首页 > 论文 > 激光与光电子学进展 > 57卷 > 24期(pp:242804--1)

基于改进3D-CNN的多源遥感数据树种识别

Tree-Species Identification of Multisource Remote-Sensing Data using Improved 3D-CNN

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

针对森林复杂冠层结构和林分高密度下遥感树种识别精度不高的问题,将能够提取高维数据立体特征的三维卷积神经网络(3D-CNN)引入到遥感影像树种识别中,并利用残差网络(ResNet)对其进行改进,提出三维残差卷积神经网络(3D-RCNN),以减小网络深度带来的误差,降低退化现象的影响。联合高分五号高光谱数据(GF-5 AHIS)和高分六号高空间分辨率数据(GF-6 PMS),辅以森林资源数据和外业调查数据构建样本集。结合3D-RCNN思想构建树种识别模型。实验结果表明:相较于传统3D-CNN,3D-RCNN将模型网络从12层增加到18层,能够深化网络结构,缓解网络退化;联合GF-5 AHIS和GF-6 PMS,3D-RCNN能够有效地识别北亚热带森林树种,且识别精度(91.72%)要优于传统3D-CNN(85.65%)和支持向量机算法(85.22%)。

Abstract

Aiming to address the low identification accuracy of remote-sensing tree species of forests with a complex canopy and high density, a three-dimensional convolution neural network (3D-CNN) that can extract the stereoscopic features of hyper-dimensional data is introduced herein, and it can identify remote-sensing images. Furthermore, it is improved through residual network (ResNet) to build a 3D residual convolution neural network (3D-RCNN) to reduce the influence of degradation phenomenon and the inaccuracy caused by network depth. The sample set is constructed by combining GF-5 hyperspectral data (GF-5 AHIS)and GF-6 high spatial resolution data (GF-6 PMS), supplemented by forest resource data and field survey data. Then, a tree species recognition model is constructed based on the concept of 3D-RCNN. The experimental results show that compared with traditional 3D-CNN, the proposed 3D-RCNN increases the model network''s density from 12 layers to 18 layers, which can deepen the network structure and alleviate network degradation. By combining GF-5 AHIS and GF-6 PMS, 3D-RCNN can effectively identify northern subtropical forest species, providing better recognition accuracy (91.72%) than traditional 3D-CNN (85.65%) and support vector machine algorithm (85.22%).

广告组1 - 空间光调制器+DMD
补充资料

中图分类号:TP751.1

DOI:10.3788/LOP57.242804

所属栏目:遥感与传感器

基金项目:安徽省高校协同创新项目、安徽省高校学科优秀拔尖人才学术培育项目、高分专项省域产业化应用项目;

收稿日期:2020-04-23

修改稿日期:2020-05-29

网络出版日期:2020-12-01

作者单位    点击查看

栗旭升:新疆农业大学草业与环境科学学院, 新疆 乌鲁木齐 830052
陈冬花:安徽师范大学地理与旅游学院, 安徽 芜湖 241000滁州学院计算机与信息工程学院, 安徽 滁州 239000
刘赛赛:滁州学院计算机与信息工程学院, 安徽 滁州 239000
张乃明:新疆师范大学地理科学与旅游学院, 新疆 乌鲁木齐 830001
李虎:安徽师范大学地理与旅游学院, 安徽 芜湖 241000

联系人作者:陈冬花(lihu2881@aliyun.com); 李虎(lihu2881@aliyun.com);

备注:安徽省高校协同创新项目、安徽省高校学科优秀拔尖人才学术培育项目、高分专项省域产业化应用项目;

【1】Cohen W B, Yang Z Q, Healey S P, et al. A LandTrendr multispectral ensemble for forest disturbance detection [J]. Remote Sensing of Environment. 2018, 205: 131-140.

【2】Clark M L, Roberts D A, Clark D B. Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales [J]. Remote Sensing of Environment. 2005, 96(3/4): 375-398.

【3】Wang Z W, Sun J J, Yu Z Y, et al. Review of remote sensing image classification based on support vector machine [J]. Computer Science. 2016, 43(9): 11-17,31.
王振武, 孙佳骏, 于忠义, 等. 基于支持向量机的遥感图像分类研究综述 [J]. 计算机科学. 2016, 43(9): 11-17,31.

【4】Zheng Y, Chen Q Q, Zhang Y J. Deep learning and its new progress in object and behavior recognition Journal of Image and Graphics[J]. 0, 2014(2): 175-184.
郑胤, 陈权崎, 章毓晋. 深度学习及其在目标和行为识别中的新进展 中国图象图形学报[J]. 0, 2014(2): 175-184.

【5】Hu W, Huang Y Y, Wei L, et al. Deep convolutional neural networks for hyperspectral image classification Journal of Sensors[J]. 0, 2015(2): 258619.

【6】Chen W X, Picard R W. Predicting perceived emotions in animated GIFs with 3D convolutional neural networks[C]∥2016 IEEE International Symposium on Multimedia (ISM), December 11-13, 2016, San Jose, CA, USA. New York: , 2016, 367-368.

【7】Yan M, Zhao H D, Li Y H, et al. Multi-classification and recognition of hyperspectral remote sensing objects based on convolutional neural network [J]. Laser & Optoelectronics Progress. 2019, 56(2): 021702.
闫苗, 赵红东, 李宇海, 等. 基于卷积神经网络的高光谱遥感地物多分类识别 [J]. 激光与光电子学进展. 2019, 56(2): 021702.

【8】Liu Y Z, Jiang Z Q, Ma F, et al. Hyperspectral image classification based on hypergraph and convolutional neural network [J]. Laser & Optoelectronics Progress. 2019, 56(11): 111007.
刘玉珍, 蒋政权, 马飞, 等. 基于超图和卷积神经网络的高光谱图像分类 [J]. 激光与光电子学进展. 2019, 56(11): 111007.

【9】Li Z Q, Zhu R F, Gao F, et al. Hyperspectral remote sensing image classification based on three-dimensional convolution neural network combined with conditional random field optimization [J]. Acta Optica Sinica. 2018, 38(8): 0828001.
李竺强, 朱瑞飞, 高放, 等. 三维卷积神经网络模型联合条件随机场优化的高光谱遥感影像分类 [J]. 光学学报. 2018, 38(8): 0828001.

【10】He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 27-30, 2016, Las Vegas, NV, USA. New York: , 2016, 770-778.

【11】Liu J W, Liu Y, Luo X L. Research and development on deep learning [J]. Application Research of Computers. 2014, 31(7): 1921-1930,1942.
刘建伟, 刘媛, 罗雄麟. 深度学习研究进展 [J]. 计算机应用研究. 2014, 31(7): 1921-1930,1942.

【12】Wang X P. Protection and utilization of medicinal plant resources in Huangfu mountain nature reserve [J]. Chinese Traditional and Herbal Drugs. 2001, 32(7): 669-670.
王晓鹏. 皇甫山自然保护区药用植物资源保护与利用 [J]. 中草药. 2001, 32(7): 669-670.

【13】Fan B, Chen X, Li B C, et al. Technical innovation of optical remote sensing payloads onboard GF-5 satellite [J]. Infrared and Laser Engineering. 2017, 46(1): 8-14.
范斌, 陈旭, 李碧岑, 等. “高分五号”卫星光学遥感载荷的技术创新 [J]. 红外与激光工程. 2017, 46(1): 8-14.

【14】Li G D, Zhang C J, Gao F, et al. Double convpool-structured 3D-CNN for hyperspectral remote sensing image classification [J]. Journal of Image and Graphics. 2019, 24(4): 639-654.
李冠东, 张春菊, 高飞, 等. 双卷积池化结构的3D-CNN高光谱遥感影像分类方法 [J]. 中国图象图形学报. 2019, 24(4): 639-654.
Li G D, Zhang C J, Gao F, et al. Double convpool-structured 3D-CNN for hyperspectral remote sensing image classification [J]. Journal of Image and Graphics. 2019, 24(4): 639-654.
李冠东, 张春菊, 高飞, 等. 双卷积池化结构的3D-CNN高光谱遥感影像分类方法 [J]. 中国图象图形学报. 2019, 24(4): 639-654.

引用该论文

Li Xusheng,Chen Donghua,Liu Saisai,Zhang Naiming,Li Hu. Tree-Species Identification of Multisource Remote-Sensing Data using Improved 3D-CNN[J]. Laser & Optoelectronics Progress, 2020, 57(24): 242804

栗旭升,陈冬花,刘赛赛,张乃明,李虎. 基于改进3D-CNN的多源遥感数据树种识别[J]. 激光与光电子学进展, 2020, 57(24): 242804

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF