Crop disease and pest monitoring device based on artificial intelligence and control method thereof

Abstract

The invention provides a crop disease and insect pest monitoring device based on artificial intelligence and a control method thereof, and the method comprises the steps: determining a historical monitoring difference sequence of a historical monitoring image set, determining a difference image weight boundary of a target crop according to the historical monitoring difference sequence, and selecting a difference weight threshold image in the difference image weight boundary, determining a monitoring gathering entropy difference base of the difference weight threshold image according to the monitoring difference sequence, repeating the steps, determining the monitoring gathering entropy difference bases of the residual difference weight threshold images in the difference image weight boundary, and determining a pest definition domain according to all the monitoring gathering entropy difference bases, determining a disease and pest decision value between each adjacent disease and pest definition image in the disease and pest definition domain, determining a disease and pest anomaly identification factor of the target crop according to all disease and pest decision values, and when the disease and pest anomaly identification factor is greater than a preset anomaly identification threshold, marking the target crop as disease and pest anomaly. The monitoring rate of plant diseases and insect pests causing crop color change can be improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence monitoring technology, and in particular to an artificial intelligence-based crop pest and disease monitoring device and its control method. Background Technology

[0002] Artificial intelligence monitoring refers to the technology of using artificial intelligence to monitor, analyze and process various information and data in real time. This monitoring can be applied to multiple fields, such as safety, health, environment, and production. The advantages of using artificial intelligence monitoring are that it can process large-scale data, has strong real-time performance, and a high degree of automation, thereby improving the perception and response speed to various situations.

[0003] Current crop pest and disease monitoring mainly relies on manual observation and experience-based judgment, using traditional agricultural techniques and methods to detect and monitor pest and disease conditions in farmland. Traditional methods include manual inspection, yellow sticky traps, and experience-based judgment. Although these methods still have some application in agricultural production, they often require a lot of time for manual monitoring and have certain limitations. Therefore, how to conduct intelligent monitoring of crop pests and diseases has become a challenge for the industry. Summary of the Invention

[0004] Therefore, in order to address the aforementioned technical problems, this application provides an artificial intelligence-based crop pest and disease monitoring device and its control method for intelligent monitoring of crop pests and diseases.

[0005] To solve the above-mentioned technical problems, this application adopts the following technical solution:

[0006] In a first aspect, this application provides a control method for a crop pest and disease monitoring device based on artificial intelligence, comprising the following steps:

[0007] Activate intelligent monitoring for crop diseases and pests, and obtain historical monitoring image sets of target crops from the intelligent monitoring video library;

[0008] The historical monitoring image set is converted into differential image weight boundaries for the target crop;

[0009] Select one differential weight threshold image from the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease boundary domain according to all the monitoring convergence entropy difference basis.

[0010] Determine the pest and disease decision values ​​between each adjacent pest and disease boundary image in the pest and disease boundary domain, and determine the pest and disease anomaly identification factor of the target crop based on all pest and disease decision values;

[0011] When the abnormal disease and pest identification factor is greater than the preset abnormal identification threshold, the target crop is marked as having abnormal disease and pest, and an early warning message is sent to the monitoring center.

[0012] In some embodiments, converting the historical monitoring image set into differential image weight boundaries for the target crop specifically includes:

[0013] Determine the historical monitoring differential sequence of the historical monitoring image set;

[0014] The differential image weight boundaries of the target crop are determined based on the historical monitoring differential sequence.

[0015] In some embodiments, determining the historical monitoring differential sequence of the historical monitoring image set specifically includes:

[0016] Determine the image difference frequency of the historical monitoring image set;

[0017] The historical monitoring image set is differentiated by the image difference frequency to obtain the historical monitoring difference sequence.

[0018] In some embodiments, determining the differential image weight boundaries of the target crop based on the historical monitoring differential sequence specifically includes:

[0019] Determine the differential image weight threshold of the historical monitoring differential sequence;

[0020] The differential weight thresholds are used to determine the corresponding differential weight threshold images in the historical monitoring differential sequence;

[0021] The set of all differential weighted images is used as the differential image weight boundary.

[0022] In some embodiments, determining the monitoring convergence entropy difference basis of the differential weighted image based on the monitoring differential sequence specifically includes:

[0023] Select a historical monitoring image from the monitoring differential sequence, determine the monitoring convergence entropy difference between the historical monitoring image and the differential weight threshold image, repeat the above steps, and determine the monitoring convergence entropy difference between the remaining historical monitoring images and the differential weight threshold image in the monitoring differential sequence.

[0024] The monitoring convergence entropy difference basis of the differential weighted threshold image is determined based on all the monitoring convergence entropy differences.

[0025] In some embodiments, determining the pest and disease boundary based on the entropy difference basis of all monitored clusters specifically includes:

[0026] Select a historical monitoring image from the monitoring differential sequence and obtain the monitoring convergence entropy difference corresponding to that historical monitoring image in each monitoring convergence entropy difference base;

[0027] Based on all the monitoring convergence entropy differences, the historical monitoring image is converged and divided. The above steps are repeated to converge and divide the remaining historical monitoring images in the monitoring difference sequence to obtain multiple pest and disease delimitation entropies.

[0028] The pest and disease boundary domain is determined based on all pest and disease boundary entropies.

[0029] In some embodiments, the method further includes: marking the target crop as normal when the pest and disease abnormality identification factor is less than or equal to a preset abnormality identification threshold.

[0030] Secondly, this application provides an artificial intelligence-based crop pest and disease monitoring device, which includes an intelligent monitoring unit, the intelligent monitoring unit comprising:

[0031] The acquisition module is used to acquire a set of historical monitoring images of the target crop from the intelligent monitoring video library after the intelligent monitoring of crop diseases and pests is started.

[0032] The conversion module is used to convert the historical monitoring image set into differential image weight boundaries of the target crop;

[0033] The domain definition module is used to select a differential weight threshold image in the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease domain based on all the monitoring convergence entropy difference basis.

[0034] An anomaly identification factor determination module is used to determine the pest decision value between each adjacent pest definition image in the pest definition domain, and to determine the pest anomaly identification factor of the target crop based on all pest decision values.

[0035] The pest and disease marking module is used to mark the target crop as having pest and disease anomalies when the pest and disease anomaly identification factor is greater than a preset anomaly identification threshold, and to send an early warning message to the monitoring center.

[0036] Thirdly, this application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the control method of the above-mentioned artificial intelligence-based crop pest and disease monitoring device.

[0037] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the control method for the above-mentioned artificial intelligence-based crop pest and disease monitoring device.

[0038] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0039] The crop pest and disease monitoring device and control method based on artificial intelligence provided in this application helps reduce redundant images by determining the historical monitoring differential sequence of the historical monitoring image set. Based on the historical monitoring differential sequence, the differential image weight boundary of the target crop is determined, and then the monitoring convergence entropy difference basis is determined. The monitoring convergence entropy difference basis represents the set of all monitoring convergence entropy differences. The monitoring convergence entropy differences in the monitoring convergence entropy difference basis reflect the degree of change between the historical monitoring image and the differential weight threshold image, more objectively reflecting the difference between the historical monitoring image and the differential weight threshold image. Thus, all monitoring convergence entropy difference bases determine the pest and disease anomaly identification factors of the target crop. The pest and disease anomaly identification factors reflect the degree of abnormal changes in the monitoring images of the target crop. Finally, the target crop is monitored for anomalies based on the pest and disease anomaly identification factors. Compared with the prior art, which requires a lot of time for manual monitoring of the target crop, this method achieves intelligent monitoring of pests and diseases that cause color changes in crops. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating the control method of an artificial intelligence-based crop pest and disease monitoring device in some embodiments of this application;

[0041] Figure 2 This is a flowchart illustrating the process of determining pest and disease decision values ​​in some embodiments of this application;

[0042] Figure 3 This is a structural block diagram of the intelligent monitoring unit in some embodiments of this application;

[0043] Figure 4 This is a diagram showing the internal structure of a computer device in some embodiments of this application. Detailed Implementation

[0044] The core of this application is to initiate intelligent monitoring for crop disease and pest identification. This involves acquiring a historical monitoring image set of the target crop from an intelligent monitoring video library, determining the historical monitoring differential sequence of the historical monitoring image set, determining the differential image weight boundaries of the target crop based on the historical monitoring differential sequence, selecting a differential weight threshold image from the differential image weight boundaries, and determining the monitoring convergence entropy difference basis of this differential weight threshold image based on the monitoring differential sequence. This process is repeated to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundaries. Finally, based on all the monitoring convergence entropy difference bases, the disease... The pest and disease boundary domain is defined by determining the pest and disease decision values ​​between adjacent pest and disease boundary images within the boundary domain. Based on all the pest and disease decision values, an anomaly identification factor for the target crop is determined. When the anomaly identification factor exceeds a preset anomaly identification threshold, the target crop is marked as having an anomaly, and an early warning message is sent to the monitoring center. Compared with the existing technology, which requires a lot of time for manual monitoring of the target crop, this scheme achieves intelligent monitoring of crop pests and diseases and can improve the monitoring rate of pests and diseases that cause changes in crop color.

[0045] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific implementation methods. (Reference) Figure 1 The figure is an exemplary flowchart of a control method for an artificial intelligence-based crop pest and disease monitoring device according to some embodiments of this application. The control method 100 for the artificial intelligence-based crop pest and disease monitoring device mainly includes the following steps:

[0046] In step 101, the intelligent monitoring for identifying crop diseases and pests is activated, and the historical monitoring image set of the target crop is obtained from the intelligent monitoring video library.

[0047] In specific implementation, after starting the intelligent monitoring of crop diseases and pests, the historical monitoring image set of the target crop is obtained from the intelligent monitoring video library. The historical monitoring image set is a collection of all historical monitoring images, and the historical monitoring images in the historical monitoring image set are each frame of the historical monitoring video of the target crop obtained from the intelligent monitoring video library.

[0048] It should be noted that in this application, the target crops can be photographed in real time using a high-definition camera, and the captured videos can be stored in the intelligent monitoring video library as historical monitoring videos.

[0049] It should be noted that the crop pests and diseases mentioned in this application refer to those that can cause changes in the color of crops. The crops mentioned in this application refer to rice seedlings, and the pest and disease is the red-spotted black dung beetle. When rice seedlings are affected by the red-spotted black dung beetle, the beetle will destroy the chlorophyll of the seedlings, thereby causing the leaves of the seedlings to change color from green to red. In other embodiments, it may also be a spider mite, rice leaf roller, or other pests and diseases that can cause changes in the color of crops, which will not be elaborated here.

[0050] In step 102, the historical monitoring image set is converted into differential image weight boundaries of the target crop.

[0051] In some embodiments, converting the historical monitoring image set into differential image weight boundaries for the target crop can be achieved using the following steps:

[0052] Determine the historical monitoring differential sequence of the historical monitoring image set;

[0053] The differential image weight boundaries of the target crop are determined based on the historical monitoring differential sequence.

[0054] In some embodiments, determining the historical monitoring differential sequence of the historical monitoring image set can be achieved using the following steps:

[0055] Determine the image difference frequency of the historical monitoring image set;

[0056] The historical monitoring image set is differentiated by the image difference frequency to obtain the historical monitoring difference sequence.

[0057] In specific implementation, the image difference frequency of the historical monitoring image set is determined. For example, the image difference frequency of the historical monitoring image set can be set based on historical experimental experience, generally set to be collected once per hour, without limitation here. The historical monitoring image set is differentiated by the image difference frequency to obtain the historical monitoring difference sequence. That is, the historical monitoring image set is collected from beginning to end according to the set difference frequency, and the collected historical monitoring images are sorted according to the order of collection time, and the sorted sequence is used as the historical monitoring difference sequence.

[0058] It should be noted that the historical monitoring differential sequence in this application refers to the image sequence acquired according to the image differential frequency. The image differential frequency reflects the parameter value of the sparsity of the historical monitoring image set. The larger the image differential frequency, the greater the sparsity of the historical monitoring image set, and the smaller the image differential frequency, the smaller the sparsity of the historical monitoring image set.

[0059] In some embodiments, determining the differential image weight boundaries of the target crop based on the historical monitoring differential sequence can be achieved using the following steps:

[0060] Determine the differential image weight threshold of the historical monitoring differential sequence;

[0061] The differential weight thresholds are used to determine the corresponding differential weight threshold images in the historical monitoring differential sequence;

[0062] The set of all differential weighted images is used as the differential image weight boundary.

[0063] In specific implementation, the differential weight threshold images in the historical monitoring differential sequence are determined by the differential image weight threshold number. That is, the historical monitoring images in the historical monitoring differential sequence are divided according to the differential image weight threshold number, and all historical monitoring images corresponding to the division points are used as differential weight threshold images. For example, if the total number of historical monitoring images in the historical monitoring differential sequence is 30 and the differential image weight threshold number is 5, then the 5th, 10th, 15th, 20th and 25th images in the historical monitoring differential sequence are used as division points, that is, the 5th, 10th, 15th, 20th and 25th historical monitoring images in the historical monitoring differential sequence are used as differential weight threshold images.

[0064] It should be noted that the differential image weight boundary in this application reflects the boundary of differential weight threshold image acquisition. The larger the differential image weight boundary, the more differential weight threshold images are contained in the differential image weight boundary. The smaller the differential image weight boundary, the fewer differential weight threshold images are contained in the differential image weight boundary.

[0065] In step 103, select one differential weight threshold image from the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease boundary domain according to all the monitoring convergence entropy difference bases.

[0066] In some embodiments, determining the monitoring convergence entropy difference basis of the differential weighted image based on the monitoring differential sequence can be achieved through the following steps:

[0067] Select a historical monitoring image from the monitoring differential sequence, determine the monitoring convergence entropy difference between the historical monitoring image and the differential weight threshold image, repeat the above steps, and determine the monitoring convergence entropy difference between the remaining historical monitoring images and the differential weight threshold image in the monitoring differential sequence.

[0068] The monitoring convergence entropy difference basis of the differential weighted threshold image is determined based on all the monitoring convergence entropy differences.

[0069] In specific implementation, the above steps are repeated, namely: the step of determining the monitoring convergence entropy difference between the remaining historical monitoring images and the differential weighted threshold image in the monitoring differential sequence is repeated to obtain the monitoring convergence entropy difference between the remaining historical monitoring images and the differential weighted threshold image in the monitoring differential sequence; the monitoring convergence entropy difference basis of the differential weighted threshold image is determined based on all the monitoring convergence entropy differences, that is: the set of all monitoring convergence entropy differences is used as the monitoring convergence entropy difference basis of the differential weighted threshold image.

[0070] In some embodiments, determining the monitoring convergence entropy difference between the historical monitoring image and the differential weighted threshold image can be achieved using the following steps:

[0071] Determine the influencing factors of convergence entropy difference of target crops ;

[0072] Determine the total number of monitored independent pixel values ​​in this historical monitoring image. ;

[0073] Determine the first [item] in this historical monitoring image. The monitoring independent display value of each monitoring independent pixel value in this historical monitoring image ;

[0074] Determine the total number of independent pixel values ​​with weight thresholds in the differentially weighted image. ;

[0075] Determine the first element in the differentially weighted threshold image. The independent display value of each weighted pixel value in this difference weighted image. ;

[0076] Based on the aggregation entropy difference influence factor of the target crop The total number of monitored independent pixel values ​​in the historical monitoring image. The first one in the historical monitoring image The monitoring independent display value of each monitoring independent pixel value in this historical monitoring image The total number of independent pixel values ​​with weight thresholds in the differentially weighted image. And the determination of the first differentially weighted image The independent display value of each weighted pixel value in this difference weighted image. The monitoring convergence entropy difference between the historical monitoring image and the differential weighted threshold image is determined, wherein the monitoring convergence entropy difference can be determined using the following formula:

[0077]

[0078] in, This represents the difference in monitoring convergence entropy between the historical monitoring image and the differential weighted threshold image. Let e ​​represent the logarithmic function with base e. , .

[0079] In practice, all pixel values ​​in the historical monitoring image are acquired, and all pixel values ​​are classified according to their values. Each category of pixel values ​​is treated as a monitoring independent pixel value. The proportion of the total number of monitoring independent pixel values ​​to the total number of pixel values ​​in the historical monitoring image is used as the monitoring independent display value of the corresponding weighted independent pixel value. For example, if all pixel values ​​in the historical monitoring image are 2, 4, 3, 3, 6, 3, 6, 3, 2, 2, then there are four categories: one with pixel value 2, one with pixel value 3, one with pixel value 4, and one with pixel value 6. Values ​​2, 3, 4, and 6 are all treated as... The independent pixel values ​​are monitored, where the independent display value of independent pixel value 2 is 0.3, the independent display value of independent pixel value 3 is 0.4, the independent display value of independent pixel value 4 is 0.1, and the independent display value of independent pixel value 6 is 0.2. Similarly, all pixel values ​​in the differential weighted image are obtained, and all pixel values ​​are classified according to their values. Each class of pixel values ​​is used as the weighted independent pixel value. The proportion of the total number of pixel values ​​in the weighted independent pixel value to the total number of pixel values ​​in the differential weighted image is used as the weighted independent display value of the corresponding weighted independent pixel value.

[0080] It should be noted that the convergence entropy difference influence factor in this application reflects the degree of influence of environmental factors on the changes between historical monitoring images and differential weighted images. In some embodiments, multiple experts can score the impact of environmental factors on the target crop, and the convergence entropy difference influence factor can be set according to the scores. The convergence entropy difference influence factor generally takes a value range of 0.5-1. In other embodiments, other methods can also be used to set the convergence entropy difference influence factor, which is not limited here.

[0081] It should be noted that the monitoring convergence entropy difference basis in this application represents the set of all monitoring convergence entropy differences. The monitoring convergence entropy differences in the monitoring convergence entropy difference basis reflect the degree of change between historical monitoring images and differential weighted threshold images. The larger the monitoring convergence entropy difference, the greater the degree of change between historical monitoring images and differential weighted threshold images. The smaller the monitoring convergence entropy difference, the smaller the degree of change between historical monitoring images and differential weighted threshold images.

[0082] In some embodiments, determining the pest and disease boundary based on all monitored convergence entropy differences can be achieved using the following steps:

[0083] Select a historical monitoring image from the monitoring differential sequence and obtain the monitoring convergence entropy difference corresponding to that historical monitoring image in each monitoring convergence entropy difference base;

[0084] Based on all the monitoring convergence entropy differences, the historical monitoring image is converged and divided. The above steps are repeated to converge and divide the remaining historical monitoring images in the monitoring difference sequence to obtain multiple pest and disease delimitation entropies.

[0085] The pest and disease boundary domain is determined based on all pest and disease boundary entropies.

[0086] In specific implementation, the historical monitoring image is merged and divided according to all the monitoring merged entropy differences. That is, all the monitoring merged entropy differences of the historical monitoring image are compared to obtain the minimum monitoring merged entropy difference. The historical monitoring image and the differential weighted threshold image corresponding to the minimum monitoring merged entropy difference are combined to form a pest and disease definition entropy difference set. The above steps are repeated to merge and divide the remaining historical monitoring images in the monitoring differential sequence to obtain multiple differential weighted threshold images corresponding to pest and disease definition entropy difference sets. The average value of all monitoring merged entropy differences in each pest and disease definition entropy difference set is taken as the pest and disease definition entropy of the corresponding pest and disease definition entropy difference set. The pest and disease definition domain is determined according to all the pest and disease definition entropies. That is, a pest and disease definition entropy is selected, and the historical monitoring image corresponding to the monitoring merged entropy difference with the value of the pest and disease definition entropy is taken as the pest and disease definition image of the pest and disease definition entropy. The above steps are repeated to determine the pest and disease definition images of the remaining pest and disease definition entropies. All pest and disease definition images are combined to form the pest and disease definition domain.

[0087] It should be noted that the pest and disease delimitation domain in this application represents the set of all pest and disease delimitation images. Each pest and disease delimitation image in the pest and disease delimitation domain can reflect the feature elements contained in different pest and disease delimitation entropy difference sets. The pest and disease delimitation entropy reflects the parameter value of the disorder of the pixel values ​​in the corresponding pest and disease delimitation entropy difference set. The larger the pest and disease delimitation entropy, the more disordered the pixel values ​​in the corresponding pest and disease delimitation entropy difference set. The smaller the pest and disease delimitation entropy, the more stable the pixel values ​​in the corresponding pest and disease delimitation entropy difference set.

[0088] In step 104, the pest decision value between each adjacent pest definition image in the pest definition domain is determined, and the pest anomaly identification factor of the target crop is determined based on all the pest decision values.

[0089] Additionally, in some embodiments, references Figure 2 As shown in the figure, this is a flowchart illustrating the process of determining pest and disease decision values ​​in some embodiments of this application. In this embodiment, determining pest and disease decision values ​​can be achieved through the following steps:

[0090] First, in step 1041, each pest and disease delineation image in the pest and disease delineation domain is acquired;

[0091] Secondly, in step 1042, all the disease and pest definition images are sorted according to the acquisition time of the corresponding differential weight threshold images, and the sorted image set is used as the disease and pest definition sequence.

[0092] Finally, in step 1043, the pest decision value between each adjacent pest definition image is determined according to the pest definition sequence.

[0093] It should be noted that the pest and disease decision value in this application reflects the degree of difference between adjacent pest and disease delineation images. The larger the pest and disease decision value, the greater the difference between the corresponding adjacent pest and disease delineation images. In some embodiments, the pest and disease decision value can be represented by the ratio of the pest and disease delineation entropy of the pest and disease delineation image with an earlier collection time to the pest and disease delineation entropy of the pest and disease delineation image with a later collection time. Adjacent pest and disease delineation images refer to two adjacent pest and disease delineation images in the pest and disease delineation sequence, which will not be elaborated here.

[0094] In some embodiments, determining the disease and pest anomaly identification factors of a target crop based on all disease and pest decision values ​​can be achieved through the following steps:

[0095] All pest and disease decision values ​​are sorted according to the order of the corresponding adjacent pest and disease boundary images in the pest and disease boundary domain to obtain the pest and disease decision sequence.

[0096] Obtain the total number of pest and disease decision values ​​in the pest and disease decision sequence. ;

[0097] Obtain the first [number] in the pest and disease decision sequence Individual pest and disease decision value ;

[0098] Obtain the first [number] in the pest and disease decision sequence Individual pest and disease decision value ;

[0099] Determine the first in the pest and disease decision sequence The decision value for pest and disease control and the first Decision correction coefficient between individual pest and disease decision values ;

[0100] Based on the total number of pest and disease decision values ​​in the pest and disease decision sequence The first in the pest and disease decision sequence Individual pest and disease decision value The first in the pest and disease decision sequence Individual pest and disease decision value and the first in the pest and disease decision sequence The decision value for pest and disease control and the first Decision correction coefficient between individual pest and disease decision values Identify the disease and pest anomaly detection factors for the target crop, wherein the disease and pest anomaly detection factors can be determined using the following formula:

[0101]

[0102] in, Factors indicating abnormal identification of pests and diseases. .

[0103] It should be noted that the decision correction coefficient in this application reflects the degree of influence of the natural growth of the target crop in a normal environment on the decision values ​​of adjacent pests and diseases. The larger the decision correction coefficient, the greater the influence of the natural growth of the target crop in a normal environment on the decision values ​​of adjacent pests and diseases. The smaller the decision correction coefficient, the smaller the influence of the natural growth of the target crop in a normal environment on the decision values ​​of adjacent pests and diseases. In some embodiments, the decision correction coefficient can be set according to previous experimental data. The decision correction coefficient generally takes the value range of 0-1. In other embodiments, other methods can also be used to set it, which is not limited here.

[0104] It should be noted that the pest and disease anomaly identification factor in this application reflects the degree of abnormal changes in the monitoring image of the target crop. The larger the pest and disease anomaly identification factor, the greater the abnormal changes in the target crop; the smaller the pest and disease anomaly identification factor, the smaller the abnormal changes in the target crop.

[0105] It should be noted that when the seedlings are growing normally, the color changes between the images collected during the normal growth process follow a certain pattern, changing from light green to dark green. However, when the seedlings are infested with the red-spotted black dung beetle, the beetle damages the chlorophyll of the seedlings, causing an abnormal color change between the images collected during the seedling growth process, changing from light green to red. This application calculates the changes in pixel values ​​during the seedling growth process to obtain the seedling disease and pest anomaly identification factor, and determines whether the seedlings are infested with the red-spotted black dung beetle based on this disease and pest anomaly identification factor.

[0106] In step 105, when the abnormal identification factor of pests and diseases is greater than the preset abnormal identification threshold, the target crop is marked as having abnormal pests and diseases, and an early warning message is sent to the monitoring center.

[0107] In addition, in some embodiments, when the pest and disease abnormality identification factor is less than or equal to a preset abnormality identification threshold, the target crop is marked as normal.

[0108] In practice, early warning information is sent to the monitoring center. Specifically, the early warning level of the early warning signal can be set based on the value of the abnormal identification factor of pests and diseases being greater than the preset abnormal identification threshold, and the early warning level is sent to the monitoring center. This will not be elaborated further here.

[0109] It should be noted that in this application, the abnormal identification threshold can be set by using historical abnormal identification factors of pests and diseases. Generally, it is the average value of all abnormal identification factors of pests and diseases in the previous year's experiment. Other methods can also be used to set it in other embodiments, which are not limited here.

[0110] It should be noted that in this application, the historical monitoring differential sequence is determined by using a set of historical monitoring images of the target crop, which reduces redundant historical monitoring images and helps to reduce the processing time of historical monitoring images. The monitoring convergence entropy difference of each historical monitoring image is calculated. The monitoring convergence entropy difference reflects the degree of change between the historical monitoring image and the differential weighted threshold image. Then, similar historical monitoring images in the historical monitoring image set are classified according to the monitoring convergence entropy difference. Thus, according to the above implementation steps, a specific historical monitoring image is used to represent a class of historical monitoring images obtained from the corresponding classification, which further reduces the processing time of historical monitoring images. Finally, the anomaly identification factor of pests and diseases is determined based on the degree of change between each historical monitoring image, and the anomaly identification factor of pests and diseases is compared with the normal threshold to determine whether the target crop has been affected by pests and diseases.

[0111] Furthermore, in another aspect of this application, in some embodiments, this application provides an artificial intelligence-based crop pest and disease monitoring device, which includes an intelligent monitoring unit, referencing... Figure 3 The figure is a schematic diagram of exemplary hardware and / or software of an intelligent monitoring unit according to some embodiments of this application. The intelligent monitoring unit 300 includes: an acquisition module 301, a conversion module 302, a boundary domain determination module 303, an anomaly identification factor determination module 304, and a pest and disease marking module 305, which are described below:

[0112] The acquisition module 301 in this application is mainly used to acquire the historical monitoring image set of the target crop from the intelligent monitoring video library after the intelligent monitoring of crop pests and diseases is started.

[0113] The conversion module 302 in this application is mainly used to determine the historical monitoring differential sequence of the historical monitoring image set, and to determine the differential image weight boundary of the target crop based on the historical monitoring differential sequence.

[0114] The domain determination module 303 in this application is mainly used to select a differential weight threshold image in the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease boundary domain according to all the monitoring convergence entropy difference bases.

[0115] The anomaly identification factor determination module 304 in this application is mainly used to determine the pest decision value between each adjacent pest definition image in the pest definition domain, and to determine the pest anomaly identification factor of the target crop based on all the pest decision values.

[0116] The pest and disease marking module 305 in this application is mainly used to mark the target crop as having a pest and disease abnormality when the abnormality identification factor of the pest and disease is greater than the preset abnormality identification threshold, and to send early warning information to the monitoring center.

[0117] Each module in the aforementioned intelligent monitoring unit can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0118] In another embodiment, this application provides a computer device, which may be a server, and its internal structure diagram may be as follows. Figure 4 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores control data for an AI-based crop pest and disease monitoring device. The network interface communicates with external terminals via a network. When the computer program is executed by the processor, it implements a control method for an AI-based crop pest and disease monitoring device.

[0119] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0120] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described embodiment of the control method for the artificial intelligence-based crop pest and disease monitoring device.

[0121] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps described in the control method embodiment of the artificial intelligence-based crop pest and disease monitoring device.

[0122] In one embodiment, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps described in the control method embodiment of the artificial intelligence-based crop pest and disease monitoring device.

[0123] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0124] In summary, the crop pest and disease monitoring device and control method based on artificial intelligence disclosed in this application firstly initiates intelligent monitoring for crop pest and disease identification, obtains a set of historical monitoring images of the target crop from the intelligent monitoring video library, determines the historical monitoring differential sequence of the historical monitoring image set, determines the differential image weight boundary of the target crop based on the historical monitoring differential sequence, selects a differential weight threshold image from the differential image weight boundary, determines the monitoring convergence entropy difference basis of the differential weight threshold image based on the monitoring differential sequence, and repeats the above steps to determine the monitoring convergence entropy difference of the remaining differential weight threshold images in the differential image weight boundary. Based on the entropy difference of all monitored data, a pest and disease boundary region is determined. Then, a pest and disease decision value is determined between adjacent pest and disease boundary images within the boundary region. Based on all pest and disease decision values, an anomaly identification factor for the target crop is determined. When the anomaly identification factor exceeds a preset anomaly identification threshold, the target crop is marked as having a pest and disease anomaly, and an early warning message is sent to the monitoring center. Therefore, compared to existing technologies that require extensive manual monitoring of target crops, this scheme achieves intelligent monitoring of crop pests and diseases, and can improve the monitoring rate of pests and diseases that cause color changes in crops.

[0125] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0126] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A control method for a crop pest and disease monitoring device based on artificial intelligence, characterized in that, Includes the following steps: Activate intelligent monitoring for crop diseases and pests, and obtain historical monitoring image sets of target crops from the intelligent monitoring video library; The historical monitoring image set is converted into differential image weight boundaries for the target crop; Select one differential weight threshold image from the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease boundary domain according to all the monitoring convergence entropy difference basis. Determine the pest and disease decision values ​​between each adjacent pest and disease boundary image in the pest and disease boundary domain, and determine the pest and disease anomaly identification factor of the target crop based on all pest and disease decision values; When the abnormal disease and pest identification factor is greater than the preset abnormal identification threshold, the target crop is marked as having abnormal disease and pest, and an early warning message is sent to the monitoring center.

2. The method as described in claim 1, characterized in that, Converting the historical monitoring image set into differential image weight boundaries for the target crop specifically includes: Determine the historical monitoring differential sequence of the historical monitoring image set; The differential image weight boundaries of the target crop are determined based on the historical monitoring differential sequence.

3. The method as described in claim 2, characterized in that, Determining the historical monitoring differential sequence of the historical monitoring image set specifically includes: Determine the image difference frequency of the historical monitoring image set; The historical monitoring image set is differentiated by the image difference frequency to obtain the historical monitoring difference sequence.

4. The method as described in claim 2, characterized in that, Determining the differential image weight boundary of the target crop based on the historical monitoring differential sequence specifically includes: Determine the differential image weight threshold of the historical monitoring differential sequence; The differential weight thresholds are used to determine the corresponding differential weight threshold images in the historical monitoring differential sequence; The set of all differential weighted images is used as the differential image weight boundary.

5. The method as described in claim 1, characterized in that, Determining the monitoring convergence entropy difference basis of the differential weighted image based on the monitoring differential sequence specifically includes: Select a historical monitoring image from the monitoring differential sequence, determine the monitoring convergence entropy difference between the historical monitoring image and the differential weight threshold image, repeat the above steps, and determine the monitoring convergence entropy difference between the remaining historical monitoring images and the differential weight threshold image in the monitoring differential sequence. The monitoring convergence entropy difference basis of the differential weighted threshold image is determined based on all the monitoring convergence entropy differences.

6. The method as described in claim 1, characterized in that, The pest and disease delimitation domain is determined based on the entropy difference of all monitored data, specifically including: Select a historical monitoring image from the monitoring differential sequence and obtain the monitoring convergence entropy difference corresponding to that historical monitoring image in each monitoring convergence entropy difference base; Based on all the monitoring convergence entropy differences, the historical monitoring image is converged and divided. The above steps are repeated to converge and divide the remaining historical monitoring images in the monitoring difference sequence to obtain multiple pest and disease delimitation entropies. The pest and disease boundary domain is determined based on all pest and disease boundary entropies.

7. The method as described in claim 1, characterized in that, Also includes: When the abnormal identification factor of the pests and diseases is less than or equal to the preset abnormal identification threshold, the target crop is marked as normal.

8. An artificial intelligence-based crop pest and disease monitoring device, characterized in that, It includes an intelligent monitoring unit, which comprises: The acquisition module is used to acquire a set of historical monitoring images of the target crop from the intelligent monitoring video library after the intelligent monitoring of crop diseases and pests is started. The conversion module is used to convert the historical monitoring image set into differential image weight boundaries of the target crop; The domain definition module is used to select a differential weight threshold image in the differential image weight boundary, determine the monitoring convergence entropy difference basis of the differential weight threshold image according to the monitoring differential sequence, repeat the above steps to determine the monitoring convergence entropy difference basis of the remaining differential weight threshold images in the differential image weight boundary, and determine the pest and disease domain based on all the monitoring convergence entropy difference basis. An anomaly identification factor determination module is used to determine the pest decision value between each adjacent pest definition image in the pest definition domain, and to determine the pest anomaly identification factor of the target crop based on all pest decision values. The pest and disease marking module is used to mark the target crop as having pest and disease anomalies when the pest and disease anomaly identification factor is greater than a preset anomaly identification threshold, and to send an early warning message to the monitoring center.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the control method for the crop disease and pest monitoring device based on artificial intelligence as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the control method for the artificial intelligence-based crop pest and disease monitoring device as described in any one of claims 1 to 7.

Images