An effective underwater image enhancement method based on CLAHE-HF

Abstract

Abstract In order to solve the problem that the unbalanced illumination and noise induce poor contrast, uneven brightness and poor clarity to underwater image, an effective underwater image enhancement method based on spatial and frequency domain is proposed in this paper. Firstly, Contrast Limited Adaptive Histogram Equalization technology is used in spatial domain to enhance the local contrast of different regions of the image according to the distribution of pixel values of each region, which can not only ensure the image texture clarity and detail characteristics, but also suppress the noise. Secondly, Homomorphic Filtering technology is used to enhance the image details and reduce the noise in the image in frequency domain. Two sets of image data that verify the effectiveness of the proposed method were provided. The experimental results show that, compared with the traditional underwater image enhancement methods, the peak signal-to-noise ratio (PSNR), mean square error (MSE) and information entropy of the proposed method are combined optimal.