A method and device for rapidly generating crop growth information

Abstract

The present application provides a kind of crop growth information fast generation method and device, utilize satellite remote sensing inversion to carry out the regular growth situation monitoring of crop, support rice, corn, rape, citrus, tea and other crop field growth sustained monitoring and growth stage growth state monitoring.First, to the obtained multi-temporal optical image, time series optical image reconstruction and data space-time fusion are carried out, and high space-time remote sensing data is generated.Then, from the high space-time remote sensing image, the remote sensing parameters related to crop growth are extracted, multiple parameters such as normalized vegetation index and normalized difference yellow index are calculated, and high space-time crop growth grid results are obtained.Under the spatial constraint of crop vector plot, the elements are converted to grid, as mask data, the mask extraction resampled high space-time crop growth grid result is obtained, the spectral characteristics of the corresponding unit crop are obtained, the band set statistical calculation is carried out to obtain the mean value of the growth index of a certain crop, and the comprehensive evaluation result of plot-level crop growth is obtained.

Description

Technical Field

[0001] This invention belongs to the field of agricultural remote sensing applications and geographic information technology, and particularly relates to a method and apparatus for rapidly generating crop growth information. Background Technology

[0002] In today's era of precision agriculture, real-time and detailed monitoring of crop growth has become a common requirement across various stages, including insurance claims and field management. Traditional ground surveys are labor-intensive and time-consuming, and struggle to cover large areas. While remote sensing can provide a "spotlight," it has long been limited by two major bottlenecks: first, the lack of a reliable historical sample database, resulting in a lack of reference points for determining whether this year's crop growth is good or bad; and second, the lack of spatial units that precisely match field boundaries, meaning that monitoring results can only remain at the level of a "grid," unable to directly correspond to the "specific plot of land" actually cultivated by farmers.

[0003] Vegetation remote sensing theory provides the key to solving the first bottleneck. The physical characteristics of crop leaves—strong absorption of red light and strong reflection of near-infrared light—lay the foundation for the development of various vegetation indices. Ratio vegetation indices highlight lush vegetation, while the Normalized Difference Vegetation Index (NDVI) is more sensitive in the early and middle growth stages. Adding blue light information can reduce interference from atmospheric aerosols, resulting in more robust atmospheric-resistant vegetation indices. These indices amplify synchronously with biomass and simultaneously reduce external noise such as solar altitude angle, topographic shadows, and atmospheric conditions, making them a natural "yardstick" for characterizing crop growth.

[0004] However, a ruler alone is not enough. For example, existing publicly available research often relies on medium- to low-resolution imagery such as MODIS, classifying crops using Fourier component similarity indices, and then evaluating growth using an anomaly model or grading model based on "current year's value vs. last year's value." Figure 1 As shown. This approach works when the sample is continuous and the land use type is constant, but it overlooks two practical problems:

[0005] The frequent occurrence of clouds and rain in time and space results in incomplete optical images, especially in rainy areas such as the southwest.

[0006] Land use changes such as crop rotation, crop rotation, and even "agricultural to forest" conversion mean that the same pixel may represent completely different crops in different years, or even no longer be arable land.

[0007] As a result, while large-scale grid products may appear to provide "full coverage," they are actually far from what farmers truly care about: "how much better is this paddy field this year compared to last year?" The error is often so large that it cannot support subsequent management actions such as precision fertilization and differentiated insurance.

[0008] In short, the industry urgently needs a technological approach that can quickly generate "plot-level" crop growth information under multiple constraints such as the lack of historical samples, dynamic changes in land use, and unfavorable weather conditions. This is the core pain point that remote sensing agriculture applications urgently need to overcome. Summary of the Invention

[0009] This invention solves the problem of remote sensing monitoring of crop growth at the fine plot scale and the lack of historical crop growth sample data, and enables the rapid generation of crop growth information.

[0010] The specific technical solution is as follows:

[0011] A method for rapidly generating crop growth information includes the following steps:

[0012] Step 1: Use multi-temporal optical remote sensing images to perform time series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images;

[0013] Step 2: Use the reconstructed sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data;

[0014] Step 3: Extract crop growth remote sensing parameters from high spatiotemporal remote sensing data, and spatially overlay the crop growth remote sensing parameters onto the crop planting structure map. Analyze the differences in crop phenology to generate high spatiotemporal crop growth raster results.

[0015] Step 4: Extract the target crop vector from the crop vector plot and convert the features to raster to generate a mask. After resampling the crop growth raster results from Step 3, extract the growth index within the plot using the mask. Calculate the average crop growth index for each plot through regional statistics. The average crop growth index is used as a historical benchmark. Calculate the growth rate of each pixel during the crop's growth period in the current year and generate a growth grading raster according to a set threshold. Assign the average value of the growth grading raster to the corresponding crop vector plot to obtain the comprehensive crop growth assessment results at the plot level.

[0016] A device for rapidly generating crop growth information includes:

[0017] The sequence reconstruction module uses multi-temporal optical remote sensing images to perform time-series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images;

[0018] The data acquisition module uses the reconstruction sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data;

[0019] The crop growth raster result generation module extracts crop growth remote sensing parameters from high spatiotemporal remote sensing data, and spatially overlays the crop growth remote sensing parameters onto the crop planting structure map, generating high spatiotemporal crop growth raster results based on crop phenological differences analysis.

[0020] The comprehensive crop growth assessment module extracts the target crop vector from the crop vector plot and converts the elements to raster to generate a mask. After resampling the crop growth raster results, the growth index within the plot is extracted using the mask. The mean crop growth index of each plot is calculated through regional statistics. The mean crop growth index is used as a historical benchmark to calculate the growth rate of each pixel during the crop's growth period in the current year, and the growth is generated into a growth classification raster according to a set threshold. The growth classification raster is then assigned to the corresponding crop vector plots in a regional statistical manner using the average value method to obtain the comprehensive crop growth assessment results at the plot level.

[0021] An electronic device includes: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method.

[0022] A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to implement the method described thereon.

[0023] The present invention has the following beneficial effects:

[0024] In situations where it is impossible to obtain the average value of crop growth over several consecutive years, a method was invented that uses vector data of a crop plot as a mask to obtain the NDVI growth value within the mask, calculates the average NDVI value within the mask range, and quickly obtains the average crop growth value.

[0025] Existing technologies use anomaly models and extreme value models to compare the growth raster results generated by the same cell, without considering the problem that changes in land use type of the same cell will lead to different growth information of different crops. This invention compares a cell of a certain crop in a certain year with the historical average of a certain crop, thus avoiding the problem of inaccurate growth information caused by different crops in different years due to changes in crop type of the same cell.

[0026] Existing technologies produce crop growth products as raster data, which cannot meet the requirements of precision agricultural production management. This invention proposes to obtain crop growth information based on the scale of crop plots, thus addressing the requirements of precision agricultural production management. Attached Figure Description

[0027] Figure 1 Flowchart of traditional crop growth monitoring methods;

[0028] Figure 2 This is a flowchart of the method for rapidly generating crop growth information according to the present invention;

[0029] Figure 3 This is a flowchart of time-series optical image reconstruction technology.

[0030] Figure 4 Flowchart of high spatiotemporal fusion and reconstruction method;

[0031] Figure 5 Flowchart of the technology for calculating the average remote sensing index of crops;

[0032] Figure 6 A crop growth grading chart;

[0033] Figure 7 To reconstruct the forward-backward LSTM network graph for missing data. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other. To achieve the above objectives, this invention adopts the following technical solution.

[0035] This invention proposes a rapid method for generating crop growth information. It utilizes satellite remote sensing inversion (Normalized Difference Vegetation Index, NDVI) technology to conduct regular monitoring of crop growth, supporting continuous field monitoring of crops such as rice, corn, rapeseed, citrus, and tea, as well as monitoring of growth status at key growth stages. First, time-series optical image reconstruction and spatiotemporal fusion are performed on the acquired multi-temporal optical images to generate high-spatial-time remote sensing data. Then, remote sensing parameters related to crop growth are extracted from the high-spatial-time remote sensing images, and multiple parameters such as the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Yellow Index (NDYI) are calculated to obtain high-spatial-time crop growth raster results. Under the spatial constraints of crop vector plots, the elements are converted to raster data and used as mask data. The mask data is then used to extract the resampled high-spatial-time crop growth raster results, obtaining the spectral characteristics of the corresponding crop units. Band set statistical calculations are performed to obtain the mean value of a certain crop growth index. Based on the meaning and method of the NDVI index, crop growth monitoring statistics are conducted to construct a fine-scale NDVI index. Obtain comprehensive assessment results of crop growth at the plot level. For example... Figure 2 As shown, the method includes the following steps:

[0036] Step 1: Use multi-temporal optical remote sensing images to perform time series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images.

[0037] First, the optical remote sensing image sequence of the study area underwent preprocessing including geometric registration, pixel alignment, and cropping of common (overlapping) regions; second, such as Figure 3As shown, cloud / shadow (mask) regions are visually drawn on each image, and the pixel values ​​of the masked regions are set to mask values ​​(these values ​​should be outside the range of pixel values, such as 9999), while keeping the pixel values ​​of non-masked (non-cloud / shadow covered) regions unchanged. A time series curve containing mask values ​​is constructed for each pixel. Next, an autoencoder network is constructed using a temporal recurrent neural network (LSTM) as the encoding core to perform unsupervised clustering on the pixel time series curves, generating multiple clusters (similar pixel regions). Then, with the reconstruction of a specific temporal image of a certain cluster region as the goal, a forward and backward recurrent neural network prediction model is constructed, and the pixel time series of the region is divided into a training sample set and a reconstruction (prediction) sample set according to whether the specific temporal value is a mask value. Then, the prediction model is trained using the training sample set, and the trained prediction model is used to predict and repair the missing temporal values ​​of the prediction sample set. Finally, the prediction data is merged with the cloudless / image regions to generate a reconstruction sequence of multi-temporal remote sensing images. The time series data output in this step will provide time series data support for functions such as "data spatiotemporal fusion" and "extraction of observation parameters, time series parameters and indices".

[0038] An autoencoder network is constructed using a Long Short-Term Memory (LSTM) recurrent neural network as the encoding core to process time series data. The LSTM-AE model consists of an encoder and a decoder. The encoder converts the input time series data into a fixed-length vector, and the decoder restores the vector back to the original time series data.

[0039] The structure of the forward and backward recurrent neural network prediction model is as follows: Figure 7 As shown, first, a pixel-by-pixel time series TS is constructed, which is then split into a forward time series fTS and a backward time series bTS. Next, in the forward part, the forward time series fTS is sequentially input into a mask layer, followed by a stacked LSTM network and a fully connected layer to generate forward time series features ff; following a similar network structure, backward time series features bf are generated. Finally, the forward time series features ff and backward time series features bf are concatenated and input into a fully connected layer to predict missing values.

[0040] Step 2: Use the reconstructed sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data.

[0041] To address the characteristics of multi-source heterogeneous data, such as different imaging mechanisms, diverse data styles, multiple scales, temporal sparsity, missing temporal data, and inconsistent data quality, a multi-level spatiotemporal fusion framework for multi-source heterogeneous remote sensing data, guided by prior knowledge, is constructed to explore the statistical characteristics of data imaging mechanisms, scale differences, and morphological differences. A point-line-area integrated remote sensing fusion and reconstruction method is used to generate a high-spatial-weighted remote sensing dataset, providing data support for subsequent calculations of crop growth monitoring parameters. For example... Figure 4 As shown, firstly, the low-resolution high-temporal remote sensing image sequence and the high-resolution low-temporal remote sensing image sequence are paired in nearest-neighbor time, and the low-resolution high-temporal image is reprojected and sampled to the same resolution as the high-resolution low-temporal image, based on the high-spatial image. Then, based on the low-resolution high-temporal remote sensing image at a certain time, the high-resolution remote sensing image at the same time is predicted, using the spatiotemporal adaptive reflectance fusion model STARFM proposed by Gao et al. (Gao et al., 2006). This algorithm fuses the high-temporal low-temporal image and the low-resolution high-temporal image with rich temporal information, considering not only the temporal differences but also using the values ​​of neighboring spectrally similar pixels to calculate the center pixel value, generating a high-temporal remote sensing image dataset.

[0042] Step 3: Extract remote sensing parameters of crop growth using high spatiotemporal remote sensing data.

[0043] First, remote sensing parameters related to crop growth (such as NDVI) are extracted from high-spatial-temporal remote sensing images. The calculation method for NDVI is obtained through the formula: Wherein, NIR represents the reflectance in the near-infrared band, and R represents the reflectance in the red band. Secondly, the high spatiotemporal remote sensing parameters are spatially overlaid onto the crop planting structure map, and the causes of differences in crop growth are analyzed based on crop phenology, generating high spatiotemporal crop growth raster results.

[0044] Step 4: Calculate the mean crop growth index using crop plot-level vector data.

[0045] Because crop growth monitoring lacks historical crop growth sample point information, it is impossible to derive the average crop growth index. Furthermore, for the same plot of land, there may be inconsistencies in crop types planted in different years or changes in land use, such as agricultural land being converted to forest land. Therefore, using the NDVI of the same plot this year and the same period last year to calculate crop growth will introduce significant errors. The following proposes a rapid method for generating the average crop growth index. The main steps are as follows: First, extract vector data of a specific crop from the crop vector plot product. Using the feature-to-raster method, convert the crop plot vector data into raster data to generate raster data containing the extracted crop growth data. Second, for the high spatiotemporal crop growth raster results obtained in step 3, including Landsat 8 / 9 data at 15 to 30 meters, meter-level GF2 data, and 10-meter Sentinel-2 data, spatial transformation is used to uniformly resample them to 10-meter resolution NDVI data. This process includes crop raster data and growth index raster data. Finally, the raster data is used as a mask for the crop index data. The mask extracts the growth index within the plot area, while values ​​outside the mask area are set to 0 or NULL. A zonal statistical method is used to calculate the crop growth index within each mask area. Then, the growth index areas of each plot are statistically summed, and finally, the mean growth index is calculated. The formula used is: ,in, The crop growth index is defined for each crop plot, where n is the total number of crop plots. This represents the average crop growth index. Its technical flowchart is as follows: Figure 5 As shown.

[0046] Step 5: Crop growth monitoring and statistics.

[0047] To monitor the growth of a specific crop in a given year, first obtain the crop's growth stamina index data for that year in the region. Then, use step 3 to calculate the historical average growth stamina index of that crop, replacing the previous years' growth stamina index for the same plot. For example... Figure 6 As shown, the growth growth rate (GGPV) is calculated by measuring the crop's growth growth rate during its growth period. The calculation method is as follows:

[0048] GGPV= ;

[0049] in, This represents the current crop growth index value, where y represents the year and m represents the crop growth stage. This represents the historical average growth of a crop, and the raster result of crop growth classification is calculated.

[0050] The growth results are divided into 4 levels:

[0051] 1) -1≤GGPV<-0.05, which means it is worse than in previous years;

[0052] 2) -0.05≤GGPV≤0.05, which means it is the same as in previous years;

[0053] 3) 0.05 < GGPV ≤ 1, which is better than in previous years;

[0054] 4) GGPV < -1 or GGPV > 1, i.e., incorrect data.

[0055] Step 6: Result Processing. Using the crop growth grading raster data obtained in Step 4, this growth grading data needs to be assigned to the crop plot vector data. A zonal statistical analysis method is used, which includes five methods: mean, mode, median, maximum, and minimum. Considering all factors, the mean is used to calculate the mean of the growth grading raster data within the crop plot area, and the zonal statistical data is then assigned to the corresponding crop vector plot. After assignment, the growth grading attribute values ​​need to be checked, and erroneous data needs to be removed.

[0056] The present invention has the following alternatives:

[0057] In step 3, when extracting the remote sensing parameter NDVI related to crop growth from high spatiotemporal remote sensing images, the Enhanced Vegetation Index (EVI) can also be used instead.

[0058] The crop growth grading standard in step 4 can also use different standards, such as a 5-level grading standard:

[0059] 1) When GGPV < -0.1, the growth is poor;

[0060] 2) When -0.1 ≤ GGPV < -0.05, growth is deviated;

[0061] 3) When -0.05 ≤ GGPV < 0.05, the growth remains the same;

[0062] 4) When 0.05 ≤ GGPV < 0.1, the plant exhibits a growth preference;

[0063] 5) When GGPV≥0.1, the growth is better.

[0064] Another aspect of the present invention provides a device for rapidly generating crop growth information, comprising:

[0065] The sequence reconstruction module uses multi-temporal optical remote sensing images to perform time-series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images;

[0066] The data acquisition module uses the reconstruction sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data;

[0067] The crop growth raster result generation module extracts crop growth remote sensing parameters from high spatiotemporal remote sensing data, and spatially overlays the crop growth remote sensing parameters onto the crop planting structure map, generating high spatiotemporal crop growth raster results based on crop phenological differences analysis.

[0068] The comprehensive crop growth assessment module extracts the target crop vector from the crop vector plot and converts the elements to raster to generate a mask. After resampling the crop growth raster results, the growth index within the plot is extracted using the mask. The mean crop growth index of each plot is calculated through regional statistics. The mean crop growth index is used as a historical benchmark to calculate the growth rate of each pixel during the crop's growth period in the current year, and the growth is generated into a growth classification raster according to a set threshold. The growth classification raster is then assigned to the corresponding crop vector plots in a regional statistical manner using the average value method to obtain the comprehensive crop growth assessment results at the plot level.

[0069] Another aspect of the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method.

[0070] Another aspect of the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to implement the method described thereon.

[0071] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0072] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0073] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0074] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0075] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0076] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for rapidly generating crop growth information, characterized in that, Includes the following steps: Step 1: Use multi-temporal optical remote sensing images to perform time series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images; Step 2: Use the reconstructed sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data; Step 3: Extract crop growth remote sensing parameters from high spatiotemporal remote sensing data, and spatially overlay the crop growth remote sensing parameters onto the crop planting structure map. Analyze the differences in crop phenology to generate high spatiotemporal crop growth raster results. Step 4: Extract the target crop vector from the crop vector plot and convert the features to raster to generate a mask. After resampling the crop growth raster results from Step 3, extract the growth index within the plot using the mask. Calculate the average crop growth index for each plot through regional statistics. The average crop growth index is used as a historical benchmark. Calculate the growth rate of each pixel during the current crop's growth period and generate a growth grading raster based on a set threshold. Assign the average value of the growth grading raster to the corresponding crop vector plot to obtain the comprehensive crop growth assessment results at the plot level. In step 3, the remote sensing parameter NDVI related to crop growth is extracted from the high spatiotemporal remote sensing image. The calculation method of NDVI is obtained by the formula: Where NIR represents the reflectivity of the near-infrared band and R represents the reflectivity of the red band; The average crop growth index is calculated using the following formula: ,in, The crop growth index is defined for each crop plot, where n is the total number of crop plots. This represents the mean of the crop growth index; Growth rate grading pixel value The calculation method is as follows: ； Where y represents the year and m represents the crop growth period.

2. The method for rapidly generating crop growth information according to claim 1, characterized in that, Step 1 includes: Preprocessing of optical remote sensing images includes geometric registration, pixel alignment, and common area cropping; Visually draw cloud / shadow mask areas on each image, set the pixel values ​​of the mask areas to the mask values, and keep the pixel values ​​of non-cloud / shadow mask areas unchanged, and construct a time series curve containing mask values ​​for each pixel; An autoencoder network is constructed to perform unsupervised clustering of pixel time series curves, generating multiple cluster regions. Then, focusing on the reconstruction of temporal images of a specific cluster region, a forward and backward recurrent neural network prediction model is constructed. The pixel time series of this region is divided into a training sample set and a reconstruction sample set based on whether the temporal phase is a mask value. The prediction model is then trained using the training sample set, and the trained model is used to predict and repair missing temporal values ​​in the prediction sample set. Finally, the predicted data is merged with cloudless / image regions to generate a reconstructed sequence of multi-temporal remote sensing images.

3. The method for rapidly generating crop growth information according to claim 1, characterized in that, Step 2 specifically includes: First, performing nearest-neighbor time pairing on the low-resolution high-temporal remote sensing image sequence and the high-resolution low-temporal remote sensing image sequence, and using the high-spatial image as the reference, reprojecting and sampling the low-resolution high-temporal image to the same resolution as the high-resolution low-temporal image. Then, based on the low-resolution high-temporal remote sensing image at a certain time, predicting the high-resolution remote sensing image at the same time.

4. The method for rapidly generating crop growth information according to claim 1, characterized in that, The growth results are divided into 4 levels: 1) -1≤GGPV<-0.05, which means it is worse than in previous years; 2) -0.05≤GGPV≤0.05, which means it is the same as in previous years; 3) 0.05 < GGPV ≤ 1, which is better than in previous years; 4) GGPV < -1 or GGPV > 1, i.e., incorrect data.

5. A device for rapidly generating crop growth information, implementing the method described in any one of claims 1-4, characterized in that, include: The sequence reconstruction module uses multi-temporal optical remote sensing images to perform time-series reconstruction to obtain the reconstructed sequence of multi-temporal remote sensing images; The data acquisition module uses the reconstruction sequence of multi-temporal remote sensing images to perform spatiotemporal fusion of data to obtain high spatiotemporal remote sensing data; The crop growth raster result generation module extracts crop growth remote sensing parameters from high spatiotemporal remote sensing data, and spatially overlays the crop growth remote sensing parameters onto the crop planting structure map, generating high spatiotemporal crop growth raster results based on crop phenological differences analysis. The comprehensive crop growth assessment module extracts the target crop vector from the crop vector plot and converts the elements to raster to generate a mask. After resampling the crop growth raster results, the growth index within the plot is extracted using the mask. The mean crop growth index of each plot is calculated through regional statistics. The mean crop growth index is used as a historical benchmark to calculate the growth rate of each pixel during the crop's growth period in the current year, and the growth is generated into a growth classification raster according to a set threshold. The growth classification raster is then assigned to the corresponding crop vector plots in a regional statistical manner using the average value method to obtain the comprehensive crop growth assessment results at the plot level.

6. An electronic device, characterized in that, include: One or more processors; A memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method of any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, cause the processor to implement the method described in any one of claims 1 to 4.

Images