Garden plant risk grading early warning method based on artificial intelligence

Abstract

The invention discloses a garden plant risk grading early warning method based on artificial intelligence, and the method comprises the following steps: 1, carrying out the high-altitude patrol of an unmanned aerial vehicle, executing the daily inspection of the unmanned aerial vehicle according to a preset route, and synchronously collecting visible light, multispectral and thermal infrared data; step 2, insect trapping: trapping mechanisms are uniformly fixed at target positions at intervals through ground nails respectively, then pheromone sex attractants or light traps corresponding to insects are put into storage bottles communicated with the trapping mechanisms, and then images of the insects in the trapping bottles are shot in a surrounding manner every 30 minutes through a camera; 3, data analysis; 4, establishing a risk grading system; and 5, fusing and uploading the multi-source data. Data integration analysis is performed on spectral data and insect image data acquired by the unmanned aerial vehicle, limitation of manual monitoring means is broken through, and biological invasion abnormity can be found in time so as to inform workers of timely processing and reduce biological invasion hazards.

Description

Technical Field

[0001] This invention relates to the field of garden plant protection technology, specifically to an artificial intelligence-based method for risk classification and early warning of garden plants. Background Technology

[0002] As an important component of the urban ecosystem, the health of garden plants is directly related to ecological balance, urban aesthetics, public safety, and sustainable economic development.

[0003] In the current technology, invasive plants (such as Canadian goldenrod and Spartina alterniflora) and insects (such as the fall webworm) are highly competitive and can quickly squeeze out the living space of native species, leading to the extinction of native plants and the imbalance of insect community structure.

[0004] Therefore, the importance of risk classification and early warning for garden plants is self-evident. However, current garden plant risk early warning technology has the following core problems: 1. Lagging monitoring of invasive plants: Traditional manual inspections rely on experience and judgment, making it difficult to detect the subtle spread of invasive plants in the early stages (such as the hidden growth of a single Canadian goldenrod in a shrub). Although remote sensing satellites can cover large areas, their spatial resolution is insufficient (usually >10 meters), making it impossible to identify small-scale invasive patches.

[0005] 2. Low efficiency of insect monitoring: Manual trapping relies on periodic sampling (e.g., once a week), resulting in low data temporal resolution; traditional image recognition systems require insects to be brought back to the laboratory for analysis, making real-time early warning impossible. For example, fall webworm larvae can devour an entire tree within 48 hours of an outbreak, and traditional methods struggle to respond in a timely manner.

[0006] Therefore, a three-dimensional, multimodal, and dynamically graded early warning method that integrates air, ground, and insect elements is needed to address the above problems and fill the gaps in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to provide an artificial intelligence-based risk classification and early warning method for garden plants to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: An artificial intelligence-based risk classification and early warning method for garden plants includes the following steps: First, high-altitude patrol by drones. The drones perform daily inspections according to preset routes (such as vertical take-off and landing multi-rotor drones covering small areas, and fixed-wing drones patrolling large areas), and simultaneously collect visible light, multispectral and thermal infrared data. The second step is insect trapping. The trapping mechanism is fixed at the target location with evenly spaced ground nails. Then, pheromone attractants or light traps for the corresponding insects are placed in the storage bottle of the trapping mechanism. Then, the camera takes pictures of the insects in the trapping bottle every 30 minutes. The third step is data analysis. The spectral and image data collected by the drone are compared with historical spectral databases to determine for any anomalies. Simultaneously, the types of insects trapped are classified and counted. Subsequently, the population density index of invasive insects is calculated. The principle behind this is using AI to compare newly collected spectral data with historical spectral databases. Visible light images monitored by the drone can visually present the color, texture, and spatial distribution of plants. For example, Canadian goldenrod displays a bright yellow during its flowering period, creating a significant contrast with surrounding vegetation. RGB image segmentation can initially locate suspected areas. Multispectral cameras can capture near-infrared and red-edge bands, reflecting physiological information such as chlorophyll content and water status. Healthy plants have high reflectivity in the near-infrared band, while invasive plants (such as jellyfish grass and wild oats) exhibit specific spectral curves due to differences in leaf structure or pigmentation. This allows for the determination of the presence of such invasive plants. Step 4: Establish a risk classification system: Level 1 warning, promptly notify staff to conduct inspections and confirmations; Level 2 warning, staff detect abnormal areas and spray corresponding biological control agents to kill invasive plants or insects in a wide area; Level 3 warning, coordinate with drone swarms to spray biological control agents over a large area to disinfect invasive plants and animals in the entire area. The fifth step involves the fusion and uploading of multi-source data. The drone identification results and insect capture data are uploaded to the cloud platform via 4G / 5G so that staff can access and query them at any time, facilitating the establishment of a database to provide data support for subsequent AI training.

[0009] Preferably, in the first step, the drone needs to be equipped with a multispectral sensor (such as Parrot Sequoia+), a thermal imaging camera (FLIR series), and an AI edge computing chip (NVIDIA Jetson) to realize the spectral feature recognition and dynamic positioning of invasive plants; Preferably, the multispectral sensor in the first step captures the vegetation reflectance spectrum, with key monitoring bands including (680-750nm), near-infrared (750-900nm) and short-wave infrared (1500-1700nm), used to identify abnormal chlorophyll content and water stress; Preferably, in the second step, the population density index (PDI) is calculated after the insects are trapped. The calculation formula is as follows: PDI=100%, where N is the daily capture amount, S is the coverage area of ​​the trap, and T is the monitoring duration. The population density index (PDI) of the invasive insects is calculated to determine the density of the invasive insects. Preferably, in the fourth step, the risk classification includes early warning for invasive plants: Level 1 Warning (Low Risk): Single invasive plant, spread area <10㎡; Level 2 Warning (Medium Risk): Cluster intrusion, area 10-100㎡ and annual spread rate >30%; Level 3 Warning (High Risk): Large-scale outbreak, area >100㎡ or threatening key protected areas.

[0010] Preferably, in the fourth step, the risk classification includes early warning for invasive insects: Level 1 warning (low risk): A single invasive insect has been detected; Level 2 Warning (Medium Risk): PDI > 5% for 3 consecutive days; Level 3 Warning (High Risk): PDI > 15% for 3 consecutive days.

[0011] A device for trapping invasive insects that pose a risk to garden plants includes a base plate. Ground stakes are fixedly installed around the outer surface of the base plate. A guide chamber is fixedly installed at the upper end of each ground stake. A trapping mechanism for catching insects is installed at the lower end of the guide chamber. The trapping mechanism includes a connecting pipe, a trapping bottle, a pheromone storage chamber, a motor, a worm gear, a worm wheel, an L-shaped connecting rod, and a camera. The connecting pipe is fixedly installed at the lower end of the guide chamber. The trapping bottle is movably screwed onto the lower end of the connecting pipe via a thread. The pheromone storage chamber is fixedly installed at the lower end of the trapping bottle. A motor is fixedly installed at the upper end of the base plate. A worm gear is fixedly installed at the power output shaft end of the motor. A worm wheel is movably installed in the middle of the upper end of the base plate via a rotating shaft. The worm wheel and the worm gear mesh with each other. An L-shaped connecting rod is fixedly installed at the upper end of the worm wheel. A camera is fixedly installed at the upper end of the L-shaped connecting rod. Preferably, the upper part of the trapping bottle is provided with a funnel-shaped inlet, which allows insects to fly in and crawl in easily, but prevents them from flying out and crawling out, so as to achieve better insect trapping.

[0012] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention integrates and analyzes spectral data collected by drones, overcoming the limitations of manual monitoring methods. It can promptly detect abnormal plant biological invasions, notify staff to handle them in a timely manner, and reduce the harm caused by biological invasions.

[0013] 2. The insect-catching device of the present invention, combined with the analysis method of the present invention, can capture the intended insects more effectively and accurately, calculate the population density index of invasive insects, establish an early warning level for invasive insects in a timely manner, and reduce the harm caused by insect invasion. Attached Figure Description

[0014] Figure 1This is a flowchart illustrating the steps of an artificial intelligence-based risk classification and early warning method for garden plants according to the present invention. Figure 2 This is an overall structural view of a device for capturing invasive insects that pose a risk to garden plants, according to the present invention. Figure 3 This is a cross-sectional view of an invasive insect trapping device for garden plants according to the present invention; Figure 4 This invention relates to a device for trapping invasive insects that pose a risk to garden plants. Figure 3 A magnified view of point A in the middle.

[0015] In the diagram: 1. Base plate; 2. Ground stake; 3. Guiding chamber; 4. Connecting pipe; 5. Trapping bottle; 6. Pheromone storage chamber; 7. Motor; 8. Worm gear; 9. Worm wheel; 10. L-shaped connecting rod; 11. Camera; 12. Funnel-shaped inlet. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1-4 The present invention provides a technical solution: a risk classification and early warning method for garden plants based on artificial intelligence, including the following steps: First, high-altitude patrol by drones. The drones perform daily inspections according to preset routes (such as vertical take-off and landing multi-rotor covering a small area, and fixed-wing patrolling a large area), and simultaneously collect visible light, multispectral and thermal infrared data. The second step is insect trapping. The trapping mechanism is fixed at the target location with evenly spaced ground nails. Then, pheromone attractants or light traps for the corresponding insects are placed in the storage bottle of the trapping mechanism. Then, the camera takes pictures of the insects in the trapping bottle every 30 minutes. The third step is data analysis. The spectral and image data collected by the drone are compared with historical spectral databases to determine for any anomalies. Simultaneously, the types of insects trapped are classified and counted. Subsequently, the population density index of invasive insects is calculated. The principle behind this is using AI to compare newly collected spectral data with historical spectral databases. Visible light images monitored by the drone can visually present the color, texture, and spatial distribution of plants. For example, Canadian goldenrod displays a bright yellow during its flowering period, creating a significant contrast with surrounding vegetation. RGB image segmentation can initially locate suspected areas. Multispectral cameras can capture near-infrared and red-edge bands, reflecting physiological information such as chlorophyll content and water status. Healthy plants have high reflectivity in the near-infrared band, while invasive plants (such as jellyfish grass and wild oats) exhibit specific spectral curves due to differences in leaf structure or pigmentation. This allows for the determination of the presence of such invasive plants. Step 4: Establish a risk classification system: Level 1 warning, promptly notify staff to conduct inspections and confirmations; Level 2 warning, staff detect abnormal areas and spray corresponding biological control agents to kill invasive plants or insects in a wide area; Level 3 warning, coordinate with drone swarms to spray biological control agents over a large area to disinfect invasive plants and animals in the entire area. The fifth step involves the fusion and uploading of multi-source data. The drone identification results and insect capture data are uploaded to the cloud platform via 4G / 5G so that staff can access and query them at any time, facilitating the establishment of a database to provide data support for subsequent AI training.

[0018] This invention integrates UAV spectral data, insect image data, and environmental sensor data, overcoming the limitations of manual monitoring methods. It can promptly detect abnormal biological invasions, allowing staff to take timely action and reduce the harm caused by biological invasions.

[0019] In the first step, the drone needs to be equipped with a multispectral sensor (such as Parrot Sequoia+), a thermal imaging camera (FLIR series), and an AI edge computing chip (NVIDIA Jetson) to realize the spectral feature identification and dynamic positioning of invasive plants; The multispectral sensor mentioned in the first step captures the vegetation reflectance spectrum, with key monitoring bands including (680-750nm), near-infrared (750-900nm) and short-wave infrared (1500-1700nm), which are used to identify abnormal chlorophyll content and water stress; In the second step, after insect trapping, the population density index (PDI) is calculated. The calculation formula is as follows: PDI=100%, where N is the daily capture amount, S is the coverage area of ​​the trap, and T is the monitoring duration. The population density index (PDI) of invasive insects is calculated to determine the density of invasive insects. In the fourth step, the risk classification includes early warning for invasive plants: Level 1 Warning (Low Risk): Single invasive plant, spread area <10㎡; Level 2 Warning (Medium Risk): Cluster intrusion, area 10-100㎡ and annual spread rate >30%; Level 3 Warning (High Risk): Large-scale outbreak, area >100㎡ or threatening key protected areas.

[0020] In the fourth step, the risk classification includes early warning for invasive insects: Level 1 warning (low risk): A single invasive insect has been detected; Level 2 Warning (Medium Risk): PDI > 5% for 3 consecutive days; Level 3 Warning (High Risk): PDI > 15% for 3 consecutive days.

[0021] A device for capturing invasive insects that pose a risk to garden plants includes a base plate 1. Ground nails 2 are fixedly installed around the outer surface of the base plate 1. A guide chamber 3 is fixedly installed at the upper end of each ground nail 2. A trapping mechanism for attracting insects is installed at the lower end of the guide chamber 3. The trapping mechanism includes a connecting pipe 4, a trapping bottle 5, a pheromone storage chamber 6, a motor 7, a worm gear 8, a worm wheel 9, an L-shaped connecting rod 10, and a camera 11. The connecting pipe 4 is fixedly installed at the lower end of the guide chamber 3. The trapping bottle 5 is screwed into the lower end of the connecting pipe 4 via a thread. The pheromone storage chamber 6 is fixedly installed inside the lower end of the trapping bottle 5. The motor 7 is fixedly installed at the upper end of the base plate 1. A worm gear 8 is fixedly installed at the power output shaft end of the motor 7. A worm wheel 9 is movably installed in the middle of the upper end of the base plate 1 via a rotating shaft. The worm wheel 9 meshes with the worm gear 8. An L-shaped connecting rod 10 is fixedly installed at the upper end of the worm wheel 9. A camera 11 is fixedly installed at the upper end of the L-shaped connecting rod 10. The upper part of the trapping bottle 5 is provided with a funnel-shaped inlet 12. The funnel-shaped inlet 12 allows insects to fly in and crawl in easily, but prevents insects from flying out and crawling out, so as to achieve better insect trapping.

[0022] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0023] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for risk classification and early warning of garden plants based on artificial intelligence, comprising the following steps, characterized in that: The first step is to conduct high-altitude patrols by drones. The drones perform daily inspections according to preset routes (such as vertical take-off and landing multi-rotor drones covering small areas, and fixed-wing drones patrolling large areas), and collect visible light, multispectral and thermal infrared data simultaneously. The second step is insect trapping. The trapping mechanism is fixed at the target location with evenly spaced ground nails. Then, pheromone attractants or light traps for the corresponding insects are placed in the storage bottle of the trapping mechanism. Then, the camera takes pictures of the insects in the trapping bottle every 30 minutes. The third step is data analysis, which compares the spectral and image data collected by the drone with the historical spectral database to determine if there are any anomalies. At the same time, the types of insects trapped are classified and counted, and the population density index of invasive insects is calculated. Step 4: Establish a risk classification system: Level 1 warning, promptly notify staff to conduct inspections and confirmations; Level 2 warning, staff detect abnormal areas and spray corresponding biological control agents to kill invasive plants or insects in a wide area; Level 3 warning, coordinate with drone swarms to spray biological control agents over a large area to disinfect invasive plants and animals in the entire area. The fifth step involves the fusion and uploading of multi-source data. The drone identification results and insect capture data are uploaded to the cloud platform via 4G / 5G so that staff can access and query them at any time, facilitating the establishment of a database to provide data support for subsequent AI training.

2. The method for risk classification and early warning of garden plants based on artificial intelligence according to claim 1, characterized in that: In the first step, the drone needs to be equipped with a multispectral sensor, a thermal imaging camera and an AI edge computing chip to realize the spectral feature recognition and dynamic positioning of invasive plants.

3. The method for risk classification and early warning of garden plants based on artificial intelligence according to claim 1, characterized in that: The multispectral sensor mentioned in the first step captures the vegetation reflectance spectrum, with key monitoring bands including (680-750nm), near-infrared (750-900nm) and short-wave infrared (1500-1700nm), which are used to identify abnormal chlorophyll content and water stress.

4. The method for risk classification and early warning of garden plants based on artificial intelligence according to claim 1, characterized in that: In the second step, after insect trapping, the population density index (PDI) is calculated. The calculation formula is as follows: PDI=100%, where N is the daily capture amount, S is the coverage area of ​​the trap, and T is the monitoring duration. The population density index (PDI) of invasive insects is calculated to determine the density of invasive insects.

5. The method for risk classification and early warning of garden plants based on artificial intelligence according to claim 1, characterized in that: In the fourth step, the risk classification includes early warning for invasive plants: Level 1 Warning (Low Risk): Single invasive plant, spread area <10㎡; Level 2 Warning (Medium Risk): Cluster intrusion, area 10-100㎡ and annual spread rate >30%; Level 3 warning (high risk): Large-scale outbreak, area > 100㎡.

6. The method for risk classification and early warning of garden plants based on artificial intelligence according to claim 1, characterized in that: In the fourth step, the risk classification includes early warning for invasive insects: Level 1 warning (low risk): A single invasive insect has been detected; Level 2 Warning (Medium Risk): PDI > 5% for 3 consecutive days; Level 3 Warning (High Risk): PDI > 15% for 3 consecutive days.

7. A device for trapping invasive insects that pose a risk to garden plants, comprising a base plate, characterized in that: Ground stakes are fixedly installed around the outer surface of the base plate. A guide chamber is fixedly installed at the upper end of each ground stake. A trapping mechanism for insects is installed at the lower end of the guide chamber. The trapping mechanism includes a connecting pipe, a trapping bottle, a pheromone storage chamber, a motor, a worm gear, a worm wheel, an L-shaped connecting rod, and a camera. A connecting pipe is fixedly installed at the lower end of the guide chamber. A trapping bottle is screwed onto the lower end of the connecting pipe. A pheromone storage chamber is fixedly installed at the lower end of the trapping bottle. A motor is fixedly installed at the upper end of the base plate. A worm gear is fixedly installed at the power output shaft end of the motor. A worm wheel is movably installed in the middle of the upper end of the base plate through a rotating shaft. The worm wheel and the worm gear mesh with each other. An L-shaped connecting rod is fixedly installed at the upper end of the worm wheel. A camera is fixedly installed at the upper end of the L-shaped connecting rod.

8. The invasive insect trapping device for garden plants according to claim 1, characterized in that: The trapping bottle has a funnel-shaped inlet at the upper part of its interior.

Images