A forest road checkpoint monitoring intelligent site selection method based on unmanned aerial vehicle intelligent identification

Abstract

The application discloses a forest road checkpoint monitoring intelligent site selection method based on unmanned aerial vehicle intelligent identification, and the method comprises the following steps: collecting data based on an unmanned aerial vehicle, acquiring forest environment multi-source data in real time, and performing multi-source data preprocessing; identifying a risk area based on the forest environment multi-source data, and positioning a high-risk area in the forest; balancing a high-risk area coverage range, a cost and dynamic adaptability based on a multi-target optimization site selection model, and generating an optimal checkpoint monitoring layout scheme. Through the unmanned aerial vehicle intelligent identification technology, the application realizes intelligent site selection of the forest road checkpoint monitoring, effectively improves monitoring site selection efficiency and accuracy, and through comprehensive consideration of the multi-target optimization site selection model, ensures that the checkpoint layout comprehensively covers the high-risk area, and also considers economic cost and flexible adjustment capacity, so that the maximization of the forest monitoring efficiency is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent site selection technology for forest road checkpoint monitoring, and in particular to an intelligent site selection method for forest road checkpoint monitoring based on UAV intelligent identification. Background Technology

[0002] The site selection and deployment of traditional forest road checkpoints have long relied on manual experience and static planning, making them ill-suited to complex terrain, dynamic risk changes, and the need for balancing multiple objectives. Existing methods mainly suffer from the following problems:

[0003] 1) Limited Coverage: Fixed checkpoints rely on manual patrols and pre-set routes, making it difficult to cover complex areas such as steep slopes and dense forests, resulting in blind spots in high-risk areas. Manual patrols require a large amount of manpower, and efficiency is often reduced in remote areas due to the inability to reach or see what is desired. Although some systems have introduced fixed cameras, they are limited by terrain and lighting conditions, making it impossible to achieve full-area dynamic monitoring. 2) Insufficient Dynamic Response Capability: Traditional monitoring systems mostly rely on static data from satellite remote sensing or fixed cameras, making it difficult to capture dynamic changes in the forest environment in real time. Although AI image recognition algorithms have been applied in some scenarios, the fixed frame rate acquisition mode of traditional cameras leads to high data redundancy and large response delays, failing to meet the needs of rapid identification and early warning. 3) Difficulty in Balancing: Checkpoint site selection needs to consider multiple dimensions such as maximizing coverage, minimizing deployment costs, and avoiding ecologically sensitive areas, but existing methods lack scientific basis. Traditional site selection often relies on experience-based judgment, which can easily lead to redundant deployments or increased maintenance costs due to ignoring terrain complexity. 4) Low Technological Integration: While drones have been initially applied in forestry patrols and resource surveys, their deep integration with intelligent identification algorithms, multi-objective optimization models, and geographic information systems (GIS) has not yet been fully developed. Existing drones are mostly used for static data collection, lacking support for real-time risk identification and dynamic adjustment; the application of multi-objective optimization algorithms in geospatial modeling is still mainly based on theoretical research, lacking practical scenario verification.

[0004] For example, invention application number 202410655324.2 discloses a method for selecting the location of fixed observation points in a near-mountain forest fire identification system. This solution can maximize the advantages of fixed observation points in fire monitoring and identification technology, thereby improving the early warning capability of forest fires and reducing the risk of forest fires. However, this solution also has limitations: insufficient site selection flexibility. This method mainly relies on pre-set fixed observation points, making it difficult to achieve comprehensive coverage and flexible adjustments in complex and changing forest environments. Especially in fire-prone areas or areas with complex terrain, the limitations of fixed observation points may lead to monitoring blind spots, affecting the timely detection and handling of fires.

[0005] To address the aforementioned issues, there is an urgent need for an intelligent site selection algorithm for monitoring forest road checkpoints based on drone-based intelligent identification. This algorithm should integrate multispectral / thermal imaging sensors, AI models, and multi-objective optimization algorithms to achieve precise deployment of dynamic data. This solution needs to overcome the spatiotemporal limitations of traditional monitoring while balancing coverage efficiency, cost control, and ecological protection requirements, providing efficient and low-cost technical support for forest fire prevention, illegal logging control, and ecological resource management. Summary of the Invention

[0006] To address the aforementioned problems, the present invention aims to provide an intelligent site selection method for monitoring checkpoints on forest roads based on UAV intelligent recognition. This method combines multispectral, thermal imaging sensing, AI recognition models, and multi-objective optimization algorithms to achieve intelligent monitoring and site selection for checkpoints on forest roads.

[0007] The first aspect: A method for intelligent site selection for monitoring checkpoints on forest roads based on unmanned aerial vehicle (UAV) intelligent recognition, including:

[0008] S1. Data collection is carried out based on drones to acquire multi-source data of the forest environment in real time, and multi-source data preprocessing is performed.

[0009] S2. Identify risk areas and locate high-risk areas in forest areas based on multi-source environmental data.

[0010] S3. Based on a multi-objective optimization site selection model, the optimal checkpoint monitoring layout scheme is generated by balancing the coverage, cost, and dynamic adaptability of high-risk areas.

[0011] Optionally, when the UAV performs data collection in S1:

[0012] Equipped with a multispectral camera, it acquires multispectral images and calculates vegetation cover characteristics through the NDVI index to help identify high-risk areas;

[0013] Equipped with a thermal imaging sensor, it acquires thermal infrared data to help detect fire hotspots and traces of illegal logging;

[0014] Equipped with lidar, it acquires terrain feature data.

[0015] Optionally, the multi-source data preprocessing includes:

[0016] Multispectral images, thermal infrared data and terrain feature data are spatiotemporally aligned to eliminate the effects of illumination changes and atmospheric interference.

[0017] By combining topographic feature data with vegetation cover features through GIS spatial analysis, a vegetation distribution map is generated.

[0018] Optionally, the NDVI index calculation formula is expressed as:

[0019]

[0020] Wherein, NIR represents the reflectance in the near-infrared band, and Red represents the reflectance in the red band.

[0021] Optionally, the risk area identification in S2 includes:

[0022] S21, Improved YOLOv8 model configuration for UAVs;

[0023] S22. While performing flight missions, the drone uses an improved YOLOv8 model to identify fire hotspots, signs of illegal logging, and risks of illegal activities.

[0024] S23. Periodically update the risk area distribution map and dynamically adjust the risk level score.

[0025] According to claim 2, the intelligent site selection method for monitoring checkpoints in forest roads based on UAV intelligent identification is characterized in that the risk level scoring formula is:

[0026]

[0027] F i The fire risk value is D. i E represents the risk value for illegal logging. i The value represents the illegal activity, and α, β, and γ are weighting coefficients.

[0028] Optionally, the objective function of the multi-objective optimization location model is expressed as:

[0029]

[0030]

[0031] Where C is the coverage weight value, R i Let w be the risk level of the i-th grid. i Here, C represents the weighting coefficient, T represents the total cost, and C represents the weighting coefficient. j The deployment cost for the j-th checkpoint.

[0032] Optionally, the C j The deployment cost of the j-th checkpoint is expressed by the formula:

[0033]

[0034] Among them, C device For equipment costs, C install For installation costs, C maintain To reduce maintenance costs, L j Let f(L) be the distance from the j-th checkpoint to the nearest maintenance station. j () is used to maintain the frequency function.

[0035] Second aspect: An electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, performs the steps of the method provided in the first aspect.

[0036] Third aspect: A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the first aspect.

[0037] The beneficial effects of this invention are:

[0038] 1. This invention uses a drone equipped with a multispectral camera, thermal imaging sensor and lidar to cover areas that are difficult for manual patrols to reach, such as steep slopes and dense forests. It combines NDVI index analysis of vegetation characteristics and thermal infrared data to capture fire hotspots, eliminating monitoring blind spots and realizing real-time perception of risks throughout the forest area, enabling rapid and accurate intelligent site selection for monitoring checkpoints on forest roads.

[0039] 2. This invention uses periodic UAV flight data collection, combined with GIS spatial analysis to generate a dynamic risk distribution map. Risk is quantified using a risk level scoring formula, supporting dynamic adjustments to checkpoint layout. Response latency is reduced from hours to minutes compared to traditional methods. The multi-objective optimization model can recalculate the optimal layout based on real-time risk changes, ensuring checkpoints always cover high-risk areas. This significantly improves early warning efficiency and enhances dynamic response capabilities compared to traditional fixed layouts.

[0040] 3. The multi-objective optimization model of this invention avoids redundant deployment by maximizing the coverage weight value of high-risk areas and minimizing the total cost. It optimizes maintenance distance and frequency, reducing checkpoint costs. Site selection incorporates rules for avoiding terrain complexity and ecologically sensitive areas, minimizing disturbance to the natural ecosystem while controlling risks, thus meeting the needs of sustainable forestry development.

[0041] 4. This invention deeply integrates UAV data collection, AI recognition (improved YOLOv8), GIS spatial analysis and multi-objective optimization algorithms to achieve full-process automation of data collection, risk identification and site selection optimization, thereby improving site selection efficiency. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating the intelligent site selection method for monitoring checkpoints on forest roads according to the present invention.

[0043] Figure 2 This is a schematic diagram illustrating the principle of the intelligent site selection method for monitoring checkpoints on forest roads according to the present invention.

[0044] Figure 3 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0045] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0046] The existing forest road checkpoint monitoring sites have the following problems: limited coverage, blind spots in high-risk areas, and inability to monitor the entire area dynamically; insufficient dynamic response, making it difficult to capture changes in the forest area in real time and failing to meet the needs for rapid identification and early warning; difficulty in balancing multiple objectives such as coverage, deployment cost, and avoidance of ecologically sensitive areas; and low degree of technological integration, lacking support for real-time risk identification and dynamic adjustment.

[0047] To address the above problems, this invention provides a method for intelligent site selection for monitoring checkpoints on forest roads based on unmanned aerial vehicle (UAV) intelligent recognition. Figure 1 This is a flowchart illustrating the intelligent site selection method for monitoring checkpoints on forest roads according to an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the principle of the intelligent site selection method for monitoring checkpoints on forest roads according to an embodiment of the present invention. The method includes:

[0048] S1. Data collection is conducted using drones to obtain real-time forest area environmental data.

[0049] Deploy automated drone airfields in key areas of the forest region, plan drone routes in advance to cover the entire area, and collect high-resolution images, vegetation spectral data, thermal signals, and environmental parameters in real time. Specifically:

[0050] The drone is equipped with multispectral cameras, thermal imaging sensors, and lidar, and the drone system is equipped with an improved YOLOv8 model.

[0051] The drone uses a multispectral camera to acquire multispectral images. For example, the multispectral image is a 5-band image, including (blue light 450nm, green light 560nm, red light 660nm, near-infrared 850nm, and thermal infrared 1050nm), with a resolution of 0.5 meters. The NDVI index is calculated using the following formula:

[0052]

[0053] NIR represents the reflectance value in the near-infrared band, and Red represents the reflectance value in the red band. The Normalized Difference Vegetation Index (NDVI) reflects the growth status and health of vegetation, making it an important indicator for vegetation cover monitoring. By calculating the NDVI, we can conduct a more precise analysis of vegetation distribution within forest areas and, combined with topographic data, further identify potential monitoring checkpoint locations. These locations are typically situated in areas with dense vegetation and road crossings, allowing us to capture traffic within the forest while ensuring the safety and concealment of monitoring equipment.

[0054] Drones, powered by thermal imaging sensors, can capture real-time thermal radiation information within forest areas. These sensors generate thermal images by detecting infrared radiation emitted from the ground and vegetation, identifying potential fire hazards, such as localized high-temperature areas caused by dryness or human activity. Furthermore, thermal imaging can aid in analyzing animal activity patterns within the forest, as different animals exhibit varying body temperature distributions at different times. Combined with multispectral image data, drones can more comprehensively assess the ecological condition of forest areas, providing more accurate information support for intelligent site selection. By comprehensively analyzing thermal imaging and multispectral data, the location of monitoring checkpoints can be further optimized, ensuring effective monitoring of anomalies within the forest while minimizing disturbance to the natural ecosystem.

[0055] For example, the thermal imaging sensor used has a resolution of 0.1℃ and can identify fire hotspots (temperature ≥600K) and traces of illegal logging (thermal signal of tree trunk cross section ≥450K).

[0056] The drone is also equipped with a lidar system to obtain terrain features, such as elevation (resolution 0.1 meters, altitude range 500-1200 meters), slope (0°-45°), and aspect (southeast slope with light intensity ≥800W / m²).

[0057] Preprocessing of multi-source data acquired by multispectral cameras, thermal imaging sensors, and lidar includes:

[0058] Multispectral images, thermal infrared data, and terrain feature data are spatiotemporally aligned to eliminate the effects of illumination variations and atmospheric interference, thereby improving data consistency and accuracy. Advanced image processing algorithms enable precise registration of multi-source data, ensuring seamless integration of various information across spatial and temporal dimensions. This process not only enhances data reliability but also lays a solid foundation for subsequent intelligent analysis.

[0059] By combining topographic feature data with vegetation cover features through spatial analysis using Geographic Information System (GIS), a vegetation distribution map is generated, enabling visualized management of forest area ecological information and providing intuitive and comprehensive basis for decision support.

[0060] S2. Identify risk areas and locate high-risk areas in forest areas based on forest environmental data;

[0061] The YOLOv8 model has been lightweighted, reducing the number of parameters to 30% of the original model (e.g., from 12 million parameters to 3.6 million parameters), and increasing inference speed to 15 frames per second (FPS). This makes it more suitable for deployment on UAV systems, reducing the demand on UAV hardware resources and improving the overall energy efficiency of the system. The lightweight YOLOv8 model maintains high detection accuracy while significantly reducing the computational burden, enabling UAVs to respond more efficiently and flexibly to various complex scenarios during mission execution.

[0062] Based on an improved YOLOv8 model, and integrating multi-source information including satellite remote sensing data, UAV aerial photography data, and ground monitoring data, this method utilizes image processing and machine learning techniques from the improved YOLOv8 model. It then combines this with information on fire hotspots and illegal logging traces identified by thermal imaging sensors, along with NDVI index data, to identify risk information such as fire sources (flame area ≥ 0.5 m²), illegal logging traces (tree trunk diameter ≥ 30 cm), and illegal activities (people carrying fire sources). This enables intelligent identification and early warning of potential risks in forest areas. This method not only improves the accuracy of risk identification but also enhances the comprehensiveness and real-time performance of the monitoring system by integrating multi-source information.

[0063] Simultaneously, the risk level is determined by combining the fire source risk value, the illegal logging risk value, and the illegal activity value. The risk level scoring formula is as follows:

[0064]

[0065] Among them, F i The fire risk value is D. i E represents the risk value for illegal logging. i The value represents the illegal activity, and α, β, and γ are weighting coefficients.

[0066] For example:

[0067] F i Fire risk value (0-10 points, e.g., fire source A scores 8 points);

[0068] D i : Risk value of illegal logging (0-10 points, such as the score of 7 points for illegal logging area B);

[0069] E i: Illegal activity value (0-10 points, e.g., illegal activity C score is 9 points).

[0070] Weighting coefficients: α=0.4 (highest weight for fire source risk), β=0.3 (second highest weight for illegal logging risk), γ=0.3 (value for illegal activities).

[0071] Substituting into the formula, the risk level is calculated to be 8.0: R i =0.4*8+0.3*7+0.3*9=3.2+2.1+2.7=8.0

[0072] S3. Based on a multi-objective optimization site selection model, the optimal checkpoint monitoring layout scheme is generated by balancing the coverage, cost, and dynamic adaptability of high-risk areas.

[0073] The objective function of the multi-objective optimization location model is expressed as:

[0074]

[0075]

[0076] Where C is the coverage weight value, R i For the risk level (0-10 points) of the i-th grid, w i The weighting coefficient of this grid (dynamically adjusted based on distance from the main road, terrain complexity, etc.).

[0077] The weight coefficient of wi is defined as:

[0078]

[0079] Where, d i t represents the distance (in meters) from the i-th grid to the main road; the closer the distance, the higher the weight (α is the normalization coefficient). i The terrain complexity of the i-th grid is (0-1), with more complex terrain having a higher weight (β is the normalization coefficient).

[0080] T represents the total cost, and C represents the total cost. j The deployment cost for the j-th checkpoint (including equipment procurement, installation, and maintenance costs) is as follows:

[0081] C j The deployment cost of the j-th checkpoint is expressed by the formula:

[0082]

[0083] Among them, C device For equipment costs, C install For installation costs, C maintain To reduce maintenance costs, L jLet f(L) be the distance from the j-th checkpoint to the nearest maintenance station. j () is used to maintain the frequency function.

[0084] f(Lj) is the maintenance frequency function, defined as f(Lj) = 1 + 0.1 * Lj. The greater the distance, the higher the maintenance frequency. The distances considered include: checkpoint spacing ≤ 500 meters, checkpoint and ecologically sensitive area buffer zone ≥ 100 meters, or checkpoint and nearest road ≤ 200 meters.

[0085] A multi-objective optimization site selection model is employed, with the core objectives of maximizing coverage, minimizing deployment costs, and avoiding ecologically sensitive areas. Inputting GIS topographic data, risk distribution maps, and transportation networks, the model generates an optimal solution set. The analytic hierarchy process (AHP) is incorporated into the model to analyze dimensions such as fire risk, illegal logging risk, and ecological sensitivity, ensuring the scientific rigor of site selection decisions. Based on the optimization results, checkpoint locations are dynamically planned, and deployment strategies are adjusted in conjunction with real-time monitoring data. Unmanned aerial vehicle (UAV) reconnaissance enables rapid response and adaptive adjustment of checkpoint locations. Checkpoint locations are dynamically adjusted to cover high-risk areas, ensuring the accuracy of checkpoint deployment and avoiding deviations caused by terrain changes.

[0086] Application examples of multi-objective optimization location selection models:

[0087]

[0088]

[0089] Coverage weight value C:

[0090] R i =8.0 (Risk level of grid A);

[0091] Where: w i =α*d i +β*t i

[0092] d i =150 meters (distance from grid A to the main road), t i =0.8 (terrain complexity, steep slopes have a higher weight), α=0.6, β=0.4 (normalization coefficients).

[0093] wi=0.6*150+0.4*0.8=90+0.32=90.32

[0094] C = 90.32 * 8.0 = 722.56

[0095] Total cost T:

[0096] C j =C device +C install *Lj+Cmaintain *f(Lj)

[0097] C device =50,000 yuan (cost of checkpoint equipment), C install ==10,000 yuan / km, C maintain =8000 yuan (average annual maintenance cost), Lj = 2 kilometers (distance from the checkpoint to the nearest maintenance station);

[0098] f(Lj) = 1 + 0.1 * 2 = 1.2 (Maintaining frequency function)

[0099] C j =50000 + 10000 * 2 + 8000 * 1.2 = 50000 + 20000 + 9600 = 79600 yuan

[0100] The above examples demonstrate that the weighting factor w i The calculation method has demonstrated its flexibility and practicality in application. It not only considers the geographical distance but also fully takes into account the impact of terrain on the difficulty and cost of site selection, making site selection decisions more scientific and comprehensive. The application of this formula also helps improve the accuracy and scientific rigor of site selection, reduces interference from human factors, and ensures that the construction of forest area road checkpoint monitoring systems achieves optimal results.

[0101] The present invention also provides an electronic device, Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 3 As shown, the electronic device may include a processor, a communications interface, memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can invoke logical instructions from the memory, for example, to execute the following method:

[0102] S1. Data collection is carried out based on drones to acquire multi-source data of the forest environment in real time, and multi-source data preprocessing is performed.

[0103] S2. Identify risk areas and locate high-risk areas in forest areas based on multi-source environmental data.

[0104] S3. Based on a multi-objective optimization site selection model, the optimal checkpoint monitoring layout scheme is generated by balancing the coverage, cost, and dynamic adaptability of high-risk areas.

[0105] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0106] This invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the methods provided in the above embodiments, including, for example:

[0107] S1. Data collection is carried out based on drones to acquire multi-source data of the forest environment in real time, and multi-source data preprocessing is performed.

[0108] S2. Identify risk areas and locate high-risk areas in forest areas based on multi-source environmental data.

[0109] S3. Based on a multi-objective optimization site selection model, the optimal checkpoint monitoring layout scheme is generated by balancing the coverage, cost, and dynamic adaptability of high-risk areas.

[0110] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0111] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent site selection for monitoring checkpoints on forest roads based on unmanned aerial vehicle (UAV) intelligent recognition, characterized in that, include: S1. Data collection is carried out based on drones to acquire multi-source data of the forest environment in real time, and multi-source data preprocessing is performed. S2. Identify risk areas and locate high-risk areas in forest areas based on multi-source environmental data. S3. Based on a multi-objective optimization location model, balance the coverage, cost, and dynamic adaptability of high-risk areas to generate the optimal checkpoint monitoring layout scheme; When the UAV collects data in S1: Equipped with a multispectral camera, it acquires multispectral images and calculates vegetation cover characteristics through the NDVI index to help identify high-risk areas; Equipped with a thermal imaging sensor, it acquires thermal infrared data to help detect fire hotspots and traces of illegal logging; Equipped with lidar to acquire terrain feature data; The risk area identification in S2 includes: S21, Improved YOLOv8 model configuration for UAVs; S22. While performing flight missions, the drone uses an improved YOLOv8 model to identify fire hotspots, signs of illegal logging, and risks of illegal activities. S23. Periodically update the risk area distribution map and dynamically adjust the risk level score; The objective function of the multi-objective optimization location model is expressed as follows: Where C is the coverage weight value, R i Let w be the risk level of the i-th grid. i Here, C represents the weighting coefficient, T represents the total cost, and C represents the weighting coefficient. j The deployment cost for the j-th checkpoint; The C j The deployment cost of the j-th checkpoint is expressed by the formula: Among them, C device For equipment costs, C install For installation costs, C maintain To reduce maintenance costs, L j Let f(L) be the distance from the j-th checkpoint to the nearest maintenance station. j () is for maintaining the frequency function; f(L j The frequency maintenance function is defined as f(L). j ) = 1 + 0.1 * L j The greater the distance, the higher the maintenance frequency.

2. The intelligent site selection method for monitoring forest road checkpoints based on UAV intelligent identification as described in claim 1, characterized in that, The multi-source data preprocessing includes: Multispectral images, thermal infrared data and terrain feature data are spatiotemporally aligned to eliminate the effects of illumination changes and atmospheric interference. By combining topographic feature data with vegetation cover features through GIS spatial analysis, a vegetation distribution map is generated.

3. The intelligent site selection method for forest road checkpoint monitoring based on UAV intelligent identification according to claim 1, characterized in that, The formula for calculating the NDVI index is as follows: Wherein, NIR represents the reflectance in the near-infrared band, and Red represents the reflectance in the red band.

4. The intelligent site selection method for monitoring forest road checkpoints based on UAV intelligent identification as described in claim 1, characterized in that, The risk level scoring formula is as follows: F i The fire risk value is D. i E represents the risk value for illegal logging. i The value represents the illegal activity, and α, β, and γ are weighting coefficients.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the intelligent site selection method for monitoring checkpoints in forest roads based on unmanned aerial vehicle (UAV) intelligent identification as described in any one of claims 1 to 4.

6. A non-transitory 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 of the intelligent site selection method for monitoring checkpoints in forest roads based on unmanned aerial vehicle (UAV) intelligent identification as described in any one of claims 1 to 4.

Images