Research on the application of super resolution reconstruction algorithm for underwater image
The Ocean is the base of modern high-tech research. High-quality underwater imaging has great applications in scientific research, economics and military. However, due to the special transmission characteristics of light in water, the quality of direct underwater imaging is poor, the ideal effect is not obtained, and the quality of underwater images is improved, which has become the research goal of this subject. This topic is mainly aimed at the characteristics of seawater and the characteristics of optical transmission in seawater. We research the degradation model of underwater images, and analyze the low resolution factors of underwater images in different situations, and compare various algorithms for super-resolution reconstruction. In particular, it is based on the research of super-resolution algorithm of deep convolutional neural network.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Articles published by TSP are under an Open Access license, which means all articles published by TSP are accessible online free of charge and as free of technical and legal barriers to everyone. Published materials can be re-used if properly acknowledged and cited Open Access publication is supported by the authors' institutes or research funding agencies by payment of a comparatively low Article Processing Charge (APC) for accepted articles.