Intelligent Monitoring System for Codling Moth Based on Multi-Source Trapping Data Fusion

Abstract

This invention discloses an intelligent monitoring system for codling moths of apples based on multi-source trapping data fusion, belonging to the field of trapping data monitoring technology. It addresses the problem of delayed monitoring strategy adjustments and increased probability of missing the optimal window for pest control. By setting multiple trapping points in the orchard under test, image sensors collect and identify codling moth trapping images at each point. Trapping points are selected and marked based on trapping efficiency. The marked trapping points and their trapping growth rates are input into a spatial diagnosis module, and spatial weights are calculated based on the distance between the marked trapping points and neighboring points. The spatial differences in trapping growth rates are analyzed to determine if local anomalies exist. An active feature module identifies the active periods of codling moths at anomaly points, detects the transpiration intensity of fruit trees and environmental noise intensity during the detection period, generates active features, and retrieves fruit tree density from the orchard database. The environmental impact coefficient is evaluated based on the active features to determine whether to adjust the trapping method, thereby improving the timeliness of response to pest anomalies.

Description

Technical Field

[0001] This invention relates to the field of trap data monitoring technology, and more specifically, to an intelligent monitoring system for codling moths based on multi-source trap data fusion. Background Technology

[0002] The codling moth is a common and highly damaging pest in orchards. Its larvae have a strong ability to penetrate and damage apple fruits and branches, seriously affecting apple yield and quality. In current orchard pest and disease control technologies, trapping and monitoring methods are often used to identify and manage the population dynamics of the codling moth. These methods mainly include hanging traps, using sex pheromone attractants, and recording the trapping results through image acquisition equipment.

[0003] The existing technology has the following shortcomings:

[0004] Currently, codling moth trapping methods mainly rely on single-point data, depending on changes in the number of traps to assess pest infestation. However, due to the complexity of orchard environmental factors, such as tree density in different areas, tree physiological state, slight climate differences, and environmental noise, the accuracy and representativeness of trapping results can be affected. Traditional methods struggle to achieve dynamic and accurate assessment of the overall pest situation in the orchard, leading to delayed adjustments in monitoring strategies and an increased probability of missing the optimal window for pest control. Therefore, this paper proposes an intelligent monitoring system for codling moths of apples based on the fusion of multi-source trapping data.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent monitoring system for codling moths based on multi-source trapping data fusion. This system addresses the problems mentioned in the background art by employing a multimodal analysis mechanism that integrates image recognition, spatial weight modeling, physiological behavior perception, and environmental factor assessment.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an intelligent monitoring system for codling moths based on multi-source trapping data fusion, comprising a trapping identification module, a spatial diagnosis module, an active feature module, and a strategy adjustment module, the functions of each module being as follows:

[0008] The trapping and identification module sets up multiple trapping points in the orchard to be tested and uses the same trapping method. It monitors the trapping images of codling moths at each trapping point through an image sensor and identifies the trapping data. Based on the trapping data, it evaluates the trapping efficiency and trapping growth rate of the trapping points. It uses the trapping efficiency to screen and mark the trapping points, and then inputs the marked trapping points and the trapping growth rate into the spatial diagnosis module.

[0009] The spatial diagnostic module detects the distance between the marked trap point and neighboring trap points and calculates the spatial weight of the marked trap point. It receives the trapping growth rate of the marked trap point and analyzes the spatial differences based on the spatial weight. Based on the spatial differences, it determines whether the marked trap point has local anomalies.

[0010] After a local anomaly occurs at the marked trapping point, the active feature module obtains the active period of the bark beetle. During the active period of the bark beetle, it detects the tree transpiration intensity and environmental noise intensity at the marked trapping point. Based on the tree transpiration intensity and environmental noise level, it generates the active features of the bark beetle at the marked trapping point and transmits them to the strategy adjustment module.

[0011] The strategy adjustment module retrieves the density of fruit trees at the marked trapping points from the orchard database, receives the activity characteristics of bark moths, and assesses the environmental impact coefficient based on the orchard density. Based on the environmental impact coefficient, it determines whether to change the trapping method.

[0012] In a preferred embodiment, in the trapping and identification module, multiple trapping points are deployed in the orchard area to be tested, and images of the moth trapping are collected by an image sensor at fixed time intervals.

[0013] After preprocessing the images of moth-trapping, the images are then subjected to target detection and classification using an image recognition algorithm. The number of moths in each frame of the images is identified, and the average number of moths trapped at each trapping point within a fixed time period is calculated as the trapping data.

[0014] The current time period of each trapping point is compared with the trapping data of the previous time period to calculate the trapping growth rate.

[0015] In a preferred embodiment, the trapping identification module calculates the average number of traps per unit time within a set time window as the trapping efficiency.

[0016] If the trapping efficiency is less than the efficiency discrimination threshold, the trapping point is judged to be of substandard trapping efficiency;

[0017] Conversely, if the trapping efficiency is not met, it is considered that the trapping efficiency has met the standard.

[0018] The trapping sites that are deemed to have substandard trapping efficiency are identified and marked.

[0019] In a preferred embodiment, the spatial diagnostic module receives marked trap points output from the trap identification module and extracts a set of neighboring trap points based on the geographical coordinates of each marked trap point in two-dimensional space.

[0020] Based on the spatial positional relationship of each trapping point in the neighborhood trapping point set, the spatial weight of each marked trapping point is calculated;

[0021] After obtaining the spatial weights, the local spatial deviation is defined by combining the trapping growth rate of the marked trapping points input from the trapping recognition module.

[0022] When the local spatial deviation exceeds the preset spatial difference threshold, it is determined that a local anomaly has occurred at the marked trapping point;

[0023] Conversely, if no local anomalies are found at the marked trapping point, it is determined that no such anomalies have occurred.

[0024] In a preferred embodiment, in the active feature module, the time period corresponding to the maximum trapping growth rate of the marked trapping point where local anomalies occur is taken as the active period of the bark beetle;

[0025] Centered on the marked trapping point where local anomalies occur, the area of ​​the affected region is calculated based on the preset spatial radius, and the fruit trees are marked within the area of ​​the affected region;

[0026] The initial and final times of the bark borer's active period were used as the collection times, and the thermal diffusion temperature difference of the marked fruit trees was detected using a thermal diffusion stem flow meter.

[0027] In a preferred embodiment, in the active feature module, the thermal diffusion temperature difference is the absolute value of the temperature difference between the heating needle and the reference needle in the thermal diffusion stem flow meter.

[0028] The maximum value of the thermal diffusion temperature difference of all marked fruit trees during the active period of the bark beetle was taken as the peak thermal diffusion temperature difference;

[0029] The transpiration intensity of the marked fruit trees was calculated using the thermal diffusion temperature difference and the peak thermal diffusion temperature difference.

[0030] For marked trapping sites where local anomalies occur, the average tree transpiration intensity at the time of collection of all marked fruit trees during the active period of the bark borer moth is taken as the tree transpiration intensity of the corresponding marked trapping site.

[0031] In a preferred embodiment, in the active feature module, based on the marked trapping points where local anomalies occur, the sound pressure level data at the time of acquisition is monitored by a sound level meter, and the average value of the sound pressure level data at the time of acquisition is used as the ambient noise intensity.

[0032] Tree transpiration intensity and environmental noise intensity were standardized, and the ratio of the standardized tree transpiration intensity to environmental noise intensity was used as a characteristic of bark beetle activity.

[0033] In a preferred embodiment, the strategy adjustment module retrieves the area of ​​the affected region and the number of marked fruit trees of the marked trapping point where a local anomaly occurs from the orchard database.

[0034] The ratio of the number of marked fruit trees to the area of ​​the circular region is taken as the fruit tree density;

[0035] After standardizing the activity characteristics of bark beetles and the density of fruit trees, the environmental impact coefficient was calculated based on an exponential function.

[0036] In a preferred embodiment, the strategy adjustment module compares a preset environmental coefficient threshold with an environmental impact coefficient to determine whether to change the trapping method.

[0037] If the environmental impact coefficient exceeds the preset environmental coefficient threshold, the trapping method will be changed.

[0038] Conversely, the current trapping strategy will be maintained.

[0039] The technical effects and advantages of this invention are as follows:

[0040] This invention sets up multiple trapping points in the orchard under test, collects images of codling moths at each trapping point using image sensors, identifies the trapping data, filters and marks trapping points based on trapping efficiency, and inputs the marked trapping points and their trapping growth rate into a spatial diagnosis module. The spatial diagnosis module calculates spatial weights by combining the distance relationship between the marked trapping points and neighboring trapping points, analyzes the spatial differences in trapping growth rates, and then determines whether there are local anomalies at the marked trapping points. For marked trapping points with local anomalies, the activity feature module determines the active period of codling moths, and detects the tree transpiration intensity and environmental noise intensity of the corresponding fruit trees during the period to generate codling moth activity features. The strategy adjustment module retrieves the fruit tree density from the orchard database, comprehensively evaluates the environmental impact coefficient by combining the codling moth activity features and fruit tree density, and determines whether it is necessary to change the trapping method, thereby realizing dynamic optimization of the codling moth trapping strategy and improving the timeliness of response to abnormal insect infestations. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the implementation of the intelligent monitoring system for codling moths of the apple tree based on multi-source trapping data fusion, as described in this invention.

[0042] Figure 2 This is a schematic diagram illustrating the steps of the intelligent monitoring system for codling moths of the apple tree based on multi-source trapping data fusion, as described in this invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] This invention sets up multiple trapping points in the orchard under test, collects images of codling moths trapped at each point using an image sensor, and identifies the trapping data. Based on the trapping efficiency, the trapping points are screened and marked, and the marked trapping points and their trapping growth rates are transmitted to a spatial diagnosis module. The spatial diagnosis module calculates spatial weights by combining the distance relationship between the marked trapping points and neighboring trapping points, analyzes the spatial differences in the trapping growth rates, and then determines whether there are local anomalies at the marked trapping points. For marked trapping points with local anomalies, the activity feature module determines the active period of codling moths, and detects the tree transpiration intensity and environmental noise intensity of the corresponding fruit trees during the period to generate codling moth activity features. The strategy adjustment module retrieves the fruit tree density from the orchard database, comprehensively evaluates the environmental impact coefficient by combining the codling moth activity features and fruit tree density, and determines whether the trapping method needs to be changed, thereby achieving dynamic optimization of the codling moth trapping strategy. Example 1

[0045] Please see Figures 1 to 2 The intelligent monitoring system for codling moths based on multi-source trapping data fusion includes a trapping identification module, a spatial diagnosis module, an activity feature module, and a strategy adjustment module. The functions of each module are as follows:

[0046] The trapping and identification module sets up multiple trapping points in the orchard to be tested and uses the same trapping method. It monitors the trapping images of codling moths at each trapping point through an image sensor and identifies the trapping data. Based on the trapping data, it evaluates the trapping efficiency and trapping growth rate of the trapping points. It uses the trapping efficiency to screen and mark the trapping points, and then inputs the marked trapping points and the trapping growth rate into the spatial diagnosis module.

[0047] The spatial diagnostic module detects the distance between the marked trap point and neighboring trap points and calculates the spatial weight of the marked trap point. It receives the trapping growth rate of the marked trap point and analyzes the spatial differences based on the spatial weight. Based on the spatial differences, it determines whether the marked trap point has local anomalies.

[0048] After a local anomaly occurs at the marked trapping point, the active feature module obtains the active period of the bark beetle. During the active period of the bark beetle, it detects the tree transpiration intensity and environmental noise intensity at the marked trapping point. Based on the tree transpiration intensity and environmental noise level, it generates the active features of the bark beetle at the marked trapping point and transmits them to the strategy adjustment module.

[0049] The strategy adjustment module retrieves the density of fruit trees at the marked trapping points from the orchard database, receives the activity characteristics of bark moths, and assesses the environmental impact coefficient based on the orchard density. Based on the environmental impact coefficient, it determines whether to change the trapping method.

[0050] The specific implementation is as follows:

[0051] In the trapping and identification module, multiple trapping points are deployed within the orchard area to be tested. These trapping points are evenly distributed spatially and all use the same type of trapping device. Each trapping point is equipped with an image sensor. After acquiring images of the trapped moths at fixed time intervals, the image sensor preprocesses the images, including image enhancement, noise filtering, and grayscale normalization. First, image enhancement improves the recognizability of the moth target in the image by adjusting the image contrast and sharpness. Then, noise filtering removes random noise introduced by changes in ambient light or the image sensor. Finally, grayscale normalization maps the image pixel grayscale values ​​uniformly to a range. This is to improve the stability of subsequent image recognition models.

[0052] Image recognition algorithms are used to detect and classify pre-processed barn moth trapping images, identify the number of barn moths in each frame of the trapping image, and calculate the average number of trappings at each trapping point within a fixed time period as the trapping data.

[0053] It should be noted that the image sensor is an image acquisition device set at each trapping point, used to periodically acquire images of the physical environment within the trapping device. Based on the imaging principle of charge-coupled devices, it converts light signals into digital image signals. The image recognition algorithm is a computing unit deployed in the trapping recognition module, used to perform target detection and classification on the trapping images of moths acquired and preprocessed by the image sensor, and to identify the number and location of individual moths in the images.

[0054] The current time period at each trapping point is compared with the trapping data from the previous time period to calculate the trapping growth rate. The calculation formula is as follows:

[0055] ;

[0056] in, To increase the trapping rate, This is the trapping data for the current time period. This is the trapping data from the previous time period. Number the trapping points. The total number of trapping points. This is the number of the current time period;

[0057] It should be noted that the acquisition of images of moth trapping is carried out in a fixed time period. Each image acquisition and processing is recorded as a time period. The current time period represents the current round of image acquisition for identification and calculation, and the previous time period represents the time period that is immediately preceding the current time period in chronological order.

[0058] The trapping efficiency of each trapping point was then evaluated. The trapping efficiency was defined as the average number of traps per unit time at that trapping point within a set time window, calculated using the following formula:

[0059] ;

[0060] in, For trapping efficiency, it means that in continuous The average number of moths trapped per cycle within a fixed period. The time window length is expressed in periods. The starting fixed period number of the time window. Trapping point In a fixed period The average number of traps within the area.

[0061] Set an efficiency threshold. If the trapping efficiency is less than the efficiency threshold, the trapping point is judged as having substandard trapping efficiency; otherwise, it is judged as having standard trapping efficiency.

[0062] The trapping sites that are deemed to have substandard trapping efficiency are identified and marked.

[0063] The efficiency discrimination threshold is a static preset parameter. By collecting the average number of traps per unit time at each trapping point in the orchard over a long period, the sample distribution of trapping efficiency under multiple periods is obtained. After excluding extreme outliers, the value corresponding to the upper quartile of the distribution is selected as the efficiency discrimination threshold.

[0064] In the spatial diagnostic module, marked trap points are received from the trap identification module. For each marked trap point, its neighborhood trap point set is extracted based on its geographic coordinates in two-dimensional space. The neighborhood trap point set is defined as the set of trap point numbers that satisfy the following distance condition:

[0065] ;

[0066] in, For the set of trapping point numbers, Indicates the trapping point with trapping point The Euclidean distance between them is calculated using the following formula: , Trapping point The horizontal coordinate of the x-axis in two-dimensional space. Trapping point The vertical coordinate of the y-axis in two-dimensional space. Trapping point The horizontal coordinate of the x-axis in two-dimensional space. Trapping point The vertical coordinate of the y-axis in two-dimensional space; This is the neighborhood search radius, which is a constant set by the system. This represents the total number of trapping points; Ensure that the neighboring points do not contain other marked trap points.

[0067] If the set of neighboring trap points corresponding to the marked trap point is empty, the spatial deviation of the trap point will not be calculated, nor will it be included in the local anomaly judgment process, so as to avoid problems of undefinable calculation or misjudgment caused by lack of reference basis.

[0068] For each marked trap point, its spatial weight is calculated based on the spatial positional relationship of each trap point in the neighboring trap point set. The spatial weight is used to measure the relative sparsity of the marked trap point and its neighboring trap points in spatial distribution. The calculation formula is as follows:

[0069] ;

[0070] in, Spatial weights; Trapping point The number of neighboring trap points; add 1 to the denominator to avoid division by zero when the distance is 0.

[0071] After obtaining the spatial weights, the spatial differences of the marked trapping points are quantitatively analyzed by combining the trapping growth rate data from the trapping identification module. A local spatial deviation is defined to represent the degree of deviation between a trapping point and its neighboring trapping points in terms of growth rate. The calculation formula is as follows:

[0072] ;

[0073] in, For local spatial deviation, Trapping point In the current time period The trapping growth rate; The average growth rate of the neighboring trap points of trap point i is represented by the formula: , Trapping point In the current time period The growth rate of trapping.

[0074] When the local spatial deviation is greater than the preset spatial difference threshold, it is determined that the marked trap point has a local anomaly; otherwise, it is determined that the marked trap point has no local anomaly.

[0075] It should be noted that the spatial variability threshold is used to determine whether the local spatial deviation of the marked trapping point has reached an abnormal level. Monitoring data from multiple historical stable insect infestation periods are selected, the local spatial deviation of all marked trapping points in each period is calculated, a set is constructed, extreme values ​​are removed, and the maximum value of the remaining samples is calculated and set as the spatial variability threshold.

[0076] In the active feature module, the time period corresponding to the maximum trapping growth rate of the marked trapping point where local anomalies occur is taken as the active period of the bark beetle;

[0077] Centered on the marked trapping point where local anomalies occur, the area of ​​the affected region is calculated using the circular area formula based on the preset spatial radius. Fruit trees are marked within the area of ​​the affected region, and the area of ​​the affected region and the number of marked fruit trees within the corresponding area of ​​the affected region are stored in the orchard database.

[0078] The initial and final times of the active period of the bark beetle were used as the collection times. The thermal diffusion temperature difference of the marked fruit trees was detected by a thermal diffusion stem flow meter. Specifically, the temperature values ​​of the heating needle and the reference needle at the collection time were obtained by the thermistor in the thermal diffusion stem flow meter. The absolute value of the difference between their temperature values ​​was taken as the thermal diffusion temperature difference, which reflects the water physiological intensity of the marked fruit trees during the active period of the bark beetle.

[0079] The maximum value of the thermal diffusion temperature difference of all marked fruit trees during the active period of the bark beetle was taken as the peak thermal diffusion temperature difference;

[0080] The transpiration intensity of the marked fruit trees was calculated using the thermal diffusion temperature difference and the peak thermal diffusion temperature difference. ,in, This represents the peak thermal diffusion temperature difference. The temperature difference due to thermal diffusion at the time of data collection. , As a preset empirical constant, The tree transpiration rate at the time of collection;

[0081] For the marked trapping sites where local anomalies occur, the average tree transpiration intensity of all marked fruit trees at the time of collection during the active period of the bark beetle is taken as the tree transpiration intensity of the corresponding marked trapping site, reflecting the overall water physiological activity level of all marked fruit trees.

[0082] Transpiration intensity measures the water metabolism and gas exchange of fruit trees, reflecting the rate at which the tree releases water vapor and volatile organic compounds into the environment. Higher transpiration intensity creates a stronger moisture gradient and odor signal on the leaves and fruit surface, enhancing the moths' ability to perceive the host tree and inducing their approach, dwelling, and foraging behavior. Simultaneously, the faster diffusion rate of pheromone molecules accompanying water transpiration increases the moths' responsiveness to sex pheromone signals during their active periods. Conversely, lower transpiration intensity weakens the odor and moisture gradients released by the fruit tree, making it less conducive to moths recognizing and dwelling on the host.

[0083] It should be noted that the preset spatial radius is used to delineate the affected area around the trapping point where a local anomaly is marked. The spatial radius is determined based on the growth habits of the fruit trees, the activity radius of the bark borers, and historical monitoring experience. The shape of the affected area is not unique and other geometric shapes can be used according to actual needs. The orchard database is a data structure system used to store, manage, and retrieve multi-source information data within the orchard, including trapping point information, fruit tree distribution information, etc. The heat diffusion stem flow meter is an instrument used to monitor the water flow rate within the plant stem. The heat diffusion stem flow meter includes a heating needle and a reference needle, which are inserted into the vascular bundle area of ​​the trunk, respectively. The temperature values ​​of the heating needle and the reference needle are obtained by an internal thermistor. The preset empirical constant is an empirical fitting parameter in the Granier heat diffusion model, used to convert the measured heat diffusion temperature difference into tree transpiration intensity. It can be set based on the fruit tree variety and the diameter of the fruit tree trunk. For example, for robust fruit trees, the value of a is set to 120-130, and the value of b is set according to the growth stage of the fruit tree. The specific settings are performed by professionals.

[0084] Based on the marked trapping points where local anomalies occur, the sound pressure level data at the time of collection is monitored by a sound level meter, and the average sound pressure level data at the time of collection is used as the environmental noise intensity.

[0085] Tree transpiration intensity and environmental noise intensity were standardized, and the ratio of the standardized tree transpiration intensity to environmental noise intensity was used as a characteristic of bark beetle activity.

[0086] The greater the tree transpiration intensity, the more active the physiological activity of the marked fruit trees around the marked trapping point where local abnormalities occur, which is conducive to the moths' lodging and foraging activities. Therefore, the greater the moths' activity characteristics, the greater the activity characteristics. The lower the environmental noise intensity, the quieter the environment, which is conducive to the moths lodging around it and their pheromone response behavior. Therefore, the greater the moths' activity characteristics.

[0087] It should be noted that a sound level meter is a specialized instrument used to measure the intensity of ambient sound. In this example, it is used to collect data from marked trapping points exhibiting local anomalies during the active period of the moths to quantify the intensity of ambient noise. Standardization is a data preprocessing method that allows different data to be compared and analyzed on the same dimensional scale. Standardization methods include, but are not limited to, standard linear transformation based on interval scaling, statistical Z-score standardization, or normalization based on nonlinear mapping functions. The application methods of standardization will not be elaborated here.

[0088] In the strategy adjustment module, the area of ​​the affected region and the number of marked fruit trees of the marked trapping point that has local anomalies are retrieved from the orchard database, and the ratio of the number of marked fruit trees to the area of ​​the circular region is used as the fruit tree density.

[0089] The activity characteristics of bark beetles and the density of fruit trees were standardized, and the standardized results were used as the bark beetle activity characteristic index and the fruit tree density index, respectively. The environmental impact coefficient was evaluated using an exponential function. ,in, This is an activity characteristic index of bark beetles. The fruit tree density index, , The preset weighting coefficients, This is the environmental impact coefficient;

[0090] It should be explained that the preset weighting coefficients are used to adjust the influence of codling moth activity characteristics and fruit tree density on the environmental impact coefficient. The values ​​range from 0 to 1 and can be set based on the orchard ecosystem, codling moth behavior characteristics, and historical monitoring data. The specific settings are determined by professionals. For example, if codling moth activity characteristics have a greater impact on the environmental impact coefficient, then the weighting coefficients will be adjusted accordingly. The value is set to 0.6.

[0091] The transpiration process of trees is closely related to the water and energy metabolism levels within the fruit trees. The greater the transpiration intensity, the more significant the odor components and water gradients released by the fruit trees to the outside world, thereby increasing the probability of bark beetles sensing the host and enhancing their activity in staying and foraging. Environmental noise interferes with the directional flight and sex pheromone recognition of bark beetles. The lower the intensity of environmental noise, the more conducive it is to the aggregation of bark beetles and pheromone response. When the density of fruit trees is high, the local ecological microenvironment formed by the tree community is more stable, providing favorable conditions for the habitat, stay and reproduction of bark beetles.

[0092] The greater the activity of the bark beetle, the more favorable the environment around the marked trapping point is for the bark beetle's metabolism. The higher the physiological activity level of the bark beetle, the greater the environmental impact coefficient. The greater the orchard density, the more dense the fruit trees around the marked trapping point are, providing habitat conditions for the bark beetle's continuous stay and reproduction, thus the greater the environmental impact coefficient.

[0093] The system compares a preset environmental coefficient threshold with an environmental impact coefficient to determine whether to change the trapping method.

[0094] If the environmental impact coefficient exceeds the preset environmental coefficient threshold, the trapping method will be changed.

[0095] Conversely, the current trapping strategy will be maintained.

[0096] When the environmental impact coefficient exceeds the preset environmental coefficient threshold, it indicates that the ecological conditions at the marked trapping sites with local anomalies have enhanced the aggregation and activity of the barn moths during their active period. Therefore, the trapping method should be changed. The trapping method is based on changes made by professionals according to the orchard type, fruit tree distribution structure, and barn moth population ecological characteristics. This includes, but is not limited to, adjusting the type of trap, optimizing the density of trapping sites, adjusting the trapping time, and physical control assistance. For example, replacing ordinary pheromone traps with traps with highly volatile lures, adding 50% more trapping sites in abnormal areas, increasing the monitoring frequency from once a day to three times a day (morning, noon, and evening), and using hanging sticky insect boards or installing sonic insect repellent devices. Conversely, if the environmental impact coefficient does not exceed the threshold, it indicates that the marked trapping sites with local anomalies are effective in trapping during the barn moths' active period, and the current trapping strategy can be maintained.

[0097] It should be noted that the preset environmental coefficient threshold is a critical value used to determine whether the trapping method needs to be adjusted. It can be based on a large amount of actual orchard trapping monitoring data and the statistical distribution of the corresponding environmental impact coefficient. The threshold value can be selected as the critical value that can effectively distinguish between normal trapping effects and abnormal impact areas. The specific setting is done by professionals. For example, the environmental coefficient threshold can be determined by analyzing the quantiles, mean and fluctuation range of the environmental impact coefficient.

[0098] For example, if in a monitoring of an orchard, the average number of borer moths trapped at the trapping point in the middle is 12 from 18:00 to 19:00 and 20 from 19:00 to 20:00, then the trapping growth rate at that point is calculated to be (20-12) / 12=0.67.

[0099] If the average growth rate of its neighboring trapping points is 0.25, the comparison difference reaches 0.42, which is higher than the preset spatial difference threshold of 0.30, thus it is determined that there is a local anomaly at the trapping point during this period.

[0100] Further detection of fruit tree transpiration intensity during the active period of this anomaly point; if the result is... Meanwhile, the ambient noise intensity was monitored using a sound level meter, and the result was 42 dB(A).

[0101] To facilitate comparison of data with different dimensions, transpiration intensity and noise intensity were processed by normalizing the maximum value, that is, each value was divided by the maximum value in the corresponding monitoring sample, resulting in a transpiration intensity index of approximately 0.76 and a noise intensity index of approximately 0.34. The ratio of the two, 2.24, was taken as the characteristic value of bark beetle activity, indicating that the environment in this area is more conducive to bark beetle activity.

[0102] Based on the orchard database, the number of fruit trees within the influence radius of this point is 45, and the area is 500m². ² The calculated fruit tree density is 0.09 trees / m². ² Substituting the fruit tree density and the codling moth activity index into the exponential function, the environmental impact coefficient is approximately 4.22, which is higher than the preset threshold of 3.5. Therefore, it is determined that the trapping method in this area needs to be adjusted, such as increasing the number of trapping points or using physical control methods.

[0103] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0104] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0105] In this document, the singular forms “a,” “an,” and “the” may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that terms such as “comprising / including” or “having” specify the presence of the stated features, integrals, steps, operations, components, parts, or combinations thereof, but do not preclude the possibility of the presence or addition of one or more other features, integrals, steps, operations, components, parts, or combinations thereof. Meanwhile, the term “and / or” as used in this specification includes any and all combinations of the associated listed items.

[0106] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0107] The above description of the disclosed embodiments will enable those skilled in the art to make or use various modifications to these embodiments. It will be readily apparent to those skilled in the art that the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent monitoring system for codling moths of the apple tree based on multi-source trapping data fusion, characterized in that: It includes a trapping and identification module, a spatial diagnosis module, an active feature module, and a strategy adjustment module. The functions of each module are as follows: The trapping and identification module sets up multiple trapping points in the orchard to be tested and uses the same trapping method. It monitors the trapping images of codling moths at each trapping point through an image sensor and identifies the trapping data. Based on the trapping data, it evaluates the trapping efficiency and trapping growth rate of the trapping points. It uses the trapping efficiency to screen and mark the trapping points, and then inputs the marked trapping points and the trapping growth rate into the spatial diagnosis module. The spatial diagnostic module detects the distance between the marked trap point and neighboring trap points and calculates the spatial weight of the marked trap point. It receives the trapping growth rate of the marked trap point and analyzes the spatial differences based on the spatial weight. Based on the spatial differences, it determines whether the marked trap point has local anomalies. After a local anomaly occurs at the marked trapping point, the active feature module obtains the active period of the bark beetle. During the active period of the bark beetle, it detects the tree transpiration intensity and environmental noise intensity at the marked trapping point. Based on the tree transpiration intensity and environmental noise level, it generates the active features of the bark beetle at the marked trapping point and transmits them to the strategy adjustment module. In the active feature module, the time period corresponding to the maximum trapping growth rate of the marked trapping point where local anomalies occur is taken as the active period of the bark beetle; Centered on the marked trapping point where local anomalies occur, the area of ​​the affected region is calculated based on the preset spatial radius, and the fruit trees are marked within the area of ​​the affected region; The initial and final times of the bark borer's active period were used as the sampling points, and the thermal diffusion temperature difference of the marked fruit trees was detected using a thermal diffusion stem flow meter. In the active feature module, the thermal diffusion temperature difference is the absolute value of the temperature difference between the heating needle and the reference needle in the thermal diffusion flow meter; The maximum value of the thermal diffusion temperature difference of all marked fruit trees during the active period of the bark beetle was taken as the peak thermal diffusion temperature difference; The transpiration intensity of the marked fruit trees was calculated using the thermal diffusion temperature difference and the peak thermal diffusion temperature difference. ,in, This represents the peak thermal diffusion temperature difference. The temperature difference due to thermal diffusion at the time of data collection. , As a preset empirical constant, The tree transpiration rate at the time of collection; For marked trapping sites exhibiting localized anomalies, the average tree transpiration intensity at the time of collection from all marked fruit trees during the active period of the bark beetle was used as the tree transpiration intensity for the corresponding marked trapping site. In the active feature module, based on the marked trapping points where local anomalies occur, the sound pressure level data at the time of collection is monitored by a sound level meter, and the average sound pressure level data at the time of collection is used as the environmental noise intensity. Tree transpiration intensity and environmental noise intensity were standardized, and the ratio of the standardized tree transpiration intensity to environmental noise intensity was used as a characteristic of bark-boring moth activity. The strategy adjustment module retrieves the density of fruit trees at the marked trapping points from the orchard database, receives the activity characteristics of bark moths, and assesses the environmental impact coefficient based on the orchard density. Based on the environmental impact coefficient, it determines whether to change the trapping method.

2. The intelligent monitoring system for codling moths based on multi-source trapping data fusion according to claim 1, characterized in that: In the trapping and identification module, multiple trapping points are deployed in the orchard area to be tested, and images of moth trapping are collected by an image sensor at fixed time intervals. After preprocessing the images of moth-trapping, the images are then subjected to target detection and classification using an image recognition algorithm. The number of moths in each frame of the images is identified, and the average number of moths trapped at each trapping point within a fixed time period is calculated as the trapping data. The current time period of each trapping point is compared with the trapping data of the previous time period to calculate the trapping growth rate.

3. The intelligent monitoring system for codling moths based on multi-source trapping data fusion according to claim 2, characterized in that: In the trapping and identification module, the average number of traps per unit time at the trapping point within a set time window is calculated as the trapping efficiency. If the trapping efficiency is less than the efficiency discrimination threshold, the trapping point is judged to be of substandard trapping efficiency; Conversely, if the trapping efficiency is not met, it is considered that the trapping efficiency has met the standard. The trapping sites that are deemed to have substandard trapping efficiency are identified and marked.

4. The intelligent monitoring system for codling moths based on multi-source trapping data fusion according to claim 1, characterized in that: In the spatial diagnostic module, the marked trap points output from the trap identification module are received, and the set of neighboring trap points is extracted based on the geographical coordinates of each marked trap point in two-dimensional space. Based on the spatial positional relationship of each trapping point in the neighborhood trapping point set, the spatial weight of each marked trapping point is calculated; After obtaining the spatial weights, the local spatial deviation is defined by combining the trapping growth rate of the marked trapping points input from the trapping recognition module. When the local spatial deviation exceeds the preset spatial difference threshold, it is determined that a local anomaly has occurred at the marked trapping point; Conversely, if no local anomalies are found at the marked trapping point, it is determined that no such anomalies have occurred.

5. The intelligent monitoring system for codling moths based on multi-source trapping data fusion according to claim 1, characterized in that: In the strategy adjustment module, the area of ​​the affected region and the number of marked fruit trees of the marked trapping points that have shown local anomalies are retrieved from the orchard database. The ratio of the number of marked fruit trees to the area of ​​the circular region is taken as the fruit tree density; After standardizing the activity characteristics of bark beetles and the density of fruit trees, the environmental impact coefficient was calculated based on an exponential function.

6. The intelligent monitoring system for codling moths based on multi-source trapping data fusion according to claim 5, characterized in that: In the strategy adjustment module, a preset environmental coefficient threshold is compared with the environmental impact coefficient to determine whether to change the trapping method. If the environmental impact coefficient exceeds the preset environmental coefficient threshold, the trapping method will be changed. Conversely, the current trapping strategy will be maintained.

Images