Intelligent agricultural data service remote supervision system

Abstract

The invention relates to the technical field of intelligent agriculture, in particular to an intelligent agricultural data service remote supervision system, which comprises an agricultural information acquisition module, a growth information monitoring module, a growth analysis module, a data processing module, a sample decision module and a sample acquisition module, a process image is collected in real time, a monomer image of a target plant is separated, the area of the monomer image is accurately recognized, the reference position, the monitoring distance and the collection angle of collection equipment are comprehensively considered, and the edge contour value of a reference image is accurately determined. The sample decision module determines a real-time acquisition area according to the growth density, the edge contour value and the equipment resolution, and then equipment parameters are adjusted accordingly, the parameter setting of the acquisition equipment is continuously optimized according to the density change of crops in different growth stages, the problems of image blurring, key information omission and the like caused by the density change are reduced, and the quality of crops is improved. And the image monitoring accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart agriculture, and particularly to a remote supervision system for smart agriculture data services. Background Art

[0002] Smart agriculture is the advanced stage of modern agricultural development. It integrates a variety of new-generation information technologies such as the Internet of Things, cloud computing, big data, and artificial intelligence, aiming to improve agricultural production efficiency, optimize resource allocation, reduce environmental pollution, and ensure the quality and safety of agricultural products.

[0003] The prior art CN116300608A discloses a smart agriculture remote monitoring system based on big data, including a remote monitoring unit, an information processing unit, an information storage unit, a crop classification and filing unit, a crop evaluation unit, an information display unit, and an alarm unit. The remote monitoring unit is used to comprehensively monitor and collect agricultural environment images in the area through multiple monitoring cameras, collect and process various data in agricultural production using big data of the Internet of Things, and provide corresponding identification, early warning, teaching guidance, and solutions for various problems encountered by farmers in all aspects of agricultural production.

[0004] However, in the current process of remote supervision of smart agriculture, the positions of sensors are fixed. Due to the different growth cycles of agriculture, the density of crops will change during the growth process. At this time, the fixed acquisition area will result in data blind spots in image monitoring, thereby reducing the accuracy of monitoring. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems in the background art, and to propose a remote supervision system for smart agriculture data services.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0007] A remote supervision system for smart agriculture data services, including:

[0008] A growth analysis module, which is used to obtain inertial development information, determine a reference growth image of crops at the current growth stage, separate the crops in the collected process images according to the reference growth image to obtain the target plants and their corresponding individual images, then identify the image areas of n target plant individual images, process the image areas of the n individual images to determine the individual development areas of the target plants, divide the individual development areas by the standard growth areas, and mark the obtained results as growth densities;

[0009] A data processing module, which is used to obtain the device reference position of the acquisition device, measure the monitoring distance according to the device reference position, and at the same time obtain the acquisition angle of the acquisition device, and perform a comprehensive operation on the monitoring distance and the acquisition angle to determine the edge contour value of the reference image;

[0010] A sample decision module, which is used to obtain the growth density and the edge contour value, and at the same time perform a combined calculation on the resolution of the acquisition device, the growth density and the edge contour value to determine the real-time acquisition area;

[0011] A sample acquisition module, which is used to obtain the real-time acquisition area, determine the real-time acquisition ratio based on the real-time acquisition area, and perform real-time adjustment on the device parameters of the acquisition device based on the real-time acquisition ratio. After the parameter adjustment is completed, the sample image of the agricultural growth area is collected.

[0012] As a further solution of the present invention, the process image is collected by the growth information monitoring module. The specific method for collecting the process image includes:

[0013] The growth information monitoring module is used to monitor the growth status of the crops in the agricultural growth area in real time. When the growth information monitoring module detects a pre-acquisition signal, the monitoring device collects the images of the crops in real time, and at the same time marks the collected images as process images. Then the growth information monitoring module transmits the process images to the growth analysis module, where the pre-acquisition signal is generated and transmitted according to the system instructions, and the system instructions include automatic instructions and manual instructions.

[0014] As a further solution of the present invention, the method for obtaining the single-body image includes:

[0015] S1: Taking the process image acquisition time point as the node time, subtracting the planting start time from the node time to obtain the growth interval time;

[0016] Extract the growth parameters, and based on the growth parameters, obtain the inertial development information of the crops at the growth interval time position, where the inertial development information includes the reference growth image at the growth interval time;

[0017] S2: Obtain the process image, use the convolutional neural network algorithm with the reference growth image as the reference image to separate the crops in the process image, and at the same time mark the separated single crop as the target plant, and mark the image of the separated target plant as the single-body image. At this time, several target plants i and the corresponding single-body images Pi are obtained in the process image, where i ∈ [1, n], and n represents the total number of target plants.

[0018] As a further solution of the present invention, the method for obtaining the growth density includes:

[0019] Obtain the monomer image of the target plant i, use the regionprops function in MATLAB software to calculate the image area of the monomer image, and label it as SQi;

[0020] Use the formula Determine the area discrete value LM, where Sp is the area mean value of the image areas of all target plants;

[0021] Compare the area discrete value LM with the discrete threshold Ly. If LM ≤ Ly, directly set the area mean value Sp here as the monomer development area. Conversely, if LM > Ly, calculate the result value of |SQi - Sp| in sequence, and arrange the obtained result values in descending order to get the position sequence combination. Then mark the first data in the position sequence combination as defective data, delete the defective data, and calculate the area discrete value LM for the remaining data according to the above method. Then compare the area discrete value with the discrete threshold. If LM > Ly, mark the second data in the position sequence combination as defective data, and at the same time delete the defective data. Continue to calculate the area discrete value LM for the remaining data, and so on until LM ≤ Ly;

[0022] When LM ≤ Ly, obtain the area threshold Sp in the final area discrete value LM calculation formula, and set this area mean value Sp as the monomer development area;

[0023] Divide the monomer development area by the standard growth area, and mark the obtained result as the growth density, where the standard growth area is a fixed value.

[0024] As a further solution of the present invention, after the monomer development area is determined, identify the number of defective data, divide the number of defective data by the total number of monomer plants in the process image, and mark the obtained result as the abnormal incidence rate. When the abnormal incidence rate is greater than or equal to the occurrence threshold A1, generate a growth abnormal signal at this time and transmit it to the device terminal. When the device terminal detects the growth abnormal signal, generate an audible and visual reminder message correspondingly, so as to give an abnormal reminder to the relevant staff. Conversely, if the abnormal incidence rate is less than the occurrence threshold A1, it means that the target plant is in a normal growth state, and at this time continue to monitor the target plant in real time.

[0025] As a further solution of the present invention, the reference image refers to the image collected by the acquisition device according to the original magnification ratio. The method for determining the edge contour value of the reference image includes:

[0026] SS1: Obtain the location of the acquisition device, and mark the location of the acquisition device as the device reference position. Then collect the monitoring distance between the device reference position and the target plant through the distance detection technology, and mark the monitoring distance as DC;

[0027] SS2: Using the monitoring surface of the acquisition device as a plane, a vertical line is set on the plane. Based on the vertical line, the acquisition angles b1 and b2 of the acquisition device are obtained, where b1 is the horizontal angle of the acquisition device and b2 is the vertical angle of the acquisition device;

[0028] Using the formula and to obtain the edge contour values LB1 and LB2 of the reference image, where LB1 is the horizontal value and LB2 is the vertical value, and tan* is the tangent value in trigonometric functions.

[0029] As a further aspect of the present invention, the method for obtaining the real-time acquisition area includes:

[0030] Obtain the resolution of the acquisition device and mark the resolution as c1×c2, where c1 represents the number of horizontal pixels and c2 represents the number of vertical pixels. Using the formula to obtain the horizontal pixel density X1 and the vertical pixel density X2 respectively;

[0031] Then, based on the formula to obtain the real-time acquisition area FJ, where k is a constant related to the image clarity, and ρ Z is the growth density.

[0032] As a further aspect of the present invention, the method for obtaining the real-time acquisition ratio includes:

[0033] Dividing the real-time acquisition area FJ by the standard area value to obtain the real-time acquisition ratio, where the standard area value is equal to the horizontal value LB1 multiplied by the vertical value LB2.

[0034] As a further aspect of the present invention, it further includes an agricultural information acquisition module for collecting agricultural growth information, where the agricultural growth information includes device information, agricultural growth areas, the crops planted in the areas, and the growth parameters of the corresponding crops. Then, a one-way communication connection is set between the agricultural information acquisition module and the growth analysis module and the data processing module respectively.

[0035] Compared with the existing technology, the advantages of the present invention are:

[0036] The present invention accurately identifies the area of the monomer image by collecting the process image in real time and separating the monomer image of the target plant, and then scientifically calculates the monomer development area and growth density of the target plant. The data processing module comprehensively considers the reference position, monitoring distance, and acquisition angle of the acquisition device to accurately determine the edge contour value of the reference image. The sample decision module determines the real-time acquisition area according to the growth density, edge contour value, and device resolution, and then adjusts the device parameters accordingly, enabling the acquisition device to dynamically adapt to the changes in different growth stages of crops. According to the density changes of crops in different growth stages, the parameter settings of the acquisition device are continuously optimized, reducing problems such as image blurring and omission of key information caused by density changes, and improving the accuracy of image monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic structural diagram of the system of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0039] Referring to Figure 1 , a remote supervision system for intelligent agricultural data services includes an agricultural information collection module, a growth information monitoring module, a growth analysis module, a data processing module, a sample decision module, and a sample collection module;

[0040] The agricultural information collection module is used to collect agricultural growth information, where the agricultural growth information includes device information, agricultural growth areas, the crops planted in the areas, and the growth parameters of the corresponding crops. Further growth parameters include growth rate and growth stage, etc. Then, a one-way communication connection is set between the agricultural information collection module and the growth analysis module and the data processing module respectively;

[0041] The growth information monitoring module is used to monitor the growth status of crops in the agricultural growth area in real time. When the growth information monitoring module detects a pre-acquisition signal, the monitoring device collects the image of the crop in real time, marks the collected image as a process image, and transmits the process image to the growth analysis module. Among them, the pre-acquisition signal is generated and transmitted according to the system instruction. Specifically, the system instruction includes an automatic instruction and a manual instruction;

[0042] It should be further noted that when monitoring the growth of crops, the monitoring device used is a camera. By installing cameras in the agricultural growth area and using video monitoring technology, the growth process of crops in the agricultural growth area is monitored in real time;

[0043] The growth analysis module is used to obtain process images, perform state analysis on the process images, and determine the growth density. The specific analysis process includes:

[0044] S1: Using the process image acquisition time point as the node time, subtract the planting start time from the node time to obtain the growth interval time;

[0045] Extract growth parameters, and based on the growth parameters, obtain the inertial development information of the crops at the position of the growth interval time. Among them, the inertial development information includes the reference growth image at the growth interval time. Further, the reference growth image in the inertial development information is set by those skilled in the art according to big data experience;

[0046] S2: Obtain the process image. Using the reference growth image as the reference image, use the image analysis method to separate the crops in the process image. At the same time, mark the separated single crop as the target plant, and mark the image of the separated target plant as the monomer image. At this time, several target plants i and the corresponding monomer images Pi are obtained in the process image, where i ∈ [1, n], and n represents the total number of target plants;

[0047] Among them, the image analysis method in this embodiment is set as the convolutional neural network algorithm, and the specific processing method of the convolutional neural network algorithm for the process image belongs to the prior art, and will not be elaborated here;

[0048] S3: Obtain the monomer image of the target plant i, use the regionprops function in the MATLAB software to calculate the image area of the monomer image, and mark it as SQi. Among them, the specific operation method of the regionprops function in the MATLAB software belongs to the prior art, and will not be elaborated here;

[0049] After that, use the formula Determine the area discrete value LM, where Sp is the area mean of the image areas of all target plants;

[0050] Compare the discrete area value LM with the discrete threshold Ly. If LM ≤ Ly, directly set the area mean Sp here as the individual development area. Conversely, if LM > Ly, calculate the result value of |SQi - Sp| in sequence, and arrange the obtained result values in descending order to get the position sequence combination. Then mark the first data in the position sequence combination as defective data, delete the defective data, and calculate the discrete area value LM for the remaining data according to the above method. Then compare the discrete area value with the discrete threshold. If LM > Ly, mark the second data in the position sequence combination as defective data, and at the same time delete the defective data. Continue to calculate the discrete area value LM for the remaining data, and so on until LM ≤ Ly. Among them, the specific value of the discrete threshold Ly is set by those skilled in the art according to big data experience;

[0051] When LM ≤ Ly, obtain the area threshold Sp in the formula for the final discrete area value LM, and set this area mean Sp as the individual development area;

[0052] S4: Divide the individual development area by the standard growth area, and mark the obtained result as the growth density. Among them, the standard growth area is a fixed value, and the specific area value of the standard growth area is obtained by those skilled in the art through big data operations;

[0053] In another embodiment of the present invention, after the individual development area is determined, identify the number of defective data, divide the number of defective data by the total number of individual plants in the process image, and mark the obtained result as the abnormal incidence rate. When the abnormal incidence rate is greater than or equal to the occurrence threshold A1, generate a growth abnormal signal at this time and transmit it to the device terminal. When the device terminal detects the growth abnormal signal, generate an audible and visual reminder message correspondingly to remind relevant staff of the abnormality. Conversely, if the abnormal incidence rate is less than the occurrence threshold A1, it means that the target plant is in a normal growth state, and at this time continue to monitor the target plant in real time. Further, the specific value of the occurrence threshold A1 is obtained by those skilled in the art through big data operations;

[0054] Then the growth analysis module transmits the growth density of the target plant to the sample decision module;

[0055] The data processing module is used to obtain agricultural growth information, process the device information based on the device information in the obtained agricultural growth information, and determine the standard area value of the reference image according to the data processing result. The specific method for determining the standard area value includes:

[0056] SS1: Obtain the location of the acquisition device, mark the location of the acquisition device as the device reference position, then collect the monitoring distance between the device reference position and the target plant through distance detection technology, and mark the monitoring distance as DC. Among them, the distance detection technology in this embodiment is set as laser ranging technology, and the specific detection process of the laser ranging technology belongs to the prior art, so it will not be elaborated here;

[0057] SS2: Taking the monitoring plane of the acquisition device as a plane, set a vertical line on the plane. Based on the vertical line, obtain the acquisition angles b1 and b2 of the acquisition device, where b1 is the horizontal angle of the acquisition device and b2 is the vertical angle of the acquisition device. Further, the result obtained by multiplying the horizontal angle b1 by 2 is the horizontal field of view angle of the acquisition device, and the result obtained by multiplying the vertical angle b2 by 2 is the vertical field of view angle of the acquisition device;

[0058] Then use the formula and obtain the edge contour values LB1 and LB2 of the reference image, where LB1 is the horizontal value and LB2 is the vertical value, and tan* is the tangent value in trigonometric functions;

[0059] It should be further noted that the image collected by the acquisition device according to the original magnification ratio is marked as the reference image. At this time, the length and width in the reference image are the horizontal value LB1 and the vertical value LB2, where the original magnification ratio is set as 1 time in this embodiment;

[0060] After that, the data processing module transmits the horizontal value LB1 and the vertical value LB2 of the reference image to the sample decision module;

[0061] The sample decision module is used to receive the growth density of the target plant and the horizontal value LB1 and the vertical value LB2 of the reference image, combine and analyze the growth density with the horizontal value LB1 and the vertical value LB2, and determine the real-time acquisition area. The specific method for determining the real-time acquisition area includes:

[0062] Obtain the resolution of the acquisition device and mark the resolution as c1×c2, where c1 represents the number of horizontal pixels and c2 represents the number of vertical pixels. First, use the formula to obtain the horizontal pixel density X1 and the vertical pixel density X2 respectively;

[0063] Then, based on the formula obtain the real-time acquisition area FJ, where k is a constant related to the image clarity, and the specific value of k is obtained by those skilled in the art through big data operations, and ρ Z is the growth density;

[0064] After that, the sample decision module transmits the real-time acquisition area FJ to the sample acquisition module;

[0065] The sample collection module is used to obtain the real-time collection area FJ, and then divide the real-time collection area FJ by the standard area value to obtain the real-time collection ratio. Among them, the standard area value is equal to the horizontal value LB1 multiplied by the vertical value LB2;

[0066] After that, according to the real-time collection ratio, the device parameters of the collection device are adjusted in real time. After the parameter adjustment is completed, the sample image of the agricultural growth area is collected again.

[0067] The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A remote supervision system for intelligent agricultural data services, characterized in that, Including: A growth analysis module, which is used to obtain inertial development information, determine a reference growth image of crops at the current growth interval time, separate the crops in the collected process images according to the reference growth image to obtain target plants and their corresponding monomer images, then identify the image areas of n target plant monomer images, process the image areas of the n monomer images to determine the monomer development area of the target plants, divide the monomer development area by the standard growth area, and mark the obtained result as the growth density; A data processing module, which is used to obtain the device reference position of the acquisition device, measure the monitoring distance according to the device reference position, and at the same time obtain the acquisition angle of the acquisition device, and perform a comprehensive operation on the monitoring distance and the acquisition angle to determine the edge contour value of the reference image; A sample decision module, which is used to obtain the growth density and the edge contour value, and at the same time combine and calculate the resolution of the acquisition device with the growth density and the edge contour value to determine the real-time acquisition area; A sample acquisition module, which is used to obtain the real-time acquisition area, determine the real-time acquisition ratio based on the real-time acquisition area, and perform real-time adjustment of the device parameters of the acquisition device based on the real-time acquisition ratio. After the parameter adjustment is completed, the sample images of the agricultural growth area are collected.

2. The remote supervision system for intelligent agricultural data services according to claim 1, characterized in that, The process images are collected by the growth information monitoring module. The specific method for collecting process images includes: The growth status of crops in the agricultural growth area is monitored in real time by the growth information monitoring module. When the growth information monitoring module detects a pre-acquisition signal, the monitoring device collects the images of the crops in real time, and at the same time marks the collected images as process images. Then the growth information monitoring module transmits the process images to the growth analysis module, where the pre-acquisition signal is generated and transmitted according to the system instructions, and the system instructions include automatic instructions and manual instructions.

3. The remote supervision system for intelligent agriculture data services according to claim 1, characterized in that The method for obtaining monomer images includes: S1: Using the process image acquisition time point as the node time, subtracting the planting start time from the node time to obtain the growth interval time; Extracting growth parameters, and based on the growth parameters, obtaining the inertial development information of the crops at the growth interval time position, where the inertial development information includes the reference growth image at the growth interval time; S2: Obtaining the process image, using the reference growth image as the reference image, and separating the crops in the process image by using the convolutional neural network algorithm. At the same time, marking the separated single crop as the target plant, and marking the image of the separated target plant as the monomer image. At this time, several target plants i and their corresponding monomer images Pi are obtained in the process image, where i ∈ [1, n], and n represents the total number of target plants.

4. The remote supervision system for smart agriculture data services according to claim 3, characterized in that, The method for obtaining the growth density includes: Obtaining the monomer image of the target plant i, calculating the image area of the monomer image by using the regionprops function in the MATLAB software, and marking it as SQi; Using the formula to determine the area discrete value LM, where Sp is the area mean of the areas of all target plant images; Compare the discrete area value LM with the discrete threshold Ly. If LM ≤ Ly, directly set the area mean Sp here as the monomer development area. Conversely, if LM > Ly, calculate the result value of |SQi - Sp| in sequence, and arrange the obtained result values in descending order to get the position sequence combination. Then mark the first data in the position sequence combination as defective data, delete the defective data, and calculate the discrete area value LM of the remaining data according to the above method. Then compare the discrete area value with the discrete threshold. If LM > Ly, mark the second data in the position sequence combination as defective data, and at the same time delete the defective data. Continue to calculate the discrete area value LM of the remaining data, and so on until LM ≤ Ly; When LM ≤ Ly, obtain the area threshold Sp in the formula for the final discrete area value LM, and set this area mean Sp as the monomer development area; Divide the monomer development area by the standard growth area, and mark the obtained result as the growth density, where the standard growth area is a fixed value.

5. The remote supervision system for intelligent agriculture data services according to claim 4, characterized in that, When the determination of the monomer development area is completed, identify the number of defective data, divide the number of defective data by the total number of monomer plants in the process image, and mark the obtained result as the abnormal incidence rate. When the abnormal incidence rate is greater than or equal to the occurrence threshold A1, generate a growth abnormal signal at this time and transmit it to the device terminal. When the device terminal detects the growth abnormal signal, generate an audible and visual reminder message correspondingly, and then give an abnormal reminder to the relevant staff. Conversely, if the abnormal incidence rate is less than the occurrence threshold A1, it means that the target plant is in a normal growth state, and at this time continue to monitor the target plant in real time.

6. The remote supervision system for smart agriculture data services according to claim 1, wherein The reference image refers to the image collected by the acquisition device according to the original magnification ratio. The method for determining the edge contour value of the reference image includes: SS1: Obtain the location of the acquisition device, mark the location of the acquisition device as the device reference position, and then collect the monitoring distance between the device reference position and the target plant through the distance detection technology, and mark the monitoring distance as DC; SS2: Take the monitoring plane of the acquisition device as the plane, set a vertical line on the plane, and based on the vertical line, obtain the acquisition angles b1 and b2 of the acquisition device, where b1 is the horizontal angle of the acquisition device and b2 is the vertical angle of the acquisition device; Using the formula and obtaining the edge contour values LB1 and LB2 of the reference image, where LB1 is the horizontal value, LB2 is the vertical value, and tan* is the tangent value in trigonometric functions.

7. The remote monitoring system for intelligent agricultural data services according to claim 6, characterized in that, The method for obtaining the real-time acquisition area includes: Obtain the resolution of the acquisition device and mark the resolution as c1×c2, where c1 represents the number of horizontal pixels and c2 represents the number of vertical pixels. Use the formula to obtain the horizontal pixel density X1 and the vertical pixel density X2 respectively; Based on the formula again the real-time acquisition area FJ is obtained, where k is a constant related to the image clarity, and ρ z is the growth density.

8. The remote supervision system for intelligent agricultural data services according to claim 1, characterized in that, The method for obtaining the real-time acquisition ratio includes: Divide the real-time acquisition area FJ by the standard area value to obtain the real-time acquisition ratio, where the standard area value is equal to the horizontal value LB1 multiplied by the vertical value LB2.

9. The remote supervision system for intelligent agriculture data service according to claim 1, characterized in that, It also includes an agricultural information acquisition module for collecting agricultural growth information, where the agricultural growth information includes device information, agricultural growth areas, the crops planted in the areas, and the growth parameters of the corresponding crops. Then set a one-way communication connection between the agricultural information acquisition module and the growth analysis module and the data processing module respectively.

Images