基于上下文感知和超像素后处理的多光谱图像分类

摘要/Abstract

摘要： 高质量的地物类别提取是大量地学应用的基础。现有的基于像素的分类方法没有充分挖掘多光谱遥感图像中的上下文关联信息，且分类后的标签图像容易产生破碎。为了提升高分辨率遥感图像的分类精度，本文提出一种基于上下文感知网络和超像素后处理的多光谱图像分类方法。该方法利用新设计的卷积神经网络模型来更好地学习多光谱图像中的空间上下文信息。超像素后处理使用小区域分割和投票的策略来合并结构上关联的区域，以避免破碎标签的产生。本文方法在高分一号卫星数据上进行测试，并与6个分类算法进行比较。实验结果表明本文方法在精度和视觉效果上都优于比对算法。另外，对基于新模型分类后的结果进行超像素后处理，不仅减少了分类结果的破碎度，也进一步提升了图像的分类精度。

关键词: 图像分类, 土地利用, 高分一号

Abstract: To extract ground content is the basis for a large number of geoscientific applications. Existing pixel-based classification methods do not fully exploit the contextual associations in multispectral remote sensing images, and fragmented labels are observed everywhere in classified images. In order to improve the classification accuracy of high-resolution multispectral images, this paper proposes a new method which is based on context-aware networks and super-pixel post-processing. The method designs a new convolutional neural network to learn the spatial contextual information in multispectral images. Super-pixel post-processing uses a strategy of small region segmentation and voting to merge structurally associated regions, which can eliminate fragmented labels. The new method is tested on the Gaofen-1 satellite data and compared with six classification algorithms. The experimental results show that the new method outperforms the competing algorithms in terms of accuracy and visual effect. Among them, the super-pixel post-processing can reduce the fragmentation of classification results as well as improve the classification accuracy.

Key words: image classification, land use, Gaofen-1

吴志平, 麻尧斌, 汤文超, 胡必伟, 胡毕炜, 刘明嘉. 基于上下文感知和超像素后处理的多光谱图像分类[J]. 计算机与现代化, 2022, 0(12): 67-73.

WU Zhi-ping, MA Yao-bin, TANG Wen-chao, HU Bi-wei, HU Bi-wei, LIU Ming-jia. Multispectral Image Classification Based on Context-aware and Super-pixel Post-processing[J]. Computer and Modernization, 2022, 0(12): 67-73.

参考文献

［1］ WANG L, ZHENG C, DAI Q, et al. A comparative study of pixel level and region level classification of land use types using QuickBird imagery［C］// 2010 2nd IITA International Conference on Geoscience and Remote Sensing. 2010:219-222.
［2］ ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012,34(11):2274-2282.
［3］ GIRAUD R, TA V, BUGEAU A, et al. SuperPatchMatch: An algorithm for robust correspondences using superpixel patches［J］. IEEE Transactions on Image Processing, 2017,26(8):4068-4078.
［4］ ZHAO W, JIAO L, MA W, et al. Superpixel-based multiple local CNN for panchromatic and multispectral image classification［J］. IEEE Transactions on Geoscience and Remote Sensing, 2017,55(7):4141-4156.
［5］ LI G, LI L, ZHU H, et al. Adaptive multiscale deep fusion residual network for remote sensing image classification［J］. IEEE Transactions on Geoscience and Remote Sensing, 2019,57(11):8506-8521.
［6］ HE Z, SHI Q, LIU K, et al. Object-oriented mangrove species classification using hyperspectral data and 3-D siamese residual network［J］. IEEE Geoscience and Remote Sensing Letters, 2020,17(12):2150-2154.
［7］ MARINONI A, GAMBA P. Unsupervised data driven feature extraction by means of mutual information maximization［J］. IEEE Transactions on Computational Imaging, 2017,3(2):243-253.
［8］ HUANG S, ZHANG H, PIURICA A. Hybrid-hypergraph regularized multiview subspace clustering for hyperspectral images［J］. IEEE Transactions on Geoscience and Remote Sensing, 2022,60. DOI:10.1109/TGRS.2021.3074184.
［9］ BREIMAN L, FRIEDMAN J H, OLSHEN R, et al. Classification and Regression Trees (CART)［M］. Wiley, 1984. DOI:10.2307/2530946.
［10］BREIMAN L. Random forests［J］. Machine Learning, 2001,45:5-32.
［11］LI T, ZHANG J, ZHANG Y. Classification of hyperspectral image based on deep belief networks［C］// 2014 IEEE International Conference on Image Processing (ICIP). 2014:5132-5136.
［12］ZHONG P, GONG Z, LI S, et al. Learning to diversify deep belief networks for hyperspectral image classification［J］. IEEE Transactions on Geoscience and Remote Sensing, 2017,55(6):3516-3530.
［13］CHEN Y, LIN Z, ZHAO X, et al. Deep learning-based classification of hyperspectral data［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014,7(6):2094-2107.
［14］XING C, MA L, YANG X. Stacked denoise autoencoder based feature extraction and classification for hyperspectral images［J］. Journal of Sensors, 2016. DOI:10.1155/2016/3632943.
［15］ZHAO C, WAN X, ZHAO G, et al. Spectral-spatial classification of hyperspectral imagery based on stacked sparse autoencoder and random forest［J］. European Journal of Remote Sensing, 2017,50(1):47-63.
［16］MAGGIORI E, TARABALKA Y, CHARPIAT G, et al. Convolutional neural networks for large-scale remote-sensing image classification［J］. IEEE Transactions on Geoscience and Remote Sensing, 2017,55(2):645-657.
［17］JIAO L, LIANG M, CHEN H, et al. Deep fully convolutional network-based spatial distribution prediction for hyperspectral image classification［J］. IEEE Transactions on Geoscience and Remote Sensing, 2017,55(10):5585-5599.〖HJ0.8mm〗
［18］JI S, ZHANG C, XU A, et al. 3D convolutional neural networks for crop classification with multi-temporal remote sensing images［J］. Remote Sensing, 2018,10(1). DOI:10.3390/rs10010075.
［19］LIU X, JIAO L C, ZHAO J Q, et al. Deep multiple instance learning-based spatial-spectral classification for PAN and MS imagery［J］. IEEE Transactions on Geoscience and Remote Sensing, 2018,56(1):461-473.
［20］HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition［C］// 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016:770-778.
［21］CHAURASIA A, CULURCIELLO E. LinkNet: Exploiting encoder representations for efficient semantic segmentation［C］// 2017 IEEE Visual Communications and Image Processing (VCIP). 2017. DOI:10.1109/VCIP.2017.8305148.
［22］MIRZAPOUR F, GHASSEMIAN H. Object-based multispectral image segmentation and classification［C］// The 7th International Symposium on Telecommunications (IST'2014). 2014:430-435.
［23］LIANG W, WU Y, LI M, et al. High-resolution SAR image classification using context-aware encoder network and hybrid conditional random field model［J］. IEEE Transactions on Geoscience and Remote Sensing, 2020,58(8):5317-5335.
［24］GU Z, CHENG J, FU H, et al. CE-Net: Context encoder network for 2D medical image segmentation［J］. IEEE Transactions on Medical Imaging, 2019,38(10):2281-2292.
［25］HU Z, ZHANG Q, ZOU Q, et al. Stepwise evolution analysis of the region-merging segmentation for scale parameterization［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018,11(7):2461-2472.
［26］HU Z, SHI T, WANG C, et al. Scale-sets image classification with hierarchical sample enriching and automatic scale selection［J］. International Journal of Applied Earth Observation and Geoinformation, 2021,105(4). DOI:10.1016/j.jag.2021.102605.
［27］SIMONYAN K, ZISSERMAN A. Very Deep Convolutional Networks for Large-Scale Image Recognition［R］. Google Research, 2014.
［28］ZHONG Z, LI J, LUO Z, et al. Spectral-spatial residual network for hyperspectral image classification: A 3-D Deep learning framework［J］. IEEE Transactions on Geoscience and Remote Sensing, 2018,56(2):847-858.
［29］HUANG Y, WEI J, TANG W, et al. Pyramid convolutional neural networks and bottleneck residual modules for classification of multispectral images［C］// IGARSS 2020-2020 IEEE International Geoscience and Remote Sensing Symposium. 2020:1949-1952.

[1]	黄庭培1, 马禄彪1, 李世宝2, 刘建航1. 基于WiFi和原型网络的手势识别方法[J]. 计算机与现代化, 2024, 0(12): 34-39.
[2]	赵小明, 潘婷, 刘伟锋. 基于图像分类的自动绘画心理分析方法[J]. 计算机与现代化, 2024, 0(08): 92-97.
[3]	许跃雯1, 李明1, 李莉2. 基于对比学习MocoV2的COVID-19图像分类#br#[J]. 计算机与现代化, 2024, 0(02): 81-87.
[4]	邱凯星, 冯广. 基于双重特征注意力的多标签图像分类模型[J]. 计算机与现代化, 2023, 0(12): 41-47.
[5]	吴甜, 刘海华, 童顺延. 基于深度反馈的卷积神经网络的图像分类[J]. 计算机与现代化, 2023, 0(09): 82-86.
[6]	马国祥, 杨凌菲, 严传波, 张志豪, 孙彬, 王晓荣. 基于深度DenseNet网络的肝包虫病超声影像诊断方法[J]. 计算机与现代化, 2023, 0(09): 100-104.
[7]	秦竹媛, 吴浩忠, 谭代庆, 韩爱庆, 臧昊, 王选, 唐燕. 基于多尺度ResNet融合注意力机制的麦冬细粒度识别[J]. 计算机与现代化, 2023, 0(07): 105-111.
[8]	朱剑波, 葛明锋, 董文飞. 基于改进EfficientNet的阿尔兹海默症图像分类[J]. 计算机与现代化, 2023, 0(06): 56-61.
[9]	黄延辉, 兰海, 魏宪. 基于可分离结构变换的轻量级Vision Transformer[J]. 计算机与现代化, 2022, 0(10): 75-81.
[10]	杨贞, 单孟姣, 殷志坚, 杨凡, 李翠梅. 细粒度图像分类的通道自适应判别性学习方法[J]. 计算机与现代化, 2022, 0(10): 68-74.
[11]	王盼红, 朱昌明. 融合CNN与交互特征的多标签图像分类方法[J]. 计算机与现代化, 2022, 0(09): 85-92.
[12]	张林鹏, 汪西原, 李强. 基于双池化特征加权结构CNN的图像分类[J]. 计算机与现代化, 2021, 0(11): 67-71.
[13]	李晓双, 韩立新, 李景仙, 周经纬. 基于优化残差网络的多模态音乐情感分类[J]. 计算机与现代化, 2020, 0(12): 83-89.
[14]	李西明, 马李晓, 曾晓银, 王璇, 孙坚, 郭玉彬. 一种高精度微孔板浑浊度识别网络[J]. 计算机与现代化, 2020, 0(10): 58-63.
[15]	赵小蕾1,许喜斌2. 双重熵快速提取ROI图像优化分类方法[J]. 计算机与现代化, 2019, 0(02): 31-.