A Distributed Sensing-Based Monitoring System and Method for Thinning in Black Locust Forests

Abstract

This invention discloses a monitoring system and method for thinning in black locust forests based on distributed sensing. The system includes distributed sensors, a drone, an edge server, and a cloud platform. Distributed sensors are strategically placed within the target black locust forest to measure light intensity. The edge server acquires measurement data from the distributed sensors within the edge areas, generates a sensor location distribution map of the edge areas, and analyzes whether the distributed sensors are malfunctioning. The cloud platform acquires images of the target black locust forest captured by the drone and divides the forest area into canopy and ground regions based on image recognition algorithms. It also calculates the canopy closure of each edge region and performs correction calculations to obtain thinning parameters. This invention, by deploying distributed sensors within the black locust forest and combining them with images captured by the drone, integrates multi-source data to identify the growth status and environment of the black locust trees, enabling accurate and reliable determination of the thinning timing.

Description

Technical Field

[0001] This invention relates to the field of thinning technology, and in particular to a monitoring system and method for thinning of locust forests based on distributed sensing. Background Technology

[0002] Thinning is an important technical measure in forest management, referring to the periodic removal of some trees in immature forests to adjust stand density and create better growing space and environmental conditions for the remaining trees. Thinning can optimize forest structure, improve timber quality, and promote ecosystem health. As a pioneer tree species with fast growth, drought resistance, and strong sprouting ability, scientific thinning is crucial for the healthy management of black locust plantations.

[0003] Early thinning techniques suffered from reliance on manual labor, low automation, and inefficiency. While some intelligent methods have been incorporated into thinning techniques with technological advancements, the accuracy of determining the optimal timing for thinning remains a concern. Current technologies primarily calculate canopy closure based on the tree's projected area, determining whether thinning is necessary based on whether the canopy closure reaches a threshold. Although canopy closure is an important indicator of locust tree growth, the accuracy of drones in identifying tree canopies is still insufficient. Furthermore, the varying density of individual canopies, coupled with their dynamic nature (being blown by the wind), makes it difficult for drones to accurately identify canopy sparseness and gaps. This leads to inaccurate calculations of the canopy projected area and canopy closure that fail to accurately reflect the locust tree's growing environment, further complicating the determination of the optimal thinning time.

[0004] The existing invention patent application CN116958813A proposes a method for designing thinning operations in artificial forests based on UAV imagery. This patent application first acquires UAV imagery of the target area and reconstructs the imagery to generate DOM and DSM data. Then, it uses computer image processing methods to identify individual trees and extract canopy information. Next, it calculates forest attributes per unit grid and conducts hot and cold spot analysis. Finally, it determines the thinning area and thinning intensity. Using this method for designing thinning operations in artificial forests overcomes the shortcomings of traditional ground surveys in terms of individual tree coordinates and stand spatial information across the entire area. It subdivides the entire forest into grids, fully considering the spatial heterogeneity within the stand and improving the precision of thinning design. It also provides a priority order determination scheme for tree thinning, reducing the subjectivity of the thinning process and improving the scientific rigor of target tree selection. However, the method in this patent application is based on a threshold of canopy closure, and the calculation of canopy closure is simply based on the projection area of ​​the tree crown. This is highly dependent on the accuracy of image recognition. However, the objects captured by drones are leaves that move with the wind, so it is difficult to accurately identify moving tree crowns with different degrees of sparseness. Therefore, it cannot accurately reflect the actual growth environment of the trees and it is difficult to accurately grasp the timing of thinning. Summary of the Invention

[0005] Purpose of the invention: To address the above problems, this invention proposes a monitoring system and method for thinning in black locust forests based on distributed sensing.

[0006] Technical solution:

[0007] In a first aspect, the present invention provides a monitoring system for thinning of locust forests based on distributed sensing, including distributed sensors, drones, edge servers, and cloud platforms;

[0008] Distributed sensors were set up in the target locust forest to measure light intensity;

[0009] The drone was used to collect images of the target locust forest and send them to the cloud platform;

[0010] The edge server includes a distribution map generation module and an anomaly analysis module;

[0011] The distribution map generation module is used to acquire measurement data sent by distributed sensors in the edge area and generate a sensor location distribution map of the edge area.

[0012] The anomaly analysis module is used to analyze whether the distributed sensors are malfunctioning;

[0013] The cloud platform includes a region division module, a canopy closure calculation module, and a thinning parameter analysis module;

[0014] The region segmentation module is used to acquire images of the target locust forest collected by the UAV and divide the target locust forest region into a canopy region and a ground region based on an image recognition algorithm.

[0015] The canopy closure calculation module is used to calculate the canopy closure of each edge region;

[0016] The thinning parameter analysis module is used to calculate and correct the canopy closure of the edge area to obtain the thinning parameters.

[0017] Preferably, the anomaly analysis module is used to analyze whether the distributed sensors are abnormal, including:

[0018] Get the generation time of the regional division map and set it as the first time;

[0019] Query the sunlight and weather information for the target locust forest area in real time;

[0020] The first and second reference ranges are obtained based on sunlight and weather information;

[0021] The edge server accepts the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, determines whether the light intensity data of a type I sensor is within the first reference range, and determines whether the light intensity data of a type II sensor is within the second reference range. If so, there is no abnormality; otherwise, it is determined to be an abnormal sensor.

[0022] Preferably, the cloud platform acquires the total number M of distributed sensors in the edge area and the number m of abnormal sensors in the edge area;

[0023] The correction weight is calculated based on the anomaly ratio m / M of the distributed sensors;

[0024] The canopy closure of each edge region is corrected based on the corrected weights and measurement data to obtain the thinning parameters.

[0025] Secondly, the present invention also provides a method for monitoring thinning in black locust forests based on distributed sensing, the method comprising:

[0026] S2. Each edge server generates a sensor location distribution map of the edge region based on the measurement data; the area where the distributed sensors under the jurisdiction of each edge server are located is an edge region.

[0027] S3. Collect images of the target locust forest area using drones and send the collected images to the cloud platform;

[0028] S4. The cloud platform extracts the tree canopy from the image based on the image recognition algorithm, divides the target locust forest area into the tree canopy area and the ground area, and obtains the target area division map.

[0029] S5, the cloud platform integrates the sensor location distribution map of the edge area with the target area division map, and classifies the distributed sensors;

[0030] S6. Determine whether there are any monitoring anomalies in Class I and Class II sensors;

[0031] S7. The cloud platform calculates the canopy closure C0 of each edge region based on images from the drone.

[0032] S8. Based on the measurement data of the distributed sensors, the canopy closure of the edge region is corrected to obtain the thinning parameters.

[0033] Preferably, S5 includes:

[0034] S51. The cloud platform obtains the sensor location distribution map of the edge area generated by each edge server and synthesizes the sensor location distribution map of the target locust forest area.

[0035] S52. Integrate the sensor location distribution map of the target locust forest area with the target area division map, and classify the distributed sensors into Class I sensors and Class II sensors according to the type of the area where the sensors are located; Class I sensors are located in the canopy area, and Class II sensors are located in the ground area.

[0036] Preferably, S6 includes:

[0037] S61. Obtain the generation time of the regional division map and set it as the first time;

[0038] S62. Query the sunlight and weather information for the target locust forest area at the first moment;

[0039] S63. The cloud platform obtains a first reference range and a second reference range based on sunlight and weather information. The first reference range is the reference range of light intensity under shadow, and the second shadow range is the reference range of light intensity without shadow.

[0040] S64. The edge server receives the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, determines whether the light intensity data of the first type of sensor is within the first reference range, and determines whether the light intensity data of the second type of sensor is within the second reference range. If so, there is no abnormality; otherwise, it is determined to be an abnormal sensor.

[0041] Preferably, S8 includes:

[0042] S81. Obtain the total number M of distributed sensors within the edge region and the number m of abnormal sensors within the edge region;

[0043] S82. Calculate the correction weight α based on the anomaly ratio m / M of the distributed sensors;

[0044] S83. Based on the corrected weights and measurement data, the canopy closure of each edge region is corrected to obtain the thinning parameter D.

[0045] Preferably, step S2 is preceded by a preliminary arrangement step S1, and step S8 is followed by a thinning execution step S9.

[0046] S1 includes: deploying several distributed sensors in the locust forest; the distributed sensors are used to measure light intensity and send the measurement data to the edge server; the measurement data includes sensor number, sensor location, light intensity data, and timestamp;

[0047] S9 includes: executing a thinning strategy when the thinning parameter is greater than or equal to a preset limit; and continuously monitoring the target locust forest when the thinning parameter is less than a threshold.

[0048] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the steps in the method for monitoring thinning of locust forests based on distributed sensing.

[0049] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, preferably wherein the computer program, when executed by a processor, implements the steps in the method for monitoring thinning of locust forests based on distributed sensing.

[0050] The present invention has the following advantages over the prior art:

[0051] 1. This invention deploys distributed sensors within a black locust forest to collect light intensity data. Combined with image data captured by drones, this multi-source data fusion method identifies and assesses the growth status and environment of the black locust trees, resulting in a more accurate and reliable assessment compared to existing technologies. Furthermore, considering that black locust forests are often uninhabited, the distributed sensors are susceptible to damage from animals and insects. To enhance data reliability, this invention uses image recognition algorithms to extract the tree canopy from the images, dividing the target black locust forest area into canopy and ground regions, creating a target area division map. The distributed sensors are then categorized, enabling further assessment of sensor malfunctions and resulting in more reliable and accurate data analysis.

[0052] 2. This invention sets up an edge server in the locust forest and cooperates with a remote cloud platform to form a cloud-edge collaborative monitoring system. The edge server is responsible for determining the distribution location of sensors and judging sensor anomalies, while the cloud platform is responsible for canopy closure identification and weight correction calculations. The allocation of computing resources is more reasonable and it is suitable for large-area locust forests and other forest types.

[0053] 3. This invention first classifies distributed sensors and then determines whether the sensors are abnormal based on the classification results. It determines the correction weight based on the overall proportion of abnormal sensors in the edge area and corrects the canopy closure based on the difference between the light intensity and the median value of the baseline range, thereby obtaining the final thinning parameters. This more realistically reflects the growth status and environment of the black locust tree and can accurately grasp the timing of thinning. Attached Figure Description

[0054] Figure 1 A schematic diagram of a locust forest thinning monitoring system based on distributed sensing provided in an embodiment of the present invention;

[0055] Figure 2 A flowchart of a method for monitoring thinning in black locust forests based on distributed sensing, provided in an embodiment of the present invention;

[0056] Figure 3 This is a flowchart of a method for determining whether a distributed sensor is abnormal, provided by an embodiment of the present invention.

[0057] Figure 4 This is a flowchart of a method for obtaining thinning parameters by correcting canopy closure according to an embodiment of the present invention;

[0058] Figure 5 This is a schematic diagram of an electronic device structure provided in an embodiment of the present invention. Detailed Implementation

[0059] Obviously, many modifications and variations made by those skilled in the art based on the spirit of this invention fall within the scope of protection of this invention.

[0060] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when an element or component is referred to as “connected” to another element or component, it may be directly connected to the other element or component, or there may be intermediate elements or components. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1:

[0063] This invention provides a distributed sensing-based monitoring system for thinning in black locust forests. Please refer to the following for details. Figure 1 , Figure 1 A schematic diagram of a locust forest thinning monitoring system based on distributed sensing provided for an embodiment of the present invention. The system includes: distributed sensors, drones, edge servers, and a cloud platform.

[0064] Distributed sensors were set up in the target locust forest to measure light intensity;

[0065] The drone was used to collect images of the target locust forest and send them to the cloud platform;

[0066] The edge server includes a distribution map generation module and an anomaly analysis module;

[0067] The distribution map generation module is used to acquire measurement data sent by distributed sensors in the edge area and generate a sensor location distribution map of the edge area.

[0068] The anomaly analysis module is used to analyze whether the distributed sensors are malfunctioning;

[0069] The cloud platform includes a region division module, a canopy closure calculation module, and a thinning parameter analysis module;

[0070] The region segmentation module is used to acquire images of the target locust forest collected by the UAV and divide the target locust forest region into a canopy region and a ground region based on an image recognition algorithm.

[0071] The canopy closure calculation module is used to calculate the canopy closure of each edge region;

[0072] The thinning parameter analysis module is used to calculate and correct the canopy closure of the edge area to obtain the thinning parameters.

[0073] Preferably, the anomaly analysis module is used to analyze whether the distributed sensors are abnormal, including:

[0074] Get the generation time of the regional division map and set it as the first time;

[0075] Query the sunlight and weather information for the target locust forest area in real time;

[0076] The first and second reference ranges are obtained based on sunlight and weather information;

[0077] The edge server accepts the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, determines whether the light intensity data of a type I sensor is within the first reference range, and determines whether the light intensity data of a type II sensor is within the second reference range. If so, there is no abnormality; otherwise, it is determined to be an abnormal sensor.

[0078] Preferably, the cloud platform's region partitioning module is used for:

[0079] Obtain the sensor location distribution map of the edge region generated by each edge server, and synthesize the sensor location distribution map of the target locust forest area;

[0080] The sensor location distribution map of the target locust forest area is integrated with the target area division map. Based on the type of area where the sensor is located, the distributed sensors are divided into Class I sensors and Class II sensors. The Class I sensors are located in the canopy area, and the Class II sensors are located in the ground area.

[0081] Preferably, the cloud platform corrects the canopy closure of the edge region based on measurement data from distributed sensors to obtain thinning parameters, including:

[0082] Obtain the total number M of distributed sensors within the edge region and the number m of abnormal sensors within the edge region;

[0083] The correction weight α is calculated based on the anomaly ratio m / M of the distributed sensors.

[0084] When m / M is greater than or equal to the first threshold, α = 0;

[0085] When m / M is greater than or equal to the second threshold and less than the first threshold, α = the first weight;

[0086] When m / M is greater than or equal to the third threshold and less than the second threshold, α = the second weight;

[0087] The canopy closure of each edge region is corrected based on the corrected weights and measurement data to obtain the thinning parameter D;

[0088] In a class of sensors located at the edge region, data from abnormal sensors are removed, and light intensity data L from normal sensors is obtained.

[0089] Subtracting the light intensity data L from the median value L0 of the first reference range yields the difference L. c ;

[0090] Calculate L for all normal sensors c And calculate the average to obtain the mean L. q ;

[0091] The canopy closure C0 of the edge region is corrected to obtain the thinning parameter D=C0-(L q / L0)×α.

[0092] Example 2:

[0093] This invention also provides a method for monitoring thinning in black locust forests based on distributed sensing; please refer to the following for details. Figure 2 , Figure 2 A flowchart of a method for monitoring thinning in black locust forests based on distributed sensing, provided in an embodiment of the present invention, is shown. The method includes the following steps:

[0094] S1. Deploy several distributed sensors in the locust forest;

[0095] The distributed sensor is used to measure light intensity and send the measurement data to the edge server;

[0096] The measurement data includes sensor number, sensor location, light intensity data, and timestamp;

[0097] S2. Each edge server generates a sensor location distribution map of the edge region based on the measurement data;

[0098] The area where the distributed sensors managed by each edge server are located is called an edge region;

[0099] Before step S1, the process also includes establishing a mapping relationship between the edge server and the distributed sensor, so that the distributed sensor sends data to the corresponding edge server;

[0100] Optionally, each distributed sensor sends measurement data to only one edge server, thus forming a many-to-one distributed sensing architecture. In this way, each edge server independently collects and controls the distributed sensors in an edge area, which can improve control efficiency.

[0101] Optionally, each distributed sensor can send measurement data to at least two edge servers, which can improve the redundancy of edge computing and improve the reliability of control.

[0102] S3. Collect images of the target locust forest area using drones and send the collected images to the cloud platform;

[0103] The target locust forest area is the locust forest area to be analyzed to determine whether thinning is necessary.

[0104] S4. The cloud platform extracts the tree canopy from the image based on the image recognition algorithm, divides the target locust forest area into the tree canopy area and the ground area, and obtains the target area division map. There are currently many recognition algorithms for the tree canopy area, but this is not the focus of this invention. In order to ensure the integrity of the technical solution of this invention, the method in the prior art is used for recognition.

[0105] S4 includes:

[0106] S41. Perform preprocessing operations on the image to remove noise and reduce interference.

[0107] S42. Perform grayscale processing on the image to obtain a gradient image;

[0108] S43, initially divided into foreground and background;

[0109] S44. Use the maximum value of the foreground region as the internal label, and perform a watershed transformation on the background region to extract the adjacent regions as the external labels.

[0110] S45. Correct the gradient image so that the image has a minimum value at the inner and outer markers and makes the image flat.

[0111] S46. Perform watershed transformation on the corrected gradient image;

[0112] S47. Merge the connected regions of the segmented image to obtain the segmented foreground canopy image.

[0113] S5, the cloud platform integrates the sensor location distribution map of the edge area with the target area division map, and classifies the distributed sensors;

[0114] S51. The cloud platform obtains the sensor location distribution map of the edge area generated by each edge server and synthesizes the sensor location distribution map of the target locust forest area.

[0115] S52. The sensor location distribution map of the target locust forest area is integrated with the target area division map. Based on the type of area where the sensors are located, the distributed sensors are divided into Class I sensors and Class II sensors. The Class I sensors are located in the canopy area, and the Class II sensors are located in the ground area.

[0116] It is particularly important to note that the acquisition time of the regional delineation map is crucial and must be consistent with the timestamps of the measurement data from the distributed sensors in order to enable comparative analysis.

[0117] S6. Determine whether there are any monitoring anomalies in Class I and Class II sensors; please refer to the following for details. Figure 3 , Figure 3 A flowchart of a method for determining whether a distributed sensor is abnormal, provided by an embodiment of the present invention, includes:

[0118] S61. Obtain the generation time of the regional division map and set it as the first time;

[0119] S62. Query the sunlight and weather information for the target locust forest area at the first moment;

[0120] S63. Based on the sunshine weather information, a first reference range and a second reference range are obtained. The first reference range is the reference range of sunshine intensity under shadow, and the second shadow range is the reference range of sunshine intensity under no shadow.

[0121] S64. The edge server receives the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, determines whether the light intensity data of the first type of sensor is within the first reference range, and determines whether the light intensity data of the second type of sensor is within the second reference range. If so, there is no abnormality; otherwise, it is determined to be an abnormal sensor.

[0122] Since the images captured by the drone are multiple images collected in a continuous period of time, it is difficult to select the time point for image generation. This invention adopts the generation time of the regional division map because the total time spent by the drone in collecting images and the cloud platform in generating the regional division map is relatively short. Within this time, it can be assumed that the weather and the light intensity have not changed.

[0123] Based on images captured by drones, the location of the tree canopy region can be determined through recognition algorithms. However, the obtained tree canopy region is not strictly accurate because the tree canopy is not static but moves with the wind, making it difficult for drone photography and image analysis to accurately determine the precise tree canopy region. Furthermore, the tree canopy region varies in sparseness, and the reference range for different sparse conditions is an interval. Therefore, this invention uses a first reference range and a second reference range to determine whether the distributed light intensity sensor has experienced serious data anomalies.

[0124] Furthermore, multiple sets of real-time data can be collected for analysis and comparison to improve the accuracy of the judgment.

[0125] S7. The cloud platform calculates canopy closure based on images from drones, and calculates the canopy closure C0 for each edge region.

[0126] For each edge region, the area of ​​the canopy region is calculated and divided by the total area of ​​the edge regions to obtain the canopy closure C0 of the edge region.

[0127] S8. Based on the measurement data from distributed sensors, the canopy closure of the edge region is corrected to obtain the thinning parameters; please refer to [reference needed]. Figure 4 , Figure 4 A flowchart of a method for obtaining thinning parameters by correcting canopy closure, provided in an embodiment of the present invention, includes:

[0128] S81. Obtain the total number M of distributed sensors within the edge region and the number m of abnormal sensors within the edge region;

[0129] S82. Calculate the correction weight α based on the anomaly ratio m / M of the distributed sensors;

[0130] When m / M is greater than or equal to the first threshold, α = 0;

[0131] When m / M is greater than or equal to the second threshold and less than the first threshold, α = the first weight;

[0132] When m / M is greater than or equal to the third threshold and less than the second threshold, α = the second weight;

[0133] S83. Based on the corrected weights and measurement data, the canopy closure of each edge region is corrected to obtain the thinning parameter D;

[0134] In a class of sensors located at the edge region, data from abnormal sensors are removed, and light intensity data L from normal sensors is obtained.

[0135] Subtracting the light intensity data L from the median value L0 of the first reference range yields the difference L. c ;

[0136] Calculate L for all normal sensorsc And calculate the average to obtain the mean L. q ;

[0137] The canopy closure C0 of the edge region is corrected to obtain the thinning parameter D=C0-(L q / L0)×α.

[0138] This invention corrects canopy closure based on edge-end calculations. Because the canopy identified by the image has sparse differences, it must be combined with light intensity to more realistically reflect the plant growth status, which is the core indicator of thinning. It should be noted that the value obtained by S8 in this invention is not the canopy closure, but an indicator after correcting the canopy closure, which is a physical parameter that can better reflect the timing of thinning.

[0139] It should be noted that if there are many abnormal sensors in an edge area, it indicates that the environment in this area is relatively poor, the failure rate is high, and the data reliability is low, so the correction weight is reduced; conversely, if there are few abnormal sensors, the data reliability is high, so the correction weight is increased.

[0140] The specific values ​​of the first threshold, second threshold, third threshold, first weight, and second weight mentioned above need to be determined according to the specific environment and tree species. Based on experiments, this invention exemplarily uses a first threshold of 0.8, a second threshold of 0.5, a third threshold of 0.2, a first weight of 0.4, and a second weight of 0.7.

[0141] Additionally, it should be noted that this invention uses a type I sensor for correction during the correction stage, but does not use a type II sensor. This does not mean that the analysis and judgment of the type II sensor in the previous steps is useless, because the division between type I and type II is dynamic. Since trees are in a cycle of growth, felling, and growth, what was once type II may now be classified as type I.

[0142] Furthermore, the main content of this invention is to grasp the timing of thinning based on the thinning parameters calculated by S8. In addition, other thinning strategies should be implemented in a targeted manner according to the characteristics of black locust forests. Taking multiple strategies into account can be more conducive to the growth of black locust.

[0143] Young black locust forests grow rapidly. If the initial planting density is too high (e.g., >2000 trees / hm²), a high canopy closure (>0.8) can form within 3-5 years, leading to a polarization of "dominant trees" and "suppressed trees" competing for light and nutrients. Practice in the sandy areas of the eastern Henan plain shows that black locust forests need their first thinning at 5-7 years old to alleviate competitive pressure. Studies on the Loess Plateau show that when the retention density is 1110 trees / hm² (thinning intensity of about 50%), forest light is significantly improved, soil moisture increases by 40%, and understory biomass and forest productivity reach optimal levels. If the density after thinning is <800 trees / hm², although there is sufficient light, it may lead to an increased risk of soil erosion (especially on slopes) and intensified competition from understory weeds.

[0144] In addition, thinning of black locust forests should also adopt the understory thinning method, prioritizing the removal of suppressed and weak trees, and retaining strong and dominant trees to promote the growth of the main trunk.

[0145] Example 3:

[0146] This invention also provides an electronic device, please refer to the following for details. Figure 5 , Figure 5 This is a schematic diagram of an electronic device structure provided in an embodiment of the present invention, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps in the method for monitoring thinning of locust forests based on distributed sensing.

[0147] Example 4:

[0148] This invention also provides a computer-readable storage medium storing a computer program thereon. Preferably, when the computer program is executed by a processor, it implements the steps in the method for monitoring thinning of locust forests based on distributed sensing.

[0149] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0150] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0151] Finally, it should be noted that in this document, relationships such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

Claims

1. A distributed sensing-based monitoring system for thinning in acacia forests, comprising distributed sensors, drones, edge servers, and a cloud platform; the distributed sensors are distributed throughout the target acacia forest to measure light intensity; the drones are used to acquire images of the target acacia forest and transmit them to the cloud platform; characterized in that, The edge server includes a distribution map generation module and an anomaly analysis module. The distribution map generation module acquires measurement data sent by distributed sensors within the edge area and generates a sensor location distribution map of the edge area. The anomaly analysis module analyzes whether the distributed sensors are abnormal. The cloud platform includes a region division module, a canopy closure calculation module, and a thinning parameter analysis module. The region division module acquires images of the target locust forest collected by UAVs and divides the target locust forest area into canopy areas and ground areas based on image recognition algorithms. The canopy closure calculation module calculates the canopy closure of each edge area. The thinning parameter analysis module corrects the canopy closure of the edge areas to obtain thinning parameters. The cloud platform obtains the total number M of distributed sensors in the edge area and the number m of abnormal sensors in the edge area; The correction weight is calculated based on the anomaly ratio m / M of the distributed sensors; The canopy closure of each edge region is corrected based on the corrected weights and measurement data to obtain the thinning parameters.

2. The black locust forest thinning monitoring system based on distributed sensing according to claim 1, characterized in that, The anomaly analysis module is used to analyze whether distributed sensors are abnormal, including: obtaining the generation time of the regional division map, which is set as the first time; querying the light and weather information of the target locust forest area at the first time; obtaining the first reference range and the second reference range based on the light and weather information; the edge server receives the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, and determines whether the light intensity data of the first type of sensor is within the first reference range and whether the light intensity data of the second type of sensor is within the second reference range. If they are, there is no anomaly; otherwise, the sensor is judged as abnormal.

3. A method for monitoring thinning in black locust forests based on distributed sensing, applied to the black locust forest thinning monitoring system based on distributed sensing as described in any one of claims 1-2, characterized in that, The method includes the following steps: S2. Each edge server generates a sensor location distribution map of the edge region based on the measurement data; the area where the distributed sensors under the jurisdiction of each edge server are located is an edge region. S3. Collect images of the target locust forest area using drones and send the collected images to the cloud platform; S4. The cloud platform extracts the tree canopy from the image based on the image recognition algorithm, divides the target locust forest area into the tree canopy area and the ground area, and obtains the target area division map. S5, the cloud platform integrates the sensor location distribution map of the edge area with the target area division map, and classifies the distributed sensors; S6. Determine whether there are any monitoring anomalies in Class I and Class II sensors; S7. The cloud platform calculates the canopy closure C0 of each edge region based on images from the drone. S8. Based on the measurement data of the distributed sensors, the canopy closure of the edge region is corrected to obtain the thinning parameters.

4. The method for monitoring thinning in black locust forests based on distributed sensing according to claim 3, characterized in that, S5 include: S51. The cloud platform obtains the sensor location distribution map of the edge area generated by each edge server and synthesizes the sensor location distribution map of the target locust forest area. S52. Integrate the sensor location distribution map of the target locust forest area with the target area division map, and classify the distributed sensors into Class I sensors and Class II sensors according to the type of the area where the sensors are located; Class I sensors are located in the canopy area, and Class II sensors are located in the ground area.

5. The method for monitoring thinning in black locust forests based on distributed sensing according to claim 4, characterized in that, S6 include: S61. Obtain the generation time of the regional division map and set it as the first time; S62. Query the sunlight and weather information for the target locust forest area at the first moment; S63. The cloud platform obtains a first reference range and a second reference range based on sunlight and weather information. The first reference range is the reference range of light intensity under shadow, and the second shadow range is the reference range of light intensity without shadow. S64. The edge server receives the first reference range and the second reference range, queries the measurement data corresponding to the timestamp and the first time, determines whether the light intensity data of the first type of sensor is within the first reference range, and determines whether the light intensity data of the second type of sensor is within the second reference range. If so, there is no abnormality; otherwise, it is determined to be an abnormal sensor.

6. The method for monitoring thinning in black locust forests based on distributed sensing according to claim 5, characterized in that, S8 includes: S81. Obtain the total number M of distributed sensors within the edge region and the number m of abnormal sensors within the edge region; S82. Calculate the correction weight α based on the anomaly ratio m / M of the distributed sensors; S83. Based on the corrected weights and measurement data, the canopy closure of each edge region is corrected to obtain the thinning parameter D.

7. The method for monitoring thinning in black locust forests based on distributed sensing according to claim 6, characterized in that, Before step S2, there is a preliminary arrangement step S1, and after step S8, there is a thinning execution step S9. S1 includes: deploying several distributed sensors in the locust forest; the distributed sensors are used to measure light intensity and send the measurement data to the edge server; the measurement data includes sensor number, sensor location, light intensity data, and timestamp; S9 includes: executing a thinning strategy when the thinning parameter is greater than or equal to a preset limit; and continuously monitoring the target locust forest when the thinning parameter is less than a threshold.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the steps of the locust forest thinning monitoring method based on distributed sensing as described in any one of claims 3-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps in the locust forest thinning monitoring method based on distributed sensing as described in any one of claims 3-7.

Images