抚仙湖太阳观测和研究基地知识库

在云南天文台拍摄的模糊太阳斑点图的超分辨率重建过程中，采用现有深度学习算法存在高频信息难以恢复、重建图不够清晰等问题，为此，提出一种结合GAN（生成对抗网络）和风格迁移网络的太阳斑点图超分辨率重建方法STYLE-NICE-GAN。首先，利用GAN获取低分辨率太阳斑点图到Level1+高分辨率太阳斑点图的映射关系，重建太阳斑点图的全局轮廓和部分细节；其次，使用风格迁移网络，对GAN的重建结果进行二次重建，在保留局部细节、高频信息和不影响后续分析的同时，提高图像的整体对比度和清晰度。实验结果表明，与现有深度学习超分辨率重建算法相比，该方法具有重建图像清晰度更高、高频信息恢复能力更强的优点，重建结果在两个有参考评价指标PSNR、SSIM和三个无参考评价指标BRISQUE、NIQE、PIQE上的评价均占有优势。 

In the process of super-resolution reconstruction of the blur solar speckle images taken by the Yunnan Observatory, the existing deep learning algorithms have problems such as difficulty in recovering high-frequency information and lack of clarity in reconstructed images. Therefore, a super -resolution reconstruction method of solar speckle image named STYLE -NICE -GAN is proposed, which combines GAN (Generative Adversarial Networks) and style transfer networks. Firstly, GAN is used to obtain the mapping relationship from the low-resolution solar speckle image to the high-resolution solar speckle image and reconstruct the global outline and some details of the solar speckle image. Secondly, style transfer networks is used to perform secondary reconstruction on the reconstructed results of GAN, which can improve the overall contrast and clarity of the image while preserving local details and high frequency information and not affecting subsequent analysis. The experimental results show that compared with the existing deep learning super - resolution reconstruction algorithms, the proposed method has the advantages of higher image definition and stronger ability of high-frequency information recovery, its reconstruction results are superior in the evaluation of two reference evaluation indicators PSNR, SSIM and three nonreference evaluation indicators BRISQUE, NIQE and PIQE.