基于迁移学习及气象卫星云图的台风等级分类研究  

Research on Transfer Learning Methods for Classification of Typhoon Cloud Image

作　　者：郑宗生[1] 胡晨雨 黄冬梅[1] 邹国良[1] 刘兆荣 宋巍 Zheng Zongsheng;Hu Chenyu;Huang Dongmei;Zou Guoliang;Liu Zhaorong;Song Wei(School of Information,Shanghai Ocean University,Shanghai 201306,China)

机构地区：[1]上海海洋大学信息学院,上海201306

出　　处：《遥感技术与应用》2020年第1期202-210,共9页Remote Sensing Technology and Application

基　　金：国家自然科学基金项目(41671431);上海市科委地方院校能力建设项目(17050501900);国家海洋局数字海洋科学技术重点实验室开放基金项目共同资助。

摘　　要：针对传统卫星云图特征提取方法复杂且深度卷积神经网络(Deep Convolutional Neural Network,DCNN)模型开发困难的问题,提出一种基于参数迁移的台风等级分类方法。利用日本气象厅发布的近40 a 10000多景台风云图数据,构建了适应于迁移学习的台风云图训练集和测试集。在大规模ImageNet源数据集上训练出3种源模型VGG16,InceptionV3和ResNet50,依据台风云图低层特征与高层语义特征的差异,适配网络最佳迁移层数并冻结低层权重,高层权重采用自适应微调策略,构建出了适用于台风小样本数据集的迁移预报模型T-typCNNs。实验结果表明:T-typCNNs模型在自建台风数据集上的训练精度为95.081%,验证精度可达91.134%,比利用浅层卷积神经网络训练出的精度高18.571%,相比于直接用源模型训练最多提高9.819%。Aiming at the complexity of traditional methods for feature extraction about satellite cloud images,and the difficulty of developing deep convolutional neural network from scratch,a parameter-based transfer learning method for classifying typhoon intensity is proposed.Take typhoon satellite cloud images published by Japan Meteorological Agency,which includes 10000 scenes among nearly 40 years to construct training and test typhoon datasets.Three deep convolutional neural networks,VGG16,InceptionV3 and ResNet50 are trained as source models on the large-scale ImageNet datasets.Considering the discrepancy between low-level features and high-level semantic features of typhoon cloud images,adapt the optimal number of transferable layers in neural networks and freeze weights of low-level network.Meanwhile,fine-tune surplus weights on typhoon dataset adaptively.Finally,a transferred prediction model which is suitable for small sample typhoon datasets,called T-typCNNs is proposed.Experimental results show that the T-typCNNs can achieve training accuracy of 95.081%and testing accuracy of 91.134%,18.571%higher than using shallow convolutional neural network,9.819%higher than training with source models from scratch.

关 键 词：台风等级 迁移学习 深度卷积神经网络 迁移层数 自适应微调 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]