A moth trapping system and method for greenhouse crops

Abstract

The present application relates to a kind of for greenhouse crop's moth trapping system and method.The system includes detection module, trapping module and early warning module, early warning module processes the data information of trapping module and detection module and feeds back corresponding control signal, to adjust operating parameter, so that trapping module and detection module at least for the same kind of pest work in collaborative mode, early warning module generates early warning information based on the data transmitted by detection module and trapping module working in collaborative mode.The method includes: determining the high-risk pest species with strong association with current environmental information;Search whether there is high-risk pest in captured pest;Analysis whether there is high-risk pest egg in key area and judge the pest situation of high-risk pest;When the number of high-risk pest egg in key area is greater than the first threshold, adjust the capture parameter of trapping module according to the species of high-risk pest;Predict the population number of high-risk pest, when the population number is greater than the second threshold, collect image data in non-key area.

Description

Technical Field

[0001] This invention relates to the field of plant protection technology, and in particular to an insect-attracting system and method for greenhouse crops. Background Technology

[0002] Frequent insect pests greatly affect crop growth. Using a monitoring system to monitor and warn of insect pests in real time can help growers predict the development of pests in advance and take certain control measures to block the growth and development of pests, remove insect eggs or larvae in advance, reduce the risk of pest outbreaks, and increase crop yields.

[0003] CN112931456A discloses a field crop insect collection device and a pest monitoring and early warning method. This early warning system uses a field trapping method to collect insect data from the field. The system includes an insect trapping device and a data collection box. The insect outlet of the insect trapping device is connected to the data collection box, which contains an insect information collection device that is communicatively connected to a pest early warning service platform. The system traps insects in the field using the insect trapping device, collects insect information in the data collection box, and finally sends the collected insect information to the pest early warning service platform for pest early warning.

[0004] This method can capture mostly highly mobile insects such as Lepidoptera. The types of insects captured are related to factors such as the height at which the trapping device is set. Different insects have their own optimal heights, so the capture results tend to be related to the height of the trapping device. Furthermore, its broad-spectrum trapping method has varying effects on different types of pests, resulting in low accuracy. The attracted and captured insects are mostly adults. When a large number of adults are captured and an alert is issued, the insect population is already quite large, making pest control difficult and time-consuming.

[0005] The grassland locust intelligent identification system and method provided by CN113688858A uses image recognition to identify insects. The grassland locust intelligent identification system includes a front-end information acquisition terminal, an image transmission system, a locust image species recognition model on a back-end cloud platform, and a back-end management cloud platform. The front-end information acquisition terminal is a mobile phone. The grassland locust intelligent identification system acquires images of locusts through the mobile phone. The locust images are transmitted to the locust image species recognition model on the back-end cloud platform through the image transmission system. The deep neural network in the locust image species recognition model learns and extracts the features of the locusts in the images, ultimately realizing the monitoring and identification of grassland locusts, providing users with an efficient and effective locust control method.

[0006] This identification method is suitable for collecting insects during the day, but its accuracy is lower for collecting and identifying insects active at night. Furthermore, during daytime collection, the number of pests collected varies depending on the device's location, and the accuracy is lower for insects with strong mobility. Additionally, ensuring high accuracy through image acquisition requires extended collection times and massive amounts of data processing. Since insect development is measured in days, with most insects reproducing in about 40 days, the extended collection and processing time is detrimental to timely pest warnings.

[0007] Because different insects have different growth habits and optimal climatic environments, combining environmental prediction with insect development trends can significantly improve monitoring accuracy. Existing methods combining environmental prediction include, for example, CN110719733A, which discloses a pest control system and related methods, including: a pest monitoring device for placement at a location and generating a signal when a pest is detected; and a pesticide dispensing unit comprising a reservoir for containing a diluent and a pesticide module for containing pesticides. A computer device remotely configured to monitor the pest monitoring device receives the signal generated therefrom, analyzes environmental and historical data of the location to determine the quantity and placement of the pest monitoring device for that location, and guides the deployment of the pest control monitoring device based on the analysis. The computer device also receives data from the pesticide dispensing unit, including the amount or ratio of pesticide substances dispensed from the pesticide module to form a treatment agent with a certain concentration of pesticide substances in the diluent, which is dispensed from the pesticide dispensing unit to treat the detected pests at the location.

[0008] The existing technology optimizes the quantity and placement of pest monitoring equipment by combining environmental data. While this can improve the accuracy of monitoring and early warning to some extent, it does not fundamentally solve the technical limitations inherent in the monitoring devices themselves.

[0009] CN107357271A discloses a method and system for controlling greenhouse crop diseases and pests. The method includes: establishing an early warning model for greenhouse crop diseases and pests; acquiring the current crop type and current environmental parameters in real time; matching the current crop type with the crop type in the early warning model to determine whether the current environmental parameters are within the early warning threshold range of the crop type in the early warning model; when the current environmental parameters are determined to be within the early warning threshold range, activating an image recognition device to acquire an image of the current crop and identifying whether the current crop is affected by diseases or pests based on the image; and activating the corresponding control mode based on the recognition result.

[0010] However, this existing technology still has some potential technical problems: First, it mainly relies on environmental parameters and image recognition for pest and disease prediction, which may not be accurate enough in complex and ever-changing real-world environments; second, although it can identify pests and diseases and take measures, it is not very targeted to specific pests and lacks specific trapping and management strategies for different pest species; third, it does not fully consider the different characteristics of each stage of the pest life cycle (such as egg, larva, and adult) and the degree of damage to crops.

[0011] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0012] In view of the shortcomings of the prior art, the present invention provides an insect-attracting system and method for greenhouse crops to solve at least some of the above-mentioned technical problems.

[0013] This invention discloses an insect-attracting system for greenhouse crops, including a detection module for collecting climate information and image information around the plants, and a trapping module for attracting insects and identifying the species and number of insects. It also includes an early warning module that is connected to the trapping module and the detection module. The early warning module comprehensively processes the data information from the trapping module and the detection module and feeds back corresponding control signals to the trapping module and the detection module to adjust the working parameters of the trapping module and the detection module, so that the trapping module and the detection module work in a cooperative mode against at least the same pest. The early warning module generates early warning information based on the data transmitted by the detection module and the trapping module working in cooperative mode to notify the operator of the pest progress in advance.

[0014] This invention utilizes a detection module to detect climate information surrounding plants and image information of key areas of highly harmful insects strongly correlated with this climate information. Based on the detected climate information, the detection module is adjusted to target highly harmful insects potentially present under the current climatic conditions, rather than performing broad-spectrum detection. This not only reduces the amount of data to be processed, improving data feedback speed, but also increases detection accuracy. Compared to methods that use broad-spectrum trapping parameters to capture pests, this invention's method of adjusting the trapping module's parameters according to the species of highly harmful insects is more targeted. It reduces capture errors caused by factors such as the trapping module's height, especially in the early stages of pest infestation when pest populations are small, increasing the probability of the trapping module capturing adult highly harmful insects and thus advancing the pest warning process. The detection module, by acquiring image information to identify pests, has high accuracy in detecting insects with poor mobility, such as eggs and larvae of highly harmful insects, but low accuracy in detecting more mobile adults. The trapping module achieves a high capture rate for highly mobile adult insects by trapping them to identify pests, but its accuracy is poor for detecting less mobile eggs and larvae. The detection module detects highly harmful insects' eggs and larvae, which are less mobile, while the trapping module detects highly mobile adults of highly harmful insects. This allows for the complementary advantages of image-based insect detection and trapping-based insect detection methods, enabling accurate prediction of insect population development and changes.

[0015] According to a preferred embodiment, the detection module includes a climate detection module and an image acquisition module. The comprehensive processing specifically includes: an early warning module predicting highly harmful insect species strongly correlated with current environmental information based on climate information collected by the climate acquisition module and data from a preset database; the early warning module determining whether highly harmful adult insects have been captured based on data of captured pests sent by the trapping module; if the early warning module determines the presence of highly harmful insects, it controls the image acquisition module to acquire images of key areas where highly harmful insects occur, and analyzes the insect situation based on the image data acquired by the image acquisition module; when adults are present, it indicates the emergence of overwintering pupae. Since most pests mate on the night of or the day after the overwintering pupae emerge, and then lay eggs about a day later, it is necessary to detect whether egg-laying behavior has already occurred as soon as adults are present. The system monitors the number of eggs laid and the number of pests. The early warning module, based on the first trigger condition that the number of eggs of highly harmful insects in key areas exceeds a first threshold, controls the trapping module to adjust its capture parameters according to the type of highly harmful insect to specifically capture them. The trapping module can attract and kill newly emerged adult insects, thus preventing mating and egg-laying. Targeted capture increases the capture rate of highly harmful insects with a higher risk of causing pest infestations, effectively reducing their damage. Simultaneously, the number of highly harmful insects killed by the trapping module provides data for the actions of the image acquisition module and the early warning module. The early warning module predicts the population size of highly harmful insects based on the types and proportions of insects captured by the trapping module. Based on the second trigger condition that the population size exceeds a second threshold, the early warning module controls the detection module to acquire images of non-key areas to accurately assess the infestation of highly harmful insects.

[0016] Preferably, the preset database is a database of the relationship between the physiological characteristics of pests and climate.

[0017] The trapping module can increase the capture rate of highly harmful insects by targeting them, while also accurately capturing the number of insects. Based on the captured area and the physiological habits of these insects, the population size can be determined. When the population size exceeds a second threshold, it indicates a large-scale reproduction and growth trend. In this case, images of the key areas are insufficient for a complete population size prediction; combining image data from non-key areas is necessary for a more accurate prediction. Furthermore, a large number of captured adult insects indicates a large population size and an expanded range of movement, posing a significant risk of damage to surrounding non-key areas. In this situation, information on plant growth and insect numbers in both key and non-key areas is needed to assess the actual damage and provide a reference for large-scale pest control operations by personnel or automated systems (such as automated pesticide spraying machines). Adjusting the collection plan according to different insect infestation stages increases the accuracy of early warnings.

[0018] According to a preferred embodiment, the early warning system includes at least a first-level warning, a second-level warning, and a third-level warning with progressively increasing intensity. The early warning module issues a first-level warning based on the fulfillment of a first triggering condition that the number of highly harmful insect eggs in a key area exceeds a first threshold. A first-level warning indicates that the risk of pest infestation is low and the difficulty of pest control is relatively low. Killing pest eggs during their peak hatching period is most efficient. Since the hatching time for most pest eggs is approximately 10 days, issuing a first-level warning when a large number of insect eggs have formed at this time prompts personnel or automated control programs to perform egg control, which can improve the effectiveness of pest control.

[0019] The early warning module issues a second-level early warning based on the cancellation of the first-level early warning and the fulfillment of a second triggering condition that the number of highly harmful insect species exceeds the second threshold.

[0020] Level 2 warning indicates a high risk of pest infestation but relatively low difficulty in pest control. Level 1 warning is lifted based on the death of a large number of insect eggs due to pesticide use and the existing number of eggs being less than the first threshold. At this time, the eggs produced by the first generation of overwintering pupae will not hatch, so the pest infestation is temporarily difficult to continue to develop. However, since the population of highly damaging insects is greater than the second threshold at this time, there is still a trend of a large number of adults laying eggs and the population of eggs hatching into adults increasing rapidly.

[0021] The early warning module issues a third-level early warning based on the failure to resolve the first-level early warning and the fulfillment of the second triggering condition that the population size exceeds the second threshold.

[0022] The Level 3 warning indicates a high risk of pest infestation and significant difficulty in pest control. After a large number of eggs hatch, new offspring will be produced. These offspring will mature in approximately 10 days and then rapidly lay their next batch of eggs. Therefore, if the eggs are not eliminated, the risk of pest infestation after hatching is high. Larvae are more mobile than eggs, making them even more difficult to kill. In addition, the presence of a large number of adults and the imminent laying of the second generation of eggs further exacerbate the high risk of infestation. Therefore, prompt action is required, hence the Level 3 warning, which is more urgent than the Level 1 and Level 2 warnings.

[0023] This invention uses different levels of early warning information to warn of different pest infestation processes, pest risks, and control difficulties. Control personnel can flexibly adjust control plans based on the received early warning information, prioritizing the handling of high-level early warnings and temporarily suspending the implementation of low-level early warnings when personnel or resources are scarce. This avoids misleading control personnel by having them spend a lot of time on low-level early warnings and delaying the control of pests under high-level early warnings, thus helping to improve the work efficiency of control personnel.

[0024] According to a preferred embodiment, the early warning module controls the image acquisition module to initiate a detection scheme and update the pest situation based on the satisfaction of a first trigger condition that the number of highly harmful insect eggs in the key area exceeds a first threshold. Specifically, the detection scheme involves the early warning module associating climate information collected after the acquisition time point used to generate the first-level early warning image with the type of pest, predicting the actual development process of the pest based on the real-time collected climate information, and generating predicted key time nodes for the growth and development of the pest in real time. The early warning module then controls the image acquisition module to acquire images again at a time node a first time interval before the key time node arrives.

[0025] According to a preferred embodiment, the critical time point is the time when the pest's eggs hatch in large numbers.

[0026] By comparing the re-collected images with the images used to generate the first-level warning, the current insect infestation progress is determined. If a large number of eggs are about to hatch or have already hatched, it is determined that the first-level warning has not been lifted and the first-level warning is repeated to prompt management personnel to eliminate the insect eggs in a timely manner to avoid the risk and spread of insect infestation. When a large number of eggs have been killed, the first-level warning is lifted.

[0027] According to a preferred embodiment, the early warning includes a control plan for the current pest infestation. The early warning module also associates the first pest infestation with the second pest infestation. Based on the existence of the first pest infestation, the early warning module adds a prevention and / or control plan for the second pest to the control plan for the first pest infestation.

[0028] According to a preferred embodiment, the key area image acquisition module switches from a first working mode to a second working mode with higher power consumption based on a first control signal sent by the early warning module, which includes information on image acquisition of key areas. The image acquired in the second working mode has a higher clarity than the image acquired in the first working mode.

[0029] According to a preferred embodiment, when a first-level warning for multiple pests is issued, the warning module allocates surrounding image acquisition modules to collect images of key areas affected by different highly harmful pests based on location parameters of multiple high-risk locations in the greenhouse environment. This allows the insect-attracting system to simultaneously detect multiple pest infestations and publish the pest infestations in the warning according to their severity level. For example, the pest infestations are ranked based on a comprehensive score considering factors such as the magnitude of economic losses caused, the development stage of the infestation (e.g., whether it is still in the stage of large-scale egg hatching or the stage of large-scale nymph development into adults), and the difficulty of pest control.

[0030] The distributed execution method is more targeted than simply executing the same task across all instances, enabling simultaneous monitoring of different pest species. When there are multiple highly damaging pest species, the same detection module can collect data on different species and transmit it to the early warning module. The early warning module then processes and analyzes the data to generate continued monitoring and control plans for the pest situation. This method offers high real-time performance and applicability.

[0031] Another aspect of the present invention provides a method for attracting insects to greenhouse crops, comprising the following steps:

[0032] Based on climate information and the physiological habits of pests, highly harmful insect species that are strongly correlated with current environmental information are identified.

[0033] Search among the captured pests for highly harmful insects;

[0034] In the presence of highly harmful insects, analyze whether there are highly harmful insect eggs in key areas and assess the insect infestation situation;

[0035] When the number of highly harmful insect eggs in a key area exceeds the first threshold, the capture parameters of the trapping module are adjusted according to the type of highly harmful insect in order to capture the highly harmful insect in a targeted manner.

[0036] The population size of highly harmful insects is predicted based on the species and proportion of the captured highly harmful insects. When the population size exceeds a second threshold, image data from non-key areas is collected to accurately determine the insect infestation situation.

[0037] Compared to existing technologies for controlling pests and diseases in greenhouse crops (such as CN107357271A), this invention not only combines climate information and image recognition technology but also introduces a specially designed trapping module to attract and capture insects. This multi-layered information collection method greatly enhances the understanding of pest activity patterns. The detection module collects climate information and image information around the plants, and the trapping module attracts insects and identifies their species and numbers. The early warning module comprehensively processes this data, enabling the system to more accurately predict pest development trends. More importantly, this invention emphasizes the monitoring and management of each stage of the pest life cycle, especially early intervention during the egg and larval stages of highly damaging insects, effectively reducing the population growth rate and avoiding the risk of large-scale outbreaks. Furthermore, this invention considers the interactions between different pests, such as the synergistic effect between the rice stem borer and the brown planthopper, providing a scientific basis for developing more comprehensive and effective control strategies. By performing correlation analysis on the pest data of the first and second pests, the system can prevent the occurrence of one pest when the first appears, enhancing the overall control effect. Meanwhile, the early warning module of this invention can dynamically adjust the recommended dosage of pesticides based on the pest population density. Combined with changes in the number of natural enemies, it achieves an intelligent integration of chemical and biological control methods, further reducing pesticide use and protecting the ecological environment. Therefore, this invention not only improves the accuracy of pest monitoring but also effectively controls the reproduction rate of pests and reduces economic losses through more targeted trapping and management strategies. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the insect-attracting system provided by the present invention;

[0039] Figure 2 This is a hardware connection diagram of the insect-attracting system provided by the present invention;

[0040] Figure 3 This is a schematic diagram of data transmission in the insect-attracting system provided by the present invention;

[0041] Figure 4 This is a schematic diagram illustrating the generation of different levels of early warning information by the early warning module provided by the present invention;

[0042] Figure 5 This is a flowchart of the insect-attracting method provided by the present invention.

[0043] List of reference numerals

[0044] 100: Detection module; 110: Climate detection module; 120: Image acquisition module; 200: Trapping module; 300: Early warning module. Detailed Implementation

[0045] The following is a detailed explanation with reference to the accompanying drawings.

[0046] "Highly harmful insects" are insect species that are extremely prone to pest infestation under current environmental conditions.

[0047] "Key areas" are areas where highly harmful insects are frequently active, such as oviposition areas and larval growth areas.

[0048] Greenhouse crops, due to the unique characteristics of their growing environment, often face unique challenges from pests and diseases. While greenhouses provide a relatively controlled environment that helps optimize crop growth conditions, this also means that once harmful organisms invade, they can multiply rapidly without natural predators, posing a serious threat to crops. For example, aphids, mites, and certain types of flying insects are particularly adapted to greenhouse environments and can proliferate in a short period, leading to reduced yields or even crop death. Furthermore, greenhouse crops typically have high economic value, and any damage caused by pests and diseases directly translates into economic losses.

[0049] Based on this, such as Figures 1-3 As shown, this invention discloses an insect-attracting system for greenhouse crops, including a detection module 100, a trapping module 200, and an early warning module 300. The trapping module 200 is used to attract insects and identify the species and quantity of insects. The detection module 100 includes a climate detection module 110 and an image acquisition module 120 that are data-connected to each other. The climate detection module 110 is used to collect surrounding climate information to correlate this information with the progress of insect pests. The image acquisition module 120 is used to collect image information around the plants to provide the early warning module 300 with data to determine the damage status of the plants and analyze the progress of insect pests around them. The image acquisition module 120 includes a low-quality working mode and a high-quality working mode. It can switch from the low-quality working mode to the high-quality working mode upon receiving a control signal, and automatically switch from the high-quality working mode back to the low-quality working mode if no control signal is received within a first time period after completing the acquisition task, thus saving energy. Preferably, the first time period is 10 minutes.

[0050] The trapping module 200 and the detection module 100 are respectively connected to the early warning module 300. The early warning module 300 comprehensively processes the data information from the detection module 100 and the trapping module 200, and then generates and sends early warning information.

[0051] Comprehensive processing includes:

[0052] S1. The early warning module 300 analyzes the climate information collected by the climate detection module 110 and identifies highly harmful insect species that are strongly correlated with the current environmental conditions. The early warning module 300 acquires the physiological habits of the strongly correlated insect species and, based on these physiological habits, confirms at least one oviposition site and at least one oviposition location on the plant for that insect.

[0053] The growth and development of pests are significantly correlated with environmental conditions. Under suitable conditions of temperature, humidity, food, and light, pests grow and develop faster and reproduce more rapidly. Different pest species have different environmental requirements. For example, cotton aphids and many spider mites thrive in environments with relative humidity below 75%, where they reproduce in large numbers, often leading to outbreaks during dry seasons. Armyworms, cotton bollworms, cotton leafminer, and many stem borers thrive in environments with relative humidity above 80%, making them prone to infestations during the rainy season. Locusts cease laying eggs when relative humidity is below 25%, and lay the most eggs at a relative humidity of 60%–70% and a temperature of 37.7°C. After laying eggs, pests require suitable conditions for successful hatching. For example, corn borer eggs require relatively high humidity to hatch; when the temperature is maintained at 25°C, the relative humidity must reach 90% for all eggs to hatch. If the relative humidity drops to 80%, the egg mortality rate reaches 6%; if the relative humidity drops to 70%, the egg mortality rate rises to 75%. First-instar larvae of the corn borer rarely die at temperatures between 20°C and 30°C and at saturated humidity; however, their development is delayed if the relative humidity drops below 95%.

[0054] Based on the correlation between surrounding environmental information and the physiological processes of different pests, such as oviposition, egg hatching, and larval growth, climate information can be used to predict the species of highly damaging pests with a large-scale reproduction trend and the current growth stage of the pest population. According to the pests' life habits, the early warning module 300 can screen out key areas for pest reproduction and growth. Detecting pests in these key areas can improve detection accuracy. For example, the first generation of rice stem borers mainly lays eggs on the surface of rice seedling leaves, about 3-6 cm from the leaf tip. The second generation lays eggs on the leaf sheath about 3 cm from the ground. The third generation lays eggs on the outer side of the leaf sheath in late-season rice. Based on the physiological habits of the rice stem borer, the key areas are: the leaf surface about 3-6 cm from the leaf tip and the leaf sheath. Detecting egg data in these key areas can reflect the risk of large-scale rice stem borer infestation. By utilizing climate information from different locations within the greenhouse environment and combining it with the life habits of highly harmful insects, detection modules 100 can be allocated and installed in different areas of the greenhouse to simultaneously detect different types of highly harmful insects based on the varying microclimates of the greenhouse. This rational allocation of detection resources allows for accurate early warning of multiple pest infestations. In spring, summer, autumn, and winter, with their distinct climatic characteristics, the insect-attracting system can target different highly harmful insects. This setup enhances the applicability of the insect-attracting system. Compared to traditional large-scale detection, this method intelligently adjusts detection strategies for different pest species in different seasons, resulting in high targeting and accuracy.

[0055] S2. The early warning module 300 sends a first control signal to the image acquisition module 120, including images of the spawning location and spawning site. At least one image acquisition module 120 receives the first control signal and switches from a low-quality working mode to a high-quality working mode. The image acquisition module 120 completes the image acquisition of the spawning location in high-quality working mode and transmits the acquired images back to the early warning module 300. The early warning module 300 sends a second control signal to the trapping module 200. The trapping module 200 adjusts its trapping parameters based on the received second control signal.

[0056] Low-quality acquisition mode can significantly reduce the amount of data that needs to be processed, while consuming less power. Low-quality acquisition mode can capture images of moving objects, such as flying insects. When it is necessary to detect the main breeding and growth areas of pests, high-quality acquisition mode can improve the accuracy of detection. Differentiated sampling between the main breeding and growth areas of pests and non-key areas where pests are not easy to grow and reproduce can reduce the amount of data that the early warning module 300 needs to transmit and process, improve the speed of data analysis and the feedback speed of pest detection, and is more conducive to real-time feedback of pest conditions and timely handling of pests.

[0057] Compared to traditional large-scale, low-target detection methods, this detection method can improve the accuracy and targeting of detection while reducing the amount of data processing, and can use limited storage space to store high-value image data, reducing unnecessary data processing.

[0058] Preferably, the trapping parameters are adjusted by adjusting at least one of the following: the setting height, heat source temperature, and emission spectrum of the trapping module 200, according to the physiological habits of the pests.

[0059] For example, when trapping locusts, the parameters of the trapping module 200 are adjusted at night to: the heat source temperature is 65℃, and at the same time, the ultraviolet light is emitted and coupled with the heat source to trap the locusts.

[0060] S3. The early warning module 300 analyzes the presence, quantity, and maturity of insect eggs based on the images of the oviposition site transmitted back by the image acquisition module 120, and generates a first pest infestation analysis result. This first pest infestation analysis result includes the predicted number of adult insects, the growth cycle of the insect eggs, and the number of eggs. The early warning module 300 issues a first-level early warning based on the analysis result indicating the presence of at least one pest egg, indicating the progress of the pest infestation. In other words, if... Figure 4 As shown, the early warning module 300 issues a first-level early warning based on the satisfaction of a first trigger condition that the number of highly harmful insect eggs in the key area exceeds a first threshold. The first trigger condition is that the number of highly harmful insect eggs in the key area exceeds the first threshold.

[0061] Preferably, the image acquisition module 120 is a camera; the camera is a 360-degree high-definition camera, and the camera is a 50x zoom camera; the cameras are distributed in the greenhouse; the camera can capture high-definition photos of plant leaves, identify the defects of plant leaves and stems, and identify adult insects, larvae and eggs on the plant.

[0062] Preferably, the image acquisition module 120 also includes a ground-mounted camera, which is used to capture images of the back of leaves and promptly detect the distribution of insects on the back of leaves.

[0063] Preferably, the steps for obtaining insect infestation analysis results include:

[0064] Convert the color image to a grayscale image, and then perform thresholding on the grayscale image to convert it into a binary image.

[0065] Label connected components in a binary graph;

[0066] Target identification is performed on adult insects and their eggs, while non-target insects are filtered out;

[0067] Record the number of insects and eggs;

[0068] Display the counting results.

[0069] Preferably, the analysis result indicating the presence of at least one pest egg is determined based on the number of eggs of at least one pest exceeding a first threshold. Preferably, the first threshold can be automatically set based on the pest's damage risk and the population's reproductive capacity. For example, for the rice stem borer, the first threshold is 60 egg masses per acre.

[0070] Preferably, the first-level warning includes the type of pest, the location where the pest is thriving in large numbers, the level of damage caused by the pest, the infestation progress, and the treatment plan for the current infestation. Monitoring personnel refer to the received first-level warning information and the treatment plan for the first-level warning to control the pest.

[0071] S4, such as Figure 4 As shown, the early warning module 300 issues a second-level early warning based on the cancellation of the first-level early warning and the satisfaction of a second triggering condition: the number of highly hazardous insect populations or the number of insects in a certain growth stage within the population exceeds a second threshold. For example, both nymphs and adults of locusts have the habit of feeding on crops, but nymphs can only jump and have not yet grown wings, so the risk level of damage to crops is relatively low; while winged locusts have strong migration capabilities and pose a greater threat to crops. The nymph stage is the optimal stage for killing locusts. The early warning module 300 generates a second-level early warning based on the satisfaction of a second triggering condition: the number of locusts in the nymph stage within the locust population exceeds a second threshold, reminding monitoring personnel that the current locust plague risk is high. Preferably, the second threshold can be changed according to the needs and plans of the users.

[0072] The second-level warning includes the type of pest, the location where the pest lives in large numbers, the level of damage caused by the pest, the progress of the pest infestation, and the treatment plan for the current infestation.

[0073] Preferably, the warning intensity of the second-level warning is greater than that of the first-level warning, to emphasize the higher risk level of the current pest infestation and remind operators to promptly eliminate the pests. For example, the first-level warning is a system message notification, which operators need to manually click to view; while the second-level warning is a system pop-up notification, which automatically appears at the top of any screen without requiring manual clicking, thus attracting the operator's attention.

[0074] Preferably, for pest pairs with synergistic relationships, cross-species association calibration can be performed: when the density of rice stem borer eggs reaches 70% of its threshold, the brown planthopper-specific monitoring protocol is activated, the image acquisition frequency of the leaf sheath is adjusted to once every 30 minutes, and a long-wave ultraviolet lamp (wavelength 365nm) is activated to enhance the visibility of fluorescent markers on the egg mass surface. Preferably, the early warning module 300 activates a targeted detection plan based on the analysis results that the number of eggs of at least one pest exceeds a first threshold, performs targeted detection on key areas where a large number of pest eggs are currently present, and promptly feeds back the detection results to update the pest situation progress.

[0075] Preferably, the detection scheme is as follows:

[0076] The early warning module 300 associates the types of pests with the climate information collected after the acquisition time point of the image used to generate the first-level early warning, predicts the actual development process of pests based on the real-time collected climate information, and generates the predicted key time nodes for the growth and development of pests in real time.

[0077] The early warning module 300 sends a second control signal, including resampling information, to the image acquisition module 120 at a time point two hours before the critical time point. The image acquisition module 120 then re-enters high-quality mode and acquires images based on the received second control signal. For example, the critical time point is the time when pest eggs hatch in large numbers, and the second time point is 2-3 days. The egg hatching prediction model integrates an optical feature analysis algorithm: after converting the acquired high-definition images to the Lab color space, it extracts the b-channel value of the central region of the egg mass (diameter ≥ 60% of the egg area). When the b-value decreases from the initial 28±2 to 22±1, it determines that the critical hatching period has begun and automatically sends an early warning message to all monitoring terminals within a 200m radius 24 hours in advance.

[0078] By comparing the re-collected images with the images used to generate the first-level warning, the current development of the pest infestation is determined. If a large number of eggs are about to hatch or have already hatched, it is determined that the first-level warning has not been lifted and the first-level warning is repeated to prompt management personnel to eliminate the eggs in a timely manner to avoid the risk and spread of pests. If a large number of eggs are killed, the first-level warning is lifted.

[0079] Preferably, the second-level early warning includes the development process of greenhouse insect infestation, values ​​of key time points, estimated number of hatched larvae, expected damage, and recommended treatment plans.

[0080] Preferably, the method for determining the treatment plan for the current pest infestation further includes: the early warning module 300 correlates the pest infestation data with the number of natural enemies, and reduces the recommended amount of pesticide when the population density of natural enemies increases. When the number of natural enemies of pests increases, the pest population size is predicted after a period of time based on the relationship between the population size of natural enemies and the population size of pests, and the recommended amount of pesticide is reduced. Combined with chemical and biological control methods, the pest control method is intelligently recommended to reduce pesticide usage. When the population density of natural enemies of pests decreases, the recommended amount of pesticide is increased after a period of time based on the relationship between the population size of natural enemies and the population size of pests, to promptly stop the increase in the pest population. Further, the improved Holling type II equation can be used to calculate the pesticide attenuation coefficient: Q=Q0×[1-(aA) / (1+aT)] h [A], where Q is the actual recommended dosage, Q0 is the baseline dosage, A is the number of natural enemies, a is the natural enemy attack efficiency coefficient, which can be 0.3, and T h The processing time (or natural enemy regulation lag time) can be set to 0.7 hours. When ≥3 adult lacewings are identified per square meter, the dosage of pyrethroid insecticides can be reduced to 55% of the baseline value, and yellow sticky traps (reflection wavelength 580-590nm) can be used to enhance natural enemy protection. Preferably, the number of natural enemies of the pest can be measured by image detection or other detection methods that do not harm natural enemies. For example, a buried camera can be installed at a 35° angle on the south side of the plant, 10-15cm from the base of the stem, and equipped with a ring light (color temperature 5500K, brightness 200lux). It takes pictures every 10 minutes from 06:00 to 18:00 every day. During image processing, the red pixel area with H value of 0-20° in the HSV color space is first extracted, and then noise is filtered out by morphological opening operation. Finally, the number of targets that meet the specified morphological characteristics (long axis 1.2-2.8mm, aspect ratio 1.1-1.5) is counted. Preferably, the increase and decrease in drug dosage can be calculated based on the current food conditions and the population relationships of natural enemies and pests in the environment. The population relationships of natural enemies and pests in the current food conditions and environment can be obtained through surveys or historical data.

[0081] For example, based on the pest population density, the current pesticide dosage is the first value. Based on the increased natural enemy population density, the resulting pest population density is calculated, leading to a lower pest population density than before, and thus a pesticide dosage lower than the first value is calculated. Similarly, when the natural enemy population density decreases, a pesticide dosage higher than the first value is calculated.

[0082] According to a preferred embodiment, the early warning module 300 associates the pest infestation data of a first pest and a second pest that have a cooperative relationship. When both the pest infestation data of the first pest and the pest infestation data of the second pest meet the first triggering condition for a first-level early warning, the first-level early warning is upgraded to a third-level early warning regarding both the first pest and the second pest. In other words, as... Figure 4 As shown, the early warning module issues a third-level early warning based on the failure to resolve the first-level early warning and the satisfaction of the second triggering condition that the population size is greater than the second threshold.

[0083] According to a preferred embodiment, when the early warning module 300 predicts that one of the first pest and the second pest is a highly harmful pest, it also lists the other pest as a highly harmful pest and controls the image acquisition module 120 to acquire image information of the key areas corresponding to the first pest and the second pest respectively.

[0084] For example, when the early warning module 300 predicts that the first pest is a highly harmful pest, the early warning module 300 also lists the second pest as a highly harmful pest; the early warning module 300 controls the image acquisition module 120 to acquire image information of the first key area corresponding to the first pest; at the same time, the early warning module 300 controls the image acquisition module 120 to acquire image information of the second key area corresponding to the second pest; the early warning module 300 analyzes the pest situation of the first pest and the second pest based on the received image information of the first key area and the second key area, respectively.

[0085] Simultaneously, the pest situation of the first pest and the second pest are analyzed. When the number of insect eggs in the key area corresponding to the first pest is greater than the first threshold corresponding to the first pest, that is, when the pest situation of the first pest meets the triggering condition of the first level warning; and when the number of insect eggs in the key area corresponding to the second pest is greater than the first threshold corresponding to the second pest, that is, when the pest situation of the second pest meets the triggering condition of the first level warning, the warning module 300 issues a third level warning for the first pest and the second pest.

[0086] Preferably, the third-level early warning includes the specific pest situation of the first pest, the specific pest situation of the second pest, the cooperation mechanism and cooperation intensity of the first and second pests, and the control methods for the first and second pests.

[0087] For example, the primary pest is the rice stem borer, and the secondary pest is the brown planthopper. Both are common pests of rice, and studies have found that they have a cooperative relationship; the presence of one pest reduces the rice's ability to control the other, thus creating favorable conditions for the survival of the other pest. After the rice stem borer infests rice, the content of free amino acids in the rice increases, while the content of defense substances such as sterols decreases. This significantly promotes the growth of the brown planthopper. Specifically, a detection threshold for free amino acid concentration can be set at 1.2–1.8 mg / g. When near-infrared spectroscopy detects an absorbance increase Δ≥0.15 at a wavelength of 1720 nm on the leaves, the monitoring weight of the brown planthopper is automatically increased. Furthermore, when the rice stem borer egg density exceeds 50 eggs / m²... 2 Three monitoring points were set up radiating outwards from the base of the plant in the designated area, with each point spaced 2 meters apart. The focus was on collecting macro images (30x magnification) at a depth of 3 cm below the leaf sheath. Simultaneously, the behavior of the rice stem borer infesting rice induces the production of volatiles in the rice plant. These volatiles significantly repel the rice planthopper's natural enemy, the rice wasp *Agrostis oryzae*, significantly reducing the risk of *Agrostis oryzae* parasitizing *Agrostis oryzae* eggs. Therefore, rice stem borer infestation creates favorable conditions for the growth of the rice planthopper. When the rice planthopper infests rice, it significantly inhibits the rice's defense response, downregulating the expression of related defense genes and significantly reducing the content of protease inhibitors. This reduces the negative impact of the rice's defense response on the rice stem borer larvae. Therefore, rice planthopper infestation creates favorable conditions for the growth of the rice stem borer. Thus, when one pest is present, the probability of the other pest appearing and causing damage is very high. Furthermore, the damage to crops is greater when both pests appear simultaneously than when either appears alone, making control more difficult and requiring preventative measures in advance.

[0088] When pests that cooperate with each other infest the same plant, they disrupt the plant's normal defense mechanisms, reducing the plant's ability to defend itself against pests. This makes it easier for pests to grow and reproduce on the plant, leading to rapid infestation. The simultaneous presence of cooperative pests poses a greater threat to the plant and increases the probability of infestation. Therefore, timely intervention is necessary to eliminate risks and prevent infestations. Through the aforementioned settings, when a primary and secondary pest with a cooperative relationship appear simultaneously, the early warning module 300 can raise the warning level. If either the primary or secondary pest individually reaches only the first warning level, the warning level will be raised to the third level, indicating the faster-growing primary and secondary pests and drawing the attention of personnel. Even personnel lacking basic pest knowledge can use the warning information from the early warning module 300 to prioritize more urgent infestations, scientifically and effectively controlling pests and avoiding losses caused by pests.

[0089] Preferably, when the early warning module 300 predicts that the first pest is a highly damaging pest, it determines the stage of the pest infestation. If the first pest is in its first stage, the early warning module 300 controls the image acquisition module 120 to acquire image information of key areas of the second pest. If the second pest is in its second stage, the early warning module 300 allocates its early warning resources to the second pest, reducing its focus on the first pest. Similarly, if the second pest is in its first stage, the early warning module 300 allocates its early warning resources to the first pest, reducing its focus on the second pest. If the first pest is in its second stage, the early warning module 300 controls the image acquisition module 120 to acquire image information of key areas of the second pest. If the second pest is in its second stage, the early warning module 300 simultaneously allocates its early warning resources to both the first and second pests, and sends a preliminary early warning message to the outside world. The preliminary early warning message serves as a reminder that there is a high probability of synergistic effects between the first and second pests in the current planting area, requiring a further increase in alert levels. The process of assessing pest infestation includes at least the degree of resistance triggered by the pest. The degree of resistance refers to the strength of natural forces that reduce the harmfulness of the pest. Natural forces include natural factors other than human intervention, such as the plant's own secretion of resistance substances and the reproduction of the pest's natural enemies.

[0090] Preferably, the first stage is a stage of mass egg production, for example, when the triggering conditions for issuing a first-level warning are met but the triggering conditions for a second-level warning are not met.

[0091] Preferably, the second stage is a stage in which a large number of adults exist, for example, the population size of the second pest is greater than a second threshold.

[0092] The advantage of this setup is that the interaction between synergistic pests is that adults prefer to feed and / or lay eggs on plants damaged by each other, thus the number of adults is closely related to the degree of synergy. By paying more attention to pests in the second stage, the synergistic effect between pests can be effectively eliminated with the same amount of labor and time.

[0093] "Reassigning early warning resources to the second pest" could, for example, increase the number of trapping modules 200 used to specifically attract the second pest and / or increase the number of image acquisition modules 120 used to acquire images of key areas of the second pest, so as to obtain timely and accurate information on changes in the second pest situation.

[0094] The first and second stages mentioned above can also be divided according to the growth stages of the pests, such as the nymph stage and the adult stage. At least some pests will have different interactions with other pests at different stages. For example, when the first pest is in the nymph stage, it competes with the second pest. When the number of the first pest is dominant, it will significantly reduce the number of the second pest. However, when the first pest enters the adult stage, the associated effects of its damage to the plant cause it to have a synergistic effect with the second pest, which will further intensify the damage to the plant. Traditional pest control tends to kill as soon as it is discovered, leaving no one behind. However, this will result in multiple and large amounts of pesticides being released into the planting environment, which will also affect the growth and quality of the plants. The natural relationship between several pests, plants and the environment is rarely considered. When a large number of human external conditions are involved, the uncontrollable risks increase, which is detrimental to controlling plant quality. This design is based on the natural relationship between pests, plants, and the environment. It selectively allocates the processing resources of the early warning module 300 by taking into account the resistance of natural forces to the greatest extent possible. This allows the pest-attracting system to focus on more important pest situations and leave some pest control to natural processes. As a result, it can improve the naturalness of the planted products, increase their economic value, and reduce the investment of monitoring and control resources.

[0095] According to a preferred embodiment, such as Figure 5 As shown, the present invention also discloses a method for attracting insects to greenhouse crops, which can be implemented using the aforementioned insect attracting system.

[0096] Preferably, the insect-attracting method may include the following steps:

[0097] Based on climate information and the physiological habits of pests, highly harmful insect species that are strongly correlated with current environmental information are identified.

[0098] Search among the captured pests for highly harmful insects;

[0099] In the presence of highly harmful insects, analyze whether there are highly harmful insect eggs in key areas and assess the insect infestation situation;

[0100] When the number of highly harmful insect eggs in a key area exceeds the first threshold, the capture parameters of the trapping module are adjusted according to the type of highly harmful insect in order to capture the highly harmful insect in a targeted manner.

[0101] The population size of highly harmful insects is predicted based on the species and proportion of the captured highly harmful insects. When the population size exceeds a second threshold, image data from non-key areas is collected to accurately determine the insect infestation situation.

[0102] It should be noted that the specific embodiments described above are exemplary. Those skilled in the art can devise various solutions inspired by the disclosure of this invention, and these solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of this invention is defined by the claims and their equivalents. This specification contains multiple inventive concepts; terms such as "preferredly," "according to a preferred embodiment," or "optionally" indicate that the corresponding paragraph discloses an independent concept. The applicant reserves the right to file divisional applications based on each inventive concept. Throughout the text, features introduced by "preferredly" are merely optional and should not be construed as mandatory. Therefore, the applicant reserves the right to abandon or delete relevant preferred features at any time.

Claims

1. An insect-attracting system for greenhouse crops, comprising a detection module (100) for collecting climate information and image information around the plant, and a trapping module (200) for attracting insects and identifying the species and number of insects. Its features are, It also includes an early warning module (300) that is data-connected to the trapping module (200) and the detection module (100). The early warning module (300) processes the data information of the trapping module (200) and the detection module (100) and feeds back corresponding control signals to the trapping module (200) and the detection module (100) to adjust the operating parameters of the trapping module (200) and the detection module (100), so that the trapping module (200) and the detection module (100) work in a cooperative mode against at least the same pest. The early warning module (300) generates early warning information for notifying operators of the pest infestation process in advance based on the data transmitted by the detection module (100) and the trapping module (200) operating in collaborative mode; The detection module (100) includes a climate detection module (110) and an image acquisition module (120). The early warning module (300) predicts highly harmful insect species that are strongly correlated with the current environmental information based on the climate information collected by the climate detection module (110) and the data in the preset database. The early warning module (300) determines whether any highly harmful adult insects have been captured based on the data of captured pests sent by the trapping module (200). When the early warning module (300) determines that there are adult insects of highly harmful insects, the early warning module (300) controls the image acquisition module (120) to acquire images of key areas where highly harmful insects occur, and the early warning module (300) analyzes the insect situation of highly harmful insects based on the image data acquired by the image acquisition module (120). The early warning module (300) controls the trapping module (200) to target and capture high-risk insects by adjusting its own capture parameters according to the type of high-risk insects, based on the satisfaction of the first triggering condition. The first triggering condition is that the number of high-risk insect eggs in the key area is greater than a first threshold. The early warning module (300) predicts the population size of highly harmful insects based on the species and quantity ratio of highly harmful insects captured by the trapping module (200). The early warning module (300) controls the detection module (100) to collect images of non-key areas to accurately determine the insect situation of highly harmful insects based on the satisfaction of a second triggering condition. The second triggering condition is that the population size of the highly harmful insects is greater than a second threshold.

2. The insect-attracting system according to claim 1, characterized in that, The early warning information includes a first-level early warning, a second-level early warning, and a third-level early warning, with the warning intensity increasing sequentially. The early warning module (300) issues a first-level early warning based on the satisfaction of a first triggering condition that the number of highly harmful insect eggs in the key area is greater than a first threshold. The early warning module (300) issues a second-level early warning based on the cancellation of the first-level early warning and the satisfaction of a second triggering condition that the number of highly harmful insect populations is greater than a second threshold. The early warning module (300) issues a third-level early warning based on the fulfillment of the second triggering condition that the first-level early warning prompt has not been lifted and the number of highly harmful insect species is greater than the second threshold.

3. The insect-attracting system according to claim 2, characterized in that, The early warning module (300) controls the image acquisition module (120) to start the detection scheme and update the insect situation based on the satisfaction of a first trigger condition that the number of highly harmful insect eggs in the key area is greater than a first threshold. The specific detection scheme is as follows: The early warning module (300) associates the climate information collected after the acquisition time point of the image used to generate the first-level early warning with the type of pest, predicts the actual development process of the pest based on the real-time collected climate information, and generates the predicted key time nodes for the growth and development of the pest in real time. The early warning module (300) controls the image acquisition module (120) to re-acquire images of the key area at a time point one hour before the arrival of the key time point.

4. The insect-attracting system according to claim 1, characterized in that, The early warning includes control measures for the current insect infestation. The early warning module (300) also associates the first pest infestation with the second pest infestation. Based on the existence of the first pest infestation, the early warning module (300) adds a prevention and / or control plan for the second pest to the control plan for the first pest infestation.

5. The insect-attracting system according to claim 1, characterized in that, The image acquisition module (120) of the key area switches from the first working mode to a second working mode with higher power consumption based on the first control signal sent by the early warning module (300) that includes information on image acquisition of the key area. The image acquired in the second working mode has a higher clarity than the image acquired in the first working mode.

6. The insect-attracting system according to claim 2, characterized in that, When there is a first-level warning for multiple highly harmful insects, the warning module (300) allocates the surrounding image acquisition module (120) to acquire images of the key areas of different highly harmful insects based on the location parameters of multiple key areas in the greenhouse environment, so that the insect-attracting system can detect the insect situation of multiple highly harmful insects at the same time, and publish the insect situation of multiple pests in the warning according to the harm level of the insect situation.

7. A method for attracting insects to greenhouse crops, comprising using the insect-attracting system as described in any one of claims 1 to 6, characterized in that, It includes the following steps: Based on climate information and the physiological habits of pests, highly harmful insect species that are strongly correlated with current environmental information are identified. Search among the captured pests for highly harmful insects; In the presence of highly harmful insects, analyze whether there are highly harmful insect eggs in key areas and assess the insect infestation situation; When the number of highly harmful insect eggs in a key area exceeds the first threshold, the capture parameters of the trapping module (200) are adjusted according to the type of highly harmful insect in order to capture the highly harmful insect in a targeted manner. The population size of highly harmful insects is predicted based on the species and proportion of the captured highly harmful insects. When the population size exceeds a second threshold, image data from non-key areas is collected to accurately determine the insect infestation situation.

Images