WO2021068176A1 - 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 - Google Patents
一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 Download PDFInfo
- Publication number
- WO2021068176A1 WO2021068176A1 PCT/CN2019/110509 CN2019110509W WO2021068176A1 WO 2021068176 A1 WO2021068176 A1 WO 2021068176A1 CN 2019110509 W CN2019110509 W CN 2019110509W WO 2021068176 A1 WO2021068176 A1 WO 2021068176A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- remote sensing
- neural network
- crop
- convolutional neural
- time series
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
Definitions
- the invention relates to a crop planting distribution prediction method based on time series remote sensing data and convolutional neural network.
- the purpose of the present invention is to provide a crop planting distribution prediction method based on time series remote sensing data and convolutional neural network to solve the problem that the accuracy of planting distribution prediction is not high due to insufficient use of time series information or context information in the prior art .
- the method for predicting crop planting distribution based on time series remote sensing data and convolutional neural network includes the following steps:
- Step 1 Ground survey and training sample establishment
- Step 2 Construct a crop planting distribution prediction model based on time series remote sensing data and a convolutional neural network.
- the convolutional neural network predicts the target pixel and its surrounding pixels in the multi-temporal image, and the input value is Multi-temporal high-resolution multi-spectral images, the output value is the classification information of crop type and crop rotation mode;
- Step 3 Input the time series remote sensing data of the statistical area into the built model to obtain the recognition result.
- the step 1 includes the following steps:
- Step 1.1 Investigate the main planting crop types, crop rotation methods, and key phenological time nodes in the statistical area;
- Step 1.2 According to the different crop types and rotation methods in step 1.1, select some representative plots, and record the location and scope of plots, planting area, planting crops, planting time, etc. in detail to obtain training samples.
- the step 2 includes the following steps:
- Step 2.1 Collect remote sensing images containing the representative plots selected in step 1.2 at different times, accurately register the remote sensing images at different times, and classify and mark the areas of different crop types and crop rotation methods pixel by pixel;
- Step 2.2 Take time series remote sensing data as the input of the model, and the classification information of crop type and crop rotation as the output of the model, and build a multi-layer convolutional neural network;
- Step 2.3 Use the training samples in step 2.1 to fully train the parameters in the multi-layer convolutional neural network using the backpropagation algorithm.
- each band data of the remote sensing image in different time phases is stored as multiple color channel images, and the prediction is based on the field size, the number of time phases, and the number of image channels per time phase. Obtain the corresponding pixel data to predict the target pixel.
- the step 3 includes the following steps:
- Step 3.1 Collect remote sensing images with the same time as the time of collecting remote sensing images during neural network training, and accurately register remote sensing images at different times;
- Step 3.2 Input the registered time series remote sensing data in step 3.1 into the multi-layer convolutional neural network model trained in step 2.3 to obtain the crop planting classification prediction results of the research area.
- the time for collecting remote sensing images is from March to November in the middle of each month.
- the present invention only needs to pass the ground survey of a small number of representative plots, combined with the high-resolution multispectral image provided by the satellite data, can give the prediction results of the crop planting distribution, and can greatly reduce the on-site survey cost of the ground investigators.
- Historical remote sensing information can also be used to predict crop planting distribution in previous years; on the other hand, a fusion of remote sensing data sequence features and remote sensing image local features is constructed to improve the accuracy of the prediction results.
- this method introduces the context information of the decision point.
- the target pixel not only the multi-phase image data is collected, the target pixel data is used for prediction, and the target pixel is collected through the set neighborhood size.
- the pixel data around the point is a point-oriented prediction, and it also combines the data of multiple image channels, which greatly improves the accuracy of the prediction result.
- Fig. 1 is a flowchart of a method for predicting crop planting distribution based on time series remote sensing data and convolutional neural network according to the present invention.
- the present invention provides a crop planting distribution prediction method based on time-series remote sensing data and convolutional neural network.
- the method includes the following steps:
- Step 1 Ground survey and training sample establishment; specifically include the following steps:
- Step 1.1 Investigate the main planting crop types, crop rotation methods, and key phenological time nodes in the statistical area;
- Step 1.2 According to the different crop types and rotation methods in step 1.1, select some representative plots, and record the location and scope of plots, planting area, planting crops, planting time, etc. in detail to obtain training samples.
- Step 2 Construct a crop planting distribution prediction model based on time series remote sensing data and a convolutional neural network.
- the convolutional neural network predicts the target pixel and its surrounding pixels in the multi-temporal image, and the input value is Multi-temporal high-resolution multi-spectral images (that is, remote sensing images, input after processing), the output value is the classification information of crop type and crop rotation mode; specifically includes the following steps:
- Step 2.1 Collect remotely sensed images of the representative plots selected in step 1.2 from March to mid-November, accurately register the remotely sensed images at different times, and compare the regions of different crop types and crop rotation methods one by one. Pixel classification mark; in the process of image processing, each band data of remote sensing images at different time phases are stored in multiple color channel images.
- Step 2.2 Take the time series remote sensing data as the input of the model, and the classification information of crop type and crop rotation as the output of the model, and construct a multi-layer convolutional neural network.
- the data of the corresponding pixel is obtained according to the field size, the number of time phases and the number of image channels per time phase to predict the target pixel.
- the commonly used image channel r (red) + g (green) + b (blue) + nir (near infrared spectrum) is used.
- the collected remote sensing images are processed to form time series remote sensing data corresponding to each image channel according to the time phase.
- the field is set to 5.
- the model used in this method needs to collect pixel data of neighborhood size * number of phases * number of image channels per time phase during prediction, which is a point-oriented prediction, which greatly improves the accuracy of the prediction result compared with the prior art.
- the prediction based on different time phases can not only predict the types of crops, but also predict whether the crops are early rice or late rice in the region according to different crop phenological periods, which is not available in the prior art.
- Step 2.3 Use the training samples in step 2.1 to fully train the parameters in the multi-layer convolutional neural network using the backpropagation algorithm.
- Step 3 Input the time series remote sensing data of the statistical area into the constructed model to obtain the recognition results; specifically, it includes the following steps:
- Step 3.1 Collect the remote sensing images of the study area from March to November in the middle of each month, and accurately register the remote sensing images at different times;
- Step 3.2 Input the registered time series remote sensing data in step 3.1 into the multi-layer convolutional neural network model trained in step 2.3 to obtain the crop planting classification prediction results of the research area.
- the present invention only needs to conduct ground surveys on a small number of representative plots, constructs a prediction model that combines the time series characteristics of remote sensing data and local characteristics of remote sensing images, introduces context information of decision points, and improves the accuracy of prediction results.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，包括下列步骤：步骤1：地面调查及训练样本建立；步骤2：构造基于时序遥感数据和卷积神经网络的作物种植分布预测模型，所述卷积神经网络通过在多时相图像中目标像素点及其周围像素点的数据对其进行预测，输入值为多时相的高分辨率多光谱图像，输出值为作物类型、轮作方式的分类信息；步骤3：将统计区域的时序遥感数据输入已构建的模型获取识别结果。只需要通过少量的代表性地块的地面调查，构建了融合遥感数据时序特征和遥感图像局部特征的预测模型，引入了决策点的上下文信息，提高了预测结果的准确性。
Description
本发明涉及一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法。
获取农作物种植分布的时空变化信息是国家掌握耕地资源状况、粮食生产态势以及生态系统保护的重要环节。种植分布信息具有时间变异性和空间变异性，使用地面调查方法获取这些信息耗费巨大的人力物力。随着卫星及传感器技术的进步，高空间分辨率遥感卫星已实现周期性地对地进行观测，所搭载的传感器能有效获取作物光谱。在各种作物生长的不同物候期，光谱会呈现不同的特征。利用机器学习模型，结合研究区域主要作物的物候期，对高分辨率多光谱时序遥感数据进行处理，能够预测作物种植情况。现有模型方法中，由于没有充分利用时序信息或上下文信息，导致种植分布预测准确度不高，因此需要一种基于时序数据和上下文信息的作物种植分布预测方法。
发明内容
本发明的目的在于提供一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，以解决现有技术中由于没有充分利用时序信息或上下文信息，导致种植分布预测准确度不高的问题。
所述的一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，包括下列步骤：
步骤1：地面调查及训练样本建立；
步骤2：构造基于时序遥感数据和卷积神经网络的作物种植分布预测模型，所述卷积神经网络通过在多时相图像中目标像素点及其周围像素点的数据对其进行预测，输入值为多时相的高分辨率多光谱图像，输出值为作物类型、轮作方式的分类信息；
步骤3：将统计区域的时序遥感数据输入已构建的模型获取识别结果。
优选的，所述步骤1包括如下步骤：
步骤1.1：调研统计区域的主要种植作物类型、轮作方式、关键物候期时间节点；
步骤1.2：针对步骤1.1中不同作物类型、轮作方式，各选取部分代表性地块，详细记录地块位置及范围、种植面积、种植作物、种植时间等，得到训练样本。
优选的，所述步骤2包括如下步骤：
步骤2.1：收集不同时期包含步骤1.2中选取的代表性地块的遥感图像，将不同时间的遥感图像进行精确配准，并将不同作物类型、轮作方式的区域进行逐像素的分类标记；
步骤2.2：将时序遥感数据作为模型的输入，作物类型、轮作方式的分类信息作为模型的输出，构建多层卷积神经网络；
步骤2.3：利用步骤2.1中的训练样本，使用反向传播算法对多层卷积神经网络中的参数进行充分训练。
优选的，所述步骤2中，图像处理过程中将不同时相遥感图像的各波段数据分别以多种颜色通道图像存储，在进行预测时根据领域大小、时相数和每时相图像通道数获取相应像素点的数据对目标像素点进行预测。
优选的，所述步骤3包括如下步骤：
步骤3.1：收集研究区域范围的时间与神经网络训练时采集遥感图像时间相同的遥感图像，将不同时间的遥感图像进行精确配准；
步骤3.2：将步骤3.1中的配准后的时序遥感数据输入到步骤2.3中训练的多层卷积神经网络模型中，获取研究区域的作物种植分类预测结果。
优选的，神经网络训练时，采集遥感图像的时间为三月到十一月每月中旬。
本发明具有如下优点：
本发明一方面只需要通过少量的代表性地块的地面调查，结合卫星数据提供的高分辨率多光谱图像就能给出作物种植分布的预测结果，可以大幅减少地面调查人员的现场调查成本，还可以利用历史遥感信息实现往年作物种植分布的预测；另一方面，构建了融合遥感遥感数据时序特征和遥感图像局部特征，提高了预测结果的准确性。相比现有预测方法，本方法引入了决策点的上下文信息，针 对目标像素点不仅采集了多时相的图像数据，以其中目标像素点数据进行预测，而且通过设定的邻域大小采集目标像素点周围的像素数据，是面对点的预测，还结合了多图像通道的数据，这大大提高了预测结果的准确性。
图1为本发明一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法的流程图。
下面对照附图，通过对实施例的描述，对本发明具体实施方式作进一步详细的说明，以帮助本领域的技术人员对本发明的发明构思、技术方案有更完整、准确和深入的理解。
如图1所示，本发明提供了一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，该方法包括下列步骤：
步骤1：地面调查及训练样本建立；具体包括如下步骤：
步骤1.1：调研统计区域的主要种植作物类型、轮作方式、关键物候期时间节点；
步骤1.2：针对步骤1.1中不同作物类型、轮作方式，各选取部分代表性地块，详细记录地块位置及范围、种植面积、种植作物、种植时间等，得到训练样本。
步骤2：构造基于时序遥感数据和卷积神经网络的作物种植分布预测模型，所述卷积神经网络通过在多时相图像中目标像素点及其周围像素点的数据对其进行预测，输入值为多时相的高分辨率多光谱图像(即遥感图像，经过处理后进行输入)，输出值为作物类型、轮作方式的分类信息；具体包括如下步骤：
步骤2.1：收集三月到十一月每月中旬包含步骤1.2中选取的代表性地块的遥感图像，将不同时间的遥感图像进行精确配准，并将不同作物类型、轮作方式的区域进行逐像素的分类标记；图像处理过程中将不同时相遥感图像的各波段数据分别以多种颜色通道图像存储。
步骤2.2：将时序遥感数据作为模型的输入，作物类型、轮作方式的分类信息作为模型的输出，构建多层卷积神经网络。
在进行预测时根据领域大小、时相数和每时相图像通道数获取相应像素点的数据对目标像素点进行预测。例如采用常用的图像通道r(红色)+g(绿色)+b(蓝色)+nir(近红外光谱)，这样采集的遥感图像处理后依据时相形成对应各个图像通道的时序遥感数据，如果神经网络建立时设定领域为5，则在对目标像素点预测时会采集各个图像中目标像素点及其周围5×5的像素点数据。因此本方法所用的模型在预测时需要采集邻域大小*时相数*每时相图像通道数的像素数据，是面对点的预测，相比现有技术大大提高了预测结果的准确性。根据时相不同进行的预测，不仅能预测作物的种类，还能根据作物物候期的不同预测区域中作物是早稻还是晚稻，这是现有技术所不具备的。
步骤2.3：利用步骤2.1中的训练样本，使用反向传播算法对多层卷积神经网络中的参数进行充分训练。
步骤3：将统计区域的时序遥感数据输入已构建的模型获取识别结果；具体包括如下步骤：
步骤3.1：收集研究区域范围在三月到十一月每月中旬的遥感图像，将不同时间的遥感图像进行精确配准；
步骤3.2：将步骤3.1中的配准后的时序遥感数据输入到步骤2.3中训练的多层卷积神经网络模型中，获取研究区域的作物种植分类预测结果。
本发明只需要通过少量的代表性地块的地面调查，构建了融合遥感遥感数据时序特征和遥感图像局部特征的预测模型，引入了决策点的上下文信息，提高了预测结果的准确性。
上面结合附图对本发明进行了示例性描述，显然本发明具体实现并不受上述方式的限制，只要采用了本发明的发明构思和技术方案进行的各种非实质性的改进，或未经改进将本发明构思和技术方案直接应用于其它场合的，均在本发明保护范围之内。
Claims (6)
- 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，其特征在于：包括下列步骤：步骤1：地面调查及训练样本建立；步骤2：构造基于时序遥感数据和卷积神经网络的作物种植分布预测模型，所述卷积神经网络通过在多时相图像中目标像素点及其周围像素点的数据对其进行预测，输入值为多时相的高分辨率多光谱图像，输出值为作物类型、轮作方式的分类信息；步骤3：将统计区域的时序遥感数据输入已构建的模型获取识别结果。
- 根据权利要求1所述的一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，其特征在于：所述步骤1包括如下步骤：步骤1.1：调研统计区域的主要种植作物类型、轮作方式、关键物候期时间节点；步骤1.2：针对步骤1.1中不同作物类型、轮作方式，各选取部分代表性地块，详细记录地块位置及范围、种植面积、种植作物、种植时间等，得到训练样本。
- 根据权利要求2所述的一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，其特征在于：所述步骤2包括如下步骤：步骤2.1：收集不同时期包含步骤1.2中选取的代表性地块的遥感图像，将不同时间的遥感图像进行精确配准，并将不同作物类型、轮作方式的区域进行逐像素的分类标记；步骤2.2：将时序遥感数据作为模型的输入，作物类型、轮作方式的分类信息作为模型的输出，构建多层卷积神经网络；步骤2.3：利用步骤2.1中的训练样本，使用反向传播算法对多层卷积神经网络中的参数进行充分训练。
- 根据权利要求3所述的一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，其特征在于：所述步骤2中，图像处理过程中将不同时相遥感图像的各波段数据分别以多种颜色通道图像存储，在进行预测时根据领域大小、时相数和每时相图像通道数获取相应像素点的数据对目标像素点进行预测。
- 根据权利要求4所述的一种基于时序遥感数据和卷积神经网络的作物种 植分布预测方法，其特征在于：所述步骤3包括如下步骤：步骤3.1：收集研究区域范围的时间与神经网络训练时采集遥感图像时间相同的遥感图像，将不同时间的遥感图像进行精确配准；步骤3.2：将步骤3.1中的配准后的时序遥感数据输入到步骤2.3中训练的多层卷积神经网络模型中，获取研究区域的作物种植分类预测结果。
- 根据权利要求3-5中任一所述的一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法，其特征在于：神经网络训练时，采集遥感图像的时间为三月到十一月每月中旬。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2019/110509 WO2021068176A1 (zh) | 2019-10-11 | 2019-10-11 | 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 |
| CN201980016521.7A CN112840348B (zh) | 2019-10-11 | 2019-10-11 | 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2019/110509 WO2021068176A1 (zh) | 2019-10-11 | 2019-10-11 | 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2021068176A1 true WO2021068176A1 (zh) | 2021-04-15 |
Family
ID=75436924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2019/110509 WO2021068176A1 (zh) | 2019-10-11 | 2019-10-11 | 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112840348B (zh) |
| WO (1) | WO2021068176A1 (zh) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361952A (zh) * | 2021-06-25 | 2021-09-07 | 河北工程大学 | 一种芦苇生长情况评估方法 |
| CN113723254A (zh) * | 2021-08-23 | 2021-11-30 | 三明学院 | 一种毛竹林分布的识别方法、装置、设备和存储介质 |
| CN113792648A (zh) * | 2021-09-13 | 2021-12-14 | 杨凌龙翔数字科技有限公司 | 一种基于多时相的卫星遥感果园分类方法 |
| CN113920421A (zh) * | 2021-07-03 | 2022-01-11 | 桂林理工大学 | 一种快速分类的全卷积神经网络模型 |
| CN114092804A (zh) * | 2021-11-06 | 2022-02-25 | 支付宝(杭州)信息技术有限公司 | 遥感图像的识别方法和装置 |
| CN114792116A (zh) * | 2022-05-26 | 2022-07-26 | 中国科学院东北地理与农业生态研究所 | 一种时间序列深度卷积网络农作物遥感分类方法 |
| WO2022267693A1 (en) * | 2021-06-21 | 2022-12-29 | Ping An Technology (Shenzhen) Co., Ltd. | System and method for super-resolution image processing in remote sensing |
| CN116167630A (zh) * | 2023-04-19 | 2023-05-26 | 昆明理工大学 | 一种选择三七轮作适宜性地块的计算机自动识别方法 |
| CN116958814A (zh) * | 2023-07-24 | 2023-10-27 | 河北工程大学 | 一种湿地典型植物生长情况实时监测方法及系统 |
| CN117094430A (zh) * | 2023-07-19 | 2023-11-21 | 青海师范大学 | 一种农作物分布预测方法、系统、设备及介质 |
| CN117593661A (zh) * | 2023-12-06 | 2024-02-23 | 农业农村部南京农业机械化研究所 | 一种基于深度学习的跨区域作物适收期监测方法 |
| CN117789025A (zh) * | 2023-12-26 | 2024-03-29 | 北京工业大学 | 一种基于物候信息的分作物秸秆露天焚烧排放估算方法 |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115661617B (zh) * | 2022-12-28 | 2023-03-14 | 成都中轨轨道设备有限公司 | 一种面向遥感大数据的动态自适应分布式协同工作方法 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063660A (zh) * | 2018-08-09 | 2018-12-21 | 成都天地量子科技有限公司 | 一种基于多光谱卫星影像的作物识别方法 |
| CN109145885A (zh) * | 2018-10-11 | 2019-01-04 | 北京师范大学 | 一种大尺度农作物遥感分类方法及系统 |
| CN109409261A (zh) * | 2018-10-11 | 2019-03-01 | 北京师范大学 | 一种农作物分类方法及系统 |
| US10430691B1 (en) * | 2019-01-22 | 2019-10-01 | StradVision, Inc. | Learning method and learning device for object detector based on CNN, adaptable to customers' requirements such as key performance index, using target object merging network and target region estimating network, and testing method and testing device using the same to be used for multi-camera or surround view monitoring |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102668899B (zh) * | 2012-03-28 | 2015-02-25 | 北京师范大学 | 一种农作物种植模式识别方法 |
| US20170161560A1 (en) * | 2014-11-24 | 2017-06-08 | Prospera Technologies, Ltd. | System and method for harvest yield prediction |
| CN106845806A (zh) * | 2017-01-06 | 2017-06-13 | 中国科学院遥感与数字地球研究所 | 农田种植状态的遥感监测方法和系统 |
| CN107292343B (zh) * | 2017-06-23 | 2020-05-22 | 中南大学 | 一种基于六层卷积神经网络和光谱-空间信息联合的高光谱遥感图像分类方法 |
| CN107392130B (zh) * | 2017-07-13 | 2020-12-08 | 西安电子科技大学 | 基于阈值自适应和卷积神经网络的多光谱图像分类方法 |
| CN108064642A (zh) * | 2018-01-05 | 2018-05-25 | 安徽省农业科学院作物研究所 | 一种麦玉轮作中提高土壤质量和作物产量的方法 |
| CN108596248B (zh) * | 2018-04-23 | 2021-11-02 | 上海海洋大学 | 一种基于改进深度卷积神经网络的遥感影像分类方法 |
| CN108932521B (zh) * | 2018-04-26 | 2021-02-09 | 中国农业大学 | 一种基于深度学习的农作物分类方法及系统 |
| CN109035286A (zh) * | 2018-06-29 | 2018-12-18 | 北京农业信息技术研究中心 | 植物三维点云数据高通量获取方法及装置 |
| CN109360117A (zh) * | 2018-10-08 | 2019-02-19 | 西充恒河农牧业开发有限公司 | 一种农作物种植模式识别方法 |
| CN109658679B (zh) * | 2018-12-24 | 2020-06-12 | 安徽中科智能感知产业技术研究院有限责任公司 | 一种用于田间作物生长数据信息采集无人小型车辆的多模式数据采集装置 |
| CN109948520B (zh) * | 2019-03-18 | 2023-01-06 | 中南大学 | 一种基于多时相双极化sar特征曲线的农作物分类方法 |
| CN110287869B (zh) * | 2019-06-25 | 2022-03-18 | 吉林大学 | 基于深度学习的高分辨率遥感影像农作物分类方法 |
-
2019
- 2019-10-11 CN CN201980016521.7A patent/CN112840348B/zh active Active
- 2019-10-11 WO PCT/CN2019/110509 patent/WO2021068176A1/zh active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063660A (zh) * | 2018-08-09 | 2018-12-21 | 成都天地量子科技有限公司 | 一种基于多光谱卫星影像的作物识别方法 |
| CN109145885A (zh) * | 2018-10-11 | 2019-01-04 | 北京师范大学 | 一种大尺度农作物遥感分类方法及系统 |
| CN109409261A (zh) * | 2018-10-11 | 2019-03-01 | 北京师范大学 | 一种农作物分类方法及系统 |
| US10430691B1 (en) * | 2019-01-22 | 2019-10-01 | StradVision, Inc. | Learning method and learning device for object detector based on CNN, adaptable to customers' requirements such as key performance index, using target object merging network and target region estimating network, and testing method and testing device using the same to be used for multi-camera or surround view monitoring |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022267693A1 (en) * | 2021-06-21 | 2022-12-29 | Ping An Technology (Shenzhen) Co., Ltd. | System and method for super-resolution image processing in remote sensing |
| CN113361952B (zh) * | 2021-06-25 | 2024-05-14 | 河北工程大学 | 一种芦苇生长情况评估方法 |
| CN113361952A (zh) * | 2021-06-25 | 2021-09-07 | 河北工程大学 | 一种芦苇生长情况评估方法 |
| CN113920421A (zh) * | 2021-07-03 | 2022-01-11 | 桂林理工大学 | 一种快速分类的全卷积神经网络模型 |
| CN113723254A (zh) * | 2021-08-23 | 2021-11-30 | 三明学院 | 一种毛竹林分布的识别方法、装置、设备和存储介质 |
| CN113792648A (zh) * | 2021-09-13 | 2021-12-14 | 杨凌龙翔数字科技有限公司 | 一种基于多时相的卫星遥感果园分类方法 |
| CN114092804A (zh) * | 2021-11-06 | 2022-02-25 | 支付宝(杭州)信息技术有限公司 | 遥感图像的识别方法和装置 |
| CN114792116B (zh) * | 2022-05-26 | 2024-05-03 | 中国科学院东北地理与农业生态研究所 | 一种时间序列深度卷积网络农作物遥感分类方法 |
| CN114792116A (zh) * | 2022-05-26 | 2022-07-26 | 中国科学院东北地理与农业生态研究所 | 一种时间序列深度卷积网络农作物遥感分类方法 |
| CN116167630B (zh) * | 2023-04-19 | 2023-06-27 | 昆明理工大学 | 一种选择三七轮作适宜性地块的计算机自动识别方法 |
| CN116167630A (zh) * | 2023-04-19 | 2023-05-26 | 昆明理工大学 | 一种选择三七轮作适宜性地块的计算机自动识别方法 |
| CN117094430A (zh) * | 2023-07-19 | 2023-11-21 | 青海师范大学 | 一种农作物分布预测方法、系统、设备及介质 |
| CN117094430B (zh) * | 2023-07-19 | 2024-04-26 | 青海师范大学 | 一种农作物分布预测方法、系统、设备及介质 |
| CN116958814A (zh) * | 2023-07-24 | 2023-10-27 | 河北工程大学 | 一种湿地典型植物生长情况实时监测方法及系统 |
| CN117593661A (zh) * | 2023-12-06 | 2024-02-23 | 农业农村部南京农业机械化研究所 | 一种基于深度学习的跨区域作物适收期监测方法 |
| CN117789025A (zh) * | 2023-12-26 | 2024-03-29 | 北京工业大学 | 一种基于物候信息的分作物秸秆露天焚烧排放估算方法 |
| CN117789025B (zh) * | 2023-12-26 | 2024-06-04 | 北京工业大学 | 一种基于物候信息的分作物秸秆露天焚烧排放估算方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112840348A (zh) | 2021-05-25 |
| CN112840348B (zh) | 2024-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021068176A1 (zh) | 一种基于时序遥感数据和卷积神经网络的作物种植分布预测方法 | |
| Shafi et al. | A multi-modal approach for crop health mapping using low altitude remote sensing, internet of things (IoT) and machine learning | |
| Khanal et al. | Integration of high resolution remotely sensed data and machine learning techniques for spatial prediction of soil properties and corn yield | |
| US11751499B2 (en) | Computer-implemented methods, computer readable medium and systems for a precision agriculture platform that detects disease in crops | |
| US10838936B2 (en) | Computer-implemented methods, computer readable medium and systems for generating an orchard data model for a precision agriculture platform | |
| US10748081B2 (en) | Computer-implemented methods, computer readable medium and systems for a precision agriculture platform that identifies generic anomalies in crops | |
| WO2020044480A1 (ja) | 作物生育ステージ判定システムのサーバ装置、生育ステージ判定方法及びプログラム | |
| Smith et al. | Scaling up high-throughput phenotyping for abiotic stress selection in the field | |
| US10395355B2 (en) | Computer-implemented methods, computer readable medium and systems for a precision agriculture platform | |
| US10410334B2 (en) | Computer-implemented methods, computer readable medium and systems for a precision agriculture platform with a satellite data model | |
| CN115481368B (zh) | 一种基于全遥感机器学习的植被覆盖度估算方法 | |
| Tanabe et al. | Winter wheat yield prediction using convolutional neural networks and UAV-based multispectral imagery | |
| Liu et al. | Estimating potato above-ground biomass by using integrated unmanned aerial system-based optical, structural, and textural canopy measurements | |
| US10943173B2 (en) | Computer-implemented methods, computer readable medium and systems for generating a satellite data model for a precision agriculture platform | |
| Hossen et al. | Total nitrogen estimation in agricultural soils via aerial multispectral imaging and LIBS | |
| Roth et al. | Repeated multiview imaging for estimating seedling tiller counts of wheat genotypes using drones | |
| Mirkouei | A cyber-physical analyzer system for precision agriculture | |
| Lu et al. | An assessment of multi-view spectral information from UAV-based color-infrared images for improved estimation of nitrogen nutrition status in winter wheat | |
| Li et al. | Estimation of nitrogen content in wheat using indices derived from RGB and thermal infrared imaging | |
| Sharma et al. | Technological advancements in automated crop pest and disease detection: A review & ongoing research | |
| Abuleil et al. | An integrated system for mapping red clover ground cover using unmanned aerial vehicles: A Case Study in Precision Agriculture | |
| Hassani et al. | Application of UAS-based remote sensing in estimating winter wheat phenotypic traits and yield during the growing season | |
| CA3232760A1 (en) | Methods and systems for use in processing images related to crops | |
| Das et al. | A high-throughput phenotyping pipeline for rapid evaluation of morphological and physiological crop traits across large fields | |
| Zang et al. | Estimation of Density and Height of Winter Wheat Varieties Using Unmanned Aerial Vehicles Images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 19948283 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 19948283 Country of ref document: EP Kind code of ref document: A1 |