农业大数据学报 ›› 2024, Vol. 6 ›› Issue (4): 546-551.doi: 10.19788/j.issn.2096-6369.100031

• • 上一篇    下一篇

2023年黑龙江水稻产量预测无人机遥感图像数据集

苑江浩1(), 郑作军1, 初昌明3, 姚鸿勋5, 刘海龙4,*(), 郭雷风1,4,*()   

  1. 1.河北农业大学,河北 保定 071001
    2.国家粮食和物资储备局科学研究院,北京 100037
    3.北大荒农业股份有限公司二九〇分公司农业技术推广中心,黑龙江 绥化 156202
    4.中国农业科学院农业信息学研究所,北京 100081
    5.哈尔滨工业大学,哈尔滨 150001
  • 收稿日期:2024-04-17 接受日期:2024-06-09 出版日期:2024-12-26 发布日期:2024-12-02
  • 通讯作者: 郭雷风,E-mail:guoleifeng@caas.cn
    刘海龙,E-mail:liuhailong@caas.cn
  • 作者简介:苑江浩,E-mail:yjh@ags.ac.cn
  • 基金资助:
    新一代人工智能国家科技重大专项(2021ZD0110901)

Rice Yield Prediction UAV Remote Sensing Image Dataset of Heilongjiang Province in 2023

YUAN JiangHao1(), ZHENG ZuoJun1, CHU ChangMing3, YAO HongXun5, LIU HaiLong4,*(), GUO LeiFeng1,4,*()   

  1. 1. Hebei Agricultural University, Baoding 071001, Hebei, China
    2. Academy of National Food and Strategic Reserves Administration, Beijing 100037, China
    3. Agricultural Technology Promotion Center of Beidahuang Agriculture Co., Ltd. 290 Branch, Suihua 156202, Heilongjiang, China
    4. Agricultural Information Institute of Chinese Academy of Agricultural Sciences, Beijing 100081, China
    5. Harbin Institute of Technology, Harbin 150001, China
  • Received:2024-04-17 Accepted:2024-06-09 Published:2024-12-26 Online:2024-12-02

摘要:

水稻是我国三大粮食作物之一,准确、高效、及时地预测水稻产量对品种选育和优化田间管理至关重要。无人机遥感系统凭借其快速、无损、成本低、通量高等优势,被广泛应用在作物病虫害识别、作物生长监测和作物表型分析等领域。为探究光谱数据在水稻产量估测方面发挥的作用,本数据集利用无人机遥感采集了水稻生长过程中的多光谱图像。选取106个1 m×1 m的样本点人工采样测产,同时在采样后采集了可见光图像,实现光谱图像和产量数据间的关联。经过人工检查和整理构建了本数据集。数据采集地点为黑龙江省,无人机在无云、光照充足的条件下进行数据采集,采集时间为2023年7月至2023年8月,共采集试验田内不同品种3天的多光谱数据和1天的可见光数据。本数据集各项数据完整,可为产量估测研究提供数据支撑。

数据摘要:

项目 描述
数据库(集)名称 2023年黑龙江水稻产量预测无人机遥感图像数据集
所属学科 农业科学
研究主题 计算机视觉
数据时间范围 2023年7月—2023年8月
时间分辨率
数据类型与技术格式 .tif,.xlsx,.jpg
数据库(集)组成 数据集由三部分数据组成,第一部分为水稻生长过程中多光谱图像数据,包括蓝(450 nm)、绿(555 nm)、红(660 nm)、红边1(720 nm)、红边2(750 nm)和近红外(840 nm)六个光谱通道,共计14226张,约32.6GB;第二部分为产量数据,以.xlsx格式保存;第三部分为用于标注采样点的可见光图像数据,共计746张,约18.9 GB。
数据量 51.5 GB
主要数据指标 梯度设置,地块标号,产量,多光谱图像,可见光图像
数据可用性 CSTR:https://cstr.cn/17058.11.sciencedb.agriculture.00131
DOI:https://doi.org/10.57760/sciencedb.agriculture.00131
NASDC访问链接: https://agri.scidb.cn/,限制性获取
经费支持 新一代人工智能国家科技重大专项(2021ZD0110901)

关键词: 无人机, 水稻, 多光谱图像, 黑龙江

Abstract:

Rice is one of the three major grain crops in China, and accurate, efficient and timely prediction of rice yield is crucial for variety selection and optimization of field management. UAV remote sensing system is widely used in crop pest and disease identification, crop growth monitoring and crop phenotyping by virtue of its advantages of fast, non-destructive, low cost and high throughput. To explore the role of spectral data in estimating rice yield, this dataset used UAV remote sensing to collect multispectral images of rice growth process, 106 sample points of 1 m×1 m were selected for manual sampling and yield measurement, and at the same time, visible images were collected after the sampling to realize the correlation between spectral images and yield data. The dataset of this paper was constructed after manual checking and organizing. The data collection location was Heilongjiang Province, and the UAV collected the data under cloudless and light-sufficient conditions, and the collection time was from July to August in 2023, and a total of 3 days of multispectral data and 1 day of visible light data were collected with different varieties in the experimental field. The dataset in this paper was complete in all data and provided data support for research on yield estimation.

Data summary:

Items Description
Dataset name Rice Yield Prediction UAV Remote Sensing Image Dataset of Heilongjiang Province in 2023
Specific subject area Agricultural Science
Research Topic computer vision
Time range July 2023- August 2023
Temporal resolution Day
Data types and technical formats .tif,.xlsx,.jpg
Dataset structure The dataset consists of three parts of data. The first part is the multispectral image data of the entire growth period of rice, including six spectral channels: blue (450nm), green (555nm), red (660nm), red edge 1 (720nm), red edge 2 (750nm), and near-infrared (840nm), with a total of 14226 images, approximately 32.6GB; The second part is production data, saved in. xlsx format; The third part is visible light image data used to annotate sampling points, totaling 746 images, approximately 18.9GB.
Volume of dataset 51.5 GB
Key index in dataset Gradient settings, plot labeling, yield, multispectral images, RGB images
Data accessibility CSTR:https://cstr.cn/17058.11.sciencedb.agriculture.00131
DOI:https://doi.org/10.57760/sciencedb.agriculture.00131
NASDC Access link: https://agri.scidb.cn/, restricted access
Financial support National Science and Technology Major Project(2021ZD0110901)

Key words: unmanned aerial vehicle (UAV), rice, multispectral imagery, Heilongjiang