Unmanned aerial vehicle based sunken greenery vegetation management system

Abstract

The present application relates to the technical field of vegetation maintenance management, and particularly relates to a concave green land vegetation management system based on a UAV, which collects current images of a concave green land area and extracts state information through a UAV collection module, filters analog images and constructs a feature explicit image set through a feature matching module, filters a dynamic representation of a local area and extracts contour performance information of vegetation leaves based on changes in pixel point brightness of the local area through a dynamic filtering module, calculates a correlation coefficient of the contour performance representation and the changes in pixel point brightness to determine whether vegetation state monitoring is effective through a growth monitoring module, and determines a vegetation management strategy for the dynamic representation of the local area according to water aggregation through a processing module. Thus, the accuracy of the determination of the growth state of vegetation under different water aggregation terrains of the concave green land is improved, and the scientificity of the execution of the vegetation management strategy is enhanced.

Description

Technical Field

[0001] This invention relates to the field of vegetation maintenance and management technology, and in particular to a vegetation management system for sunken green spaces based on unmanned aerial vehicles (UAVs). Background Technology

[0002] With the advancement of urbanization in Xiong'an New Area, sunken green spaces, as an important new form of urban green space, play a vital role in the urban ecosystem. These spaces can collect and store rainwater, reduce surface runoff, alleviate urban flooding, and provide a favorable environment for vegetation growth. Traditional management of sunken green space vegetation relies mainly on manual inspections and experience-based judgment, which is not only inefficient but also lacks accuracy. Furthermore, it lacks precise monitoring and control methods for key factors affecting vegetation growth, such as sunlight, water, and nutrients. In addition, with the construction of the new area, the utilization of urban water resources has undergone new changes, making it difficult for traditional management methods to achieve refined management tailored to local conditions.

[0003] For example, Chinese Patent Publication No. CN116773458A discloses a method and apparatus for investigating vegetation stability. The method includes: selecting a multispectral standard light source based on the species of the investigated vegetation community; collecting growth data of the investigated vegetation community using the selected multispectral standard light source; comparing the growth data of the investigated vegetation community with the growth data of standard vegetation communities in a vegetation database to obtain ecological information of the investigated vegetation community; and analyzing the ecological information of the investigated vegetation community through an expert system to obtain factors affecting the ecological information of the investigated vegetation community.

[0004] The following problems still exist in the existing technology: Existing technologies cannot verify the effectiveness of growth monitoring by quantifying the correlation between vegetation outline characteristics and environmental factors. This leads to inaccurate judgments on vegetation growth status under different water accumulation terrains in sunken green spaces, affecting the scientific nature of management strategies and impacting vegetation growth quality and the ecological function of sunken green spaces. Summary of the Invention

[0005] To address this issue, the present invention provides a vegetation management system for sunken green spaces based on unmanned aerial vehicles (UAVs), which overcomes the problem that existing technologies cannot verify the effectiveness of growth monitoring by quantifying the correlation between vegetation outline features and environmental factors, leading to inaccurate judgments on vegetation growth status under different drainage micro-topographic conditions in sunken green spaces.

[0006] To achieve the above objectives, the present invention provides a drone-based vegetation management system for sunken green spaces, comprising: The drone data acquisition module is used to acquire current images of the sunken green area in order to extract status information; The feature matching module is connected to the UAV acquisition module and is used to filter analog images in the historical image database based on the status information, and to construct a feature explicit image set by combining the current image with the filtered analog images according to the time dimension. A dynamic filtering module, which is connected to the feature matching module, is used to filter dynamically represented local regions based on the pixel brightness changes in local regions of the feature-visible image set, and to extract the outline representation information of vegetation leaves in the dynamically represented local regions. The growth monitoring module is connected to the dynamic screening module and the UAV acquisition module respectively. It is used to calculate the contour performance representation quantity based on the contour performance information of adjacent time points in the time series, and calculate the correlation coefficient between the contour performance representation quantity and the pixel brightness change, so as to determine whether the monitoring of vegetation status in the local area is effective. The processing module, which is connected to the growth monitoring module, is used to obtain the dynamic characterization of water accumulation in a local area based on the effective judgment results of vegetation status monitoring, and to determine the vegetation management strategy for the dynamic characterization of the local area based on the water accumulation. The vegetation management strategy includes water level management and application of pesticides and fertilizers.

[0007] Furthermore, the status information includes the location captured by the drone and direct solar radiation.

[0008] Furthermore, the feature matching module is used to construct a set of feature-rich images, wherein, The feature matching module is used to filter images in the historical image database that meet the vegetation status screening criteria as analog images, and to construct the feature-explicit image set by arranging the current image and the analog images in chronological order. The vegetation status screening criterion is that the status information of the image is the same as the status information of the current image.

[0009] Furthermore, the dynamic filtering module is used to determine the changes in pixel brightness, wherein, The dynamic filtering module is used to divide the images in the feature explicit image set into several local regions of the same size, and to determine the average brightness of all pixels in each local region as the dynamic representation quantity of the corresponding local region.

[0010] Furthermore, the dynamic filtering module is used to filter dynamically represented local regions, wherein, The dynamic filtering module calculates the standard deviation of the dynamic representation quantity of each local region in the time dimension, and compares the standard deviation of the dynamic representation quantity with the preset standard deviation reference value. If the standard deviation of the dynamic characterization quantity is greater than the reference value of the standard deviation, the dynamic screening module will screen the local region corresponding to the standard deviation of the dynamic characterization quantity as the local region of dynamic characterization.

[0011] Furthermore, the growth monitoring module is used to extract the outlines of vegetation leaves in a dynamic local area, and the total length of the vegetation leaf outlines is determined as the outline representation quantity.

[0012] Furthermore, the growth monitoring module is used to calculate the correlation coefficient between contour representation parameters and pixel brightness changes, wherein, The growth monitoring module is used to record the difference in contour performance and dynamic performance between adjacent time points in the feature dominant image set, and the Pearson coefficient of the difference in contour performance and dynamic performance is determined as the correlation coefficient.

[0013] Furthermore, the growth monitoring module is used to determine whether the monitoring of vegetation status in a dynamic local area is effective, wherein, The growth monitoring module is used to compare the correlation coefficient with a preset correlation coefficient threshold. If the correlation coefficient is greater than or equal to the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the dynamic local area is effective. If the correlation coefficient is less than the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the local area is invalid.

[0014] Furthermore, the processing module is used to obtain dynamic characteristics of water accumulation in local areas, wherein, The processing module is used to obtain topographic point cloud feature parameters that dynamically represent a local area, and to determine the water accumulation characteristics of the dynamic local area based on the topographic point cloud feature parameters. The water accumulation characteristics include a first water accumulation characteristic and a second water accumulation characteristic.

[0015] Furthermore, the processing module is used to determine vegetation management strategies for dynamically characterized local areas based on the water accumulation situation, wherein... If the water accumulation characteristics of all dynamic characterization local areas are either the first water accumulation characteristic or the second water accumulation characteristic, then the processing module adopts a water level management strategy for the dynamic characterization local areas. If the dynamic characterization of a local area exhibits both a first water accumulation characteristic and a second water accumulation characteristic, then the processing module adopts a strategy of applying pesticides and fertilizers to the dynamically characterized local area.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention uses a drone acquisition module to collect current images of sunken green areas and extract status information; a feature matching module to filter analogous images and construct a set of feature-rich images; a dynamic filtering module to filter dynamically characterized local areas based on pixel brightness changes and extract the outline appearance information of vegetation leaves; a growth monitoring module to calculate the correlation coefficient between outline appearance and pixel brightness changes to determine the effectiveness of vegetation status monitoring; and a processing module to determine vegetation management strategies for dynamically characterized local areas based on water accumulation. Furthermore, this invention achieves the verification of the effectiveness of growth monitoring by quantifying the correlation between vegetation outline features and environmental factors, improving the accuracy of judging vegetation growth status under different water accumulation terrains in sunken green areas, and enhancing the scientific nature of vegetation management strategy implementation.

[0017] Furthermore, by selecting images from the historical image database that have the same status information as the current image as analog images, and then integrating the current image and these analog images in chronological order into a set of characteristic explicit images, this invention can construct a time-series image dataset under a unified illumination benchmark, providing a reliable image basis for accurately analyzing the dynamic changes in vegetation growth at different time points.

[0018] Furthermore, by dividing the image into several local regions of the same size, the present invention can achieve refined segmentation and targeted analysis of green areas. The average brightness of each local region's pixels can reflect the overall brightness characteristics of that region. Calculating the standard deviation of the dynamic representation quantity of each local region in the time dimension can accurately reflect the change in brightness of that region over time, thereby selecting regions where vegetation growth status changes as dynamic representation local regions.

[0019] Furthermore, the total length of the vegetation leaf outline in this invention can directly quantify the vegetation growth status. The degree of leaf lushness will change the intensity of reflected light in the area, thereby affecting the average brightness. The changes of the two are related. The Pearson coefficient can accurately measure the degree of linear correlation between the two variables. By calculating the Pearson coefficient of the difference between the outline performance characteristics and the difference between the dynamic characteristics at adjacent time points, the closeness of the correlation between vegetation growth changes and regional brightness changes is quantified. This allows us to determine whether brightness changes are dominated by vegetation growth changes, and avoid brightness fluctuations caused by other irrelevant factors from interfering with the vegetation growth monitoring results.

[0020] Furthermore, the Pearson coefficient directly reflects the degree of linear correlation between the change in the total length of the vegetation leaf outline and the change in the brightness of regional pixels at adjacent time points. When the coefficient is greater than or equal to the threshold, it indicates that the two are closely related, and the change in regional brightness is mainly driven by the change in vegetation growth. The previously screened dynamic characterization of local areas does indeed correspond to the dynamics of vegetation growth, and the relevant monitoring data is effective. When the coefficient is less than the threshold, it indicates that the brightness change may originate from other interfering factors that are not related to vegetation growth. The monitoring results in this area cannot accurately reflect the vegetation status, so the monitoring is deemed invalid. Thus, the validity of the growth monitoring judgment is verified.

[0021] Furthermore, the topographic point cloud feature parameters in this invention can accurately reflect the core topographic features such as topographic undulation, slope, and elevation of a local area. These topographic features directly determine the direction of water flow, convergence capacity, and retention. Therefore, the topographic point cloud feature parameters can accurately determine the water accumulation characteristics of the area. When the water accumulation characteristics of a local area are uniformly of one type, it indicates that the water conditions in the area are consistent, and the core limiting factor for vegetation growth is concentrated on water content. At this time, water management strategies such as water replenishment and drainage can specifically solve the water adaptation problem. However, when two water accumulation characteristics exist simultaneously in the area, it means that the topography leads to uneven water distribution. Simply adjusting the water level is insufficient to meet the growth needs of vegetation in different areas. Uneven water distribution caused by micro-topographic differences will lead to the migration and accumulation of soil nutrients in the horizontal and vertical directions, thus forming fertility differences in local areas. Therefore, the application of pesticides and fertilizers can supplement soil nutrients, improve soil conditions, take into account the growth needs of vegetation under different water conditions, ensure the overall growth quality of vegetation, and enhance the scientific nature of vegetation management strategies. Attached Figure Description

[0022] Figure 1 This is a system block diagram of a UAV-based vegetation management system for sunken green spaces, as described in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the logic of the dynamic filtering module in this embodiment of the invention for filtering dynamically characterized local regions. Figure 3 This is a flowchart illustrating the logic of determining the effectiveness of dynamic characterization of vegetation status monitoring in a local area, as described in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0024] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0025] It should be noted that in the description of this invention, the terms "upper," "lower," "inner," "outer," etc., which indicate the direction or positional relationship, are based on the direction or positional relationship shown in the drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0026] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation" and "connection" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0027] Please see Figure 1 The diagram shown is a system block diagram of a drone-based sunken green space vegetation management system according to an embodiment of the present invention. The drone-based sunken green space vegetation management system of the present invention includes: The drone data acquisition module is used to acquire current images of the sunken green area in order to extract status information; This invention does not limit the drone data acquisition module; it can be a drone equipped with a laser scanner and a high-definition camera. Such drones are widely used in forest, soil, and water monitoring scenarios, and will not be elaborated here.

[0028] The feature matching module is connected to the UAV acquisition module and is used to filter analog images in the historical image database based on the status information, and to construct a feature explicit image set by combining the current image with the filtered analog images according to the time dimension. This invention does not limit the construction of a historical image database, which can be a log record during the inspection process, including information such as the location of drone shots and direct solar radiation.

[0029] A dynamic filtering module, which is connected to the feature matching module, is used to filter dynamically represented local regions based on the pixel brightness changes in local regions of the feature-visible image set, and to extract the outline representation information of vegetation leaves in the dynamically represented local regions. The growth monitoring module is connected to the dynamic screening module and the UAV acquisition module respectively. It is used to calculate the contour performance representation quantity based on the contour performance information of adjacent time points in the time series, and calculate the correlation coefficient between the contour performance representation quantity and the pixel brightness change, so as to determine whether the monitoring of vegetation status in the local area is effective. The processing module, which is connected to the growth monitoring module, is used to obtain the dynamic characterization of water accumulation in a local area based on the effective judgment results of vegetation status monitoring, and to determine the vegetation management strategy for the dynamic characterization of the local area based on the water accumulation. The vegetation management strategy includes water level management and application of pesticides and fertilizers.

[0030] Specifically, the present invention does not limit the specific structure of the feature matching module, dynamic screening module, growth monitoring module and processing module. They or their units can be constructed using logic components, such as field-programmable logic components, microprocessors, processors used in computers, etc., which will not be elaborated here.

[0031] Specifically, the status information includes the location captured by the drone and direct solar radiation.

[0032] In this invention, the location of the drone's camera is determined using BeiDou positioning.

[0033] This invention does not limit the method of obtaining direct solar radiation. The intensity of direct solar radiation is related to factors such as solar altitude angle, atmospheric transparency, cloud conditions, and altitude. The calculation method of direct solar radiation is well known in the meteorological field and will not be elaborated here.

[0034] Specifically, the feature matching module is used to construct a set of feature-rich images, wherein, The feature matching module is used to filter images in the historical image database that meet the vegetation status screening criteria as analog images, and to construct the feature-explicit image set by arranging the current image and the analog images in chronological order. The vegetation status screening criterion is that the status information of the image is the same as the status information of the current image.

[0035] It is understandable that illumination angle and other state information directly affect the visual characteristics of vegetation and its surrounding environment, such as pixel brightness and color rendering. Images under different illumination conditions are difficult to use directly for comparative analysis of vegetation growth dynamics. However, images with the same state information can eliminate illumination interference to the greatest extent and ensure the consistency and comparability of vegetation growth-related characteristics in the images. Therefore, by selecting images with the same state information as the current image from the historical image database as analog images, and then integrating the current image and these analog images in chronological order into a set of feature-explicit images, a time-series image dataset under a unified illumination benchmark can be constructed, providing a reliable image basis for accurately analyzing the dynamic changes of vegetation growth at different time points.

[0036] Specifically, the dynamic filtering module is used to determine the changes in pixel brightness, wherein, The dynamic filtering module is used to divide the images in the feature explicit image set into several local regions of the same size, and to determine the average brightness of all pixels in each local region as the dynamic representation quantity of the corresponding local region.

[0037] In this invention, a grid division method is adopted, and the actual surface area corresponding to a local area is 0.5m×0.5m.

[0038] In this invention, the pixel brightness is obtained based on image grayscale processing, with the unit being dimensionless grayscale value. The RGB three-channel values ​​are converted into single-channel grayscale values ​​through a grayscale conversion algorithm. The grayscale conversion algorithm is existing technology and will not be described in detail here.

[0039] Please see Figure 2 The diagram shown is a flowchart illustrating the logic of the dynamic filtering module in an embodiment of the present invention for filtering dynamic representation local regions. The dynamic filtering module is used to filter dynamic representation local regions. The dynamic filtering module calculates the standard deviation of the dynamic representation quantity of each local region in the time dimension, and compares the standard deviation of the dynamic representation quantity with the preset standard deviation reference value. If the standard deviation of the dynamic characterization quantity is less than or equal to the standard deviation reference value, the dynamic filtering module will not filter the local area. If the standard deviation of the dynamic characterization quantity is greater than the reference value of the standard deviation, the dynamic screening module will select the local region corresponding to the standard deviation of the dynamic characterization quantity as the local region of dynamic characterization.

[0040] In the implementation of this invention, the standard deviation of the dynamic characterization quantity is specifically calculated by taking all dynamic characterization quantities of a local area in the feature explicit image set in the time series as samples, and using the standard deviation calculation formula to calculate it, retaining two decimal places of precision.

[0041] In this invention, the preset standard deviation reference value is derived from the statistical analysis of historical monitoring data of the concave vegetation. Under the condition that the actual surface area of ​​the local area is 0.5m×0.5m, the standard deviation reference value ranges from [3.5, 7.5]. Preferably, when the vegetation type is herbaceous, the standard deviation reference value can be 5.5. Technicians can also set it according to monitoring needs.

[0042] Understandably, feature-explicit image sets exclude external environmental interference such as lighting. The pixel brightness changes in local areas of the image are more likely to be related to internal dynamics such as vegetation growth. Dividing the image into several local areas of the same size can achieve fine-grained segmentation and targeted analysis of green areas. The average pixel brightness of each local area can reflect the overall brightness characteristics of that area, serving as a core dynamic representation quantity for measuring regional dynamic changes. Standard deviation is an effective indicator for quantifying the degree of data fluctuation. Calculating the standard deviation of the dynamic representation quantity of each local area in the time dimension can accurately reflect the magnitude of brightness change in that area over time, distinguishing between areas with significant brightness changes and areas with gradual changes, and thus selecting areas with changes in vegetation growth status as dynamic representation local areas.

[0043] Specifically, the growth monitoring module is used to extract the outlines of vegetation leaves in a dynamic local area and to determine the total length of the vegetation leaf outlines as the outline representation quantity.

[0044] In this invention, the Canny edge detection algorithm can be used to identify the outline of vegetation leaves. The total length of the outline is obtained by calculating the sum of the Euclidean distances of the edge pixels of the leaf outline in pixels. If there is overlap of leaf outlines, only the outline length of the visible leaf is calculated. The Canny edge detection algorithm and the calculation of the Euclidean distance of pixels are existing technologies and will not be described in detail here.

[0045] Specifically, the growth monitoring module is used to calculate the correlation coefficient between contour representation parameters and pixel brightness changes, wherein, The growth monitoring module is used to record the difference in contour performance and dynamic performance between adjacent time points in the feature dominant image set, and the Pearson coefficient of the difference in contour performance and dynamic performance is determined as the correlation coefficient.

[0046] In this invention, the difference in contour representation is the dimensionless value obtained by subtracting the total contour length of the previous time point from the total contour length of the subsequent time point, and the difference in dynamic representation is the dimensionless value obtained by subtracting the average brightness of the local area of ​​the previous time point from the average brightness of the local area of ​​the corresponding adjacent time point.

[0047] This invention does not limit the calculation method of the Pearson coefficient. The Pearson coefficient is a statistical indicator used to measure the degree of linear correlation between two variables. Its value ranges from -1 to 1. The calculation method of the Pearson coefficient is existing technology and will not be described in detail here.

[0048] For example: There are six sets of contour performance difference (Δy) and dynamic performance difference (Δx); Using the above data, the correlation coefficient r was calculated using the formula for the Pearson coefficient, and the correlation coefficient r≈0.93 was obtained.

[0049] Those skilled in the art will understand that the total length of the vegetation leaf outline can directly quantify the vegetation growth status, such as the degree of lushness, which is the core outline performance characteristic reflecting the changes in vegetation growth. The dynamic performance characteristic is affected by the vegetation growth status. The degree of leaf lushness will change the intensity of reflected light in the area, thus affecting the average brightness. The changes of the two are related. The Pearson coefficient can accurately measure the degree of linear correlation between the two variables. By calculating the Pearson coefficient of the difference between the outline performance characteristic and the difference between the dynamic performance characteristic at adjacent time points, the closeness of the correlation between vegetation growth changes and regional brightness changes is quantified. This allows us to determine whether brightness changes are dominated by vegetation growth changes and avoid brightness fluctuations caused by other irrelevant factors from interfering with the vegetation growth monitoring results.

[0050] Please see Figure 3 The diagram shown is a flowchart illustrating the logic of determining the effectiveness of dynamic characterization of vegetation status monitoring in a local area according to an embodiment of the present invention. The growth monitoring module is used to determine the effectiveness of dynamic characterization of vegetation status monitoring in a local area. The growth monitoring module is used to compare the correlation coefficient with a preset correlation coefficient threshold. If the correlation coefficient is greater than or equal to the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the dynamic local area is effective. If the correlation coefficient is less than the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the local area is invalid.

[0051] In this invention, the preset correlation coefficient threshold is derived from the statistical analysis of historical monitoring data of the concave vegetation. Its purpose is to distinguish whether there is a correlation between changes in brightness and changes in vegetation growth. The value range of the correlation coefficient threshold is [0.88, 0.92]. Preferably, the correlation coefficient threshold in this invention can be 0.9.

[0052] Understandably, the Pearson coefficient directly reflects the degree of linear correlation between the change in the total length of the vegetation leaf outline and the change in the brightness of regional pixels at adjacent time points. When the coefficient is greater than or equal to the threshold, it indicates that the two are closely related, and the change in regional brightness is mainly driven by the change in vegetation growth. The previously screened dynamic characterization of local areas does indeed correspond to the dynamics of vegetation growth, and the relevant monitoring data is valid. When the coefficient is less than the threshold, it indicates that the brightness change may originate from other interfering factors that are not related to vegetation growth. The monitoring results in this area cannot accurately reflect the vegetation status, so the monitoring is deemed invalid. Thus, the validity of the growth monitoring judgment is verified.

[0053] Specifically, the processing module is used to obtain dynamic characteristics of water accumulation in local areas, wherein... The processing module is used to obtain topographic point cloud feature parameters that dynamically represent a local area, and to determine the water accumulation characteristics of the dynamic local area based on the topographic point cloud feature parameters. The water accumulation characteristics include a first water accumulation characteristic and a second water accumulation characteristic.

[0054] In this invention, the terrain point cloud feature parameters are collected by a laser scanner mounted on a drone. The acquisition of terrain point cloud feature parameters adopts a combination of overhead and side-view photography to ensure data integrity and accuracy. Acquiring point cloud data of terrain surfaces by drone is a common method of drone surveying, which will not be elaborated here.

[0055] It is understandable that the first water accumulation feature corresponds to areas with strong surface water collection capacity, and the second water accumulation feature corresponds to areas with strong surface drainage capacity. If the average height coordinate value in the topographic point cloud feature parameters of the dynamic representation of a local area in the direction perpendicular to the horizontal ground is less than the average height coordinate value of the topographic point cloud feature parameters in all local areas, then the water accumulation feature of the dynamic representation of the local area is determined to be the first water accumulation feature. If the average height coordinate value in the topographic point cloud feature parameters of the dynamic representation of a local area in the direction perpendicular to the horizontal ground is greater than or equal to the average height coordinate value of the topographic point cloud feature parameters in all local areas, then the water accumulation feature of the dynamic representation of the local area is determined to be the second water accumulation feature.

[0056] Specifically, the processing module is used to determine vegetation management strategies for dynamically characterized local areas based on the water accumulation situation, wherein... If the water accumulation characteristics of all dynamic characterization local areas are either the first water accumulation characteristic or the second water accumulation characteristic, then the processing module adopts a water level management strategy for the dynamic characterization local areas. If the dynamic characterization of a local area exhibits both a first water accumulation characteristic and a second water accumulation characteristic, then the processing module adopts a strategy of applying pesticides and fertilizers to the dynamically characterized local area.

[0057] In this invention, if all the dynamic characterization local areas have the first water accumulation characteristic, the processing module implements a drainage strategy to lower the water level in the dynamic characterization local areas. If all the dynamic characterizations of the local area show the second water accumulation characteristic, then the processing module will implement a water replenishment strategy to raise the water level in the local area of ​​the dynamic characterization.

[0058] In practice, this invention does not limit the implementation of drainage strategies for lowering water levels or water replenishment strategies for raising water levels. It can utilize the water conveyance channels at the bottom of the depression vegetation in a sponge city to replenish and lower the water level of the depression vegetation. Water level management of depression vegetation in a sponge city is a well-known technical means in the art, and will not be described in detail here.

[0059] In practice, the types and amounts of pesticides and fertilizers applied are determined by technicians based on the vegetation type, and are not specified here.

[0060] Understandably, topographic point cloud feature parameters can accurately reflect the core topographic features such as topographic undulation, slope, and elevation of a local area. These topographic features directly determine the direction of water flow, convergence capacity, and retention. Therefore, topographic point cloud feature parameters can accurately determine the water accumulation characteristics of a region. When a local area exhibits only one type of water accumulation characteristic, it indicates that the water conditions in the area are consistent, and the core limiting factor for vegetation growth is concentrated on water content. In this case, water management strategies such as water replenishment and drainage can specifically address the water matching problem. However, when two types of water accumulation characteristics exist simultaneously in a region, it means that the topography leads to uneven water distribution. Simply adjusting the water level is insufficient to meet the growth needs of vegetation in different areas, and water differences can easily lead to an imbalance in soil nutrient distribution. Therefore, applying pesticides and fertilizers can supplement soil nutrients, improve soil conditions, take into account the growth needs of vegetation under different water conditions, ensure the overall growth quality of vegetation, and enhance the scientific nature of vegetation management strategies.

[0061] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A drone-based vegetation management system for sunken green spaces, characterized in that, include: The drone data acquisition module is used to acquire current images of the sunken green area in order to extract status information; The feature matching module is connected to the UAV acquisition module and is used to filter analog images in the historical image database based on the status information, and to construct a feature explicit image set by combining the current image with the filtered analog images according to the time dimension. A dynamic filtering module, which is connected to the feature matching module, is used to filter dynamically represented local regions based on the pixel brightness changes in local regions of the feature-visible image set, and to extract the outline representation information of vegetation leaves in the dynamically represented local regions. The growth monitoring module is connected to the dynamic screening module and the UAV acquisition module respectively. It is used to calculate the contour performance representation quantity based on the contour performance information of adjacent time points in the time series, and calculate the correlation coefficient between the contour performance representation quantity and the pixel brightness change, so as to determine whether the monitoring of vegetation status in the local area is effective. The processing module, which is connected to the growth monitoring module, is used to obtain the dynamic characterization of water accumulation in a local area based on the effective judgment results of vegetation status monitoring, and to determine the vegetation management strategy for the dynamic characterization of the local area based on the water accumulation. The vegetation management strategy includes water level management and application of pesticides and fertilizers.

2. The UAV-based vegetation management system for sunken green spaces according to claim 1, characterized in that, The status information includes the location captured by the drone and direct solar radiation.

3. The UAV-based vegetation management system for sunken green spaces according to claim 1, characterized in that, The feature matching module is used to construct a set of feature-rich images, wherein, The feature matching module is used to filter images in the historical image database that meet the vegetation status screening criteria as analog images, and to construct the feature-explicit image set by arranging the current image and the analog images in chronological order. The vegetation status screening criterion is that the status information of the image is the same as the status information of the current image.

4. The UAV-based vegetation management system for sunken green spaces according to claim 3, characterized in that, The dynamic filtering module is used to determine the changes in pixel brightness, wherein... The dynamic filtering module is used to divide the images in the feature explicit image set into several local regions of the same size, and to determine the average brightness of all pixels in each local region as the dynamic representation quantity of the corresponding local region.

5. The UAV-based vegetation management system for sunken green spaces according to claim 4, characterized in that, The dynamic filtering module is used to filter dynamically represented local regions, wherein... The dynamic filtering module calculates the standard deviation of the dynamic representation quantity of each local region in the time dimension, and compares the standard deviation of the dynamic representation quantity with the preset standard deviation reference value. If the standard deviation of the dynamic characterization quantity is greater than the reference value of the standard deviation, the dynamic screening module will screen the local region corresponding to the standard deviation of the dynamic characterization quantity as the local region of dynamic characterization.

6. The UAV-based vegetation management system for sunken green spaces according to claim 5, characterized in that, The growth monitoring module is used to extract the outlines of vegetation leaves in a dynamic local area and to determine the total length of the vegetation leaf outlines as the outline representation quantity.

7. The UAV-based vegetation management system for sunken green spaces according to claim 6, characterized in that, The growth monitoring module is used to calculate the correlation coefficient between contour representation parameters and pixel brightness changes, wherein, The growth monitoring module is used to record the difference in contour performance and dynamic performance between adjacent time points in the feature dominant image set, and the Pearson coefficient of the difference in contour performance and dynamic performance is determined as the correlation coefficient.

8. The UAV-based vegetation management system for sunken green spaces according to claim 7, characterized in that, The growth monitoring module is used to determine whether the monitoring of the dynamic characterization of vegetation status in a local area is effective. The growth monitoring module is used to compare the correlation coefficient with a preset correlation coefficient threshold. If the correlation coefficient is greater than or equal to the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the dynamic local area is effective. If the correlation coefficient is less than the correlation coefficient threshold, the growth monitoring module determines that the monitoring of the vegetation status of the local area is invalid.

9. The UAV-based vegetation management system for sunken green spaces according to claim 8, characterized in that, The processing module is used to obtain dynamic characteristics of water accumulation in local areas, wherein... The processing module is used to obtain topographic point cloud feature parameters that dynamically represent a local area, and to determine the water accumulation characteristics of the dynamic local area based on the topographic point cloud feature parameters. The water accumulation characteristics include a first water accumulation characteristic and a second water accumulation characteristic.

10. The UAV-based vegetation management system for sunken green spaces according to claim 9, characterized in that, The processing module is used to determine vegetation management strategies for dynamically characterized local areas based on the water accumulation situation, wherein... If the water accumulation characteristics of all dynamic characterization local areas are either the first water accumulation characteristic or the second water accumulation characteristic, then the processing module adopts a water level management strategy for the dynamic characterization local areas. If the dynamic characterization of a local area exhibits both a first water accumulation characteristic and a second water accumulation characteristic, then the processing module adopts a strategy of applying pesticides and fertilizers to the dynamically characterized local area.

Images