基于Bi-LSTM模型的时间序列遥感作物分类研究

doi:10.3864/j.issn.0578-1752.2022.21.005

Abstract

Abstract:

【Objective】Timely and accurate crop classification mapping is an important basis for agricultural situation monitoring. This study explores the potential of deep learning in time series remote sensing crop classification and early identification based on a bidirectional long short-term memory network model.【Method】In this paper, Yellow River Delta region was chosen as an example and a time-series NDVI dataset were constructed by using Sentinel-2 year-round available satellite images as the data source. A recurrent neural network architecture is used to build a bidirectional long short-term memory (Bi-LSTM) model for structured time-series remote sensing data to carry out crop classification, then the generalization ability of the model is evaluated. Through adjusting the length of time series, we explore the earliest identifiable time of different crops under the condition of satisfying certain mapping accuracy. 【Result】 Growth characteristics represented by time series remote sensing images have great potential to discriminate different crops. The overall accuracy of the Bi-LSTM model reached 90.9% with a Kappa coefficient of 0.892. By testing the effects of different time series lengths on crop classification, the earliest identifiable time of typical crops was obtained. The accuracy of crops such as winter-wheat and rice could improve significantly after the emergence of unique characteristics. Crops such as cotton and spring maize required complete growth sequences to ensure classification accuracy.【Conclusion】The structured feature information embedded in satellite image time series could effectively reduce crop spectral confusion at specific time periods. The Bi-LSTM model was able to consider both forward and backward temporal state information and could learn the spectral change characteristics of crops, which was excellent in the identification of confusing crops such as rice, cotton and spring maize. In addition, the deep learning model could effectively capture the variation trend on the sample in general, and showed better generalization ability and robustness in the crop multi-classification task. This study provided a feasible idea for regional crop mapping with high accuracy by integrating deep learning and remote sensing time series.

Key words: crop classification, early identification, time-series remote sensing, Bi-LSTM model, model generalization

HUANG Chong,HOU XiangJun. Crop Classification with Time Series Remote Sensing Based on Bi-LSTM Model[J].Scientia Agricultura Sinica, 2022, 55(21): 4144-4157.

0
/ / Recommend

Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks

URL: https://www.chinaagrisci.com/EN/10.3864/j.issn.0578-1752.2022.21.005

https://www.chinaagrisci.com/EN/Y2022/V55/I21/4144

Figures/Tables 13

Fig. 1

Table 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Table 2

Fig.7

Fig. 8

Table 3

Table 4

Fig. 9

References 36

[1]	张佳华, 胡小夏, 刘学锋, 何贞铭. 基于MODIS数据提取华北典型区冬小麦种植面积. 中国科学院大学学报, 2013, 30(5): 637-643.
	ZHANG J H, HU X X, LIU X F, HE Z M. Extraction of winter wheat planting areas based on the time-series MODIS data in typical region of North China. Journal of University of Chinese Academy of Sciences, 2013, 30(5): 637-643. (in Chinese)
[2]	史舟, 梁宗正, 杨媛媛, 郭燕. 农业遥感研究现状与展望. 农业机械学报, 2015, 46(2): 247-260.
	SHI Z, LIANG Z Z, YANG Y Y, GUO Y. Status and prospect of agricultural remote sensing. Transactions of the Chinese Society for Agricultural Machinery, 2015, 46(2): 247-260. (in Chinese)
[3]	陈仲新, 任建强, 唐华俊, 史云, 冷佩, 刘佳, 王利民, 吴文斌, 姚艳敏, 哈斯图亚. 农业遥感研究应用进展与展望. 遥感学报, 2016, 20(5): 748-767.
	CHEN Z X, REN J Q, TANG H J, SHI Y, LENG P, LIU J, WANG L M, WU W B, YAO Y M, HASI T Y. Progress and perspectives on agricultural remote sensing research and applications in China. Journal of Remote Sensing, 2016, 20(5): 748-767. (in Chinese)
[4]	DE CASTRO A I, TORRES-SÁNCHEZ J, PENA J M, JIMENEZ- BRENES F M, CSILLIK O, LOPEZ-GRANADOS F. An automatic Random Forest-OBIA algorithm for early weed mapping between and within crop rows using UAV Imagery. Remote Sensing, 2018, 10(2): 285.
[5]	郝鹏宇, 唐华俊, 陈仲新, 牛铮. 基于历史增强型植被指数时序的农作物类型早期识别. 农业工程学报, 2018, 34(13): 179-186.
	HAO P Y, TANG H J, CHEN Z X, NIU Z. Early season crop type recognition based on historical EVI time series. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(13): 179-186. (in Chinese)
[6]	AZAR R, VILLA P, STROPPIANA D, CREMA A, BOSCHETTI M, BRIVIO P A. Assessing in-season crop classification performance using satellite data: A test case in Northern Italy. European Journal of Remote Sensing, 2016, 49(1): 361-380.
[7]	赵丽花, 李卫国, 杜培军. 基于多时相HJ卫星的冬小麦面积提取. 遥感信息, 2011, 26(2): 41-45, 50.
	ZHAO L H, LI W G, DU P J. The area extraction of winter wheat based on multi-temporal HJ remote sensing satellite images. Remote Sensing Information, 2011, 26(2): 41-45, 50. (in Chinese)
[8]	YOU N S, DONG J W. Examining earliest identifiable timing of crops using all available Sentinel 1/2 imagery and Google Earth Engine. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 161: 109-123.
[9]	REN T W, LIU Z, ZHANG L, LIU D Y, XI X J, KANG Y H, ZHAO Y Y, ZHANG C, LI S M, ZHANG X D. Early identification of seed maize and common maize production fields using sentinel-2 images. Remote Sensing, 2020, 12(13): 2140.
[10]	VRIELING A, MERONI M, DARVISHZADEH R, SKIDMORE A K, WANG T J, ZURITA-MILLA R, OOSTERBEEK K, O’CONNOR B, PAGANINI M. Vegetation phenology from Sentinel-2 and field cameras for a Dutch barrier island. Remote Sensing of Environment, 2018, 215: 517-529.
[11]	SUN Z H, DI L P, FANG H. Using long short-term memory recurrent neural network in land cover classification on Landsat and Cropland data layer time series. International Journal of Remote Sensing, 2019, 40(2): 593-614.
[12]	WANG Y M, ZHANG Z, FENG L W, MA Y C, DU Q Y. A new attention-based CNN approach for crop mapping using time series Sentinel-2 images. Computers and Electronics in Agriculture, 2021, 184: 106090.
[13]	QU Y, ZHAO W Z, YUAN Z L, CHEN J G. Crop mapping from sentinel-1 polarimetric time-series with a deep neural network. Remote Sensing, 2020, 12(15): 2493.
[14]	杨丽, 吴雨茜, 王俊丽, 刘义理. 循环神经网络研究综述. 计算机应用, 2018, 38(S2): 1-6, 26.
	YANG L, WU Y X, WANG J L, LIU Y L. Research on recurrent neural network. Journal of Computer Applications, 2018, 38(S2): 1-6, 26. (in Chinese)
[15]	LUO C, MENG S Y, HU X, WANG X Y, ZHONG Y F. Cropnet: Deep Spatial-Temporal-Spectral Feature Learning Network for crop classification from Time-Series Multi-Spectral Images// Proceedings of the IGARSS 2020-2020 IEEE International Geoscience and Remote Sensing Symposium, 2020, 4187-4190.
[16]	KUSSUL N, LAVRENIUK M, SHUMILO L. Deep Recurrent Neural Network for crop classification task based on Sentinel-1 and Sentinel-2 imagery// Proceedings of the IGARSS 2020-2020 IEEE International Geoscience and Remote Sensing Symposium, 2020, 6914-6917.
[17]	ZHOU Y N, LUO J C, FENG L, YANG Y P, CHEN Y H, WU W. Long-short-term-memory-based crop classification using high-resolution optical images and multi-temporal SAR data. GIScience & Remote Sensing, 2019, 56(8): 1170-1191.
[18]	SCHUSTER M, PALIWAL K K. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing, 1997, 45(11): 2673-2681.
[19]	KWAK G H, PARK M G, PARK C W, LEE K D, NA S I, AHN H Y, PARK N W. Combining 2D CNN and bidirectional LSTM to consider spatio-temporal features in crop classification. Korean Journal of Remote Sensing, 2019, 35(51): 681-692.
[20]	毕恺艺, 牛铮, 黄妮, 康峻, 裴杰. 基于Sentinel-2A时序数据和面向对象决策树方法的植被识别. 地理与地理信息科学, 2017, 33(5): 16-20, 27, 127.
	BI K Y, NIU Z, HUANG N, KANG J, PEI J. Identifying vegetation with decision tree model based on object-oriented method using multi-temporal sentinel-2A images. Geography and Geo-Information Science, 2017, 33(5): 16-20, 27, 127. (in Chinese)
[21]	SANCHEZ A H, PICOLI M C A, CAMARA G, ANDRADE P R, CHAVES M E D, LECHLER S, SOARES A R, MARUJO R E B, SIMBES R E O, FERREIRA K R, QUEIROZ G R. Comparison of Cloud cover detection algorithms on sentinel-2 images of the amazon tropical forest. Remote Sensing, 2020, 12(8): 1284.
[22]	JIMéNEZ-MUñOZ J C, SOBRINO J A, PLAZA A, GUANTER L, MORENO J, MARTINEZ P. Comparison between fractional vegetation cover retrievals from vegetation indices and spectral mixture analysis: Case study of PROBA/CHRIS data over an agricultural area. Sensors, 2009, 9(2): 768-793.
[23]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory. Neural Computation, 1997, 9(8): 1735-1780.
[24]	CHO K, VAN MERRIENBOER B V, BAHDANAU D, BENGIO Y. On the properties of neural machine translation:Encoder-decoder approaches// Proceedings of SSST-8, Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation, 2014, 103-111.
[25]	赵红伟, 陈仲新, 刘佳. 深度学习方法在作物遥感分类中的应用和挑战. 中国农业资源与区划, 2020, 41(2): 35-49.
	ZHAO H W, CHEN Z X, LIU J. Deep learning for crop classification of remote sensing data: Applications and challenges. Chinese Journal of Agricultural Resources and Regional Planning, 2020, 41(2):35-49. (in Chinese)
[26]	ZHAO S, LIU X N, DING C, LIU S Y, WU C S, WU L. Mapping rice paddies in complex landscapes with convolutional neural networks and phenological metrics. GIScience & Remote Sensing, 2020, 57(1): 37-48.
[27]	KWAK G H, PARK C W, AHN H Y, NA S I, LEE K D, PARK N W. Potential of bidirectional long short-term memory networks for crop classification with multitemporal remote sensing images. Korean Journal of Remote Sensing, 2020, 36(4): 515-525.
[28]	RUSSWURM M, KÖRNER M. Self-attention for raw optical satellite time series classification. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 169: 421-435.
[29]	GOODFELLOW I, BENGIO Y, COURVILLE A. Deep Learning. Massachusetts: MIT Press, 2016.
[30]	QUIñONERO-CANDELA J, SUGIYAMA M, SCHWAIGHOFER A, LAWRENCE N D. Dataset Shift in Machine Learning. Massachusetts: MIT Press, 2009.
[31]	潘海珠. 基于作物多模型遥感数据同化的区域冬小麦生长模拟研究[D]. 北京: 中国农业科学院, 2020.
	PAN H Z. Winter wheat growth simulation based on multiple crop models and remote sensing data assimilation[D]. Beijing: Chinese Academy of Agricultural Sciences, 2020. (in Chinese)
[32]	ZENG L L, WARDLOW B D, XIANG D X, HU S, LI D R. A review of vegetation phenological metrics extraction using time-series, multispectral satellite data. Remote Sensing of Environment, 2020, 237: 111511.
[33]	XU J F, ZHU Y, ZHONG R H, LIN Z X, XU J L, JIANG H, HUANG J F, LI H F, LIN T. DeepCropMapping: A multi-temporal deep learning approach with improved spatial generalizability for dynamic corn and soybean mapping. Remote Sensing of Environment, 2020, 247: 111946.
[34]	LAKSHMINARAYANAN S K, MCCRAE J P. A comparative study of SVM and LSTM deep learning algorithms for stock market prediction// Proceedings of the AICS, 2019, 446-457.
[35]	庞敏. 基于LSTM混合模型的时间序列预测[D]. 武汉: 华中科技大学, 2019.
	PANG M. Time series forecasting based on LSTM hybrid model[D]. Wuhan: Huazhong University of Science & Technology, 2019. (in Chinese)
[36]	CHEN P H, LIN C J, SCHÖLKOPF B. A tutorial on ν-support vector machines. Applied Stochastic Models in Business and Industry, 2005, 21(2): 111-136.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Comments

Recommended 0

No Suggested Reading articles found!

可用影像 Data available	月份Month
可用影像 Data available	1月Jan.	2月Feb.	3月Mar.	4月Apr.	5月May	6月Jun.	7月Jul.	8月Aug.	9月Sept.	10月Oct.	11月Nov.	12月Dec.
数量Quantity	10	6	8	9	6	11	9	9	10	9	11	12

地类 Class	SVM		Bi-LSTM
地类 Class	用户精度 User accuracy	制图精度 Cartographic accuracy	用户精度 User accuracy	制图精度 Cartographic accuracy
人工林 Plantation forest	0.970	0.941	0.971	0.971
水稻 Rice	0.957	0.759	0.962	0.879
棉花 Cotton	0.660	0.875	0.923	0.900
春玉米 Spring maize	0.800	0.830	0.847	0.943
荒地 Uncultivated land	0.886	0.899	0.930	0.957
冬小麦-夏大豆 Winter wheat-summer soybeans	0.667	0.714	0.769	0.714
冬小麦-夏玉米 Winter wheat-summer maize	0.887	0.833	0.879	0.879
总体精度 Overall accuracy	0.848		0.909
Kappa系数 Kappa coefficient	0.821		0.892

泛化能力测试区 Generalization ability test region	典型作物 Typical crop	Bi-LSTM	SVM	错分类型 Misclassification type
河口区种植区 Hekou test region	水稻 Rice	0.894	0.830	春玉米、荒地 Spring corn, Uncultivate land
广饶县种植区 Guangrao test region	春玉米 Spring corn	0.903	0.838	棉花 Cotton
广饶县种植区 Guangrao test region	冬小麦-夏玉米 Winter wheat & summer corn	0.824	0.824	冬小麦-夏大豆 Winter wheat & summer soybeans
冀州区种植区 Jizhou test region	冬小麦-夏玉米 Winter wheat & summer corn	0.836	0.787	冬小麦-夏大豆 Winter wheat & summer soybeans

评价指标 Index	作物类别 Crop types	最早可识别时间 Earliest identifiable timing (EIT)	到达该月份时的精度 Accuracy (as of EIT)
F1-分数 F1-score	水稻 Rice	6月 Jun.	0.870
	棉花 Cotton	10月 Oct.	0.897
	春玉米 Spring corn	9月 Sept.	0.857
	冬小麦-夏大豆 Winter wheat & summer soybean	—	—
	冬小麦-夏玉米 Winter wheat & summer corn	10月 Oct.	0.855
	冬小麦 Winter wheat	4月 Apr.	0.874
总体准确率 Overall Accuracy		9月 Sept.	0.864
总体Kappa（9月份）Overall Kappa (Sep.)		—	0.839

Crop Classification with Time Series Remote Sensing Based on Bi-LSTM Model

RichHTML

PDF

Abstract

Cite this article

share this article

Figures/Tables 13

References 36

Related Articles 2

Metrics

Comments

Recommended 0

[1]	MI XiaoTian,SHI Lei,HE Gang,WANG ZhaoHui. Fertilizer Reduction Potential and Economic Benefits of Crop Production for Smallholder Farmers in Shaanxi Province [J]. Scientia Agricultura Sinica, 2021, 54(20): 4370-4384.
[2]	LIU HuanJun, YU ShengNan, ZHANG XinLe, GUO Dong, YIN JiXian. Timeliness Analysis of Crop Remote Sensing Classification One Crop A Year [J]. Scientia Agricultura Sinica, 2017, 50(5): 830-839.