基于车载相机和HLS时序遥感数据的作物分类研究

doi:10.19788/j.issn.2096-6369.000098

农业大数据学报 ›› 2025, Vol. 7 ›› Issue (2): 161-172.doi: 10.19788/j.issn.2096-6369.000098

基于车载相机和HLS时序遥感数据的作物分类研究

钱涛¹^,²^,³^,⁴(), 詹雅婷⁵^,⁶^,⁷, 李胤⁵^,⁶^,⁷, 宋珂⁵^,⁶^,⁷, 邵明超¹^,²^,³^,⁴, 虞钟直¹^,²^,³^,⁴, 程涛¹^,²^,³^,⁴, 姚霞¹^,²^,³^,⁴, 郑恒彪¹^,²^,³^,⁴, 朱艳¹^,²^,³^,⁴^,⁷, 曹卫星¹^,²^,³^,⁴, 江冲亚¹^,²^,³^,⁴^,⁷^,^*()

1.南京农业大学国家信息农业工程技术中心，南京 210095
2.智慧农业教育部工程研究中心，南京 210095
3.农业农村部农作物系统分析与决策重点实验室，南京 210095
4.江苏省信息农业重点实验室，南京 210095
5.江苏省地质调查研究院，南京 210018
6.自然资源江苏省卫星应用技术中心，南京 210018
7.江苏省自然资源厅卫星遥感应用重点实验室，南京 210018

收稿日期:2025-01-28 接受日期:2025-03-19 出版日期:2025-06-26 发布日期:2025-06-23
通讯作者: 江冲亚，E-mail: chongya@njau.edu.cn。
作者简介:钱涛，E-mail: qiantao@stu.njau.edu.cn。
基金资助:
国家重点研发计划课题(2023YFD2000103);中央高校基本科研业务费(QTPY2025010);江苏特聘教授;江苏省自然资源厅2024年度科技计划项目(2024023);江苏省自然资源厅2024年度科技计划项目(2024007)

Crop Classification Research Based on Vehicle Images and HLS Time-series Remote Sensing Data

QIAN Tao¹^,²^,³^,⁴(), ZHAN YaTing⁵^,⁶^,⁷, LI Yin⁵^,⁶^,⁷, SONG Ke⁵^,⁶^,⁷, SHAO MingChao¹^,²^,³^,⁴, YU ZhongZhi¹^,²^,³^,⁴, CHENG Tao¹^,²^,³^,⁴, YAO Xia¹^,²^,³^,⁴, ZHENG HengBiao¹^,²^,³^,⁴, ZHU Yan¹^,²^,³^,⁴^,⁷, CAO WeiXing¹^,²^,³^,⁴, JIANG ChongYa¹^,²^,³^,⁴^,⁷^,^*()

1. National Engineering Technology Center for Information Agriculture, Nanjing 210095, China
2. Intelligent Agriculture Engineering Research Center, Ministry of Education, Nanjing 210095, China
3. Key Laboratory for Crop System Analysis and Decision Making, Ministry of Agriculture and Rural Affairs, Nanjing 210095, China
4. Jiangsu Key Laboratory for Information Agriculture, Nanjing 210095, China
5. Jiangsu Institute of Geological Survey, Nanjing 210018, China
6. Jiangsu Satellite Application Technology Center of natural resources, Nanjing 210018, China
7. Key Laboratory of Satellite Remote Sensing Application, Jiangsu Provincial Department of Natural Resources, Nanjing 210018, China

Received:2025-01-28 Accepted:2025-03-19 Published:2025-06-26 Online:2025-06-23

摘要/Abstract

摘要：

旨在探讨基于车载相机和HLS时序遥感数据相结合的作物分类方法，以提高作物分类的效率和精度，解决传统方法中地面样本采集效率低、遥感物候特征利用不充分等问题。研究以江苏省秋粮分类为例，验证了此方法的可行性和应用潜力。通过车载相机采集道路两侧作物图像并构建大量人工标注样本，结合2023年和2024年的HLS时序数据，采用高斯滤波重构时间连续的地表反射率，提取多维特征，构建随机森林分类模型。研究结果表明，水稻、玉米和大豆在HLS时序数据中表现出较为明显的差异。水稻的分类精度最高，生产者精度与用户精度均超过90%，而玉米和大豆因物候特征的相似性，精度相对较低（74%-85%）。模型在独立验证县的总体分类精度为89%，验证县内的水稻主要分布于全县东南区域，玉米和大豆则集中于西北区域，且分布特征清晰。车载相机结合HLS时序数据可实现高效的作物分类，随机森林模型对高维特征整合与抗样本不平衡性具有显著优势。尽管模型总体表现优良，但在破碎地块和高云量区域仍有改进空间。未来需融合多源遥感数据缓解云干扰，并扩展作物类型以增强模型泛化能力。

关键词: 车载相机, HLS, 农作物, 遥感分类, 农业大数据

Abstract:

This study aims to develop a crop classification method by integrating vehicle images with HLS time-series remote sensing data. The goal is to enhance classification efficiency and accuracy, addressing the limitations of traditional methods such as low efficiency in ground sample collection and insufficient utilization of remote sensing phenological features. A vehicle-mounted camera system was deployed to collect manually annotated crop samples along road networks, combined with HLS time-series data from 2023 and 2024. Gaussian filtering was applied to reconstruct the time-series imagery, and the Random Forest classification method was employed to classify three major crops: rice, maize, and soybean. Results demonstrated significant differences in the characteristics of rice, maize, and soybean in the HLS time-series data. Among these crops, rice achieved the highest classification accuracy, with both producer's and user's accuracy exceeding 90%, whereas maize and soybean had lower accuracies (74%-85%) due to their similar phenological characteristics. The overall classification accuracy in the validation area was 89%. The rice in the verification area is mainly distributed in the southeast region of the county, while corn and soybeans are concentrated in the northwest region, and their distribution characteristics are clear. The integration of vehicle images and HLS time-series data proves effective for crop classification, with the Random Forest model demonstrating superior performance in handling high-dimensional features and sample imbalance. However, challenges remain in fragmented farmland and cloud-covered areas. Future improvements should focus on incorporating multi-source data to address cloud contamination and mixed-pixel effects in fragmented areas, while expanding crop categories to enhance model generalizability for broader agricultural applications.

Key words: vehicle images, HLS, crops, remote sensing classification, agricultural big data

钱涛, 詹雅婷, 李胤, 宋珂, 邵明超, 虞钟直, 程涛, 姚霞, 郑恒彪, 朱艳, 曹卫星, 江冲亚. 基于车载相机和HLS时序遥感数据的作物分类研究[J]. 农业大数据学报, 2025, 7(2): 161-172.

QIAN Tao, ZHAN YaTing, LI Yin, SONG Ke, SHAO MingChao, YU ZhongZhi, CHENG Tao, YAO Xia, ZHENG HengBiao, ZHU Yan, CAO WeiXing, JIANG ChongYa. Crop Classification Research Based on Vehicle Images and HLS Time-series Remote Sensing Data[J]. Journal of Agricultural Big Data, 2025, 7(2): 161-172.

图/表 10

图1

图2

图3

表1

图4

图5

表2

表3

图6

图7

参考文献 35

[1]	黄翀, 侯相君. 基于Bi-LSTM模型的时间序列遥感作物分类研究. 中国农业科学, 2022, 55(21): 4144-4157. doi: 10.3864/j.issn.0578-1752.2022.21.005
[2]	赵龙才, 李粉玲, 常庆瑞. 农作物遥感识别与单产估算研究综述. 农业机械学报, 2023, 54(2): 1-19.
[3]	YI Z, JIA L, CHEN Q. Crop classification using Multi-temporal Sentinel-2 Data in the Shiyang River Basin of China. Remote Sensing, 2020, 12(24): 4052.
[4]	LUO K, LU L, XIE Y, et al. Crop type mapping in the central part of the North China Plain using Sentinel-2 time series and machine learning. Computers and Electronics in Agriculture, 2023, 205: 107577.
[5]	LI J, LEI H. Tracking the spatio-temporal change of planting area of winter wheat-summer maize cropping system in the North China Plain during 2001-2018. Computers and Electronics in Agriculture, 2021, 187: 106222.
[6]	YAN Y, RYU Y. Exploring Google Street View with deep learning for crop type mapping. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 171: 278-296.
[7]	WU F, WU B, ZHANG M, et al. Identification of crop type in crowdsourced road view photos with Deep Convolutional Neural Network. Sensors, 2021, 21(4): 1165.
[8]	JIANG C, GUAN K, HUANG Y, et al. A vehicle imaging approach to acquire ground truth data for upscaling to satellite data: A case study for estimating harvesting dates. Remote Sensing of Environment, 2024, 300: 113894.
[9]	PALIYAM M, NAKALEMBE C, LIU K, et al. Street2sat: A machine learning pipeline for generating ground-truth geo-referenced labeled datasets from street-level images[C]. ICML 2021 Workshop on Tackling Climate Change with Machine Learning. 2021.
[10]	PILGER N, BERG A, JOOSSE P. Semi-automated roadside image data collection for characterization of agricultural land management practices. Remote Sensing, 2020, 12(14): 2342.
[11]	JIA D, GAO P, CHENG C, et al. Multiple-feature-driven co-training method for crop mapping based on remote sensing time series imagery. International Journal of Remote Sensing, 2020, 41(20): 8096-8120.
[12]	TIAN X, CHEN Z, LI Y, et al. Crop classification in mountainous areas using Object-oriented methods and multi-source data: A case study of Xishui County, China. Agronomy, 2023, 13(12): 3037.
[13]	CLAVERIE M, JU J, MASEK J G, et al. The Harmonized Landsat and Sentinel-2 surface reflectance data set. Remote Sensing of Environment, 2018, 219: 145-161.
[14]	LI B, TI C, YAN X. Estimating rice paddy areas in China using multi-temporal cloud-free normalized difference vegetation index (NDVI) imagery based on change detection. Pedosphere, 2020, 30(6): 734-746.
[15]	JONSSON P, EKLUNDH L. Seasonality extraction by function fitting to time-series of satellite sensor data. IEEE Transactions on Geoscience and Remote Sensing, 2002, 40(8): 1824-1832.
[16]	LIANG J, REN C, LI Y, et al. Using Enhanced Gap-Filling and Whittaker Smoothing to Reconstruct High Spatiotemporal Resolution NDVI Time Series Based on Landsat 8, Sentinel-2, and MODIS Imagery. ISPRS International Journal of Geo-Information, 2023, 12(6): 214.
[17]	WU G, JIANG C, KIMM H, et al. Difference in seasonal peak timing of soybean far-red SIF and GPP explained by canopy structure and chlorophyll content. Remote sensing of environment, 2022, 279: 113104.
[18]	BREIMAN L. Random forests. Machine Learning, 2001, 45(1): 5-32.
[19]	WANG S, AZZARI G, LOBELL D B. Crop type mapping without field-level labels: Random forest transfer and unsupervised clustering techniques. Remote Sensing of Environment, 2019, 222: 303-317.
[20]	LI H, SONG X P, HANSEN M C, et al. Development of a 10-m resolution maize and soybean map over China: Matching satellite- based crop classification with sample-based area estimation. Remote Sensing of Environment, 2023, 294: 113623.
[21]	XU J, CHEN C, ZHOU S, et al. Land use classification in mine- agriculture compound area based on multi-feature random forest: A case study of Peixian. Frontiers in Sustainable Food Systems, 2024, 7: 1335292.
[22]	CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 2002, 16: 321-357.
[23]	DOUZAS G, BACAO F, FONSECA J, et al. Imbalanced learning in land cover classification: Improving minority classes’ prediction accuracy using the geometric SMOTE algorithm. Remote Sensing, 2019, 11(24): 3040.
[24]	张馨予, 蔡志文, 杨靖雅, 等. 时序滤波对农作物遥感识别的影响. 农业工程学报, 2022, 38(4): 215-224.
[25]	HUETE A, DIDAN K, MIURA T, et al. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sensing of Environment, 2002, 83(1-2): 195-213.
[26]	XIAO X, ZHANG Q, HOLLINGER D, et al. Modeling gross primary production of an evergreen needleleaf forest using MODIS and climate data. Ecological Applications, 2005, 15(3): 954-969.
[27]	GITELSON A A, KAUFMAN Y J, STARK R, et al. Novel algorithms for remote estimation of vegetation fraction. Remote Sensing of Environment, 2002, 80(1): 76-87.
[28]	ZHAI Y, ROY D P, MARTINS V S, et al. Conterminous United States Landsat-8 top of atmosphere and surface reflectance tasseled cap transformation coefficients. Remote Sensing of Environment, 2022, 274: 112992.
[29]	熊曦柳, 胡月明, 文宁, 等. 耕地遥感识别研究进展与展望. 农业资源与环境学报, 2020, 37(6): 856-865.
[30]	ORYNBAIKYZY A, GESSNER U, MACK B, et al. Crop type classification using fusion of Sentinel-1 and Sentinel-2 data: Assessing the impact of feature selection, optical data availability, and parcel sizes on the accuracies. Remote Sensing, 2020, 12(17): 2779.
[31]	何真, 胡洁, 蔡志文, 等. 协同多时相国产GF-1和GF-6卫星影像的艾草遥感识别. 农业工程学报, 2022, 38(1): 186-195.
[32]	JIN M, WANG P, LI Y. HyA-GAN: Remote sensing image cloud removal based on hybrid attention generation adversarial network. International Journal of Remote Sensing, 2024, 45(6): 1755-1773.
[33]	WEI P, YE H, QIAO S, et al. Early crop mapping based on Sentinel-2 time-series data and the random forest algorithm. Remote Sensing, 2023, 15(13): 3212.
[34]	WANG S, DI TOMMASO S, FAULKNER J, et al. Mapping crop types in southeast India with smartphone crowdsourcing and deep learning. Remote Sensing, 2020, 12(18): 2957.
[35]	SÁNCHEZ A, NANCLARES R, PELAGIO U, et al. Agave crop segmentation and maturity classification with deep learning data- centric strategies using very high-resolution satellite imagery. International Journal of Remote Sensing, 2023, 44(22): 7017-7032.

样本类型	获取年份	地区	样本数量（个）
样本类型	获取年份	地区	水稻	玉米	大豆	其他
训练样本	2024	淮安市涟水县唐集镇	254	28	88	115
		连云港市东海县驼峰乡	296	42	4	107
		南京市溧水区白马镇	113	-	-	172
		泰州市兴化市昌荣镇	16	-	1	132
		徐州市丰县师寨镇	23	550	68	235
		盐城市滨海县五汛镇	453	7	20	84
验证样本	2023	盐城市滨海县	127	93	65	180

类别		分类结果				生产者精度/%
类别		水稻	玉米	大豆	其他	生产者精度/%
验证样本	水稻	121	1	1	4	95.28
	玉米	3	71	13	6	76.34
	大豆	3	9	49	4	75.38
	其他	1	3	3	173	96.11
用户精度/%		94.53	84.52	74.24	92.51
总体精度89.03%，Kappa系数0.85

类别	SVM			RF			CNN
类别	UA (%)	PA (%)	F1-score	UA (%)	PA (%)	F1-score	UA (%)	PA (%)	F1-score
水稻	78.52	83.46	0.81	94.53	95.28	0.95	82.31	95.28	0.89
玉米	59.69	82.80	0.69	84.52	76.34	0.80	75.49	82.80	0.79
大豆	72.73	49.23	0.59	74.24	75.38	0.75	85.71	36.92	0.52
其他	94.27	82.22	0.88	92.51	96.11	0.94	92.55	96.67	0.95
总体精度 OA (%)	78.06			89.03			85.16
Kappa系数	0.69			0.85			0.79

基于车载相机和HLS时序遥感数据的作物分类研究

Crop Classification Research Based on Vehicle Images and HLS Time-series Remote Sensing Data

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 35

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	高卓君, 张丹丹, 陈荣宇. 2016—2023年广东省主要农作物审定品种知识图谱构建数据集[J]. 农业大数据学报, 2025, 7(2): 261-268.
[2]	岳瑞君, 何亮, 汤敏睿, 严威, 刘胜全, 杨婉霞, 孙卫红, 黄永峰. 基于云链融合的农业大数据安全治理模型研究[J]. 农业大数据学报, 2024, 6(3): 333-350.
[3]	吴云坤, 杨莹, 李豪, 熊健, 陈湘灵. 智慧农业领域大数据安全问题探索[J]. 农业大数据学报, 2024, 6(3): 380-391.
[4]	陶耘, 成谢锋. 区域大数据发展与区域农业大数据建设水平比较研究[J]. 农业大数据学报, 2022, 4(1): 125-135.
[5]	汪汇涓, 徐倩, 周爱莲, 梁晓贺, 谢能付, 李小雨, 吴赛赛. 区块链的发展历程及在农业领域的应用展望[J]. 农业大数据学报, 2021, 3(3): 76-86.
[6]	李强, 高懋芳, 方莹. 农业大数据信息平台构建方法初探[J]. 农业大数据学报, 2021, 3(2): 24-30.
[7]	陈富桥, 凌晨. 茶叶全产业链大数据中心功能设计与开发进展[J]. 农业大数据学报, 2021, 3(2): 54-66.
[8]	蒋锐, 黄凤洪, 吴渝, 霍梦佳, 刘华威. 油料（油菜、花生）全产业链大数据的建设[J]. 农业大数据学报, 2021, 3(2): 67-74.
[9]	顾君, 贾暑花, 曾庆鸿. 基于知识中台的农业单品全产业链大数据平台建设研究[J]. 农业大数据学报, 2021, 3(1): 25-32.
[10]	张杰, 刘升平, 岳慧丽, 吕纯阳, 洪葳. 智慧蜂业大数据平台建设与应用[J]. 农业大数据学报, 2021, 3(1): 3-13.
[11]	温晗秋子, 郑永强, 刘杨. 柑橘大数据研究与应用[J]. 农业大数据学报, 2021, 3(1): 33-44.
[12]	刘海燕, 杨榕, 侯彤瑜, 赵维, 姚兆群, 王海江, 张泽, 高攀, 吕新. 新疆棉田土壤微生物资源大数据平台建设与可视化分析[J]. 农业大数据学报, 2021, 3(1): 45-55.
[13]	段青玲, 刘怡然, 周新辉, 任妮, 李道亮. 大闸蟹养殖大数据分析模型和应用进展[J]. 农业大数据学报, 2021, 3(1): 56-65.
[14]	董春岩, 牛明雷, 姚艳, 常晓燕, 吕凌峰, 李楠. 蔬菜全产业链大数据平台建设与应用研究——以大白菜为例[J]. 农业大数据学报, 2021, 3(1): 66-72.
[15]	顾君, 齐晓军, 苑青微, 曾庆鸿. 农业单品全产业链大数据平台设计与实现[J]. 农业大数据学报, 2021, 3(1): 73-80.