A method and device for monitoring and early warning of forest pests and diseases

Abstract

The present application provides a method and device for monitoring and warning of pests and diseases in forests, which relates to the field of pest and disease monitoring technology, including: obtaining a first image set; obtaining regional greening characteristics, including greening density, tree species distribution, and tree age and height; obtaining a second image set; obtaining regional soil characteristics, including soil type and soil moisture; analyzing the layout positions of multiple acoustic wave sensors to obtain multiple matching layout positions; obtaining a sensor integration module; and monitoring and warning of pests and diseases in the target area according to the sensor integration module. The present application can solve the technical problem in the prior art that the lack of analysis of regional trees and soil leads to inaccurate sensor settings, thereby affecting the accuracy of monitoring and warning. By analyzing the greening characteristics and soil characteristics, the layout positions of acoustic wave sensors are optimized, and pest and disease disaster warnings are then carried out, thereby achieving the technical effect of improving monitoring accuracy and ensuring the health of forestry.

Description

Technical Field

[0001] The present application relates to the technical field of pest monitoring, and in particular to a method and device for monitoring and warning of forest pest disasters. Background Art

[0002] Forest land is an important part of the earth's ecosystem and plays an important role in maintaining biodiversity, regulating climate, and preventing soil erosion. Pest and disease disasters are extremely destructive to forest resources, and timely monitoring and early warning can effectively protect forest resources.

[0003] Traditional pest and disease monitoring methods mainly rely on manual inspections and empirical judgments, which have problems such as low efficiency, limited monitoring range, and poor real-time performance. In order to effectively monitor and warn of forest pests and diseases and improve the level of forestry management, it is necessary to study an efficient and accurate monitoring and early warning method.

[0004] Existing technologies usually monitor pests and diseases based on image processing and sensor technology, but lack analysis of regional trees and soil, resulting in inaccurate sensor placement, which in turn affects the accuracy of monitoring and early warning. Summary of the invention

[0005] The purpose of this application is to provide a forest pest and disease monitoring and early warning method and device to solve the technical problem in the prior art that the lack of analysis of local trees and soil leads to inaccurate sensor settings, thereby affecting the accuracy of monitoring and early warning.

[0006] In view of the above problems, the present application provides a method and device for monitoring and early warning of forest pests and diseases.

[0007] In a first aspect, the present application provides a method for monitoring and early warning of forest pests and diseases, which is implemented by a device for monitoring and early warning of forest pests and diseases, wherein the method comprises: collecting overhead images of a target area to obtain a first image set; performing feature analysis on the first image set to obtain regional greening characteristics, including greening density, tree species distribution, and tree age and height; collecting ground images of the target area to obtain a second image set; performing feature analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture; analyzing the layout positions of multiple acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics to obtain multiple matching layout positions; setting the multiple acoustic wave sensors according to the multiple matching layout positions to obtain a sensor integrated module; and monitoring and early warning of pests and diseases in the target area according to the sensor integrated module.

[0008] In the second aspect, the present application also provides a forest pest and disease monitoring and early warning device for executing the method as described in the first aspect, wherein the device includes: a bird's-eye view image acquisition unit, used to perform bird's-eye view image acquisition on the target area to obtain a first image set; a regional greening feature analysis unit, used to perform feature analysis on the first image set to obtain regional greening features, including greening density, tree species distribution, and tree age and height; a ground image acquisition unit, used to perform ground image acquisition on the target area to obtain a second image set; a regional soil feature analysis unit, used to perform feature analysis on the second image set to obtain regional soil features, including soil type and soil moisture; a layout position analysis unit, used to analyze the layout positions of multiple acoustic wave sensors according to the regional greening features and the regional soil features, and obtain multiple matching layout positions; a sensor setting unit, used to set the multiple acoustic wave sensors according to the multiple matching layout positions to obtain a sensor integrated module; a monitoring and early warning unit, used to monitor and warn of pest and disease disasters in the target area according to the sensor integrated module.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] Collect overhead images of the target area to obtain a first image set; perform feature analysis on the first image set to obtain regional greening characteristics, including greening density, tree species distribution, and tree age and height; collect ground images of the target area to obtain a second image set; perform feature analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture; analyze the layout positions of multiple acoustic wave sensors based on the regional greening characteristics and the regional soil characteristics to obtain multiple matching layout positions; set the multiple acoustic wave sensors based on the multiple matching layout positions to obtain a sensor integrated module; monitor and warn of pests and diseases in the target area based on the sensor integrated module. By analyzing the greening characteristics and soil characteristics, optimize the layout positions of the acoustic wave sensors, and then conduct pest and disease warnings, so as to achieve the technical effect of improving monitoring accuracy and ensuring the health of forestry.

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0013] Figure 1 A flowchart of a method for monitoring and early warning of pests and diseases in forest land is provided for this application;

[0014] Figure 2 This is a structural schematic diagram of a forest pest and disease monitoring and early warning device for the present application.

[0015] Description of reference numerals:

[0016] Bird's-eye view image acquisition unit 11, regional greening feature analysis unit 12, ground image acquisition unit 13, regional soil feature analysis unit 14, deployment position analysis unit 15, sensor setting unit 16, monitoring and early warning unit 17. DETAILED DESCRIPTION

[0017] This application provides a method and device for monitoring and early warning of forest pests and diseases, which solves the technical problem in the prior art that the lack of analysis of regional trees and soil leads to inaccurate sensor settings, thus affecting the accuracy of monitoring and early warning. By analyzing the greening characteristics and soil characteristics, the layout of the acoustic sensor is optimized, and then early warning of pests and diseases is carried out, achieving the technical effect of improving monitoring accuracy and ensuring the health of forestry.

[0018] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0019] For example, please refer to the attached Figure 1 The present application provides a forest pest and disease disaster monitoring and early warning method, wherein the method is applied to a forest pest and disease disaster monitoring and early warning device, and the method specifically comprises the following steps:

[0020] Step 1: Collect overhead images of the target area to obtain a first image set.

[0021] Specifically, the target area refers to the pre-designated forest land to be monitored for pests and diseases. The target area is photographed from top to bottom by a camera mounted on a high-altitude platform such as a drone or satellite to obtain a series of overhead images to form the first image set. The first image set will serve as the basis for subsequent analysis and processing to provide early warning of pests and diseases.

[0022] Step 2: Perform feature analysis on the first image set to obtain regional greening features, including greening density, tree species distribution, and tree age and height.

[0023] Specifically, regional greening feature analysis can be performed based on NDVI (normalized difference vegetation index). The images in the first image set can be fused in three dimensions, the NDVI value of each pixel in the fused image can be calculated, and an NDVI index map can be generated based on the NDVI value of each pixel. The NDVI value of each pixel represents the health of the vegetation in the area. The NDVI value is usually between -1 and 1. Positive values ​​indicate vegetation coverage, and the larger the value, the more lush the vegetation; negative values ​​or values ​​close to zero may indicate non-vegetated areas such as water bodies, bare land or rocks. By performing feature analysis based on NDVI (normalized difference vegetation index), the vegetation coverage ratio can be extracted from the image, and the greening density of the forest land can be estimated.

[0024] Trees of different species, ages, and heights may appear in different colors, shapes, and textures in images, and their corresponding NDVI values ​​will be different. Through machine learning algorithms, models can be trained to identify and distinguish different tree species. The models are trained with a large amount of labeled data so that they can accurately identify various tree species from images and annotate their locations, distributions, ages, and heights. This helps construct regional greening features and provide support for subsequent sensor deployment, thereby ensuring the accuracy of sensor deployment and, in turn, ensuring the accuracy of pest and disease warnings.

[0025] Step three: collect ground images of the target area to obtain a second image set.

[0026] Specifically, to collect images of the target area from the ground perspective, first, it is necessary to select appropriate ground image collection points according to the scale and terrain of the forest. These points should be able to represent different areas and features of the forest to ensure that the collected data is comprehensive and representative. Specifically, professional and technical personnel in this field can conduct shooting tests on the ground image collection points to ensure that the captured images can cover the ground of the entire target area. Ground image collection is usually carried out using a handheld camera, a tripod-fixed camera, or a camera installed on a vehicle. At the selected collection point, the ground of the target area is photographed to obtain multiple ground images to form a second image set.

[0027] Step 4: Perform feature analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture.

[0028] Specifically, soil type identification usually relies on soil characteristics such as color, texture, and structure. By analyzing the appearance of soil in the ground image in the second image set, different soil types can be identified. For example, sandy soil is usually lighter in color and has larger particles, while clay soil is darker in color and has a finer texture. Soil moisture is an indicator of the water content in the soil, which is crucial for plant growth and soil health. Soil moisture can be estimated by analyzing the color and reflectivity of the soil. Wet soil is usually darker in color and has lower reflectivity, while dry soil is lighter in color and has higher reflectivity. Using image processing technology, the color and reflectivity information of the soil area in the image can be extracted to estimate the soil moisture. All the collected ground images and related soil feature data in the second image set are integrated together for comprehensive analysis, including drawing the distribution map of soil types and the spatial distribution map of soil moisture as regional soil characteristics to provide support for subsequent sensor deployment and pest and disease disaster warning.

[0029] Step 5: Analyze the layout positions of multiple acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics to obtain multiple matching layout positions.

[0030] Specifically, first, it is necessary to analyze the greening characteristics of the forest land, including greening density, tree species distribution, and tree age and height. This information can help determine which areas are the focus of monitoring. For example, areas with higher greening density or older trees may need closer monitoring. Next, analyze soil characteristics, including soil type and soil moisture. Areas with different types of soil and moisture levels have different susceptibility to pests and diseases. For example, a reduction in soil moisture may make some areas more susceptible to pests and diseases. Therefore, different weights are set for different areas to configure sensors, and the matching layout positions of multiple acoustic wave sensors are determined. The matching layout positions can effectively cover the monitoring area, ensuring that the layout positions of the acoustic wave sensors are both scientific and reasonable, and can meet the monitoring needs, thereby improving monitoring efficiency and accuracy, and ensuring that forest pests and diseases can be effectively monitored and managed.

[0031] Step six: Arrange the plurality of acoustic wave sensors according to the plurality of matching layout positions to obtain a sensor integrated module.

[0032] Specifically, multiple acoustic wave sensors are deployed according to multiple matching deployment positions, and the installed acoustic wave sensors need to be connected into a network, which usually requires the use of existing wireless communication technologies such as Wi-Fi, Bluetooth, LoRa, etc. to ensure that data can be effectively transmitted between sensors. Through the above steps, multiple acoustic wave sensors are integrated together to form a complete sensor network, namely a sensor integration module, to conduct comprehensive monitoring of the forest land.

[0033] Step 7: Perform pest and disease disaster monitoring and early warning on the target area according to the sensor integrated module.

[0034] Specifically, the acoustic wave sensor in the sensor integration module continuously collects acoustic wave signals in the forest. These signals may include the activity sounds of insects, the friction sounds of leaves, etc., which may be related to the occurrence of pests and diseases. The collected data is transmitted to the data processing center through wireless communication technology. The integrity and real-time nature of the data need to be ensured during data transmission. In the data processing center, the collected acoustic wave signals are analyzed. This usually involves technologies such as signal processing, pattern recognition, and machine learning. Through analysis, specific sound patterns related to pests and diseases can be identified. When a sound pattern related to pests and diseases is detected, an early warning signal is issued. The early warning signal can notify forestry managers or relevant agencies so that timely measures can be taken. Through these steps, the sensor integration module can effectively monitor and warn of pests and diseases in the target area, thereby protecting the health and ecological balance of the forest.

[0035] Further, step 2 of this application includes:

[0036] Perform three-dimensional fusion on the first image set to obtain a first fused image; calculate the NDVI value of each pixel in the first fused image according to the NDVI index, generate an NDVI index map according to the NDVI value of each pixel, and calculate the expression of the NDVI value of each pixel NDVI=(NIR-RED) / (NIR+RED), wherein NIR is the reflectivity of vegetation in the near-infrared band, and RED is the reflectivity of vegetation in the red band; obtain the greening characteristics of the region according to the NDVI index map.

[0037] Furthermore, the present application also includes the following steps:

[0038] The target area is divided according to the NDVI value in the NDVI index map, and the greening density is marked; the tree species in the target area are classified according to the NDVI index map, and the tree species distribution is marked; the target area is analyzed according to the changing trend of the NDVI index map to obtain the tree age and tree height.

[0039] Specifically, the three-dimensional fusion of the first image set is to convert multiple two-dimensional overhead images into a three-dimensional model, so as to more realistically and comprehensively display the spatial structure and characteristics of the forest. First, the three-dimensional model of the forest is reconstructed by using multiple overhead images in the first image set through existing computer vision technologies such as stereo matching, structure from motion (SFM) or optical flow. The reconstructed three-dimensional model is converted into a fused image, which shows the three-dimensional structure of the target area and can more intuitively observe and analyze the spatial characteristics of the forest. Next, the NDVI value of each pixel in the first fused image is calculated based on NDVI (normalized difference vegetation index). NDVI is an important indicator for measuring the health of vegetation. Its calculation formula is NDVI=(NIR-RED) / (NIR+RED), where NIR is the reflectivity of vegetation in the near-infrared band and RED is the reflectivity of vegetation in the red band. These two values ​​can be extracted from the fused image. Specifically, the existing atmospheric correction algorithm (such as FLAASH) can be used to eliminate the atmospheric impact, convert the radiation brightness value into the surface reflectance value, and extract the reflectance values ​​of the red band and the near-infrared band from the corrected image respectively. This can be done using existing image processing software (such as ENVI, ERDAS IMAGINE).

[0040] According to the NDVI value of each pixel, an NDVI index map is generated. In the NDVI index map, the NDVI value of each pixel represents the health of vegetation in the area. NDVI values ​​are usually between -1 and 1. Positive values ​​indicate vegetation coverage, and the larger the value, the more luxuriant the vegetation; negative values ​​or values ​​close to zero may indicate non-vegetated areas such as water bodies, bare land or rocks. Then, by analyzing the NDVI index map, the greening characteristics of the forest land, such as greening density, tree species distribution, tree age and height, etc., can be obtained. In general, through three-dimensional fusion and NDVI analysis, the spatial structure and vegetation status of the forest land can be extracted from the bird's-eye view image, providing a basis for monitoring pests and diseases.

[0041] Specifically, according to the NDVI index map, the step of obtaining the greening characteristics of the area includes: according to the size of the NDVI value in the NDVI index map, the target area can be divided into different categories, that is, similar NDVI values ​​are clustered into one area, such as a high vegetation density area, a medium vegetation density area, and a low vegetation density area. These categories can be determined by setting a threshold value of the NDVI value. For example, an area with an NDVI value greater than 0.7 is considered to be a high vegetation density area. After the division is completed, these areas can be marked with different colors or patterns on the map to intuitively display the greening density of different areas.

[0042] Specifically, when dividing the NDVI values ​​and marking the green density, a large number of sample NDVI values ​​and corresponding sample green density can be collected, and the corresponding relationship between the NDVI value and the green density can be established. The deviation of the NDVI values ​​that can be considered to have the same green density can be used as the division standard to divide the NDVI index map. Exemplarily, with 0.2 as the division distance, the green density of the area with an NDVI value less than 0.2 is 0%; the green density of the area with an NDVI value greater than or equal to 0.2 and less than 0.4 is 25%; the green density of the area with an NDVI value range greater than or equal to 0.4 and less than 0.6 is 50%; the green density of the area with an NDVI value greater than or equal to 0.6 is 75%, thereby obtaining the green density of different locations in the target area.

[0043] Next, the tree species in the target area are classified according to the NDVI index map, and the tree species distribution is marked. The NDVI index map can reflect the growth conditions of different tree species. Different tree species can be distinguished by analyzing the differences in NDVI values. For example, some tree species may show higher NDVI values. Specifically, it is necessary to combine the existing NDVI index distribution map of tree species for classification, obtain the tree species in different locations for marking, and use different symbols or colors to mark the distribution of different tree species on the map of the target area to obtain the tree species distribution.

[0044] Finally, the target area is analyzed according to the changing trend of the NDVI index map to obtain information on tree age and tree height. The changing trend of the NDVI value can reflect the growth and health of the trees. For example, a continuous decline in NDVI values ​​may indicate that the trees are aging. By analyzing the time series data of NDVI values, the age and height of the trees can be estimated. Specifically, NDVI value change samples corresponding to tree height samples of different ages can be collected, and a tree age and tree height assessment model can be trained based on the existing machine learning model. According to the NDVI index map, NDVI index change data is established, and the data is input into the tree age and tree height assessment model for analysis to obtain the tree age and tree height.

[0045] The regional greening characteristics are composed of greening density, tree species distribution, tree age and tree height, providing data support for subsequent sensor deployment.

[0046] Further, step 4 of this application includes:

[0047] The second image set is fused in three dimensions to obtain a second fused image; spectral characteristics and texture characteristics are obtained by analyzing the second fused image; the soil in the target area is classified according to the spectral characteristics and the texture characteristics to determine the soil type; the soil in the target area is analyzed according to the reflectance change of the spectral characteristics to determine the soil moisture.

[0048] Specifically, similar to the three-dimensional fusion of overhead images, the three-dimensional fusion of ground images is also to create a more realistic three-dimensional model, so that the characteristics of the soil can be observed and analyzed from multiple angles. The second image set is three-dimensionally fused using the same method as the three-dimensional fusion of overhead images to obtain a second fused image. Spectral characteristics refer to the reflectivity or absorptivity of the soil in different bands. These characteristics can reflect the physical and chemical properties of the soil, such as mineral composition, organic matter content, etc. The reflectivity of the soil area in different bands in the second fused image can be analyzed by existing multispectral or hyperspectral image analysis technology to extract spectral characteristics. Texture characteristics refer to the texture structure of the soil surface, such as particle size, arrangement, etc. These characteristics can reflect the texture and structure of the soil. Texture characteristics can be extracted by analyzing the texture pattern of the soil area in the fused image by existing gray level co-occurrence matrix, wavelet transform and other methods.

[0049] Using the extracted spectral features and texture features, the soil can be classified and the soil type can be determined. This usually requires the use of machine learning algorithms, such as support vector machines (SVM), random forests, etc., to train known soil samples to establish a classification model. Soil moisture can be determined by analyzing the reflectance changes of spectral features. Wet soil usually has a lower reflectance, while dry soil has a higher reflectance. By comparing the reflectance of different areas, the moisture of the soil can be estimated. Specifically, known soil moisture samples and reflectance change samples can be obtained to construct the relationship between reflectance and humidity, thereby analyzing soil moisture.

[0050] Through the above steps, regional soil characteristics, including soil type and soil moisture, can be extracted from ground images to provide data support for pest and disease early warning.

[0051] Further, step five of this application includes:

[0052] According to the correlation between the regional greening characteristics and the regional soil characteristics and pests and diseases, the greening weight and the soil weight are obtained; according to the greening weight, the weights corresponding to the greening density, the tree species distribution and the tree age and height are configured; according to the soil weight, the weights corresponding to the soil type and the soil moisture are configured; according to the weight configuration result, each pixel is calculated to obtain a pixel entropy distribution map, and according to the pixel entropy distribution map, a plurality of matching layout positions are obtained.

[0053] Furthermore, the present application also includes the following steps:

[0054] Acquire the coverage range of the acoustic wave sensor; identify pixels in the pixel entropy distribution map that are greater than a preset threshold and mark them as high entropy areas; initialize the position of the acoustic wave sensor, perform iterative calculations based on the initialized position of the acoustic wave sensor, and obtain the coverage rate of each iteration, where the coverage rate is the ratio of the acoustic wave sensor covering the high entropy area; optimize the positions of the multiple acoustic wave sensors according to the coverage rate until the coverage rate meets the preset coverage rate, and obtain the multiple matching layout positions.

[0055] Specifically, first, the greening weight needs to be determined based on the correlation between pests and diseases and greening characteristics. For example, if certain pests and diseases are more likely to occur on trees of a specific species or age, then these characteristics should have higher weights. The greening weight can be obtained by statistically analyzing historical pest and disease disaster data, and the correlation between pests and diseases and greening characteristics is analyzed by existing correlation analysis methods to obtain the greening weight. Similarly, the soil weight is determined based on the correlation between pests and diseases and soil characteristics. For example, if certain pests and diseases are more likely to occur in specific types of soil or soil moisture conditions, then these characteristics should have higher weights. The soil weight can also be determined by statistically analyzing historical pest and disease disaster data.

[0056] According to the determined greening weight, the weights corresponding to greening density, tree species distribution, tree age and tree height are configured. Specifically, the correlation between pests and diseases and greening density, tree species distribution, tree age and tree height can be obtained by analyzing historical pest and disease disaster data, and the greening weight is divided according to the correlation ratio to generate the weights corresponding to the greening density, the tree species distribution and the tree age and tree height. The soil weight is divided in the same way to generate the weights corresponding to the soil type and the soil moisture.

[0057] Using the configured weights, the greening characteristics and soil characteristics of each pixel are weighted and calculated to generate a pixel entropy distribution map. Entropy is an indicator of information uncertainty, which can be used here to indicate the possibility of pests and diseases at each pixel. The higher the pixel entropy, the greater the possibility of pests and diseases. Based on the pixel entropy distribution map, areas with high incidence of pests and diseases can be determined, thereby determining the matching layout locations of the acoustic wave sensors. These locations should be concentrated in areas with higher entropy values ​​to ensure that pests and diseases can be effectively monitored and prevented.

[0058] Specifically, according to the pixel entropy distribution map, the specific steps of obtaining multiple matching layout positions include: first, it is necessary to determine the coverage of each acoustic wave sensor, which can be read through the user manual of the acoustic wave sensor. Next, it is necessary to identify pixels in the pixel entropy distribution map that are greater than a preset threshold. These pixels represent areas where pests and diseases are more likely to occur and can be identified as high entropy areas. The preset threshold can be determined by professional and technical personnel in this field based on the severity of the pests and diseases and monitoring needs. After the high entropy area is determined, the position of the acoustic wave sensor can be initialized, usually based on some preliminary assumptions or experience, such as uniform distribution throughout the area or concentrated in the center of the high entropy area.

[0059] According to the initialized acoustic wave sensor position, iterative calculation is performed to obtain the coverage rate of each iteration. Coverage rate refers to the ratio of the acoustic wave sensor covering the high entropy area, which can be calculated by simulating the overlapping part of the acoustic wave sensor coverage and the high entropy area. Specifically, the coverage area of ​​the initialized acoustic wave sensor position can be simulated according to the coverage range of each acoustic wave sensor, and the proportion of the overlapping part of the coverage area and the high entropy area in the high entropy area can be calculated to obtain the coverage rate. If the coverage rate does not meet the preset coverage rate, the position of the acoustic wave sensor is adjusted. The adjustment strategy may include adjusting the position of the sensor, increasing or decreasing the number of sensors, etc. The goal of optimization is to increase the coverage rate until the preset coverage rate is reached. Then recalculate the coverage rate until the coverage rate meets the preset coverage rate after multiple iterations. Among them, the preset coverage rate is set by professional and technical personnel in this field, such as 95%.

[0060] When the coverage rate meets the preset requirements, it can be considered that the layout of the acoustic sensor has been optimized, and the location at this time is a plurality of matching layout locations. These locations can ensure that the sensor network can effectively cover the areas with high incidence of pests and diseases, thereby improving monitoring efficiency and data accuracy.

[0061] Further, step seven of this application includes:

[0062] According to the sensor integrated module, real-time monitoring is performed to obtain multiple sound wave sensor signals, and the sound wave frequency, sound wave amplitude and duration are recorded; the multiple sound wave sensor signal sets are filtered to output multiple filtered sensor signal sets; a sound wave sensor sample set is constructed, and pattern recognition is performed with the multiple filtered sensor signal sets and the sound wave sensor sample set to output a signal abnormality; if the signal abnormality is greater than a preset signal abnormality, a warning signal is generated.

[0063] Specifically, the acoustic wave sensor in the sensor integration module collects acoustic wave signals in the forest in real time. These signals may include the activity sounds of insects, the friction sounds of leaves, etc. During the collection process, the characteristics of the acoustic wave signal, such as frequency, amplitude and duration, are recorded. These characteristics are crucial for subsequent pattern recognition and anomaly detection. The multiple sets of acoustic wave sensor signals collected are filtered to remove noise and interference. The filtered signal is called a filtered sensor signal set, which can be filtered by an existing median filter, mean filter, etc. Then, based on historical data or expert knowledge, an acoustic wave sensor sample set is constructed. The acoustic wave sensor sample set contains the characteristics of acoustic wave signals under normal conditions and can be used as a benchmark for pattern recognition. The filtered sensor signal is used for pattern recognition with the acoustic wave sensor sample set. Pattern recognition usually uses machine learning algorithms, such as support vector machines (SVMs) and random forests, to compare the similarity between the filtered signal and the sample set.

[0064] The greater the similarity between the filtered signal and the sample set, the smaller the corresponding signal abnormality. In other words, the signal abnormality is inversely proportional to the similarity. The signal abnormality can be obtained by subtracting the similarity from 1. This is an indicator to measure the degree of signal abnormality. The higher the abnormality, the greater the difference between the signal and the normal signal.

[0065] If the signal abnormality is greater than the preset signal abnormality threshold, it indicates that there may be abnormal conditions such as pests and diseases. At this time, an early warning signal is generated to notify relevant personnel to conduct further investigation and processing. Thus, the sensor integration module can monitor the acoustic wave activity in the forest in real time and issue a warning in time when an abnormal situation is detected, thereby effectively providing early warning of pests and diseases.

[0066] In summary, the forest pest and disease monitoring and early warning method provided by this application has the following technical effects:

[0067] Collect overhead images of the target area to obtain a first image set; perform feature analysis on the first image set to obtain regional greening characteristics, including greening density, tree species distribution, and tree age and height; collect ground images of the target area to obtain a second image set; perform feature analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture; analyze the layout positions of multiple acoustic wave sensors based on the regional greening characteristics and the regional soil characteristics to obtain multiple matching layout positions; set the multiple acoustic wave sensors based on the multiple matching layout positions to obtain a sensor integrated module; monitor and warn of pests and diseases in the target area based on the sensor integrated module. By analyzing the greening characteristics and soil characteristics, optimize the layout positions of the acoustic wave sensors, and then conduct pest and disease warnings, so as to achieve the technical effect of improving monitoring accuracy and ensuring the health of forestry.

[0068] Embodiment 2: Based on the forest pest and disease monitoring and early warning method in the above embodiment, the present application also provides a forest pest and disease monitoring and early warning device, please refer to the attached Figure 2 , the device comprises:

[0069] The overhead image acquisition unit 11 is used to acquire overhead images of the target area to obtain a first image set.

[0070] The regional greening feature analysis unit 12 is used to perform feature analysis on the first image set to obtain regional greening features, including greening density, tree species distribution, and tree age and height.

[0071] The ground image acquisition unit 13 is used to acquire ground images of the target area to obtain a second image set.

[0072] The regional soil characteristic analysis unit 14 is used to perform characteristic analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture.

[0073] The deployment position analysis unit 15 is used to analyze the deployment positions of the plurality of acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics, and obtain a plurality of matching deployment positions.

[0074] The sensor setting unit 16 is used to set the multiple acoustic wave sensors according to the multiple matching layout positions to obtain a sensor integrated module.

[0075] The monitoring and early warning unit 17 is used to monitor and warn the target area of ​​pests and diseases based on the sensor integrated module.

[0076] Furthermore, the monitoring and early warning unit 17 in the device is also used for:

[0077] According to the sensor integrated module, real-time monitoring is performed to obtain multiple sound wave sensor signals, and the sound wave frequency, sound wave amplitude and duration are recorded; the multiple sound wave sensor signal sets are filtered to output multiple filtered sensor signal sets; a sound wave sensor sample set is constructed, and pattern recognition is performed with the multiple filtered sensor signal sets and the sound wave sensor sample set to output a signal abnormality; if the signal abnormality is greater than a preset signal abnormality, a warning signal is generated.

[0078] Furthermore, the regional greening feature analysis unit 12 in the device is also used for:

[0079] Perform three-dimensional fusion on the first image set to obtain a first fused image; calculate the NDVI value of each pixel in the first fused image according to the NDVI index, generate an NDVI index map according to the NDVI value of each pixel, and calculate the expression of the NDVI value of each pixel NDVI=(NIR-RED) / (NIR+RED), wherein NIR is the reflectivity of vegetation in the near-infrared band, and RED is the reflectivity of vegetation in the red band; obtain the greening characteristics of the region according to the NDVI index map.

[0080] Furthermore, the regional greening feature analysis unit 12 in the device is also used for:

[0081] The target area is divided according to the NDVI value in the NDVI index map, and the greening density is marked; the tree species in the target area are classified according to the NDVI index map, and the tree species distribution is marked; the target area is analyzed according to the changing trend of the NDVI index map to obtain the tree age and tree height.

[0082] Furthermore, the regional soil characteristic analysis unit 14 in the device is also used for:

[0083] The second image set is fused in three dimensions to obtain a second fused image; spectral characteristics and texture characteristics are obtained by analyzing the second fused image; the soil in the target area is classified according to the spectral characteristics and the texture characteristics to determine the soil type; the soil in the target area is analyzed according to the reflectance change of the spectral characteristics to determine the soil moisture.

[0084] Furthermore, the regional soil characteristic analysis unit 14 in the device is also used for:

[0085] According to the correlation between the regional greening characteristics and the regional soil characteristics and pests and diseases, the greening weight and the soil weight are obtained; according to the greening weight, the weights corresponding to the greening density, the tree species distribution and the tree age and height are configured; according to the soil weight, the weights corresponding to the soil type and the soil moisture are configured; according to the weight configuration result, each pixel is calculated to obtain a pixel entropy distribution map, and according to the pixel entropy distribution map, a plurality of matching layout positions are obtained.

[0086] Further, the sensor setting unit 16 in the device is also used for:

[0087] Acquire the coverage range of the acoustic wave sensor; identify pixels in the pixel entropy distribution map that are greater than a preset threshold and mark them as high entropy areas; initialize the position of the acoustic wave sensor, perform iterative calculations based on the initialized position of the acoustic wave sensor, and obtain the coverage rate of each iteration, where the coverage rate is the ratio of the acoustic wave sensor covering the high entropy area; optimize the positions of the multiple acoustic wave sensors according to the coverage rate until the coverage rate meets the preset coverage rate, and obtain the multiple matching layout positions.

[0088] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The forest pest and disease monitoring and early warning method and specific examples in the first embodiment are also applicable to the forest pest and disease monitoring and early warning device in the present embodiment. Through the above detailed description of the forest pest and disease monitoring and early warning method, those skilled in the art can clearly know the forest pest and disease monitoring and early warning device in the present embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0089] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0090] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for monitoring and early warning of forest pests and diseases, characterized in that: The method comprises: Capturing overhead images of the target area to obtain a first image set; Performing feature analysis on the first image set to obtain regional greening features, including greening density, tree species distribution, and tree age and height; Collecting ground images of the target area to obtain a second image set; Performing feature analysis on the second image set to obtain regional soil characteristics, including soil type and soil moisture; Analyzing the layout positions of the plurality of acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics to obtain a plurality of matching layout positions; Arrange the plurality of acoustic wave sensors according to the plurality of matching arrangement positions to obtain a sensor integrated module; The target area is monitored and warned of pests and diseases according to the sensor integrated module.

2. The method according to claim 1, characterized in that The method for monitoring and warning pests and diseases in the target area according to the sensor integrated module includes: Perform real-time monitoring according to the sensor integrated module, obtain multiple sound wave sensor signals, and record the sound wave frequency, sound wave amplitude and duration; Filtering a plurality of acoustic wave sensor signal sets and outputting a plurality of filtered sensor signal sets; Constructing an acoustic wave sensing sample set, performing pattern recognition with the plurality of filtered sensing signal sets and the acoustic wave sensing sample set, and outputting a signal abnormality degree; If the signal abnormality is greater than a preset signal abnormality, a warning signal is generated.

3. The method according to claim 1, characterized in that Performing feature analysis on the first image set to obtain regional greening features, the method comprising: Performing three-dimensional fusion on the first image set to obtain a first fused image; Calculating the NDVI value of each pixel in the first fused image according to the NDVI index, generating an NDVI index map according to the NDVI value of each pixel, and calculating the expression of the NDVI value of each pixel NDVI=(NIR-RED) / (NIR+RED), wherein NIR is the reflectance of vegetation in the near-infrared band, and RED is the reflectance of vegetation in the red band; According to the NDVI index map, the greening characteristics of the region are obtained.

4. The method according to claim 3, characterized in that According to the NDVI index map, the greening characteristics of the region are obtained, and the method includes: Divide the target area according to the NDVI value in the NDVI index map and mark the green density; Classifying the tree species in the target area according to the NDVI index map and marking the distribution of tree species; The target area is analyzed according to the changing trend of the NDVI index map to obtain the tree age and height.

5. The method according to claim 1, characterized in that Performing feature analysis on the second image set to obtain regional soil features, the method comprising: Performing three-dimensional fusion on the second image set to obtain a second fused image; By analyzing the second fused image, spectral features and texture features are acquired; Classifying the soil in the target area based on the spectral characteristics and the texture characteristics to determine the soil type; The soil in the target area is analyzed based on the reflectance change of the spectral feature to determine the soil moisture.

6. The method according to claim 1, characterized in that Analyzing the layout positions of the plurality of acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics to obtain a plurality of matching layout positions, the method further includes: According to the correlation between the regional greening characteristics and the regional soil characteristics and pests and diseases, the greening weight and the soil weight are obtained; According to the greening weight, the weights corresponding to the greening density, the tree species distribution, and the tree age and tree height are configured; According to the soil weight, weights corresponding to the soil type and the soil moisture are configured; Calculate each pixel according to the weight configuration result to obtain a pixel entropy distribution map, and obtain multiple matching layout positions according to the pixel entropy distribution map.

7. The method according to claim 6, characterized in that According to the pixel entropy distribution map, a plurality of matching layout positions are obtained, the method comprising: Get the coverage of the acoustic sensor; Identify pixels in the pixel entropy distribution map that are greater than a preset threshold and mark them as high entropy areas; Initialize the position of the acoustic wave sensor, perform iterative calculation according to the initialized position of the acoustic wave sensor, and obtain the coverage rate of each round of iteration, wherein the coverage rate is the ratio of the acoustic wave sensor covering the high entropy area; The positions of the plurality of acoustic wave sensors are optimized according to the coverage rate until the coverage rate meets a preset coverage rate, and the plurality of matching layout positions are obtained.

8. A forest pest and disease monitoring and early warning device, characterized in that: For implementing the steps of the method according to any one of claims 1 to 7, the device comprises: A bird's-eye view image acquisition unit, used to acquire a bird's-eye view image of a target area to obtain a first image set; A regional greening feature analysis unit, configured to perform feature analysis on the first image set to obtain regional greening features, including greening density, tree species distribution, and tree age and height; A ground image acquisition unit, used to acquire ground images of the target area to obtain a second image set; A regional soil feature analysis unit, used for performing feature analysis on the second image set to obtain regional soil features, including soil type and soil moisture; A layout position analysis unit, used to analyze the layout positions of multiple acoustic wave sensors according to the regional greening characteristics and the regional soil characteristics, and obtain multiple matching layout positions; A sensor setting unit, configured to set the plurality of acoustic wave sensors according to the plurality of matching arrangement positions to obtain a sensor integrated module; A monitoring and early warning unit is used to monitor and warn of pests and diseases in the target area according to the sensor integrated module.

Images