基于SRGAN的栅格海图超分辨率提升方法

Super-resolution improvement method for raster charts based on SRGAN

针对海图制图领域中低分辨率栅格海图难以满足制图需求的问题,提出了基于超分辨率生成对抗网络(super resolution generative adversarial network,SRGAN)模型的栅格海图超分辨率提升方法。该方法通过以下步骤实现:首先对低分辨率栅格海图进行切分,以满足SRGAN模型输入要求;然后使用构建的训练集对SRGAN模型进行训练,并利用该模型对低分辨率栅格海图进行超分辨率提升处理;最后利用输出拼接模型对输出结果进行重组,将原有的低分辨率栅格海图转化为高分辨率栅格海图。实验结果表明:相较于其他重采样方法,本文方法具有更优的图像还原效果。本文方法为海图制图技术的发展提供了新的思路和方向,并为今后相关领域的研究提供了一定的参考和借鉴价值。

This paper proposes a method for improving the resolution of low-resolution raster charts in the field of nautical charting,using Super Resolution Generative Adversarial Network(SRGAN)models.The method involves the following steps:firstly,dividing the low-resolution raster charts to meet SRGAN's input requirements;secondly,training the SRGAN model with a constructed training set and using it to improve the resolution of the low-resolution raster charts;and finally,using an output splicing model to reorganize the output results and convert the original low-resolution raster charts into high-resolution ones.Experimental results show that this method is more effective than other resampling methods in terms of image restoration.In conclusion,this method provides new ideas and directions for the development of nautical charting technology and serves as a reference for future research in related fields.