Mangrove forest ecological management system and method based on unmanned aerial vehicle
By using two identification models to identify vegetation growth state and pest state respectively in mangrove ecological management, and conducting quantitative analysis, the problems of inaccurate identification and lack of quantitative standards in decision-making in the existing technology are solved, and more efficient and scientific mangrove ecological management is achieved.
Patent Information
- Application Number
- CN202411850485.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-16
- Publication Date
- 2025-04-22
AI Technical Summary
The existing mangrove ecological management methods based on single model identification have inaccuracies in identifying multiple vegetation states and pests and diseases, and lack of quantitative standards for management decisions, resulting in the management plan being not scientific and reasonable enough.
Two identification models were used to identify the vegetation growth status and pest status of mangroves, and scores of vegetation growth status and pest status were obtained through quantitative analysis. The scores were comprehensively obtained to obtain the health status score of mangroves, thereby formulating a scientific and reasonable management plan.
It improves the recognition efficiency, reduces the image data processing volume of a single model, provides more scientific analysis results and more targeted management plans, and realizes the effectiveness and scientific nature of mangrove ecological management.
Smart Images
Figure CN119863718A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ecological management, and in particular to a mangrove ecological management system and method based on unmanned aerial vehicle (UAV). Background Art
[0002] In order to better protect and manage mangrove resources, modern scientific and technological means, especially the application of drone technology, have played an important role in mangrove ecological monitoring and management. Traditional mangrove ecological management methods usually rely on ground surveys and satellite remote sensing, which are not only time-consuming and labor-intensive, but also have limitations in the real-time and accuracy of data acquisition. In recent years, drone technology has been widely used in mangrove ecological monitoring due to its high flexibility, low cost, and high resolution.
[0003] At present, the mangrove ecological management method based on drones mainly includes the following steps: first, use drones equipped with multispectral cameras or other sensor equipment to take aerial photos of the mangrove area; second, use image processing software to preliminarily process the collected data and extract key parameters such as vegetation index; finally, input the processed data into a pre-trained machine learning model or deep learning model to identify the growth status, pest and disease conditions of mangrove vegetation and other information. Based on the output results of the model, managers can formulate management plans for different health states, thereby achieving effective management of mangrove ecology. However, the existing methods based on single model recognition have some significant defects: First, it is difficult for a single model to accurately identify multiple types of vegetation conditions and pests and diseases at the same time. Especially in complex and changing natural environments, a single model may have inaccurate recognition due to the limitations of the data set; Second, due to the high complexity and diversity of the mangrove ecosystem itself, a single model is prone to bottlenecks when processing large amounts of image data. In particular, when the amount of data is huge, the computational load of the model increases, resulting in reduced recognition efficiency. 3. Although existing methods can make management decisions through model output results, these decisions often lack clear quantitative standards, making the management plans formulated not scientific and reasonable enough to accurately reflect the actual health status of mangroves.
[0004] Therefore, the prior art urgently needs a technical solution for a mangrove ecological management system and method based on drones. Summary of the invention
[0005] In order to solve the above technical problems, the present invention provides a mangrove ecological management method based on drones, which specifically includes the following steps: Step S1, collecting image data of the target mangrove area by using a drone, and preprocessing the image data; Step S2, inputting the preprocessed image data into the recognition model to obtain the vegetation growth status and the vegetation pest status of the target mangrove area; Step S2a, inputting the preprocessed image data into the first recognition model to obtain the vegetation growth status of the target mangrove area; Step S2a1, performing a first segmentation on the preprocessed image data to obtain a target mangrove area within the boundary; Step S2a2: using an image feature extraction algorithm to extract features of the target mangrove area within the boundary; Step S2a3, constructing a first recognition model, and using historical image data to train the first recognition model until the first recognition model converges, thereby obtaining a trained first recognition model; Step S2a4, inputting the characteristics of the target mangrove area within the boundary into the first recognition model to obtain the vegetation growth status of the target mangrove area; Step S2b, inputting the preprocessed image data into a second recognition model to obtain the vegetation pest status of the target mangrove area; Step S2b1, performing secondary segmentation on the image data after the primary segmentation to obtain the suspected damaged plants in the target mangrove area within the boundary; Step S2b2, performing morphological processing on the suspected damaged plants to obtain the characteristics of the suspected damaged plants in the target mangrove area within the boundary; Step S2b3, constructing a second recognition model, and using historical image data to train the second recognition model until the second recognition model converges, thereby obtaining a trained second recognition model; Step S2b4, inputting the characteristics of the suspected damaged plants in the target mangrove area within the boundary into the second recognition model to obtain the vegetation pest status of the target mangrove area; Step S3, comprehensively analyzing the vegetation growth status and vegetation pest status to obtain a health status score of the target mangrove area; Step S3a, further analyzing the vegetation growth status of the target mangrove area obtained by the first recognition model to obtain a vegetation growth status score of the target mangrove area; Step S3a1, extracting key indicators according to the characteristics of the target mangrove area within the boundary; Step S3a2: setting at least two first thresholds for each key indicator, setting at least two levels for each key indicator according to the at least two first thresholds, and setting different scores for each key indicator according to different levels, wherein the scores include an upper score limit and a lower score limit; Step S3a3, comprehensively analyzing the upper and lower limits of the scores of each key indicator at different levels, and establishing a quantitative scoring table for vegetation growth status, wherein the quantitative scoring table for vegetation growth status includes the vegetation growth status level; Step S3a4: Analyze each current key indicator to obtain the indicator value of each key indicator; Step S3a5: Analyze the influence of each current key indicator on the vegetation growth state, rank each key indicator according to the analysis result, and assign a weight to each key indicator according to the ranking result; Step S3a6: Obtain a vegetation growth status score of the target mangrove area according to the indicator value of each key indicator and the weight of each key indicator; The calculation formula for obtaining the vegetation growth status score of the target mangrove area is: ; in, A score representing the vegetation growth status of the target mangrove area; Represents the maximum number of key indicators; Represents the weight of the i-th key indicator; Represents the indicator value of the i-th key indicator; Step S3b, further analyzing the vegetation pest status of the target mangrove area obtained by the second recognition model to obtain a vegetation pest status score of the target mangrove area; Step S3c, comprehensively analyzing the vegetation growth status score and the vegetation pest status score of the target mangrove forest area to obtain a health status score of the target mangrove forest area; The calculation formula for obtaining the health status score of the target mangrove area is: ; in, A score representing the health status of the target mangrove area; A score representing the vegetation growth status of the target mangrove area; A score representing the vegetation pest status of the target mangrove area; Represents the preset weight coefficient, 0< <1, used to balance the impact of vegetation growth status score and vegetation pest status score on health status score; Step S4: Develop a management plan based on the health status score of the target mangrove area.
[0006] A mangrove ecological management system based on drones, using a mangrove ecological management method based on drones as described in any of the above items, comprising the following modules: Data acquisition and preprocessing module: used to collect image data of the target mangrove area through drones and preprocess the image data; State recognition module: connected to the data acquisition and preprocessing module, used to input the preprocessed image data into the recognition model to obtain the vegetation growth state and the vegetation pest state of the target mangrove area; Analysis and evaluation module: connected to the state recognition module, used to comprehensively analyze the vegetation growth state and vegetation pest state to obtain the health status score of the target mangrove area; Plan formulation module: connected to the analysis and evaluation module, used to formulate a management plan based on the health status score of the target mangrove area.
[0007] The embodiments of the present invention have the following technical effects: The present invention sets a first recognition model and a second recognition model to respectively identify the vegetation growth status and suspected damaged plants in the target mangrove area. Compared with the existing method of using only a single model for recognition, this can effectively reduce the image data processing amount of each recognition model, thereby achieving the effect of improving recognition efficiency, and each recognition model can avoid the situation where the single model recognition accuracy is poor under the premise of being able to effectively perform targeted recognition; at the same time, the present invention quantifies the recognition results of each recognition model, so as to obtain a more data-driven and scientific analysis result. Specifically, by further quantitatively analyzing the vegetation growth status, a quantitative score of the vegetation growth status of the target mangrove area is obtained, and similarly, a quantitative score of the vegetation pest status of the target mangrove area is obtained. Based on these two quantitative scores, a quantitative score that more accurately reflects the health status of the target mangrove area after quantification can be obtained, so that the formulation of the subsequent corresponding management plan can be carried out based on a more scientific quantitative basis, so that the ecological management of mangroves can be more effective and scientific; At the same time, the present invention divides the management plan into a first management plan and a second management plan under two situations based on the second threshold, wherein the first management plan is mainly set based on the health status score of the target mangrove area being greater than or equal to the second threshold, that is, it is mainly set when the health status of the target mangrove area is good, and the setting of the second management plan is similar. Therefore, the specific formulation process of the first management plan and the second management plan is completely formulated based on the health status score of the target mangrove area, which makes the management plan diverse. In addition, in order to further enhance the pertinence of the formulation of the management plan, in the specific formulation process, the management plan can be further optimized and formulated based on the vegetation growth status score, the vegetation pest status score and even the corresponding levels, so that the management plan can be more targeted. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0009] Figure 1 is a flow chart of a mangrove ecological management method based on drones provided in an embodiment of the present invention; Figure 2 It is a framework diagram of a mangrove ecological management system based on drones provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0010] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.
[0011] Embodiment 1: Figure 1 As shown, the present invention provides a mangrove ecological management method based on drones, comprising the following steps: Step S1, collecting image data of the target mangrove area by using a drone, and preprocessing the image data; In terms of drone selection, drones suitable for long-term flight and with strong load-bearing capacity are preferred to ensure that they can carry the necessary sensors. The configured sensors include but are not limited to visible light cameras, multispectral cameras and thermal imaging cameras. Regarding the limitations on drone flight altitude, route and time: first determine the appropriate flight altitude based on the required image resolution and drone performance, then plan a flight route covering the entire target mangrove area to ensure that the image can fully cover the target area, and choose a time period with good lighting conditions to fly in order to obtain high-quality image data; The collected image data includes but is not limited to visible light images, multispectral images, and thermal imaging images; Among them, visible light images are collected by flying a visible light camera over the target mangrove area to observe the overall appearance of the mangroves; multispectral images are collected by a multispectral camera to capture spectral information in different bands for analyzing the health of vegetation; thermal imaging images are collected by a thermal imaging camera to detect potential abnormal temperature distribution, such as local temperature increases that may occur in pest and disease areas; Preprocessing of the above image data includes but is not limited to: image format conversion, image quality inspection, image correction, etc.; When converting image formats, the collected images in various formats are uniformly converted into a standard format that is easy to process; when checking image quality, the main focus is on checking whether the image is clear, whether the exposure is normal, and whether there are any problems such as cloud obstruction; when correcting the image, it includes geometric correction and radiation correction, among which geometric correction is mainly to correct image distortion caused by changes in the drone's attitude, and radiation correction is mainly to correct image radiation differences caused by factors such as atmospheric scattering and reflection.
[0012] Step S2, inputting the preprocessed image data into the recognition model to obtain the vegetation growth status and the vegetation pest status of the target mangrove area; Step S2a, inputting the preprocessed image data into the first recognition model to obtain the vegetation growth status of the target mangrove area; Step S2a1, performing a first segmentation on the preprocessed image data to obtain a target mangrove area within the boundary; The purpose of this embodiment is to separate the target mangrove area from other background areas in the preprocessed image data through the first segmentation, so as to focus on the analysis of the target mangrove area; specifically, when the preprocessed image data is segmented by the threshold segmentation method, since the mangrove area usually presents a distinct green color, the target mangrove area and the background are distinguished by setting a green range threshold, and the green range threshold needs to be selected based on the hue value in the HSV color space, and then the target mangrove area can be separated from the background based on the green range threshold, and then the Canny edge detection algorithm is applied to detect the edge of the segmented target mangrove area, and finally, the image segmentation algorithm is further used to refine the boundary of the target mangrove area by combining the results of threshold segmentation and edge detection. Through the above steps, the target mangrove area within the boundary can be successfully obtained, that is, the mangrove area is successfully segmented from the original image; Step S2a2: using an image feature extraction algorithm to extract features of the target mangrove area within the boundary; The purpose of this embodiment is to extract key features that can reflect the growth status of mangroves from the target mangrove area within the boundary through a feature extraction algorithm, wherein the feature extraction algorithm includes but is not limited to: texture features, using gray level co-occurrence matrix to calculate the texture features of the mangrove area, such as contrast, homogeneity, etc.; color features: calculating the color features of the mangrove area, such as average color value, etc.; shape features: calculating the geometric properties of the mangrove area, such as area, perimeter, aspect ratio, etc.; Specifically, when extracting texture features, the gray-level co-occurrence matrix is used to calculate the texture features of the mangrove area, and the contrast, homogeneity and other features are extracted by analyzing the matrix in different directions; when extracting color features, the image of the mangrove area is converted to the HSV color space, and then the average hue value, average saturation value and average brightness value of the mangrove area are calculated to extract the hue, saturation and brightness features; when extracting shape features, the functions provided by the OpenCV library are used to calculate the area, perimeter and other geometric properties of the mangrove area, and further calculate the shape features of the mangrove area such as the aspect ratio and roundness; through the above steps, the key features reflecting the growth status of the target mangrove area within the boundary can be successfully extracted, and these features will be used for subsequent analysis; Step S2a3, constructing a first recognition model, and using historical image data to train the first recognition model until the first recognition model converges, thereby obtaining a trained first recognition model; The purpose of this embodiment is to construct and train a first recognition model so that it can identify the growth status of mangroves based on the extracted features; wherein the selection of the first recognition model includes but is not limited to: random forest, support vector machine or neural network, etc.; then the first recognition model is trained. During the training process, a large amount of historical image data is first collected, and the growth status of the mangrove area in each image is marked, such as "good", "average", "poor", etc., that is, the historical image data with labels are used as part of the first recognition model training set, and the features extracted in step S2a2 are combined with the above training set to form a training set for the first recognition model training, and then the first recognition model is trained using the training set. Training is performed to obtain a first recognition model that has been trained. It is worth noting that during the training process, the first recognition model needs to be cross-validated to evaluate the performance of the model, ensure that the model has good generalization ability, and adjust model parameters, such as learning rate, regularization term, etc., to improve the accuracy and convergence speed of the model. At the same time, in the process of constructing the above training set, it is necessary to further extract one quarter of the training set from the training set at a ratio of 4:1 as a validation set. The main function of the validation set is that when the performance of the first recognition model on the validation set is no longer significantly improved, it is considered that the model has converged. Through the above steps, a first recognition model that can identify the growth status of mangroves according to the extracted features can be constructed and trained.
[0013] Step S2a4, inputting the characteristics of the target mangrove area within the boundary into the first recognition model to obtain the vegetation growth status of the target mangrove area; The purpose of this embodiment is to identify the vegetation growth status of the target mangrove area by inputting the characteristics of the target mangrove area within the boundary into the trained first recognition model; specifically, the characteristics of the target mangrove area within the boundary extracted in step S2a2 are used and input into the trained first recognition model, and the first recognition model outputs the vegetation growth status of the target mangrove area based on the input characteristics.
[0014] Step S2b, inputting the preprocessed image data into a second recognition model to obtain the vegetation pest status of the target mangrove area; Step S2b1, performing secondary segmentation on the image data after the primary segmentation to obtain the suspected damaged plants in the target mangrove area within the boundary; The purpose of this embodiment is to further refine the mangrove area obtained by the first segmentation through secondary segmentation, identify suspected damaged plants, and analyze their pest status later; specifically, firstly, morphological operations are used to refine the boundaries of the mangrove area and highlight the characteristics of the damaged plants. Preferably, a connected domain analysis method is then applied to separate the suspected damaged plants from other healthy plants. Preferably, finally, thresholds are set for specific characteristics of the damaged plants, such as color and shape, to further segment the suspected damaged plants. In the specific operation, first, when applying the morphological operation, the erosion operation is used to remove small particles or noise on the boundary, and then the dilation operation is used to fill the internal holes or connect adjacent plants; preferably, next, when applying the connected domain analysis operation, the connected domain analysis method is applied to identify independent connected areas in the image, and each connected area is analyzed to identify suspected damaged plants; preferably, finally, when applying the feature threshold segmentation operation, a threshold is set according to the color characteristics of the damaged plants, and the suspected damaged plants are screened out, and the shape characteristics of the screened areas are further analyzed, such as irregular shapes, holes, etc., to confirm the suspected damaged plants; through the above steps, the suspected damaged plants in the target mangrove area within the boundary can be successfully obtained; Step S2b2, performing morphological processing on the suspected damaged plants to obtain the characteristics of the suspected damaged plants in the target mangrove area within the boundary; The purpose of this embodiment is to highlight the characteristics of suspected damaged plants through morphological processing to facilitate subsequent analysis; since morphological operations have been applied in the above segmentation process, it is worth noting that in the above morphological operation process, small particles or noise on the boundary are removed, and then the internal holes are filled with dilation operations, which belong to the opening operation and closing operation in morphology respectively. Therefore, the contour extraction algorithm is applied next to extract the contour of the suspected damaged plant, and the characteristics of the suspected damaged plant are calculated according to the extracted contour, such as calculating the area, perimeter, circularity and other geometric characteristics of the suspected damaged plant, calculating the color characteristics of the suspected damaged plant, such as the average color value, standard deviation, etc., calculating the texture characteristics of the suspected damaged plant, such as the contrast and homogeneity in the grayscale co-occurrence matrix; through the above steps, the characteristics of the suspected damaged plant can be successfully extracted from the target mangrove area within the boundary, and these characteristics will be used for subsequent analysis; Step S2b3, constructing a second recognition model, and using historical image data to train the second recognition model until the second recognition model converges, thereby obtaining a trained second recognition model; Step S2b4, inputting the characteristics of the suspected damaged plants in the target mangrove area within the boundary into the second recognition model to obtain the vegetation pest status of the target mangrove area; It is worth noting that the selection of the second recognition model here also includes but is not limited to: random forest, support vector machine or neural network, etc. The specific training process and recognition process are consistent with the operation process of the first recognition model, so they will not be repeated one by one.
[0015] Step S3, comprehensively analyzing the vegetation growth status and vegetation pest status to obtain a health status score of the target mangrove area; Step S3a, further analyzing the vegetation growth status of the target mangrove area obtained by the first recognition model to obtain a vegetation growth status score of the target mangrove area; Step S3a1, extracting key indicators according to the characteristics of the target mangrove area within the boundary, wherein the key indicators include but are not limited to vegetation coverage, average chlorophyll content, vegetation density, average plant height, etc.; Step S3a2: setting at least two first thresholds for each key indicator, setting at least two levels for each key indicator according to the at least two first thresholds, and setting different scores for each key indicator according to different levels, wherein the scores include an upper score limit and a lower score limit; This embodiment will be described based on the above key indicators, wherein the purpose of setting the first threshold and level is to set the first threshold and level for each key indicator so as to quantify the vegetation growth state; illustratively, when the first threshold is set, at least two first thresholds set for the vegetation coverage rate include but are not limited to 60% and 80%, which means that when the vegetation coverage rate is greater than or equal to 80%, the level is defined as excellent, when the vegetation coverage rate is greater than or equal to 60% and less than 80%, the level is defined as good, and when the vegetation coverage rate is less than 60%, the level is defined as general, and so on. Further including but not limited to the definition of poor and extremely poor grades; based on this, the upper and lower limits of the vegetation coverage rate score are obvious, when it is excellent, the score is [80, 100], when it is good, the score is [60, 80], and when it is general, the score is [1, 60]. Of course, the definition standard of the specific score and the setting of the first threshold value need to be based on the actual situation, and the opinions of the expert group need to be referred to to ensure the scientific setting of the specific values and the accuracy of the results; at the same time, the first threshold setting and grade definition of other key indicators are consistent with the above; the following Table 1 is a schematic diagram of the key indicator grade division and scoring standard; Table 1: Schematic diagram of key indicator classification and scoring criteria ; Step S3a3, comprehensively analyzing the upper and lower limits of the scores of each key indicator at different levels, and establishing a quantitative scoring table for vegetation growth status, wherein the quantitative scoring table for vegetation growth status includes the vegetation growth status level; It is worth noting that the setting of the entire vegetation growth state quantitative scoring table needs to be further measured and set after obtaining accurate and scientific scores of key indicators; in the specific setting process, it is necessary to combine the upper and lower limits of the scores of each key indicator at different levels. For example, assuming that only vegetation coverage and average chlorophyll content are used as the setting standards of the vegetation growth state quantitative scoring table, then in the vegetation growth state quantitative scoring table, the score for when the vegetation growth state level is excellent should be based on the upper and lower limits of the vegetation coverage when the level is excellent, and the upper and lower limits of the average chlorophyll content when the level is excellent. Only after the combination can the upper and lower limits of the score when the vegetation growth state level is excellent be obtained. When specifically combined, the highest value of the upper limit and the lowest value of the lower limit are obtained first; Step S3a4: Analyze each current key indicator to obtain the indicator value of each key indicator; Step S3a5: Analyze the influence of each current key indicator on the vegetation growth state, rank each key indicator according to the analysis result, and assign a weight to each key indicator according to the ranking result; Step S3a6: obtaining a vegetation growth status score of the target mangrove area according to the indicator value of each key indicator and the weight of each key indicator, and determining the vegetation growth status grade of the target mangrove area with reference to the vegetation growth status grade in the vegetation growth status quantitative scoring table; The calculation formula for obtaining the vegetation growth status score of the target mangrove area is: ; in, A score representing the vegetation growth status of the target mangrove area; Represents the maximum number of key indicators; Represents the weight of the i-th key indicator; Represents the indicator value of the i-th key indicator; It is worth noting that after obtaining the index value of each key indicator, the vegetation coverage rate is determined by calculating the ratio of the number of pixels of vegetation in the target mangrove area to the number of pixels of the entire image. After the vegetation coverage rate is calculated, assuming that the calculated vegetation coverage rate is Y%, the Y value in the Y% is used as the score of the current vegetation coverage rate, and after comparing with the corresponding Table 1, the level of the vegetation coverage rate is determined. Similarly, the index value of each key indicator is calculated, and through the above formula and after combining the weight of each key indicator, the vegetation growth status score of the target mangrove area can be obtained, and after comparing with the vegetation growth status quantitative score table, the vegetation growth status level of the target mangrove area can be further obtained.
[0016] Step S3b, further analyzing the vegetation pest status of the target mangrove area obtained by the second recognition model to obtain a vegetation pest status score of the target mangrove area; It is worth noting that the method of obtaining the vegetation pest status score of the target mangrove area is similar to the method of obtaining the vegetation growth status score of the target mangrove area. The differences lie in the different key indicators and specific values. The tabulation and calculation processes are basically the same, so I will not go into details.
[0017] Step S3c, comprehensively analyzing the vegetation growth status score and the vegetation pest status score of the target mangrove forest area to obtain a health status score of the target mangrove forest area; The calculation formula for obtaining the health status score of the target mangrove area is: ; in, A score representing the health status of the target mangrove area; A score representing the vegetation growth status of the target mangrove area; A score representing the vegetation pest status of the target mangrove area; Represents the preset weight coefficient, 0< <1, used to balance the impact of vegetation growth status score and vegetation pest status score on health status score; Step S4: setting a second threshold based on the health status score of the target mangrove area, and formulating a first management plan when the health status score of the target mangrove area is greater than or equal to the second threshold; and formulating a second management plan when the health status score of the target mangrove area is less than the second threshold; It is worth noting that when the health status score of the target mangrove area is greater than or equal to the second threshold, the first management plan formulated is divided into multiple situations. This is mainly based on how much the health status score of the target mangrove area exceeds the second threshold when it is greater than or equal to the second threshold. If it exceeds enough, it proves that the health status of the target mangrove area is good enough, and the first management plan formulated at this time should focus on long-term maintenance of this status; if it exceeds less, it proves that although the health status of the target mangrove area is already very good, it is still not the best state. At this time, the first management plan formulated can focus on maintaining the existing status and try to further improve it through slight intervention; Similarly, when formulating the second management plan, if the score is much lower than the second threshold, it proves that the health status of the target mangrove area is poor and urgent restoration measures are needed; if the score is much lower than the second threshold, it proves that although the health status of the target mangrove area is acceptable, there are still some serious problems. At this time, the second management plan formulated can focus on solving these problems and promoting restoration; Furthermore, in the process of formulating the first and second management plans, further reference can be made to the vegetation growth status score of the target mangrove area and the vegetation pest status score of the target mangrove area, and an intuitive analysis can be made as to which of the two scores the current score is specifically affected by, so as to formulate a targeted management plan. For example, a targeted pest and disease control plan can be introduced into the above management plan to address the situation where the health status score is affected when the vegetation pest status score is low or relatively low. Furthermore, a more reasonable and scientific pest and disease control plan can be formulated based on the current vegetation pest status level associated with the current vegetation pest status score.
[0018] Embodiment 2: Figure 2 As shown, the present invention also proposes a mangrove ecological management system based on drones, which implements the above-mentioned mangrove ecological management method based on drones, including the following modules: It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude the presence of other identical elements in the process, method or device including the elements.
[0019] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0020] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.
Claims
1. A mangrove ecological management method based on drones, characterized in that: The following steps are involved: Step S1, collecting image data of the target mangrove area by using a drone, and preprocessing the image data; Step S2, inputting the preprocessed image data into the recognition model to obtain the vegetation growth status and the vegetation pest status of the target mangrove area; Step S3, comprehensively analyzing the vegetation growth status and vegetation pest status to obtain a health status score of the target mangrove area; Step S4: setting a second threshold based on the health status score of the target mangrove area, and formulating a first management plan when the health status score of the target mangrove area is greater than or equal to the second threshold; When the health status score of the target mangrove area is less than the second threshold, a second management plan is formulated.
2. The mangrove ecological management method based on drone according to claim 1 is characterized in that: The step S2, inputting the pre-processed image data into the recognition model, respectively obtaining the vegetation growth status of the target mangrove area and the vegetation pest status of the target mangrove area, includes: Step S2a, inputting the preprocessed image data into the first recognition model to obtain the vegetation growth status of the target mangrove area; Step S2b: input the preprocessed image data into the second recognition model to obtain the vegetation pest status of the target mangrove area.
3. The mangrove ecological management method based on drone according to claim 2 is characterized in that: The step S2a, inputting the pre-processed image data into the first recognition model to obtain the vegetation growth status of the target mangrove area, includes: Step S2a1, performing a first segmentation on the preprocessed image data to obtain a target mangrove area within the boundary; Step S2a2: using an image feature extraction algorithm to extract features of the target mangrove area within the boundary; Step S2a3, constructing a first recognition model, and using historical image data to train the first recognition model until the first recognition model converges, thereby obtaining a trained first recognition model; Step S2a4: input the characteristics of the target mangrove area within the boundary into the first recognition model to obtain the vegetation growth status of the target mangrove area.
4. The mangrove ecological management method based on drone according to claim 2 is characterized in that: The step S2b, inputting the pre-processed image data into the second recognition model to obtain the vegetation pest status of the target mangrove area, includes: Step S2b1, performing secondary segmentation on the image data after the primary segmentation to obtain the suspected damaged plants in the target mangrove area within the boundary; Step S2b2, performing morphological processing on the suspected damaged plants to obtain the characteristics of the suspected damaged plants in the target mangrove area within the boundary; Step S2b3, constructing a second recognition model, and using historical image data to train the second recognition model until the second recognition model converges, thereby obtaining a trained second recognition model; Step S2b4: input the characteristics of the suspected damaged plants in the target mangrove area within the boundary into the second recognition model to obtain the vegetation pest status of the target mangrove area.
5. The mangrove ecological management method based on drone according to claim 1 is characterized in that: The step S3 comprehensively analyzes the vegetation growth status and the vegetation pest status to obtain the health status index of the target mangrove area, including: Step S3a, further analyzing the vegetation growth status of the target mangrove area obtained by the first recognition model to obtain a vegetation growth status score of the target mangrove area; Step S3b, further analyzing the vegetation pest status of the target mangrove area obtained by the second recognition model to obtain a vegetation pest status score of the target mangrove area; Step S3c, comprehensively analyzing the vegetation growth status score of the target mangrove area and the vegetation pest status score of the target mangrove area to obtain a health status score of the target mangrove area.
6. The mangrove ecological management method based on drone according to claim 5 is characterized in that: The step S3a further analyzes the vegetation growth status of the target mangrove area obtained by the first recognition model to obtain a vegetation growth status score of the target mangrove area, including: Step S3a1, extracting key indicators according to the characteristics of the target mangrove area within the boundary; Step S3a2: setting at least two first thresholds for each key indicator, setting at least two levels for each key indicator according to the at least two first thresholds, and setting different scores for each key indicator according to different levels, wherein the scores include an upper score limit and a lower score limit; Step S3a3, comprehensively analyzing the upper and lower limits of the scores of each key indicator at different levels, and establishing a quantitative scoring table for vegetation growth status, wherein the quantitative scoring table for vegetation growth status includes the vegetation growth status level; Step S3a4: Analyze each current key indicator to obtain the indicator value of each key indicator; Step S3a5: Analyze the influence of each current key indicator on the vegetation growth state, rank each key indicator according to the analysis result, and assign a weight to each key indicator according to the ranking result; Step S3a6: Obtain a vegetation growth status score for the target mangrove area according to the indicator value of each key indicator and the weight of each key indicator.
7. The method for mangrove ecological management based on drone according to claim 6, characterized in that: in, The calculation formula for obtaining the vegetation growth status score of the target mangrove area is: ； in, A score representing the vegetation growth status of the target mangrove area; Represents the maximum number of key indicators; Represents the weight of the i-th key indicator; Represents the indicator value of the i-th key indicator.
8. The mangrove ecological management method based on drone according to claim 5 is characterized in that: in, The calculation formula for obtaining the health status score of the target mangrove area is: ； in, A score representing the health status of the target mangrove area; A score representing the vegetation growth status of the target mangrove area; A score representing the vegetation pest status of the target mangrove area; Represents the preset weight coefficient, 0< <1, used to balance the impact of vegetation growth status score and vegetation pest status score on health status score.
9. A mangrove ecological management system based on drones, used to implement a mangrove ecological management method based on drones as claimed in any one of claims 1 to 8, characterized in that: The system comprises: Data acquisition and preprocessing module: used to collect image data of the target mangrove area through drones and preprocess the image data; State recognition module: connected to the data acquisition and preprocessing module, used to input the preprocessed image data into the recognition model to obtain the vegetation growth state and the vegetation pest state of the target mangrove area; Analysis and evaluation module: connected to the state recognition module, used to comprehensively analyze the vegetation growth state and vegetation pest state to obtain the health status score of the target mangrove area; Plan formulation module: connected to the analysis and evaluation module, used to formulate a management plan based on the health status score of the target mangrove area.