Defoliating agent application time inversion method, system, equipment and medium

Abstract

The invention discloses a defoliant application time inversion method, system and device and a medium, and relates to the technical field of remote sensing observation. The method includes: acquiring a satellite remote sensing digital image of a to-be-measured cotton field; determining a vegetation index of each time point in the satellite remote sensing digital image of the cotton field to be detected, and summarizing the vegetation indexes to obtain a time sequence Sentinel-NDVI of the vegetation indexes; concave points in the Sentinel-NDVI of the time sequence serve as abnormal points, the NDVI average value of the adjacent time points is used for replacing the NDVI value of the abnormal points for preliminary denoising, and a denoised time sequence is obtained; and filtering the de-noising time sequence to generate a growth curve representing the growth trend and rule of the cotton, taking a local change rate maximum point of the growth curve as a time to-be-determined point, and taking a time point of the time to-be-determined point before a set period as the application time of the defoliating agent. According to the invention, the growth conditions of crops in different growth stages can be more comprehensively known.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing observation, and in particular to a defoliant application time inversion method, system, equipment and medium. Background Art

[0002] Defoliation of cotton fields is an important prerequisite for mechanical harvesting. Reasonable selection of defoliation time can coordinate the growth and development process of cotton and reduce the difficulty and cost of cotton field management. If the defoliant is used too early, the cotton will not mature enough, which will lead to a decline in cotton quality. On the contrary, if the defoliant is used too late and the temperature drops, it will affect the defoliation effect and is also not conducive to improving the quality of cotton. Therefore, reasonable selection of defoliation time can ensure that cotton is defoliated at an appropriate stage, thereby improving the quality of cotton.

[0003] In the prior art, the determination of the application time of defoliants mainly relies on the traditional site-based visual observation method, and its implementation process includes: first, it is necessary to select representative farmland sites to ensure that the observation environment is relatively stable and can reflect the regional climate characteristics. Then, the key phenological periods of cotton (such as sowing, emergence, flowering, maturity, etc.) are selected. Observers go to the sites regularly and directly observe the growth status and phenological phenomena of cotton with the naked eye according to the predetermined time and method, and record the weather conditions, soil conditions, growth characteristics of cotton and the appearance time of its phenological phase in detail. When the fruit branches at the bottom of the cotton plants begin to open the cotton bolls / cracking, and the cotton boll shells are hard, and 30% of the cotton plants in the field begin to open the cotton bolls naturally, it can be regarded as a suitable time window for spraying defoliants. The observed data is sorted and analyzed to obtain the application time of the defoliant.

[0004] The defects of the above-mentioned prior art are: the visual observation method is limited by the visual perception ability and judgment experience of the observer, and it is often difficult to accurately capture the subtle changes in cotton growth. For the observation of indicators such as cotton growth period, density, and seedling growth, it is unable to provide sufficiently accurate data support, resulting in inaccurate application time of the defoliant. Summary of the invention

[0005] Based on this, it is necessary to provide a defoliant application time inversion method, system, equipment and medium to address the above technical problems.

[0006] An embodiment of the present invention provides a defoliant application time inversion method, comprising: Obtain satellite remote sensing digital images of the cotton field to be tested; Determine the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested, and summarize the vegetation index at each time point to obtain the time series of vegetation index Sentinel-NDVI; The concave points in the time series Sentinel-NDVI are regarded as outliers, and the NDVI value of the outliers is replaced by the average NDVI value of the adjacent time points for preliminary denoising to obtain the denoised time series; The denoised time series is filtered to generate a growth curve that represents the growth trend and law of cotton. The point with the maximum local change rate of the growth curve is taken as the time-to-be-determined point, and the time point before a set time period of the time-to-be-determined point is taken as the application time of the defoliant.

[0007] Optionally, after obtaining the satellite remote sensing digital image of the cotton field to be measured, the satellite remote sensing digital image is pre-processed by correction, cloud removal and screening.

[0008] Optionally, determining the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested specifically includes: Define the study area by specifying coordinates, set the year for analysis, and define the time range; Load satellite remote sensing digital image collection, and use cloud mask function to remove the influence of clouds and cloud shadows in satellite remote sensing digital images; According to the reflectivity R of the red band and the reflectivity NIR of the near infrared band, the vegetation index at each time point in the satellite remote sensing digital image is calculated using the formula NDVI = (NIR-R) / (NIR+R).

[0009] The embodiment of the present invention also provides a defoliant application time reversal system, comprising: A data acquisition module is used to acquire satellite remote sensing digital images of the cotton field to be tested; The sequence construction module is used to determine the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested, and summarize the vegetation index at each time point to obtain the time series of vegetation index Sentinel-NDVI; The denoising module is used to treat the concave points in the time series Sentinel-NDVI as outliers, and replace the NDVI value of the outlier with the average NDVI value of the adjacent time points for preliminary denoising to obtain the denoised time series; The inversion module is used to filter the denoised time series to generate a growth curve that characterizes the growth trend and law of cotton, and to take the maximum point of the local change rate of the growth curve as the time to be determined point, and the time point before the set period of the time to be determined point as the application time of the defoliant.

[0010] An embodiment of the present invention further provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the defoliant application time inversion method are implemented.

[0011] An embodiment of the present invention further provides a storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the defoliant application time inversion method are implemented.

[0012] Compared with the prior art, the above-mentioned defoliant application time inversion method, system, device and medium provided by the embodiment of the present invention have the following beneficial effects: The present invention uses satellite remote sensing technology to obtain digital images of cotton fields to be tested, calculates the time series Sentinel-NDVI and performs preliminary denoising and filtering to generate a growth curve that can characterize the growth trend and law of cotton. By identifying the maximum point of the local change rate of the growth curve, the time point to be determined is determined, and the time point before a set time period of the time point to be determined is used as the application time of the defoliant. Compared with the traditional visual observation method, the inversion of the defoliant application time using remote sensing data not only has a wide coverage, high spatial resolution, and complete time series, but also can more comprehensively understand the growth of cotton at different growth stages, and can provide more accurate and reliable information support for agricultural production and management, and the obtained defoliant application time is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The figure is a schematic flow chart of a time reversal method for defoliant application provided in one embodiment. DETAILED DESCRIPTION

[0014] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0015] In one embodiment, a defoliant application time reversal method is provided, the method comprising: Obtain satellite remote sensing digital images of the cotton field to be tested.

[0016] The vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested is determined, and the vegetation index at each time point is summarized to obtain the time series of vegetation index Sentinel-NDVI.

[0017] The concave points in the time series Sentinel-NDVI are regarded as outliers, and the NDVI value of the outliers is replaced by the average NDVI value of adjacent time points for preliminary denoising to obtain the denoised time series.

[0018] The denoised time series is filtered to generate a growth curve that represents the growth trend and law of cotton. The point with the maximum local change rate of the growth curve is taken as the time-to-be-determined point, and the time point before a set time period of the time-to-be-determined point is taken as the application time of the defoliant.

[0019] Exemplarily, first, determine the cotton planting area, extract the cotton planting coordinate information based on Arcmap, and give priority to areas where cotton is connected.

[0020] 2. Obtain satellite remote sensing digital images of cotton fields, perform preprocessing and calculate the time series of vegetation index.

[0021] Select a suitable satellite data source. The present invention adopts the Sentinel series to ensure that the data source covers the study area and the required time range. Preprocess the acquired remote sensing digital image, and use the cloud mask algorithm to remove clouds and shadow pixels to ensure the accuracy and reliability of subsequent analysis. Select the vegetation index (the vegetation index used in the present invention is NDVI), and calculate the vegetation index at each time point based on the band data of the remote sensing image. Summarize the vegetation index at each time point to obtain the time series of the vegetation index, Sentinel-NDVI.

[0022] (1) The process of calculating the vegetation index at each time point includes: First, define the region and time range, identify the area of ​​interest, define the study area by specifying coordinates, set the year for analysis, and define the time range; Secondly, the satellite remote sensing digital image of the cotton field to be tested is loaded, and the influence of clouds and cloud shadows in the satellite remote sensing digital image is removed through the cloud mask function to ensure the image quality.

[0023] Finally, the vegetation index at each time point in the satellite remote sensing digital image is calculated using the formula NDVI = (NIR-R) / (NIR+R) according to the reflectivity R of the red band and the reflectivity NIR of the near infrared band. A time-based exponential change curve is generated based on the obtained vegetation index to show the vegetation changes at each time point, and the vegetation index data at each time point and the corresponding time series graph are exported.

[0024] (2) The specific operation of calculating the vegetation index at each time point is: 1. Data preparation Select Image Collection: Load data from the Sentinel-2 image collection.

[0025] Filter time range: Set the time range of images.

[0026] Filter cloud cover: Filter out images with cloud cover less than 20% to reduce the impact of clouds on the data.

[0027] Set Region of Interest (ROI): Specify a region of interest to limit the geographic extent of the image.

[0028] 2. Apply Cloud Mask Define Cloud Mask Function: Create a function to remove clouds and cloud shadows based on cloud mask data (such as the QA60 band).

[0029] Apply Cloud Mask: Apply a cloud mask function to an image collection to ensure that the image data is properly masked over clouds and cloud shadows.

[0030] 3. Select the band Extract Bands: Select the near infrared band and the red band from each image. These two bands are used to calculate NDVI.

[0031] 4. Calculate NDVI Define a function to calculate NDVI: Create a function to calculate NDVI using a formula.

[0032] 5. Analysis and visualization Analyze NDVI results: Analyze the calculated NDVI bands, which may include statistical calculations, time series analysis, etc.

[0033] Visualize results: Present NDVI data using charts or maps for further interpretation and application.

[0034] 3. Time series reconstruction of NDVI data The Sentinel data were preprocessed through correction, cloud removal, and screening, and then the Sentinel-NDVI time series was denoised.

[0035] The specific steps are as follows: the concave points in the time series Sentinel-NDVI are taken as abnormal points, and the NDVI values ​​of the abnormal points are replaced by the average NDVI values ​​of the adjacent time points for preliminary denoising to obtain the denoised time series. Then the denoised time series is filtered using the SG filtering method to generate a growth curve that characterizes the growth trend and law of cotton. The point with the maximum local change rate of the growth curve is taken as the time point to be determined, and the time point before a set time period of the time point to be determined is taken as the application time of the defoliant.

[0036] By visually observing the spatial differences in the application time of cotton defoliants, it can be found that due to differences in climate, soil and other factors in different regions, there are significant differences in the growth cycle and maturity of cotton, which in turn affects the application time of defoliants.

[0037] The SG filtering method is a digital signal processing technique that is mainly used to smooth data and reduce noise. In each given data window, a polynomial curve is fitted using the least squares method, and the smoothed data is obtained by calculating the polynomial derivative.

[0038] The specific process is as follows: First, select a window size of 7 and a polynomial order of 3, and use the least squares method to fit the order of the polynomial in the window. For polynomial fitting in the sliding window, for each data point in the time series, select a centrally symmetric sliding window, fit a polynomial in this window using the least squares method, and use the fitted polynomial to calculate the value of the center point in the window, which replaces the original center point data. The sliding window moves, moving the window backward from the start position of the sequence, one data point at a time, and repeating the fitting and replacement process until the entire time series is traversed.

[0039] NDVI is the most widely used vegetation index at present. It can partially eliminate the noise caused by observation conditions such as solar altitude angle and atmosphere. The characteristics of its formula itself make it sensitive to chlorophyll and have a large curve amplitude, so it can sensitively detect the growth characteristics of vegetation. After spraying defoliants, cotton leaves gradually aged and turned yellow, the chlorophyll content decreased gradually, the NDVI index decreased, and the rate of change reached the maximum in 3-6 days. On the NDVI curve, it is manifested as the absolute value of the second-order derivative of the NDVI curve reaching a local maximum. The point corresponding to the local maximum value is taken as the time-to-be-determined point, which is located at the period when the second-order derivative begins to change from positive to negative. The time-to-be-determined point is taken as the time point before a set time period as the application time of the defoliant.

[0040] 4. Evaluating the accuracy of the algorithm 1. Obtain the actual defoliation time and inverse defoliation time of the cotton field to be tested Actual Defoliation Time: Obtain data on the actual time of defoliant application for multiple cotton fields.

[0041] Inverse defoliation time: The defoliant application time obtained by algorithm for the same cotton field sample.

[0042] Among them, the actual defoliation time and the object cotton field of the inverted defoliation time are one-to-one corresponding.

[0043] 2. Build data pairs Construct an array of actual defoliation time based on the actual defoliation time of the cotton field to be tested Y =[Y1, Y2, ..., Yn], construct the inverse defoliation time array based on the inverse defoliation time of the cotton field to be tested X =[X1,X2,…,Xn].

[0044] 3. Array according to actual defoliation time Y and invert the defoliation time array X Determine the Pearson correlation coefficient r : ; ; ; ; ; ; in, is the standard deviation of the inverted defoliation time array, is the standard deviation of the actual defoliation time array, is the covariance of the inverted defoliation time array and the actual defoliation time array, n is the total number of elements in the array, To invert the elements in the defoliation time array, is the element in the array of actual defoliation time, is the average value of the inverted defoliation time array, is the average value of the actual defoliation time array.

[0045] When the Pearson correlation coefficient r When it is greater than or equal to 0.8, the actual defoliation time and the inverted defoliation time are strongly correlated; When the Pearson correlation coefficient r When it is greater than or equal to 0.5 and less than 0.8, the actual defoliation time has a strong correlation with the inverted defoliation time; When the Pearson correlation coefficient r When it is greater than or equal to 0.3 and less than 0.5, the actual defoliation time and the inverted defoliation time are moderately correlated; When the Pearson correlation coefficient r When it is less than 0.3, the actual defoliation time and the inverted defoliation time are weakly correlated.

[0046] Weak correlation usually means that there is a large deviation between the prediction model or observation data and the actual situation. This deviation may be largely due to the influence of interference factors such as environment and terrain. In the process of predicting the application time of cotton defoliants, if the observation frequency is reduced and the data quality deteriorates, weak correlation problems will be encountered, which brings great challenges to the analysis.

[0047] Based on cotton mapping, representative planting areas are selected as new observation points to recollect and analyze data. In the process of reanalysis, more attention needs to be paid to changes in interference factors such as environment and topography, as well as their impact on cotton growth and the application time of defoliants. In this way, the application time of cotton defoliants can be predicted more accurately, providing more reliable support for agricultural production. This also provides an important reference for us to optimize the prediction model and improve data quality in the future.

[0048] Based on the same inventive concept, the present invention also provides a defoliant application time reversal system, comprising: The data acquisition module is used to acquire satellite remote sensing digital images of the cotton field to be tested.

[0049] The sequence construction module is used to determine the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested, and summarize the vegetation index at each time point to obtain the time series of vegetation index Sentinel-NDVI.

[0050] The denoising module is used to take the concave points in the time series Sentinel-NDVI as outliers, and replace the NDVI value of the outlier with the average NDVI value of the adjacent time points for preliminary denoising to obtain the denoised time series.

[0051] The inversion module is used to filter the denoised time series to generate a growth curve that characterizes the growth trend and law of cotton, and to take the maximum point of the local change rate of the growth curve as the time to be determined point, and the time point before the set period of the time to be determined point as the application time of the defoliant.

[0052] Based on the same inventive concept, the present invention also provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the defoliant application time inversion method are implemented.

[0053] Based on the same inventive concept, the present invention also provides a computer-readable storage medium containing instructions, and a computer program is stored on the storage medium. For example, a memory containing instructions, the above instructions can be executed by a processor of a computer device to complete the above method. For example, a non-temporary computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device. When the computer program is executed by the processor, the steps in the embodiment of the defoliant application time inversion method can be implemented. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0054] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A defoliant application time reversal method, characterized in that: include: Obtain satellite remote sensing digital images of the cotton field to be tested; Determine the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested, and summarize the vegetation index at each time point to obtain the time series of vegetation index Sentinel-NDVI; The concave points in the time series Sentinel-NDVI are regarded as outliers, and the NDVI value of the outliers is replaced by the average NDVI value of the adjacent time points for preliminary denoising to obtain the denoised time series; The denoised time series is filtered to generate a growth curve that represents the growth trend and law of cotton. The point with the maximum local change rate of the growth curve is taken as the time-to-be-determined point, and the time point before a set time period of the time-to-be-determined point is taken as the application time of the defoliant.

2. A defoliant application time reversal method as claimed in claim 1, characterized in that: After obtaining the satellite remote sensing digital image of the cotton field to be tested, the satellite remote sensing digital image is preprocessed by correction, cloud removal and screening.

3. A defoliant application time reversal method as claimed in claim 1, characterized in that: The step of determining the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested specifically includes: Define the study area by specifying coordinates, set the year for analysis, and define the time range; Load the satellite remote sensing digital image of the cotton field to be tested, and use the cloud mask function to remove the influence of clouds and cloud shadows in the satellite remote sensing digital image; According to the reflectivity R of the red band and the reflectivity NIR of the near infrared band, the vegetation index at each time point in the satellite remote sensing digital image is calculated using the formula NDVI = (NIR-R) / (NIR+R).

4. A defoliant application time reversal system, characterized in that: include: A data acquisition module is used to acquire satellite remote sensing digital images of the cotton field to be tested; The sequence construction module is used to determine the vegetation index at each time point in the satellite remote sensing digital image of the cotton field to be tested, and summarize the vegetation index at each time point to obtain the time series of vegetation index Sentinel-NDVI; The denoising module is used to treat the concave points in the time series Sentinel-NDVI as outliers, and replace the NDVI value of the outlier with the average NDVI value of the adjacent time points for preliminary denoising to obtain the denoised time series; The inversion module is used to filter the denoised time series to generate a growth curve that characterizes the growth trend and law of cotton. The point with the maximum local change rate of the growth curve is used as the time to be determined point, and the time point before the set period of the time to be determined point is used as the application time of the defoliant in the cotton field to be tested.

5. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the defoliant application time reversal method according to any one of claims 1 to 3 are implemented.

6. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the defoliant application time reversal method according to any one of claims 1 to 3 are implemented.

Images