Liu Ke, Liang Yongquan
Opt Express. 2021 Aug 30;29(18):28307-28328. doi: 10.1364/OE.428626.
Underwater optical imaging technology plays a vital role in humans' underwater activities. However, the serious quality degradation of underwater optical images hinders further development of such technology. This phenomenon is mainly caused by the absorption and scattering of light in the underwater medium. The blurred image formation model is widely used in the field of optical images and depends on two optical parameters: background light (BL) and the transmission map (TM). Therefore, we propose an underwater optical image enhancement method in the context of underwater optical image restoration and color correction. First, BL estimation based on the gray close operation, which can avoid the influence of white objects while accurately calculating BL, is proposed. Then, an improved adaptive transmission fusion (IATF) method is proposed, and the adjusted reversed saturation map (ARSM) method is applied to compensate for and refine the estimated TMs to obtain the final TMs. This paper also proposes a new underwater light attenuation prior (NULAP) method. Finally, to enhance color saturation and edge details, a statistical colorless slant correction fusion smoothing filter method is proposed. Experimental results demonstrate that the proposed method outperforms state-of-the-art methods for dehazing, color and detail enhancement, and (uneven) light intensity.
水下光学成像技术在人类水下活动中发挥着至关重要的作用。然而，水下光学图像严重的质量退化阻碍了该技术的进一步发展。这种现象主要是由光在水下介质中的吸收和散射引起的。模糊图像形成模型在光学图像领域被广泛使用，它依赖于两个光学参数：背景光（BL）和透射图（TM）。因此，我们在水下光学图像恢复和色彩校正的背景下提出了一种水下光学图像增强方法。首先，提出了基于灰度闭运算的背景光估计方法，该方法在准确计算背景光的同时能够避免白色物体的影响。然后，提出了一种改进的自适应透射融合（IATF）方法，并应用调整后的反向饱和度图（ARSM）方法对估计的透射图进行补偿和细化，以获得最终的透射图。本文还提出了一种新的水下光衰减先验（NULAP）方法。最后，为了增强色彩饱和度和边缘细节，提出了一种统计无色倾斜校正融合平滑滤波方法。实验结果表明，所提出的方法在去雾、色彩和细节增强以及（不均匀）光强方面优于现有方法。
