Invasive pest intelligent monitoring and trapping system and method for tropical fruits
By integrating piezoelectric acoustic sensors, LED spectral light sources, and chemical attractant release components through multimodal signal acquisition and intelligent recognition technology, and combining them with a database of tropical pest behavior characteristics, the problem of environmental interference in traditional monitoring technologies has been solved, enabling accurate monitoring and efficient trapping of tropical fruit pests.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-21
- Publication Date
- 2026-03-13
AI Technical Summary
Traditional tropical fruit pest monitoring technologies are susceptible to environmental interference, have low identification accuracy, and lack targeted trapping methods, leading to resource waste and ecological imbalance, and failing to meet the needs of precise monitoring.
Employing multimodal signal acquisition and intelligent recognition technology, this system integrates piezoelectric acoustic sensors, LED precision spectral light sources, chemical attractant release components, and high-definition visual acquisition components. Combined with a database of tropical invasive pest behavior characteristics, it adaptively adjusts recognition priorities and eliminates environmental interference through the calculation of multimodal signal effectiveness index and modal dynamic weight coefficient, thereby achieving accurate classification and statistics of target pests.
It improves the accuracy of pest identification, reduces interference with non-target beneficial insects, provides data support for trapping efficiency index and beneficial insect false trapping rate, and ensures data reliability and trapping effect.
Smart Images
Figure CN121658997A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of pest control technology, specifically to an intelligent monitoring and trapping system and method for invasive pests on tropical fruits. Background Technology
[0002] In the tropical fruit cultivation industry, the monitoring and control of invasive pests is a key link in ensuring yield and quality. The hot and rainy climate provides a suitable living environment for invasive pests, and the pests are diverse with significant differences in behavior characteristics. For example, some pests are phototactic, chemotactic, or acoustic. In addition, their host range is wide, making it difficult for traditional monitoring methods to accurately capture pest dynamics. It is necessary to deploy intelligent monitoring equipment and combine it with Internet of Things technology to track the distribution of pest populations in real time, providing data support for early warning and precise control.
[0003] However, existing tropical fruit pest monitoring technologies mainly rely on single-modal monitoring methods, such as visual image recognition or chemical attractant monitoring. These methods have significant drawbacks in highly complex environments: single-modal data is easily affected by environmental interference, such as wind noise, changes in light, or the activity of non-target organisms, leading to a decrease in recognition accuracy; at the same time, traditional trapping methods lack specificity, easily trapping beneficial insects and disrupting the ecological balance; in addition, trapping devices with static signal output cannot dynamically adjust the trapping strategy according to the behavioral characteristics of pests, resulting in low trapping efficiency and serious waste of resources. In tropical fruit growing areas, due to the diversity of pest species and large differences in behavior, traditional technologies are difficult to meet the needs of accurate monitoring.
[0004] Therefore, a smart monitoring and trapping system and method for invasive pests on tropical fruits were developed. Summary of the Invention
[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an intelligent monitoring and trapping system and method for invasive pests in tropical fruits. This invention improves the accuracy and environmental adaptability of invasive pest monitoring in tropical fruits through multimodal signal acquisition and intelligent recognition. By integrating a piezoelectric acoustic sensor, a precise LED spectral light source, a chemical attractant release component, and a high-definition visual acquisition component, it simultaneously acquires acoustic vibration, phototaxis, chemotaxis, and morphological characteristic data of pests. Based on a database of behavioral characteristics of invasive tropical pests, it quantifies the effective proportion of acoustic, optical, chemical, and visual signals through multimodal signal effectiveness index calculation, eliminating environmental interference and ensuring data reliability. Simultaneously, by combining phototaxis in moths and chemotaxis in fruit flies, and using a modal dynamic weighting coefficient calculation formula, it adaptively adjusts the recognition priority of each modality, completing the classification and statistics of target pests. This solves the problem of traditional single-modal monitoring being easily affected by environmental interference and improves the accuracy of pest identification in complex scenarios.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On one hand, an intelligent monitoring and trapping system for invasive pests on tropical fruits, the system comprising:
[0007] Multimodal signal acquisition module: integrates piezoelectric miniature acoustic sensor, LED precision spectral light source, chemical attractant release component and high-definition visual acquisition component to simultaneously acquire multimodal data of pests;
[0008] Data purification and identification module: Through the database of behavioral characteristics of tropical invasive pests, and based on the acquired multimodal data of pests, the module calculates the multimodal signal effectiveness index and modality dynamic weight coefficient, and then performs data processing and species identification, and outputs the number of identified target pests.
[0009] Targeted signal modulation module: Based on species identification results and dynamic weight coefficients of each modality, calculates the target signal enhancement intensity and adjusts the signal intensity of each modality;
[0010] Precision trapping execution module: Deploys trapping components to cover the detection area, catches target pests attracted by the target signal, has an internal trapping unit to adsorb the insects that enter, and integrates an image recognition and infrared sensing counting unit to distinguish between target pests and beneficial insects, and counts the number of target pests and beneficial insects captured.
[0011] The monitoring report generation module calculates the trapping efficiency index and the false trapping rate of beneficial insects based on the number of target pests captured, the number of beneficial insects captured, and generates a monitoring report. It also transmits multimodal data, species identification results, trapping statistics and monitoring reports to a remote terminal for remote monitoring through an IoT network-based communication unit.
[0012] Furthermore, in the multimodal signal acquisition module,
[0013] Piezoelectric miniature acoustic sensor: frequency response range of 100~5000Hz, sensitivity of -40~-35dBV / Pa, for collecting real-time acoustic signal intensity of insect wing vibrations. ;
[0014] LED Precision Spectral Light Source: Outputs phototactic peak light in the 450~520nm blue light band, with a power adjustment range of 0~5W and a emission angle of 120~140°, collecting real-time optical reflection response intensity of phototactic pests. ;
[0015] Chemical attractant release assembly: includes a 50-60 mL reservoir, a quantitative release valve, an adjustable attractant release rate of 0.1-1 mL / h, and outputs the real-time chemical attractant concentration of the monitored area. ;
[0016] High-definition visual acquisition component: Equipped with an infrared supplementary light unit, the shooting frame rate is 1 frame every 3-5 seconds, and the visual feature matching degree of the output pest image is high. .
[0017] Furthermore, in the data purification and identification module, the tropical invasive pest behavior characteristic database is constructed using historical data and stores behavioral characteristic parameters and multimodal data benchmark values of tropical invasive pests, based on the acquired acoustic signal intensity. Optical reflection response intensity Chemical attractant concentration Visual feature matching degree The multimodal signal effectiveness index is calculated using the multimodal signal effectiveness index calculation formula. Quantify the proportion of effective target signals through a multimodal signal effectiveness index. Environmental interferences such as wind noise and leaf friction were eliminated. Furthermore, considering the behavioral differences among various invasive tropical pests, behavioral parameters such as phototaxis in moths, chemotaxis in fruit flies, and phonotaxis in thrips from the built-in database were used to calculate dynamic weighting coefficients using the modal dynamic weighting coefficient calculation formula. The system dynamically adapts the weight ratio of each modality, then integrates the features of each modality to identify pest species, and outputs the number of identified target pests. .
[0018] Furthermore, in the data purification and identification module, the formula for calculating the multimodal signal validity index is as follows: ,in, It is a multimodal signal effectiveness index. For real-time acoustic signal strength, As the acoustic effective signal reference value, For real-time optical reflection response intensity, This serves as a baseline value for the response intensity of phototactic pests. This represents the real-time concentration of the chemical attractant. This is the baseline concentration of the pheromone for the target pest. For visual feature matching degree, As the baseline value for image features, The basic weights for each modality type.
[0019] Furthermore, in the data purification and identification module, the formula for calculating the modal dynamic weight coefficient is as follows: ,in, for The dynamic weighting coefficients of the modality, These are acoustic, optical, chemical, and visual modalities, respectively. for The signal effectiveness index of the modality, This represents the behavioral characteristic coefficient of the target pest. For acoustic, optical, chemical and visual modalities Perform summation.
[0020] Furthermore, in the targeted signal modulation module, based on species identification results and modal dynamic weighting coefficients... The target signal enhancement intensity is calculated using the target signal enhancement intensity calculation formula. The formula for calculating the enhancement intensity of the target signal is: ,in, for The enhancement intensity of the target signal in each modality for The basic output value of each modality For the enhancement coefficient, for Dynamic weighting coefficients for modalities; enhancement of acoustic signal output for sound-attracting pests; and strengthening intensity based on target signal. Determine the output value of each mode; for phototactic pests, increase the light intensity of the light source in the specified wavelength; for chemotactic pests, increase the amount of attractant released.
[0021] Furthermore, in the monitoring report generation module, the number of captured target pests, the number of beneficial insects, and the number of identified target pests are considered. The trapping efficiency index is calculated using the trapping efficiency index calculation formula. Meanwhile, the beneficial insect miscible rate was calculated using the formula for calculating the beneficial insect miscible rate. And based on the trapping efficiency index and the rate of accidental capture of beneficial insects This generates monitoring reports that quantify the trapping effect and the proportion of non-target organisms accidentally trapped.
[0022] Furthermore, in the monitoring report generation module, the formula for calculating the trapping efficiency index is: ,in, The trapping efficiency index. The number of target pests captured. The number of target pests identified. For the first The dynamic weighting coefficients of each modality for The enhancement intensity of the target signal in each modality for The basic output value of each modality.
[0023] Furthermore, in the monitoring report generation module, the formula for calculating the beneficial insect false capture rate is: ,in, To reduce the rate of accidental capture of beneficial insects, The number of beneficial insects captured. The number of target pests captured.
[0024] On the other hand, a smart monitoring and trapping method for invasive pests on tropical fruits includes the following steps:
[0025] Multimodal signal acquisition: Integrating piezoelectric miniature acoustic sensors, LED precision spectral light sources, chemical attractant release components, and high-definition visual acquisition components, it simultaneously acquires multimodal data of pests;
[0026] Data purification and identification: Through the database of behavioral characteristics of tropical invasive pests, and based on the acquired multimodal data of pests, the multimodal signal effectiveness index and modality dynamic weight coefficient are calculated, and then data processing and species identification are performed, and the number of identified target pests is output.
[0027] Targeted signal modulation: Based on species identification results and dynamic weighting coefficients of each modality, the intensity of the targeted signal enhancement is calculated, and the intensity of each modality signal is adjusted.
[0028] Precise trapping execution: Deploy trapping components to cover the detection area, catch target pests attracted by the target signal, and have built-in trapping units to adsorb the insects that enter. It also integrates image recognition and infrared sensing counting units to distinguish between target pests and beneficial insects, and count the number of target pests and beneficial insects captured.
[0029] Monitoring report generation: Based on the number of target pests captured, the number of beneficial insects, and the number of identified target pests, the trapping efficiency index and the false trapping rate of beneficial insects are calculated, and a monitoring report is generated. Furthermore, through a communication unit based on an Internet of Things network, multimodal data, species identification results, trapping statistics, and monitoring reports are transmitted to a remote terminal for remote monitoring.
[0030] Compared with existing technologies, this intelligent monitoring and trapping system and method for invasive pests on tropical fruits has the following advantages:
[0031] I. This invention improves the accuracy and environmental adaptability of monitoring pests infesting tropical fruits through multimodal signal acquisition and intelligent recognition. By integrating piezoelectric acoustic sensors, LED precision spectral light sources, chemical attractant release components, and high-definition visual acquisition components, it simultaneously acquires data on the acoustic vibration, phototaxis, chemotaxis, and morphological characteristics of pests. Based on a database of behavioral characteristics of tropical infesting pests, it quantifies the effective proportion of acoustic, optical, chemical, and visual signals through multimodal signal effectiveness index calculation, eliminating environmental interference and ensuring data reliability. At the same time, by combining behavioral characteristic parameters such as phototaxis in moths and chemotaxis in fruit flies, and using the modal dynamic weight coefficient calculation formula, it adaptively adjusts the recognition priority of each modality to complete the classification and statistics of target pests. This solves the problem of traditional single-modal monitoring being susceptible to environmental interference and improves the accuracy of pest identification in complex scenarios.
[0032] II. This invention dynamically adjusts the output intensity of acoustic, optical, and chemical trapping signals by using species identification results and modal dynamic weighting coefficients, and employing a formula for calculating the enhancement intensity of targeted signals. For acoustically attracted pests, the acoustic signal output is enhanced; for phototactic pests, the light intensity of the light source is increased; and for chemotactic pests, the release of chemical attractants is increased, forming a trapping mode tailored to the specific pest. This mechanism reduces interference with non-target beneficial insects through the directional enhancement of modal signals. Simultaneously, combined with the adsorption function of the trapping unit, it captures target pests attracted by the targeted signals. Based on the number of captured target pests, the number of beneficial insects, and the identification results, the trapping efficiency index and the false capture rate of beneficial insects are quantified, providing data support for optimizing signal control strategies.
[0033] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description
[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.
[0035] Figure 1 A flowchart of an intelligent monitoring and trapping method for invasive pests on tropical fruits;
[0036] Figure 2 This is a framework diagram of an intelligent monitoring and trapping system for invasive pests on tropical fruits. Detailed Implementation
[0037] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.
[0038] Example 1:
[0039] Multimodal signal acquisition module: In a mango orchard monitoring and trapping scenario involving fruit flies and thrips, during the mango fruiting period, when the trees are densely covered with branches and leaves and the fruit is ripening, fruit flies are easily attracted by the smell of the fruit, and thrips often gather on the tender shoots and fruit surface. In this environment, a piezoelectric miniature acoustic sensor with a frequency response range of 100~5000Hz and a sensitivity of -40~-35dBV / Pa is integrated to collect the real-time acoustic signal intensity generated by thrips wing vibrations and crawling close to the tender shoots of the mango tree; it outputs a 450~520nm blue light band... A precise LED spectral light source, installed in the lower middle part of the tree, acquires the real-time optical reflectance response intensity of phototactic pests. A chemical attractant release component is fixed next to branches with dense fruit, releasing fruit fly-sensitive attractants at a rate of 0.1-1 mL / h through a 50-60 mL storage chamber and a quantitative release valve, and outputs the real-time chemical attractant concentration. A high-definition visual acquisition component equipped with an infrared supplementary lighting unit is set up high between rows, capturing an image covering the tree canopy and fruit every 3-5 seconds, and outputting the visual feature matching degree of the pest image, such as... Figure 1 As shown.
[0040] Data purification and identification module: A gentle breeze often rustles through the branches and leaves around the mango orchard, occasionally accompanied by the sound of leaves rubbing together, which can easily interfere with signal acquisition. In this case, the module utilizes behavioral characteristic parameters such as chemotaxis in fruit flies and acoustic attraction in thrips, as well as multimodal data baseline values stored in the tropical invasive pest behavior characteristic database. Based on the acquired acoustic signal intensity, optical reflection response intensity, chemical attractant concentration, and visual feature matching degree, the module calculates the multimodal signal effectiveness index using the following formula: ,in, It is a multimodal signal effectiveness index. For real-time acoustic signal strength, As the acoustic effective signal reference value, For real-time optical reflection response intensity, This serves as a baseline value for the response intensity of phototactic pests. This represents the real-time concentration of the chemical attractant. This is the baseline concentration of the pheromone for the target pest. For visual feature matching degree, As the baseline value for image features, The basic weights for each modality are established; the proportion of effective target signals is quantified, and environmental interference from wind noise and leaf friction is eliminated; simultaneously, considering the behavioral differences between fruit flies preferring to move around fruits and thrips preferring to crawl on tender shoots, the dynamic weight coefficients for each modality are calculated using the modal dynamic weight coefficient calculation formula. The modal dynamic weight coefficient calculation formula is as follows: ,in, for The dynamic weighting coefficients of the modality, These are acoustic, optical, chemical, and visual modalities, respectively. for The signal effectiveness index of the modality, This represents the behavioral characteristic coefficient of the target pest. For acoustic, optical, chemical and visual modalities The system performs summation; after dynamically adapting the weight ratio of each modality, it integrates multimodal features to complete the species identification of fruit flies and thrips, and outputs the number of identified fruit flies and thrips.
[0041] Targeted signal modulation module: Considering that fruit flies are sensitive to fruit odor during the mango fruiting period, while thrips are mostly active near tender shoots, based on the identification results of fruit flies and thrips and the dynamic weight coefficients of each mode, the enhancement intensity of the chemical mode and the acoustic mode are calculated separately using the targeted signal enhancement intensity calculation formula. The targeted signal enhancement intensity calculation formula is as follows: ,in, for The enhancement intensity of the target signal in each modality for The basic output value of each modality For the enhancement coefficient, for Dynamic weighting coefficients for modalities; enhancement of acoustic signal output for sound-attracting pests; and strengthening intensity based on target signal. The output values for each mode were determined. For phototactic pests, the light intensity of the light source in the specified wavelength was increased; for chemotactic pests, the amount of attractant released was increased; for the chemotactic characteristics of fruit flies, the amount of chemical attractant released was increased so that the attractant odor could more easily cover the area around the fruit; for the acoustic characteristics of thrips, the acoustic signal output was enhanced to ensure that the signal could be transmitted to the dense area of tender shoots; and the LED precision spectral light source and chemical attractant release components were adapted and adjusted according to the basic output values.
[0042] Precision Trapping Module: To cover key areas of the mango orchard, such as the canopy, fruit, and tender shoots, trapping components are evenly distributed in the lower and middle parts of the mango canopy and between rows. This ensures that the attraction range of each trapping component can seamlessly cover the entire detection area. The trapping components directly capture fruit flies and thrips attracted by the target signal. The built-in insect trapping unit adsorbs the insects that enter, preventing pests from escaping. At the same time, an image recognition and infrared sensing counting unit is integrated. Image recognition captures the body shape characteristics of fruit flies and the tiny size of thrips. Combined with the thermal signals of the insects captured by infrared sensing, the dual feature fusion distinguishes between target pests and beneficial insects, and automatically counts the number of target pests such as fruit flies and thrips captured, as well as the number of beneficial insects such as ladybugs that are accidentally captured.
[0043] The monitoring report generation module combines the acquired number of target pests, beneficial insects, and identified fruit flies and thrips pests to calculate the trapping efficiency index using the trapping efficiency index calculation formula. The trapping efficiency index calculation formula is as follows: ,in, The trapping efficiency index. The number of target pests captured. The number of target pests identified. For the first The dynamic weighting coefficients of each modality for The enhancement intensity of the target signal in each modality for The basic output value of each modality; the beneficial insect false capture rate is calculated using the beneficial insect false capture rate calculation formula, which is: ,in, To reduce the rate of accidental capture of beneficial insects, The number of beneficial insects captured. The system measures the number of target pests captured and generates a monitoring report that clearly quantifies the trapping effect of fruit flies and thrips in mango orchards, as well as the proportion of beneficial insects accidentally caught, based on the trapping efficiency index and the beneficial insects accidental trapping rate. Furthermore, through an IoT-based communication unit, the system transmits multimodal data, species identification results, trapping statistics, and monitoring reports to a remote terminal for remote monitoring, providing data support for subsequent adjustments to targeted pest control measures in mango orchards.
[0044] In summary, by combining the multimodal signal acquisition module with the environment of dense tree foliage and ripening fruit, and by deploying the acquisition components in tender shoots, the lower and middle parts of the tree, branches with dense fruit, and high places between rows, multimodal data of two types of pests can be obtained. The data purification and identification module effectively removes interference from wind noise and leaf friction in the orchard, and completes species identification by combining the differences in pest behavior. The targeted signal regulation module specifically enhances chemical and acoustic signals to adapt to the chemotaxis of fruit flies and the acoustic attraction of thrips. The precise trapping execution module covers key areas such as the canopy and fruit to stably capture pests. The monitoring report generation module quantifies the trapping effect and the proportion of beneficial insects accidentally trapped, providing data support for adjusting pest control measures in mango orchards and fully adapting to the monitoring and trapping needs during the fruiting period.
[0045] Example 2:
[0046] Multimodal signal acquisition module: In a banana plantation moth and fruit fly monitoring and trapping scenario, the banana plantation is in the budding stage, with vigorous plant growth, large leaves, and high field humidity. Moths tend to move around the banana buds at night, while fruit flies often approach ripe banana fruits. Under these conditions, an integrated piezoelectric miniature acoustic sensor is installed in the middle of the banana plantation to collect the real-time acoustic signal intensity generated by the crawling and wing vibrations of pests in the plantation. Its frequency response range is 100~5000Hz, and its sensitivity is -40~-35dBV / Pa. A precise LED spectral light source is fixed on a bracket near the banana buds, outputting 4... The phototactic peak light in the 50-520nm blue light band is used to obtain the real-time optical reflection response intensity of nocturnal moths by adjusting the power range from 0 to 5W and the emission angle from 120 to 140°. The chemical attractant release component is placed under the banana fruit bunch and releases fruit fly pheromones at a rate of 0.1 to 1 mL / h through a 50-60 mL storage chamber and a quantitative release valve, outputting the real-time chemical attractant concentration. The high-definition visual acquisition component is equipped with an infrared supplementary light unit and is set up on a field post to capture an image covering the banana buds, fruit, and base of the plant every 3-5 seconds to obtain the visual feature matching degree of the pest images.
[0047] Data purification and identification module: Occasionally, wind blowing through banana leaves can generate interference signals. In such cases, the module utilizes behavioral parameters related to phototaxis in moths and chemotaxis in fruit flies from the tropical invasive pest behavior database, along with multimodal data baselines. Based on the collected acoustic signal intensity, optical reflectance response intensity, chemical attractant concentration, and visual feature matching degree, the module calculates the multimodal signal effectiveness index using the following formula: ; Environmental interference signals caused by leaf movement and humidity were eliminated; then, considering the behavioral differences between moths' phototaxis at night and fruit flies' chemotaxis during the day, the dynamic weight coefficients of each mode were calculated using the modal dynamic weight coefficient calculation formula. The modal dynamic weight coefficient calculation formula is as follows: After dynamically adjusting the weights of each modality and integrating the multimodal features, the species identification of moths and fruit flies is completed, and the number of identified moths and fruit flies is output, such as... Figure 2 As shown.
[0048] Targeted signal modulation module: Considering that moths are mostly active at night and sensitive to light, while fruit flies are easily attracted by the simulated scent of ripe bananas, the targeted signal enhancement intensity of the optical and chemical modes is calculated based on the identification results of moths and fruit flies and the corresponding modal dynamic weighting coefficients, using the targeted signal enhancement intensity calculation formula: To address the phototaxis of moths, the intensity of the LED precision spectral light source is increased at night, allowing the light to attract moths more effectively. To address the chemotaxis of fruit flies, the release of chemical attractants is increased to ensure that the attractant odor covers the fruit area. The integrated piezoelectric micro acoustic sensor and chemical attractant release component are adjusted according to the appropriate basic output values to suit the environmental characteristics of banana plantations.
[0049] Precision Trapping Module: To cover the activity areas of banana buds, fruits, and the surrounding area of the plant during the banana budding stage, trapping components are arranged along the rows of banana plants and near the base of the plants. This ensures that the trapping range covers the key parts of each banana plant. The trapping components attract moths and fruit flies that are drawn by the target signal. The built-in insect trapping unit adsorbs the insects that enter. It also integrates an image recognition and infrared sensing counting unit. Through image recognition, it captures the wing span characteristics of moths and the body color characteristics of fruit flies. Combined with the thermal signals of insect movement captured by infrared sensing, it accurately distinguishes between target pests and beneficial insects, and automatically counts the number of moths and fruit flies captured, as well as the number of beneficial insects such as lacewings that are accidentally captured.
[0050] The monitoring report generation module calculates the trapping efficiency index based on the number of target pests (such as moths and fruit flies) captured, the number of beneficial insects, and the number of identified pests. The trapping efficiency index calculation formula is as follows: The beneficial insect false-catching rate is calculated using the formula: Based on the trapping efficiency index and the beneficial insect accidental trapping rate, a monitoring report is generated that quantifies the trapping effect of moths and fruit flies in banana orchards, as well as the proportion of non-target organisms accidentally trapped. Furthermore, through a communication unit based on an Internet of Things network, multimodal data, species identification results, trapping statistics, and monitoring reports are transmitted to a remote terminal for remote monitoring. This helps banana orchards adjust their pest control strategies based on the trapping situation, ensuring the normal growth of bananas during the budding stage.
[0051] In summary, the multimodal signal acquisition module installs various acquisition components in the middle of the plant, near the banana buds, below the fruit bunches, and on field posts, adapting to the activity areas of moths and fruit flies. The data purification and identification module removes interference from leaf movement and humidity, and identifies species by combining the diurnal behavior differences of pests. The targeted signal regulation module enhances optical signals and continuously strengthens chemical signals at night, matching the phototaxis of moths and the chemotaxis of fruit flies. The precise trapping execution module covers banana buds, fruits, and the base of the plant, efficiently capturing pests. The monitoring report generation module quantifies the trapping effect and the proportion of non-target organisms accidentally trapped, helping to optimize pest control strategies during the banana plantation's budding stage and fully adapting to the monitoring and trapping needs of this stage.
[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. An intelligent monitoring and trapping system for invasive pests on tropical fruits, characterized in that, The system includes: Multimodal signal acquisition module: integrates piezoelectric miniature acoustic sensor, LED precision spectral light source, chemical attractant release component and high-definition visual acquisition component to simultaneously acquire multimodal data of pests; Data purification and identification module: Through the database of behavioral characteristics of tropical invasive pests, and based on the acquired multimodal data of pests, the module calculates the multimodal signal effectiveness index and modality dynamic weight coefficient, and then performs data processing and species identification, and outputs the number of identified target pests. Targeted signal modulation module: Based on species identification results and dynamic weight coefficients of each modality, calculates the target signal enhancement intensity and adjusts the signal intensity of each modality; Precision trapping execution module: Deploys trapping components to cover the detection area, catches target pests attracted by the target signal, has an internal trapping unit to adsorb the insects that enter, and integrates an image recognition and infrared sensing counting unit to distinguish between target pests and beneficial insects, and counts the number of target pests and beneficial insects captured. The monitoring report generation module calculates the trapping efficiency index and the false trapping rate of beneficial insects based on the number of target pests captured, the number of beneficial insects captured, and generates a monitoring report. It also transmits multimodal data, species identification results, trapping statistics and monitoring reports to a remote terminal for remote monitoring through an IoT network-based communication unit.
2. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 1, characterized in that, In the multimodal signal acquisition module, Piezoelectric miniature acoustic sensor: frequency response range of 100~5000Hz, sensitivity of -40~-35dBV / Pa, for collecting real-time acoustic signal intensity of insect wing vibrations. ; LED Precision Spectral Light Source: Outputs phototactic peak light in the 450~520nm blue light band, with a power adjustment range of 0~5W and a emission angle of 120~140°, collecting real-time optical reflection response intensity of phototactic pests. ; Chemical attractant release assembly: includes a 50-60 mL reservoir, a quantitative release valve, an adjustable attractant release rate of 0.1-1 mL / h, and outputs the real-time chemical attractant concentration of the monitored area. ; High-definition visual acquisition component: Equipped with an infrared supplementary light unit, the shooting frame rate is 1 frame every 3-5 seconds, and the visual feature matching degree of the output pest image is high. .
3. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 1, characterized in that, In the data purification and identification module, the tropical invasive insect behavior characteristic database is constructed using historical data and stores behavioral characteristic parameters and multimodal data benchmark values of tropical invasive insects, based on the acquired acoustic signal intensity. Optical reflection response intensity Chemical attractant concentration Visual feature matching degree The multimodal signal effectiveness index is calculated using the multimodal signal effectiveness index calculation formula. Quantify the proportion of effective target signals through a multimodal signal effectiveness index. Environmental interferences such as wind noise and leaf friction were eliminated. Furthermore, considering the behavioral differences among various invasive tropical pests, behavioral parameters such as phototaxis in moths, chemotaxis in fruit flies, and phonotaxis in thrips from the built-in database were used to calculate dynamic weighting coefficients using the modal dynamic weighting coefficient calculation formula. The system dynamically adapts the weight ratio of each modality, then integrates the features of each modality to identify pest species, and outputs the number of identified target pests. .
4. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 3, characterized in that, In the data cleansing and identification module, the formula for calculating the multimodal signal effectiveness index is as follows: ,in, It is a multimodal signal effectiveness index. For real-time acoustic signal strength, As the acoustic effective signal reference value, For real-time optical reflection response intensity, This serves as a baseline value for the response intensity of phototactic pests. This represents the real-time concentration of the chemical attractant. This is the baseline concentration of the attractant for the target pest. For visual feature matching degree, As the baseline value for image features, The basic weights for each modality type.
5. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 3, characterized in that, In the data purification and identification module, the formula for calculating the modal dynamic weight coefficient is as follows: ,in, for The dynamic weighting coefficients of the modality, These are acoustic, optical, chemical, and visual modalities, respectively. for The signal effectiveness index of the modality, This represents the behavioral characteristic coefficient of the target pest. For acoustic, optical, chemical and visual modalities Perform summation.
6. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 1, characterized in that, In the targeted signal modulation module, based on species identification results and modal dynamic weighting coefficients... The target signal enhancement intensity is calculated using the target signal enhancement intensity calculation formula. The formula for calculating the enhancement intensity of the target signal is: ,in, for The enhancement intensity of the target signal in each modality for The basic output value of each modality For the enhancement coefficient, for Dynamic weighting coefficients for modalities; enhancement of acoustic signal output for sound-attracting pests; and strengthening intensity based on target signal. The output values of each mode were determined. For phototactic pests, the light intensity of the light source in the specified wavelength was increased; for chemotactic pests, the amount of attractant released was increased.
7. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 1, characterized in that, The monitoring report generation module generates reports based on the number of captured target pests, the number of beneficial insects, and the number of identified target pests. The trapping efficiency index is calculated using the trapping efficiency index calculation formula. Meanwhile, the beneficial insect miscible rate was calculated using the formula for calculating the beneficial insect miscible rate. And based on the trapping efficiency index and the rate of accidental capture of beneficial insects This generates monitoring reports that quantify the trapping effect and the proportion of non-target organisms accidentally trapped.
8. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 7, characterized in that, The formula for calculating the trapping efficiency index in the monitoring report generation module is as follows: ,in, The trapping efficiency index. The number of target pests captured. The number of target pests identified. For the first Dynamic weighting coefficients for each modality for The enhancement intensity of the target signal in each modality for The basic output value of each modality.
9. The intelligent monitoring and trapping system for invasive pests on tropical fruits according to claim 7, characterized in that, The formula for calculating the beneficial insect false capture rate in the monitoring report generation module is as follows: ,in, To reduce the rate of accidental capture of beneficial insects, The number of beneficial insects captured. The number of target pests captured.
10. A method for intelligent monitoring and trapping invasive pests of tropical fruits, the method being applicable to the intelligent monitoring and trapping system for invasive pests of tropical fruits as described in any one of claims 1-9, characterized in that, The method includes: Multimodal signal acquisition: Integrating piezoelectric miniature acoustic sensors, LED precision spectral light sources, chemical attractant release components, and high-definition visual acquisition components, it simultaneously acquires multimodal data of pests; Data purification and identification: Through the database of behavioral characteristics of tropical invasive pests, and based on the acquired multimodal data of pests, the multimodal signal effectiveness index and modality dynamic weight coefficient are calculated, and then data processing and species identification are performed, and the number of identified target pests is output. Targeted signal modulation: Based on species identification results and dynamic weighting coefficients of each modality, the intensity of the targeted signal enhancement is calculated, and the intensity of each modality signal is adjusted. Precise trapping execution: Deploy trapping components to cover the detection area, catch target pests attracted by the target signal, and have built-in trapping units to adsorb the insects that enter. It also integrates image recognition and infrared sensing counting units to distinguish between target pests and beneficial insects, and count the number of target pests and beneficial insects captured. Monitoring report generation: Based on the number of target pests captured, the number of beneficial insects, and the number of identified target pests, the trapping efficiency index and the false trapping rate of beneficial insects are calculated, and a monitoring report is generated. Furthermore, through a communication unit based on an Internet of Things network, multimodal data, species identification results, trapping statistics, and monitoring reports are transmitted to a remote terminal for remote monitoring.