A Hybrid Image Enhancement Framework for Underwater 3D Reconstruction
Tengyue Li, Chen Ma · OCEANS 2022, Hampton Roads · 2022
Underwater images captured by optical cameras are usually degraded due to the absorption and scattering effects. This kind of light attenuation often leads to deteriorated visual image quality. In this paper, a hybrid underwater image enhancement framework is proposed, which focuses on correcting color distortion and removing haze for degraded underwater images. The proposed framework consists of color restoration and haze removal. We first restore the color distortion based on combining underwater white balance and contrast enhancement, and then take a further step to remove the haze and enhance the details using the haze-line technique. Experimental results show that the proposed framework outperforms several state-of-the-art methods on two publicly available datasets. Additionally, we conduct the real-world underwater 3D reconstruction using the processed underwater images. The results show that the 3D reconstruction result using the processed images by our proposed framework is able to display more three-dimensional details.