Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2020, Vol. 46 ›› Issue (12): 254-261,269. doi: 10.19678/j.issn.1000-3428.0056209

• Graphics and Image Processing • Previous Articles     Next Articles

Adaptive Remote Sensing Scene Classification Based on Complexity Clustering

LIANG Wentao, KANG Yan, LI Hao, LI Jinyuan, NING Haoyu   

  1. School of Software, Yunnan University, Kunming 650500, China
  • Received:2019-10-08 Revised:2019-12-25 Published:2020-01-07

基于复杂度聚类的自适应遥感场景分类

梁文韬, 康雁, 李浩, 李晋源, 宁浩宇   

  1. 云南大学 软件学院, 昆明 650500
  • 作者简介:梁文韬(1995-),男,硕士研究生,主研方向为深度学习、图像处理;康雁,副教授、博士;李浩,教授、博士、博士生导师;李晋源、宁浩宇,硕士研究生。
  • 基金资助:
    国家自然科学基金(61762092,61762089);云南省软件工程重点实验室开放基金(2017SE204)。

Abstract: Compared with common image classification tasks,the classification of remote sensing images has a wider feature range and more complex distribution,which makes it difficult to achieve accurate classification.In view of the adaptive relationship between the feature distribution of remote sensing images and the structure of neural network,this paper proposes a remote sensing scene classification model using adaptive neural network based on complexity-adaptive clustering.The complexity evaluation matrix of remote sensing images is constructed,which includes multiple features including color moment,gray level co-occurrence matrix,information entropy,information gain and line ratio.Image subsets with different complexity degrees are obtained by calculating image similarity.The image complexity is divided into high,medium and low levels by using hierarchical clustering.Then the complexity-adaptive image subsets are trained by using DenseNet,CapsNet and SENet to obtain the adaptive remote sensing scene classification model.Experimental results show that compared with DenseNet,CapsNet,SENet and other models,this model has better performance in extracting image features with different complexity degrees,and has higher accuracy of remote sensing scene classification.

Key words: remote sensing image, scene classification, image complexity, adaptive neural network, deep learning

摘要: 遥感图像场景分类任务较普通图像分类任务的特征范围更广且分布更复杂,难以实现精准分类。针对遥感图像特征分布与神经网络结构存在一定适应性关系的情况,提出一种利用复杂度适配聚类的自适应神经网络遥感场景分类模型。构建含有颜色矩、灰度共生矩阵、信息熵、信息增益、线占比等多重特征的遥感图像复杂度评价矩阵,通过计算图像相似性得到不同复杂度的图像子集,采用层次聚类方式将图像复杂度分为高、中、低等级,并分别使用DenseNet、CapsNet和SENet神经网络对复杂度适配的图像子集进行训练,最终获得自适应遥感场景分类模型。实验结果表明,与DenseNet、CapsNet、SENet等模型相比,该模型能更有针对性地提取不同复杂度的图像特征,具有更高的遥感场景分类准确率。

关键词: 遥感图像, 场景分类, 图像复杂度, 自适应神经网络, 深度学习

CLC Number: