Soil fertility prediction system and method

Abstract

The present invention discloses a soil fertility prediction system and method comprising a data acquisition module, a data prediction module, a decision execution module, and a results display module, wherein the data prediction module analyzes collected images to construct a fertility prediction model by combining image data and a fertility index, and realizes the prediction of a fertilization strategy for the detected area.

Description

[Technical Field]

[0001] The present invention relates to the technology of data analysis and prediction, and more specifically, to a soil fertility prediction system and method. [Background technology]

[0002] Soil fertility is the ability of soil to provide and harmonize the nutritional and environmental conditions necessary for plant growth. It is a comprehensive expression of various fundamental soil properties and has a crucial impact on crop yield and quality. Accurately predicting soil fertility contributes to rational fertilization, improved fertilizer utilization, reduced environmental pollution, and the realization of sustainable agricultural development.

[0003] The invention title refers to the patent document "Soil Fertility Prediction Method and System, Electronic Device and Storage Medium" (Patent Publication Number: CN116562134A, Publication Date: 2023-08-08). The method includes determining initial sampling points and central sampling points for a target plot; constructing a sampling point array for each initial sampling point; determining preliminary average soil fertility data and preliminary soil fertility change data for each sampling point array based on the fertility index of the soil sample corresponding to the sampling point ranked in the top first rank; predicting the fertility of the soil sample corresponding to the sampling point not ranked in the top first rank; updating the preliminary average soil fertility data and preliminary soil fertility change data based on the prediction results to obtain target average soil fertility data and target soil fertility change data; and obtaining soil fertility prediction data for the target plot based on the target average soil fertility data for all sampling point arrays.

[0004] Based on the above-mentioned literature, existing soil fertility prediction processes often include both planted and unplanted areas, resulting in significant differences in fertility data between the two. Furthermore, if the location of the sampling points and the soil collection depth are not accurate, the final fertility prediction will be inconsistent. Additionally, the current process of achieving fertility prediction involves large variations in the fertility index within the area, leading to problems with uneven fertilization rates. Therefore, the present invention provides a soil fertility prediction system and method. [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] In response to the shortcomings of conventional technology, the present invention provides a soil fertility prediction system and method. In existing soil fertility prediction processes, the relevant area often includes planted and non-planted areas, and there is a significant difference in fertility data between the two. If the positioning of the sampling points and the soil collection depth are not accurate, the final fertility prediction will be inconsistent. Furthermore, in the process of realizing current fertility predictions, there is a large variation in the fertility index within the area, resulting in unevenness in the planned fertilizer application. This invention solves these problems. Means for solving the invention

[0006] To achieve the above objective, the present invention is accomplished by the following technical solution: The soil fertility prediction system is A data acquisition module that uses an image acquisition module mounted on a drone device to position the area to be detected, and after positioning is complete, performs image acquisition operations on the area to be detected, collects data on soil fertility in the image via a sensor acquisition module mounted on the drone device, and transmits and stores the collected data in a storage database. A data prediction module that analyzes collected images to determine planting area features, extracts planting parameters from a storage database to perform image segmentation, determines sampling points, calculates a fertility index for each collection point by combining collected soil fertility data, constructs a fertility prediction model by combining image data and the fertility index, and realizes a prediction of fertilization strategies for the detected area. A decision-making execution module that formulates an execution strategy based on the predicted results and completes fertilization operations in areas with different levels of fertility in cooperation with fertilization equipment, It includes a results display module that displays the result data obtained from the analysis process as text and graphs.

[0007] Preferably, the operation of image acquisition in the data acquisition module is: Determine the boundary of the area to be detected, set the drone's acquisition height, define the actual size of the image at the current height as A1 (length) and A2 (width), plan the drone's flight path, and perform the image acquisition operation. Set the collection start point for the detection region, where the start point lies on the horizontal boundary parallel to the horizontal direction of the image and the detection region, and the distance between the position of the start point and the left vertical boundary of the detection region is less than A2 / 2. Set the collection endpoint for the detection region, and the endpoint lies on the horizontal boundary of the detection region or an extension of the horizontal boundary, and the distance between the endpoint and the right vertical boundary of the detection region is less than A2 / 2. The drone's movement path first moves along the vertical boundary of the detection area, collecting images at intervals of B1. It then moves until the lateral boundary behind the detection area is visible in the image, thereby achieving a lateral movement distance of B2, thus realizing an S-shaped path for image acquisition. After data collection is complete, the image data is subjected to stitching.

[0008] Preferably, the operation of stitching the image data is performed as follows: Using the starting point image as a reference, the starting point image is divided by combining the directions of the path, and the dividing line is set at a point (A1-B1) away from the lateral boundary behind the starting point image. Then, the division operation is performed, the image with the largest area after division is retained, and images adjacent in the path direction are stitched together with the retained image. When sequentially stitching image data until it is stored at the lateral boundary behind the detected region, the path is combined so that the dividing line is set at a location (A2-B2) away from the right vertical boundary of the current image. Then the division operation is performed, the image with the largest area after division is retained, and the image stitching continues until the stitching of the image at the endpoint is completed by further combining the path, forming complete image data.

[0009] Preferably, in the data prediction module, the operation of analyzing the collected images to determine the characteristics of the planting area is: By performing a grayscale conversion operation on the entire complete image data, multiple comparison points are set at equal intervals between the midpoints of the boundary lines located at the relative boundaries of the image data. Furthermore, the direction of the planting area features is determined based on the change in the grayscale value of the comparison points. Then, based on the different grayscale values ​​of the planting area features and the separation ditch features, the feature spacing lines are determined. If adjacent comparison points experience a change in grayscale value, the midpoints of the adjacent comparison points are taken and connected sequentially as spacing lines. After that, the planting area features and the separation ditch features are distinguished.

[0010] Preferably, in the data prediction module, the operation of extracting planting parameters from the storage database to perform image segmentation and determining sampling points is: Based on the planting parameters, the spacing distance to be maintained between plants is denoted as M. Based on the spacing lines of both sides of the planting area features and the separation groove features, the planting area features are divided into multiple grid regions at equal intervals according to the distance M, and the position marks of the grid regions are realized sequentially. Randomly select any grid area as the base sampling area. After connecting the diagonal points of the current base sampling area, the intersection points generated by the intersection are used as the base sampling points. After all four sides located at the base sampling points are spaced by one grid area, set the sampling points, and set all sampling points according to the current method.

[0011] Preferably, the operation of calculating the fertility index of each collection point by combining the soil fertility data collected by the data prediction module is Extract the planting parameters to determine the pre-embedding depth of the plant seeds, that is, when collecting the soil fertility, the sensor of the device realizes deeply penetrating into the pre-embedding depth directly below the collection point. Let each fertility data value of one collection point be P i and P i represents the detected value P of the i-th fertility data category. Perform a weighting operation on each fertility data category, and the sum after weighting is the fertility index. The specific formula is

Equation

[0012] Preferably, in the data prediction module, the operation of constructing a prediction model by combining the image data and the fertility index to realize the prediction of the fertilization strategy for the detection area is . Combine the planting area characteristics and the fertility index of each sampling point to construct a fertility prediction model for the current detection area. Use the fertility prediction model to calculate the deviation value of the fertility index of the sampling points of the horizontal planting area characteristics and the vertical planting area characteristics in the detection area, and determine the fertilization direction strategy located in the planting area based on the deviation value. After that, planting parameters are extracted, the fertility interval requirements of plant seeds are determined as "F1, F2", the average fertility index in the planting area is compared with the fertility interval requirements, and a fertilization amount strategy is determined. The fertilization direction strategy and the fertilization amount strategy are combined to obtain and transmit the predicted fertilization strategy result.

[0013] Preferably, the calculation operation of the deviation value is The calculation formula for the deviation value of the horizontal planting area characteristics is

Number

Number

Number

Number

[0014] Preferably, the operation of comparing the average fertility index in the planting area with the fertility interval requirement is Q c ∈ [F1, F2] or Q d If ∈ [F1, F2], the fertility index in the current planting area is predicted to meet the fertility interval requirement, Q c > [F1, F2] or Q d If > [F1, F2], it is necessary to reduce the fertilization amount, Q c < [F1, F2] or Q d If < [F1, F2], the fertility index in the current planting area is predicted to be insufficient, and it is necessary to increase the fertilization amount.

[0015] The present invention further discloses a soil fertility prediction method. Specifically, Performing an image collection operation on the detection area by a drone device, determining the characteristics of the planting area to be sampled, determining sampling points, and realizing the detection operation of soil fertility in step 1; Based on the image data and the data of fertility detection, constructing a fertility prediction model and realizing the prediction of the fertilization strategy for the detection area in step 2; Based on the results obtained from the prediction, realizing the execution operation of the planned direction and planned usage amount in step 3.

Effects of the Invention

[0016] The present invention provides a soil fertility prediction system and method. Compared with the prior art, the present invention has the following beneficial effects.

[0017] 1. The soil fertility prediction system and method analyze collected images to determine planting area characteristics, extract planting parameters from a storage database to perform image segmentation, determine sampling points, combine collected soil fertility data to calculate a fertility index for each collection point, and construct a fertility prediction model by combining image data and the fertility index to predict a fertilization strategy for the detected area. This effectively and efficiently predicts soil fertility, enables accurate prediction operations by constructing the model synchronously, and ensures high accuracy of the fertilization strategy within the area.

[0018] 2. The soil fertility prediction system and method can improve the efficiency of data collection by determining the boundaries of the area to be detected, setting the path for image collection of the detection area, and performing stitching operations on the image data after collection is complete. Furthermore, it ensures the accuracy of content features within the images, enables the determination of sampling points after better determination of planting area features, provides a foundation for achieving accuracy in subsequent prediction operations, reduces the problem of area omissions, and better completes soil fertility detection and prediction.

[0019] 3. The soil fertility prediction system and method extract planting parameters to determine the spacing distance to be maintained between planted plants, and, based on the spacing lines between both sides of the planting area features and the separation trench features, divides the planting area features into multiple grid regions at equal intervals according to the distance, and sequentially realizes position marks for the grid regions. The intersections generated by connecting the diagonal points of the current basic sampling area are used as the basic sampling points, thereby ensuring that the sampling points become the subsequent planting points, ensuring the accuracy of data positioning, and improving the fault tolerance rate.

[0020] 4. The soil fertility prediction system and method combine the characteristics of the planting area and the fertility index of each sampling point to construct a fertility prediction model for the current detection area. Using the fertility prediction model, the system calculates the deviation values ​​of the fertility index of the sampling points for the horizontal and vertical planting area characteristics within the detection area. Based on the deviation values, it determines the fertilization direction strategy within the planting area. By comparing the average fertility index within the planting area with the fertility interval requirement, the system determines the fertility status within the area, reduces deviations, and simultaneously forms a fertilization strategy, thereby completing efficient prediction and planning operations. [Brief explanation of the drawing]

[0021] [Figure 1] This is a block diagram illustrating the principle of the soil fertility prediction system according to the present invention. [Figure 2] This is a flowchart illustrating the soil fertility prediction method according to the present invention. [Figure 3] This is a schematic diagram of the image acquisition operation according to the present invention. [Figure 4] This is a schematic diagram of the operation according to the present invention, which involves analyzing collected images to determine the characteristics of a planted area. [Modes for carrying out the invention]

[0022] The technical proposals in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiments of the present invention. It will be clear that the embodiments described are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without performing work worthy of inventive step will also fall within the scope of the protection of the present invention.

[0023] Referring to Figures 1 and 2, the present invention provides three technical solutions.

[0024] Example 1 The soil fertility prediction system is A data acquisition module that uses an image acquisition module mounted on a drone device to position the area to be detected, and after positioning is complete, performs image acquisition operations on the area to be detected, collects data on soil fertility in the image via a sensor acquisition module mounted on the drone device, and transmits and stores the collected data in a storage database. A data prediction module that analyzes collected images to determine planting area features, extracts planting parameters from a storage database to perform image segmentation, determines sampling points, calculates a fertility index for each collection point by combining collected soil fertility data, constructs a fertility prediction model by combining image data and the fertility index, and realizes a prediction of fertilization strategies for the detected area. A decision-making execution module that formulates an execution strategy based on the predicted results and completes fertilization operations in areas with different levels of fertility in cooperation with fertilization equipment, It includes a results display module that displays the result data obtained from the analysis process as text and graphs.

[0025] Here, after collecting images, planting area features are determined by analysis, planting parameters are extracted from a storage database to perform image segmentation, sampling points are determined, the fertility index for each collection point is calculated by combining the collected soil fertility data, a fertility prediction model is constructed by combining the image data and the fertility index, and a fertility strategy for the detected area is predicted. This effectively and efficiently predicts soil fertility, enables accurate prediction operations by constructing the model synchronously, and ensures high accuracy of the fertilization strategy within the area.

[0026] In an embodiment of the present invention, the operation of performing image acquisition in the data acquisition module is: Determine the boundary of the area to be detected, set the drone's acquisition height, define the actual size of the image at the current height as A1 (length) and A2 (width), plan the drone's flight path, and perform the image acquisition operation. Set the collection start point for the detection region, where the start point lies on the horizontal boundary parallel to the horizontal direction of the image and the detection region, and the distance between the position of the start point and the left vertical boundary of the detection region is less than A2 / 2. Set the collection endpoint for the detection region, and the endpoint lies on the horizontal boundary of the detection region or an extension of the horizontal boundary, and the distance between the endpoint and the right vertical boundary of the detection region is less than A2 / 2. The drone's movement path first moves along the vertical boundary of the detection area, collecting images at intervals of B1. It then moves until the lateral boundary behind the detection area is visible in the image, thereby achieving a lateral movement distance of B2, and thus enabling an S-shaped path for image acquisition. After data collection is complete, the image data is subjected to stitching.

[0027] In embodiments of the present invention, the operation of performing stitching on image data is: Using the starting point image as a reference, the starting point image is divided by combining the directions of the path, and the dividing line is set at a point (A1-B1) away from the lateral boundary behind the starting point image. Then, the division operation is performed, the image with the largest area after division is retained, and images adjacent in the path direction are stitched together with the retained image. When sequentially stitching image data until it is stored at the lateral boundary behind the detected region, the path is combined so that the dividing line is set at a location (A2-B2) away from the right vertical boundary of the current image. Then the division operation is performed, the image with the largest area after division is retained, and the image stitching continues until the stitching of the image at the endpoint is completed by further combining the path, forming complete image data.

[0028] Here, by determining the boundaries of the detection area, setting the image acquisition path for the detection area, and performing stitching operations on the image data after the acquisition is complete, it is possible to not only improve the efficiency of data acquisition, but also ensure the accuracy of content features within the images, enable the determination of sampling points after better determining the characteristics of the planted area, provide a foundation for achieving accuracy in subsequent prediction operations, reduce the problem of area omissions, and better complete soil fertility detection and prediction.

[0029] In an embodiment of the present invention, the operation of analyzing images collected in the data prediction module to determine the characteristics of the planted area is: By performing a grayscale conversion operation on the entire complete image data, multiple comparison points are set at equal intervals between the midpoints of the boundary lines located at the relative boundaries of the image data. Furthermore, the direction of the planting area features is determined based on the change in the grayscale value of the comparison points. Then, based on the different grayscale values ​​of the planting area features and the separation ditch features, the feature spacing lines are determined. If adjacent comparison points experience a change in grayscale value, the midpoints of the adjacent comparison points are taken and connected sequentially as spacing lines. After that, the planting area features and separation ditch features are distinguished.

[0030] Here, the grayscale value is a numerical value obtained by the grayscale conversion operation of the image, and is an existing processing technique. When dealing with different features, the numerical values ​​will differ. For example, planting area features and separation trench features do not match in terms of height, shape, etc., so their grayscale values ​​will be different. If the grayscale values ​​located at the set comparison points are nearly similar, it indicates that the current planting direction is the horizontal direction connecting the midpoints of the current boundary lines.

[0031] In an embodiment of the present invention, the operation of extracting planting parameters from a storage database in the data prediction module to perform image segmentation and determine sampling points is as follows: Based on the planting parameters, the spacing distance to be maintained between plants is denoted as M. Based on the spacing lines of both sides of the planting area features and the separation groove features, the planting area features are divided into multiple grid regions at equal intervals according to the distance M, and the position marks of the grid regions are realized sequentially. Randomly select an arbitrary grid region as the base sampling region, connect the diagonal points of the current base sampling region, and set the intersection points generated by the intersections as the base sampling points. Set sampling points after each of the four faces located at the base sampling points is spaced by one grid region, and set all sampling points according to the current method.

[0032] Here, planting parameters are extracted to determine the spacing distance to be maintained between planted plants. Based on the spacing lines between both sides of the planting area features and the separation groove features, the planting area features are divided into multiple grid regions at equal intervals according to the distance, and position markers for the grid regions are sequentially realized. After connecting the diagonal points of the current basic sampling area, the intersection points generated by the intersections are used as the basic sampling points. This ensures that the sampling points become the subsequent planting points, thereby ensuring the accuracy of data positioning and improving the fault tolerance rate.

[0033] In an embodiment of the present invention, the operation of calculating the fertility index for each collection point by combining the soil fertility data collected by the data prediction module is as follows: By extracting planting parameters and determining the pre-burial depth of plant seeds, and by ensuring that the device's sensor penetrates deeply to the pre-burial depth directly below the collection point when collecting soil fertility data, P i P i represents the detected value P for the i-th fertility data category. A weighting operation is performed on each fertility data category, and the weighted sum is the fertility index. The specific formula is:

number

[0034] In an embodiment of the present invention, the operation of constructing a prediction model by combining image data and fertility index in the data prediction module and realizing the prediction of a fertilization strategy for the detected area is: By combining the characteristics of the planting area and the fertility index of each sampling point, a fertility prediction model for the current detection area is constructed. Using this fertility prediction model, the standard deviation of the fertility index of the sampling points for the horizontal and vertical planting area characteristics within the detection area is calculated, and the fertilization direction strategy located within the planting area is determined based on the standard deviation. Subsequently, planting parameters were extracted to determine the fertility interval requirements of the plant seeds as "F1" and "F2". The average fertility index in the planting area was then compared with the fertility interval requirements to determine the fertilizer application strategy. The system combines fertilization direction strategies and fertilization amount strategies to obtain and transmit predicted fertilization strategy results.

[0035] Here, a fertility prediction model for the current detection area is constructed by combining the characteristics of the planting area and the fertility index of each sampling point. Using the fertility prediction model, the deviation values ​​of the fertility index of the sampling points for the horizontal and vertical planting area characteristics within the detection area are calculated. Based on the deviation values, a fertilization direction strategy within the planting area is determined. The fertilization strategy is then determined by comparing the average fertility index within the planting area with the fertility spacing requirements. In this way, the influence of multiple categories of elements on fertility is combined to determine the fertility situation within the area, reduce deviations, and simultaneously form a fertilization strategy, completing efficient prediction and planning operations.

[0036] In the embodiment of the present invention, the calculation operation of the standard score is as follows: The formula for calculating the standard score of the characteristics of the horizontal planting area is:

number

Equation

Equation

[0037] In the embodiments of the present invention, the operation of comparing the average fertility index in the planting area with the fertility interval requirement is Q c ∈ [F1, F2] or Q d ∈ [F1, F2], it is predicted that the fertility index in the current planting area meets the fertility interval requirement, and the fertilization operation with an appropriate amount is realized. Q c > [F1, F2] or Q d > [F1, F2], it is necessary to reduce the fertilization amount, and the operation of equalizing the fertility is realized by digging trenches. Q c < [F1, F2] or Q d < [F1, F2], it is predicted that the fertility index in the current planting area is insufficient, and it is necessary to increase the fertilization amount.

[0038] Example 2 The difference between Example 2 and Example 1 is that the present invention further discloses a method for predicting soil fertility, specifically, Step 1 involves using a drone device to collect images from the detection area, determining the characteristics of the planted area to be sampled, and determining the sampling point to perform soil fertility detection. Step 2 involves constructing a fertility prediction model based on image data and fertility detection data, and realizing the prediction of a fertilization strategy for the detected area. This includes step 3, which involves implementing the planned direction and planned usage based on the results obtained from the prediction.

[0039] Example 3 The difference between Example 3 and Examples 1 and 2 is that it further discloses a practical example in which a detection operation is performed on a known area of ​​soil fertility, the originally collected data parameters are stored, the operation is performed according to the application method of the present invention, and the data results are recorded and compared.

[0040] The established soil fertility standards are shown in Table 1. Table 1 Gradual fertility table [Table 1]

[0041] The corresponding data collected is shown in Table 2. Table 2 Numerical Table [Table 2] The soil fertility index was calculated to be 0.757, and after comparing it with a graded fertility table, it was found to be of medium-high fertility.

[0042] Synchronized detection was achieved using an existing soil fertility prediction system and the soil fertility prediction system of the present invention. By comparing the calculated results with initial known parameters, the required time and accuracy of the results for both were determined and recorded. The results are shown in Table 3.

[0043] Table 3 Comparison result table [Table 3]

[0044] As described above, by using the soil fertility prediction system and method of the present invention in actual application operations, the time required to complete the detection operation is shortened, and the accuracy of the obtained results is increased, thus enabling better application to actual operations.

[0045] At the same time, anything not described in detail herein is the prior art known to those skilled in the art.

[0046] In this specification, relational terms such as "first" and "second" are used solely to distinguish one entity or operation from another, and do not necessarily imply or require any actual relationship or order between these entities or operations. The terms "include," "contain," or any other variation thereof are intended to include non-exclusive inclusion, such that a process, method, article, or apparatus containing a set of elements includes not only those elements but also other elements not explicitly listed, or may include elements specific to such a process, method, article, or apparatus.

[0047] Although embodiments of the present invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited to equivalents by the appended claims.

Claims

1. A soil fertility prediction system, A data acquisition module that uses an image acquisition module mounted on a drone device to position the area to be detected, and after positioning is complete, performs image acquisition of the area to be detected, collects data on soil fertility in the image via a sensor acquisition module mounted on the drone device, and transmits and stores the collected data in a storage database. A data prediction module that analyzes collected images to determine planting area features, extracts planting parameters from a storage database to perform image segmentation, determines sampling points, calculates a fertility index for each collection point by combining collected soil fertility data, constructs a fertility prediction model by combining image data and the fertility index, and realizes a prediction of fertilization strategies for the detected area. A decision-making execution module that formulates an execution strategy based on the predicted results and completes fertilization operations in areas with different levels of fertility in cooperation with fertilization equipment, It includes a results display module that displays the result data obtained from the analysis process as text and graphs, A soil fertility prediction system characterized by the following features.

2. The operation of combining the soil fertility data collected by the aforementioned data prediction module to calculate the fertility index for each collection point is as follows: By extracting planting parameters and determining the pre-burial depth of plant seeds, and by ensuring that the equipment's sensors penetrate deeply to the pre-burial depth directly below the collection point when collecting soil fertility data, P ｉ P ｉ P represents the detected value P for the i-th fertility data category. A weighting operation is performed on each fertility data category, and the weighted sum is the fertility index. The specific formula is: [Math 1] And, Q ｊ Q is the fertility index at the j-th collection point, and W ｉ is the weighted value W for the i-th fertility data category, and n represents the total number of fertility data categories. The soil fertility prediction system according to feature 1.

Images