基于语义分割深度学习的地基云图分割方法

随着深度学习技术的迅速发展，更复杂更先进的语义分割深度学习模型在地基云图检测分割任务中得到广泛研究和应用。首先开创性地对新发布的地基云图数据集进行整理概括；然后阐述了基于深度学习语义分割模型在地基云图分割方面的研究进展，详细地介绍了典型的语义分割网络模型；接着选取了部分优秀性能的语义分割模型在标准的数据集上训练和验证，系统性评估其在地基云图分割的性能，验证了语义分割模型在地基云图分割领域的适用性;最后提出对基于语义分割的自适应地基云图像素级分割研究的总结和展望。

Research on ground-based cloud graph segmentation method based on deep learning semantic segmentation network

With the rapid development of deep learning, more complex and advanced semantic segmentation models for deep learning have been proposed and applied in ground-based cloud image detection and segmentation tasks. Firstly, the newly released ground-based cloud image datasets are sorted out and summarized unprecedentedly. Then, the research progress of ground-based cloud image segmentation based on deep learning semantic segmentation model is expounded, and typical semantic segmentation network models are introduced in detail. Next, some models with excellent performance are selected for training on standard datasets, and their performance in cloud image segmentation is systematically evaluated, which verifies the applicability of the model in this field. Finally, the summary and prospect of the research on pixel-level segmentation of adaptive ground-based cloud image based on semantic segmentation are proposed.