Underwater Image Preprocessing Technology Research

With the development and exploration of the ocean, underwater detection technology has played an important role in the underwater data acquisition, data acquisition, pollution monitoring, meteorological, and ocean mining. Underwater image is one of the important means to obtain underwater information. At present, more researches are focused on underwater image processing. Because of the suspended solids in the underwater environment, and light absorption and scattering, underwater image has the weaknesses of illumination uneven, low contrast, various kinds of noise ect. In this paper, image preprocessing technology for underwater image is studied, and it mainly includes the following three aspects:(1) The feature of the underwater images is analyzed, the various methods of image preprocessing in terms of space and frequency domain is described, and at the same time, analyses of their advantages and weaknesses are given. Finally image quality evaluation method is introduced from the subjective and objective aspect.(2) The wavelet thresholding denoising theory and step are described. The wavelet function and threshold rule selected function are analyzed. According to the weaknesses of traditional threshold denoising method, an improved threshold function and threshold rule is given based on the wavelet scale correlation. The simulation shows that the improved method has better results compared to the traditional method in the situation of the underwater images of different sizes and types.(3) Histogram equalization methods and algorithms are introduced such as Global histogram equalization, local area histogram equalization, adaptive histogram equalization etc. Based on the wavelet transform and CLAHE, a new image enhancement method is proposed, the performances of these methods are compared in underwater image, and the result shows that the new method has better performance.