摘 要：针对遥感影像场景分类提出一种改进的中心聚类的深度学习模型，该模型通过改进不同类型特征的距离间隔，提高遥感图像场景分类的性能。与现有其他深度学习模型相比，该模型通过添加聚类中心以及特征与聚类中心的距离间隔约束，设计新的目标函数。新目标函数由交叉熵损失和中心聚类间隔损失构成。通过在两个公共基准数据集上评估所提出的目标函数，分类结果获得明显提升。
关键词：遥感图像；中心聚类；深度学习；卷积神经网络
DOI:10.19850/j.cnki.2096-4706.2021.10.006
基金项目：浙江省教育厅科研项目（Y2019 41374）
中图分类号：TP183 文献标识码：A 文章编号：2096-4706（2021）10-0027-04
Deep Learning Remote Sensing Image Scene Classification Method Based on Central Clustering
LIAO Jianping
（School of Information Engineering，Quzhou College of Technology，Quzhou 324000，China）
Abstract：For remote sensing image scene classification，an improved central clustering deep learning model is proposed，which improves the performance of remote sensing image scene classification by improving the distance interval of different types of features. Compared with other existing deep learning models，this model designs a new objective function by adding cluster center and the distance interval constraint between feature and cluster center. The new objective function consists of cross entropy loss and central clustering interval loss. By evaluating the proposed objective function on two common benchmark datasets，the classification results are significantly improved.
Keywords：remote sensing image；central clustering；deep learning；convolutional neural network
参考文献：
[1] 李德仁，童庆禧，李荣兴，等 . 高分辨率对地观测的若 干前沿科学问题 [J]. 中国科学：地球科学，2012，42（6）：805- 813.
[2] SIMONYAN K，ZISSERMAN A. Very Deep Convolutional Networks for Large-Scale Image Recognition [J/OL].arXiv：1409.1556 [cs.CV].（2015-04-10）.https://arxiv.org/abs/1409.1556v4.
[3] CHENG G，YANG C，YAO X，et al. When deep learning meets metric learning: remote sensing image scene classification via learning discriminative CNNs [J]. IEEE transactions on geoscience and remote sensing，2018，56（5）：2811-2821.
[4] YANG Y，NEWSAM S. Bag-of-visual-words and spatial extensions for land-use classification [C]//GIS’10：Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems.New York：Association for Computing Machinery，2010：270-279.
[5] XIA G S，HU J W，HU F，et al. AID：A Benchmark Data Set for Performance Evaluation of Aerial Scene Classification [J].IEEE Transactions on Geoscience and Remote Sensing，2017，55（7）： 3965-3981.
作者简介：廖建平（1978—），男，汉族，浙江衢州人，教务 处副处长，副教授，研究方向：大数据技术应用、智能信息处理。