The Research Of Underwater Image Denoising Based On Inhomogeneous Optic Field

As the development of the society, the demands of resources are growing. But lack of land resources has become increasingly evident. In recent years, with the scientific and technological progress, people began to exploit the ocean resources which open a new world to people. But our understanding of the oceans is still very limited, so the precondition of reasonable and effective development of the ocean resources is to understand the marine environment and seabed conditions, and it is also an urgent real problem which needs to address.The marine environment is very complex, and the condition of imaging is also very poor. The pictures capturing from the sea are often submerged in the complex noise of the sea, it makes the understanding of the sea difficult. So how to get clearer underwater images with less noise that is to find a more reasonable and effective way to make the underwater images becomes one of the most important issue facing the ocean.In this paper, we firstly analyzes the complex nature of seawater, especially its optical properties, we also analyzes the noise which affects the quality of underwater images, and on this basis, we proposed a new theory of non-uniform square, in this condition,we can get more clearly underwater images,and then process it. Secondly, we introduced several existing classical methods which can reduce the noise of images, we analyzed the characteristics of each method, and these methods could also be used to deal with the images of underwater successfully. Again, we introduced the basic theory of wavelet transform and also analyzed the characteristics of it, and then we compared several major methods of wavelet transform. We mainly introduced the threshold denoising method, and gave a detailed description of some of the classical threshold estimation methods. We can drown conclusion from the experiment that in the non-uniform light field condition, the underwater images could become clearer under the wavelet denoising. Finally, we showed the shortcomings of our study and put forward the problems that should pay attention to and solve in the future.