation,it can produce a natural color appearance and can be applied to scenes with varying degrees of degradation.2.Color correction cannot handle the problem of visibility,a straightforward framework for detail restoration of underwater images is presented,consisting of two steps:The first step is single-scale decomposition based on edge-preserving technique,effectively processing the different decomposition layers to enhance the detailed features of underwater images and suppress some undesirable noises.The second step is to re-normalize the s-channel after the color space is transformed to improve the global contrast of the image.3.Propose a method of using color transfer to make underwater image datasets,and compared with the state-of-the-art algorithms,a fast and straightforward method not only can solve the limitations of the current algorithm,but also carry out relatively quickly on a common facility and is a good solution for engineering applications.Experiments have shown that our approach produces more natural color appearances and more valuable detail information than several existing underwater image enhancements.In addition,our approach improves the visibility of underwater images captured by low-light scenes and different cameras.