A Novel Underwater Image Enhancement Method Based on the Dual-Image Fusion

In underwater target recognition and tracking, it is necessary to utilize high-quality underwater image. However, there are a series of degraded problems with underwater image, such as low contrast, color distortion, hazy, blurry and so on. To solve the above problems, this paper proposes a fusion framework based on the dual-image. Firstly, the chroma component and illuminance component of the degraded image are improved respectively to obtain the chroma improved image and the illuminance improved image. Secondly, we use optimized adaptive gamma correction to find an intermediate image with good exposure between the exposure of the improved chroma image and the exposure of the improved illumination image. Thirdly, we obtain the hierarchical structures of the illumination improved image and the intermediate virtual image respectively by the Laplace pyramid decomposition. According to the normalized weights, the Gaussian pyramid mapping for each weight is formed. Finally, the images are reconstructed by the pyramid fusion method. Experimental results show that compared with other three underwater image enhancement algorithms, the proposed algorithm has certain advantages in underwater image enhancement.