A fruit growth monitoring system

Abstract

The present invention relates to the technical field of cluster analysis, and specifically discloses a fruit growth monitoring system, comprising: an acquisition module: obtaining a target value at a time point, obtaining target data based on the target value, and generating a target point; a cluster analysis module: setting a cluster radius, and clustering the target points according to the density of the target points within the cluster radius; a cluster optimization module: when the density of the target points in the overlapping area meets the conditions, adjusting the cluster clusters according to the theoretical center of gravity and the maximum distance; a result module: obtaining standard points, and judging the growth status of the fruit based on the standard points. The present invention can improve the efficiency of monitoring the growth status of the fruit.

Description

Technical Field

[0001] The invention relates to the technical field of cluster analysis, and in particular to a fruit growth monitoring system. Background Art

[0002] Fruit growth monitoring refers to tracking and measuring fruits at various stages of their growth process in order to understand their developmental status, health status and the impact of the environment on fruit growth. Fruit growth monitoring can help improve the efficiency and quality of agricultural production.

[0003] With the continuous development of technologies such as smart sensors, these sensors are used to obtain the required data and analyze them to determine the growth status of the fruit, which greatly improves the efficiency and accuracy of fruit growth monitoring. However, this method still evaluates the growth status of individual fruits in turn, and the number of fruits is often as high as thousands or even tens of thousands, so this method still has the problems of large workload and low efficiency. Summary of the invention

[0004] The object of the present invention is to provide a fruit growth monitoring system to solve the following technical problems:

[0005] When there are a large number of fruits, the method of analyzing the growth status of individual fruits one by one is labor-intensive and inefficient.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A fruit growth monitoring system includes a collection module, a cluster analysis module, a cluster optimization module and a result module, specifically:

[0008] Acquisition module: setting a number of time points at preset time intervals within a preset monitoring period, obtaining the size of the fruit based on the sensor at the time point, recording it as the target value, obtaining target data based on the target value, the target data including the maximum target value A1, the average change speed of the target value A2 and the change amount of the target value A3, and generating target points (A1, A2, A3) in the coordinate system;

[0009] Cluster analysis module: Set the cluster radius r, take the target point k as the center, and calculate the target point density P within the cluster radius r k , when the target point density P k When the density is greater than or equal to the preset density threshold Pys, a cluster K is generated with the target point k as the center; if the target point density corresponding to the target point a exists in the cluster K , generating cluster A with the target point a as the center, and merging it with cluster K to generate a new cluster KA;

[0010] Cluster optimization module: Draw circles k and a with radius r with the target point k and target point a as the center, and obtain the target point density P in the overlapping area DCOI of the circle k and the circle a. coi ;

[0011] When the target point density When the theoretical center of gravity Z of the overlapping area DCOI is obtained, η is the preset correction coefficient, S coi represents the area of ​​the overlapping region DCOI;

[0012] Get the maximum distance d from the theoretical center of gravity Z to the boundary of the overlapping area DCOI, and calculate the target point density within the maximum distance d with the theoretical center of gravity Z as the center. , slowly increase the maximum distance d at a preset speed, and record the target point density corresponding to each increase of the maximum distance, until the corresponding target point density after a certain increase of the maximum distance is less than the target point density after the last increase of the maximum distance, take the theoretical center of gravity as the center, and the maximum distance at this time is the clustering radius, and obtain a new cluster to replace the cluster KA;

[0013] Result module: obtain the target point corresponding to the center of the cluster, use it as the standard point, and judge the growth status of the fruit based on the standard point.

[0014] As a further solution of the present invention: the process of judging the growth state of the fruit specifically includes:

[0015] Obtain the target data corresponding to the standard point, use it as the standard data, calculate the standard score y based on the superior and inferior solution distance method, set the standard score thresholds Y1 and Y2, and Y1<Y2;

[0016] When the standard score y<Y1, it is determined that the growth state of the fruit is poor;

[0017] When the standard score Y1≤y<Y2, the growth state of the fruit is judged to be average;

[0018] When the standard score Y2≤y, it is determined that the growth state of the fruit is good.

[0019] As a further solution of the present invention: in the acquisition module, the process of acquiring the target data specifically includes:

[0020] Generate a coordinate point (i, Ni), where Ni represents the target value at time point i, and fit the coordinate point to obtain a fitting curve f(t), where t represents time;

[0021] Calculate the maximum target value ;

[0022] Calculate the average change rate of the target value ,in, , Respectively represent the starting point and end point of the monitoring period;

[0023] Calculate the target value change .

[0024] As a further solution of the present invention: In the cluster analysis module, the target point density within the cluster radius r is calculated with the target point k as the center. , m represents the number of target points within the cluster radius r with the target point k as the center.

[0025] As a further solution of the present invention: in the cluster optimization module, the process of obtaining the theoretical center of gravity Z of the overlapping area DCOI specifically includes:

[0026] Get the points inside the overlapping area DCOI The total number of target points on X j , calculate the proportionality coefficient of point j , Indicates the total number of target points in the overlapping area DCOI, and obtains the standard coordinates of point j , calculate the mean of all standard coordinates in each dimension and determine the theoretical center of gravity Z.

[0027] As a further solution of the present invention: in the acquisition module, when the maximum target value is less than a preset target value threshold, subsequent steps are not executed and it is determined that the growth state of the fruit is poor.

[0028] As a further solution of the present invention: the target value threshold is determined based on the total growth time of the fruit, and the greater the total growth time, the greater the corresponding target value threshold.

[0029] Beneficial effects of the present invention: In this solution, fruit size data is automatically acquired by sensors, which reduces the time and labor cost of manual collection. Through systematic and automated data collection, high-quality and rich basic data are provided for subsequent cluster analysis, ensuring the comprehensiveness and accuracy of monitoring; thereafter, a large number of target points are effectively classified into several clusters by setting the cluster radius and density threshold, significantly reducing the number of fruits that need to be evaluated individually, and by identifying high-density areas and classifying fruits with similar growth states into one category, thereby simplifying subsequent analysis work, helping to reduce the overall workload and improve evaluation efficiency, while maintaining accurate monitoring of the growth state of the fruit, providing a feasible solution for large-scale orchard management; then, by dynamically adjusting the cluster radius and recalculating the density of target points in the overlapping area, the preliminary clustering results are optimized to ensure the compactness and accuracy of the clusters and avoid possible errors in the preliminary clustering process. By calculating the theoretical center of gravity and the maximum distance, the new cluster radius is accurately determined, thereby generating a more reasonable cluster. This optimization process The reliability and accuracy of the clustering results are enhanced, and the efficiency and effect of the monitoring system in processing a large amount of data are further improved, ensuring the accuracy of the final evaluation results; it is worth noting that in actual situations, the distribution of target points may be uneven. In cluster K, the density of target points may be very high at a certain location, resulting in the overall target point density of cluster K being greater than the preset requirement. Target point a may also be due to the relatively dense location, resulting in the target point density of cluster A meeting the requirements of being a cluster. Therefore, by determining the area and target point density of the overlapping area, it is determined whether to update cluster KA. If the overlapping area meets the requirements, a new cluster is regenerated according to the theoretical center of gravity of the overlapping area, thereby improving the compactness of the cluster; finally, by calculating the standard score and setting the threshold, the automatic determination of the fruit growth state corresponding to the cluster center is realized, and the growth status of the fruit is quickly classified. Moreover, in the same cluster, the growth state is determined by the standard score corresponding to the standard data, avoiding the problem of large workload and low efficiency caused by separate analysis. The present invention improves the overall efficiency and accuracy of fruit growth status monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present invention will be further described below in conjunction with the accompanying drawings.

[0031] Figure 1 It is a structural schematic diagram of a fruit growth monitoring system of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] See also Figure 1 As shown, the present invention is a fruit growth monitoring system, including a collection module, a cluster analysis module, a cluster optimization module and a result module, specifically:

[0034] Acquisition module: setting a number of time points at preset time intervals within a preset monitoring period, obtaining the size of the fruit based on the sensor at the time point, recording it as the target value, obtaining target data based on the target value, the target data including the maximum target value A1, the average change speed of the target value A2 and the change amount of the target value A3, and generating target points (A1, A2, A3) in the coordinate system;

[0035] Cluster analysis module: Set the cluster radius r, take the target point k as the center, and calculate the target point density P within the cluster radius r k , when the target point density P k When the density is greater than or equal to the preset density threshold Pys, a cluster K is generated with the target point k as the center; if the target point density corresponding to the target point a exists in the cluster K , generating cluster A with the target point a as the center, and merging it with cluster K to generate a new cluster KA;

[0036] Cluster optimization module: Draw circles k and a with radius r with the target point k and target point a as the center, and obtain the target point density P in the overlapping area DCOI of the circle k and the circle a. coi ;

[0037] When the target point density When the theoretical center of gravity Z of the overlapping area DCOI is obtained, η is the preset correction coefficient, S coi represents the area of ​​the overlapping region DCOI;

[0038] Get the maximum distance d from the theoretical center of gravity Z to the boundary of the overlapping area DCOI, and calculate the target point density within the maximum distance d with the theoretical center of gravity Z as the center. , slowly increase the maximum distance d at a preset speed, and record the target point density corresponding to each increase of the maximum distance, until the corresponding target point density after a certain increase of the maximum distance is less than the target point density after the last increase of the maximum distance, take the theoretical center of gravity as the center, and the maximum distance at this time is the clustering radius, and obtain a new cluster to replace the cluster KA;

[0039] Result module: obtain the target point corresponding to the center of the cluster, use it as the standard point, and judge the growth status of the fruit based on the standard point.

[0040] It should be noted that the automatic acquisition of fruit size data by sensors reduces the time and labor cost of manual collection. Through systematic and automated data collection, high-quality and rich basic data are provided for subsequent cluster analysis, ensuring the comprehensiveness and accuracy of monitoring. Afterwards, by setting the cluster radius and density threshold, a large number of target points are effectively classified into several clusters, significantly reducing the number of fruits that need to be evaluated individually. By identifying high-density areas and grouping fruits with similar growth states into one category, subsequent analysis work is simplified, which helps to reduce the overall workload and improve evaluation efficiency. At the same time, accurate monitoring of fruit growth status is maintained, providing a feasible solution for large-scale orchard management. Then, by dynamically adjusting the cluster radius and recalculating the density of target points in overlapping areas, the preliminary clustering results are optimized to ensure the compactness and accuracy of the clusters and avoid errors that may occur in the preliminary clustering process. By calculating the theoretical center of gravity and the maximum distance, the new cluster radius is accurately determined, thereby generating more reasonable clusters. This optimization process enhances the clustering The reliability and accuracy of the classification results further improve the efficiency and effect of the monitoring system when processing large amounts of data, ensuring the accuracy of the final evaluation results; it is worth noting that in actual situations, the distribution of target points may be uneven. In cluster K, the target point density may be very high in a certain place, resulting in the overall target point density of cluster K being greater than the preset requirement. Target point a may also be in this relatively dense position, resulting in the target point density of cluster A meeting the requirements of being a cluster. Therefore, by determining the area and target point density of the overlapping area, it is determined whether to update cluster KA. If the overlapping area meets the requirements, a new cluster is regenerated according to the theoretical center of gravity of the overlapping area, thereby improving the compactness of the cluster; finally, by calculating the standard score and setting the threshold, the automatic judgment of the fruit growth status corresponding to the cluster center is realized, and the growth status of the fruit is quickly classified. Moreover, in the same cluster, the growth status is judged by the standard score corresponding to the standard data, avoiding the problems of large workload and low efficiency caused by separate analysis.

[0041] In another preferred embodiment of the present invention, the process of determining the growth state of the fruit specifically includes:

[0042] Obtain the target data corresponding to the standard point, use it as the standard data, calculate the standard score y based on the superior and inferior solution distance method, set the standard score thresholds Y1 and Y2, and Y1<Y2;

[0043] When the standard score y<Y1, it is determined that the growth state of the fruit is poor;

[0044] When the standard score Y1≤y<Y2, the growth state of the fruit is judged to be average;

[0045] When the standard score Y2≤y, it is determined that the growth state of the fruit is good.

[0046] It is worth noting that the growth status of the fruit is divided into three levels, from low to high, namely poor, general and good. The growth status of the fruit is determined according to the size of the standard score y and the standard score threshold.

[0047] In another preferred embodiment of the present invention, in the acquisition module, the process of acquiring the target data specifically includes:

[0048] Generate a coordinate point (i, Ni), where Ni represents the target value at time point i, and fit the coordinate point to obtain a fitting curve f(t), where t represents time;

[0049] Calculate the maximum target value ;

[0050] Calculate the average change rate of the target value ,in, , Respectively represent the starting point and end point of the monitoring period;

[0051] Calculate the target value change .

[0052] In another preferred embodiment of the present invention, in the cluster analysis module, the target point density within the cluster radius r is calculated with the target point k as the center. , m represents the number of target points within the cluster radius r with the target point k as the center.

[0053] In another preferred embodiment of the present invention, in the cluster optimization module, the process of obtaining the theoretical center of gravity Z of the overlapping area DCOI specifically includes:

[0054] Get the points inside the overlapping area DCOI The total number of target points on X j , calculate the proportionality coefficient of point j , Indicates the total number of target points in the overlapping area DCOI, and obtains the standard coordinates of point j , calculate the mean of all standard coordinates in each dimension and determine the theoretical center of gravity Z.

[0055] It should be noted that by accurately locating the concentrated points of the data in the overlapping area and ensuring that the calculation of the center of gravity reflects the true distribution of all target points in the area, it helps to optimize the clustering results and make the final clustering clusters more compact and accurate.

[0056] In another preferred embodiment of the present invention, in the acquisition module, when the maximum target value is less than a preset target value threshold, subsequent steps are not executed and it is determined that the growth state of the fruit is poor.

[0057] In another preferred embodiment of the present invention, the target value threshold is determined based on the total growth time of the fruit, and the greater the total growth time, the greater the corresponding target value threshold.

[0058] It is worth noting that the size of fruits is different at different growth stages. Mature fruits are often larger than immature fruits. Therefore, the longer the total growth time (indicating more mature), the larger the corresponding target value threshold.

[0059] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A fruit growth monitoring system, characterized in that: It includes acquisition module, cluster analysis module, cluster optimization module and result module, specifically: Acquisition module: setting a number of time points at preset time intervals within a preset monitoring period, obtaining the size of the fruit based on the sensor at the time point, recording it as the target value, obtaining target data based on the target value, the target data including the maximum target value A1, the average change speed of the target value A2 and the change amount of the target value A3, and generating target points (A1, A2, A3) in the coordinate system; Cluster analysis module: Set the cluster radius r, take the target point k as the center, and calculate the target point density P within the cluster radius r k , when the target point density P k When the density is greater than or equal to the preset density threshold Pys, a cluster K is generated with the target point k as the center; if the target point density corresponding to the target point a exists in the cluster K , generating cluster A with the target point a as the center, and merging it with cluster K to generate a new cluster KA; Cluster optimization module: Draw circles k and a with radius r with the target point k and target point a as the center, and obtain the target point density in the overlapping area DCOI of the circle k and the circle a. ; When the target point density When the theoretical center of gravity Z of the overlapping area DCOI is obtained, η is the preset correction coefficient, S coi represents the area of ​​the overlapping region DCOI; Get the maximum distance d from the theoretical center of gravity Z to the boundary of the overlapping area DCOI, and calculate the target point density within the maximum distance d with the theoretical center of gravity Z as the center. , slowly increase the maximum distance d at a preset speed, and record the target point density corresponding to each increase of the maximum distance, until the corresponding target point density after a certain increase of the maximum distance is less than the target point density after the last increase of the maximum distance, take the theoretical center of gravity as the center, and the maximum distance at this time is the clustering radius, and obtain a new cluster to replace the cluster KA; Result module: obtain the target point corresponding to the center of the cluster, use it as the standard point, and judge the growth status of the fruit based on the standard point.

2. A fruit growth monitoring system according to claim 1, characterized in that: In the result module, the process of judging the growth status of the fruit specifically includes: Obtain the target data corresponding to the standard point, use it as the standard data, calculate the standard score y based on the superior and inferior solution distance method, set the standard score thresholds Y1 and Y2, and Y1<Y2; When the standard score y<Y1, it is determined that the growth state of the fruit is poor; When the standard score Y1≤y<Y2, the growth state of the fruit is judged to be average; When the standard score Y2≤y, it is determined that the growth state of the fruit is good.

3. A fruit growth monitoring system according to claim 1, characterized in that: In the acquisition module, the process of acquiring the target data specifically includes: Generate a coordinate point (i, Ni), where Ni represents the target value at time point i, and fit the coordinate point to obtain a fitting curve f(t), where t represents time; Calculate the maximum target value ; Calculate the average change rate of the target value ,in, , Respectively represent the starting point and end point of the monitoring period; Calculate the target value change .

4. A fruit growth monitoring system according to claim 1, characterized in that: In the cluster analysis module, the target point density within the cluster radius r is calculated with the target point k as the center. , m represents the number of target points within the cluster radius r with the target point k as the center.

5. A fruit growth monitoring system according to claim 1, characterized in that: In the cluster optimization module, the process of obtaining the theoretical center of gravity Z of the overlapping area DCOI specifically includes: Get the points inside the overlapping area DCOI The total number of target points on X j , calculate the proportionality coefficient of point j , Indicates the total number of target points in the overlapping area DCOI, and obtains the standard coordinates of point j , calculate the mean of all standard coordinates in each dimension and determine the theoretical center of gravity Z.

6. A fruit growth monitoring system according to claim 2, characterized in that: In the acquisition module, when the maximum target value is less than a preset target value threshold, subsequent steps are not executed and it is determined that the growth state of the fruit is poor.

7. A fruit growth monitoring system according to claim 6, characterized in that: The target value threshold is determined based on the total growth time of the fruit. The longer the total growth time is, the larger the corresponding target value threshold is.

Images