Monitoring and evaluation system for identifying relapsed spartina alterniflora seedlings based on multi-spectral aerial photography of unmanned aerial vehicle

Abstract

The invention discloses a monitoring and evaluation system for identifying relapsed spartina alterniflora seedlings based on unmanned aerial vehicle multispectral aerial photography, and relates to the technical field of ecological environment monitoring, the system comprises a multispectral acquisition module, a data preprocessing module, a seedling identification module, a prevention and treatment evaluation module and a result display module; according to the method, through a spartina alterniflora spectrum recognition algorithm and a recurrent seedling recognition confidence coefficient verification algorithm, the unique spectral characteristics of low red light band absorptivity and high near-infrared band reflectivity of spartina alterniflora and the capability of inverting the photosynthetic pigment content by a red edge region spectral derivative are fully utilized; the method comprises the following steps of: extracting complete spectral data of each pixel of an image region by region, calculating an exclusive spectral recognition index of the spartina alterniflora, and performing high-precision distinguishing on target pixels of suspected spartina alterniflora recurrent seedlings in combination with a recurrent seedling recognition confidence verification algorithm; the innovation point effectively solves the identification problem caused by the fact that the relapsed spartina alterniflora seedlings are similar to local vegetation spectral characteristics in the initial growth stage.

Description

Technical Field

[0001] This invention relates to the field of ecological environment monitoring technology, specifically a monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography. Background Technology

[0002] As an invasive alien plant, Spartina alterniflora has spread rapidly in many coastal areas around the world in recent years, posing a serious threat to local ecological balance, fishery resources and coastal security. Its strong adaptability and reproductive capacity make it difficult for traditional manual monitoring methods to meet the monitoring needs of large areas and complex terrains. In particular, monitoring Spartina alterniflora is a challenge in areas that are difficult to reach, such as the intertidal zone.

[0003] However, in the process of identifying recurrent Spartina alterniflora seedlings, existing technologies often struggle to accurately distinguish Spartina alterniflora from native vegetation, especially in the early stages of seedling growth when its spectral characteristics are similar to those of the surrounding vegetation. This leads to a high rate of misidentification and missed identification. Furthermore, existing technologies lack in-depth analysis of the unique spectral characteristics of Spartina alterniflora when processing multispectral data, such as its low absorption rate in the red band, high reflectivity in the near-infrared band, and the ability to invert photosynthetic pigment content from the spectral derivative in the red edge region. This limits further improvement in identification accuracy. At the same time, in terms of evaluating control effectiveness, existing methods rely heavily on qualitative descriptions and lack a scientific evaluation system, making it difficult to accurately reflect the actual effectiveness of control efforts. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography. This invention fully utilizes the unique spectral characteristics of Spartina alterniflora—low absorption in the red band and high reflectivity in the near-infrared band—and the ability to invert photosynthetic pigment content from the spectral derivative in the red-edge region, through a Spartina alterniflora spectral identification algorithm and a recurrent seedling identification confidence verification algorithm. By extracting complete spectral data of each pixel in the image region by region and calculating the Spartina alterniflora-specific spectral identification index, combined with the recurrent seedling identification confidence verification algorithm, high-precision differentiation of suspected Spartina alterniflora recurrent seedling target pixels is achieved. This innovation effectively solves the identification problem caused by the similarity of Spartina alterniflora recurrent seedlings to the spectral characteristics of native vegetation in the early stages of growth, improving the accuracy and reliability of identification and providing technical support for the control of Spartina alterniflora.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on multispectral aerial photography by unmanned aerial vehicles, the system comprising:

[0006] Multispectral acquisition module: Using a multi-rotor UAV as a carrier, equipped with multispectral imaging equipment and auxiliary sensors, it collects multi-band spectral data of the Spartina alterniflora control area according to a preset route, and records the operating environment data simultaneously, thereby forming the original basic data;

[0007] Data preprocessing module: Based on the original basic data, it performs noise filtering, radiometric correction, geometric correction and data verification operations on the spectral image, and outputs standardized spectral image data;

[0008] Seedling identification module: Based on standardized spectral image data, the module uses a Spartina alterniflora spectral identification algorithm and a recurring seedling identification confidence verification algorithm to distinguish and spatially calibrate recurring seedlings;

[0009] Prevention and control assessment module: Based on spatial calibration data, the module uses a comprehensive recurrence assessment algorithm in the control area, combined with the recurrence seedling coverage rate, to classify the control effectiveness level and generate assessment conclusions.

[0010] Results Display Module: Based on the evaluation conclusions, generate a thematic map of relapsed vaccine distribution and a standardized monitoring report.

[0011] Furthermore, the multispectral acquisition module acquires multi-band spectral data in the 500-600nm green band, 600-700nm red band, and 700-1300nm near-infrared band using a multispectral imaging device, and is equipped with a 12-channel spectral sensor for simultaneous multi-band imaging.

[0012] Furthermore, the multispectral acquisition module records operational environment data through auxiliary sensors, including a tide level monitoring sensor, a light intensity sensor, a wind force sensor, a temperature and humidity sensor, a three-axis attitude sensor, a barometric altimeter, and a speed sensor. Specifically, the tide level monitoring sensor collects real-time tide level values ​​for the monitored area; the light intensity sensor collects light intensity data; the wind force sensor collects wind force level data; the temperature and humidity sensor collects air temperature and humidity data; the three-axis attitude sensor collects flight attitude data (pitch, roll, and yaw angles) during UAV aerial photography; the barometric altimeter collects flight altitude data; and the speed sensor collects flight speed data. The preset flight path is a flight altitude of 80-120m, a flight speed ≤5m / s, an image forward overlap rate ≥85%, and a lateral overlap rate ≥80%.

[0013] Furthermore, the seedling identification module, based on standardized spectral image data and considering the characteristics of Spartina alterniflora—low absorption in the red band, high reflectivity in the near-infrared band, and photosynthetic pigment content derived from the spectral derivative in the red-edge region—is used to identify recurrent seedlings. Complete spectral data for each pixel in the standardized spectral image is extracted region by region, and a specific spectral identification index for Spartina alterniflora is calculated using a Spartina alterniflora spectral identification algorithm. ,when The target pixel is marked as a suspected recurrence of Spartina alterniflora seedlings.

[0014] Furthermore, in the seedling identification module, the calculation formula for the Spartina alterniflora spectral identification algorithm is as follows: ,in, A spectral identification index specific to Spartina alterniflora, The first derivative of the spectrum at 723 nm in the red-edge region is obtained from the analysis of the acquired multi-band spectral data. The average reflectance in the near-infrared band is 700-1300nm. The average reflectance in the 600-700nm red light band is... The single-point reflectivity is at 723nm. The reflectivity at a single point in the 703nm band is [value missing].

[0015] Furthermore, in the seedling identification module, the confidence level for identifying recurrent seedlings is calculated using a recurrent seedling identification confidence verification algorithm. The formula for calculating the confidence level of the relapse vaccine identification algorithm is as follows: ,in, Confidence level for identifying relapsed vaccines. A spectral identification index specific to Spartina alterniflora, The average reflectance in the near-infrared band is 700-1300nm. To monitor the average red light reflectance of native vegetation in the region, The average reflectance in the 600-700nm red light band is... This is the baseline value for the red light band reflectance of Spartina alterniflora.

[0016] Furthermore, in the seedling identification module, when At that time, pixels were identified as high-confidence candidate pixels for recurrent seedlings. For these high-confidence candidate pixels, spectral reflectance data of the 723nm and 703nm red-edge bands were extracted from standardized spectral images. Two types of red-edge feature parameters were constructed: one type was the red-edge difference feature parameter obtained by differentiating the reflectance of the 723nm and 703nm bands, and the other type was the red-edge normalized feature parameter obtained by calculating the difference and sum of the reflectance of the 723nm and 703nm bands and then performing ratio processing on the difference and sum. The chlorophyll a and chlorophyll b photosynthetic pigment content of the plants corresponding to the target pixels were inverted using the red-edge difference feature parameter and the red-edge normalized feature parameter. At the same time, the fluctuation range of the spectral data in the 723nm red-edge region was combined to distinguish the vegetation species corresponding to the high-confidence candidate pixels for recurrent seedlings, thus completing the distinction between the recurrent Spartina alterniflora seedlings and native vegetation. Finally, the spatial location of the high-confidence candidate pixels for recurrent seedlings was determined to clarify the specific spatial location and distribution pattern of the recurrent seedlings in the treated area.

[0017] Furthermore, in the prevention and control assessment module, based on spatial calibration data, the total area of ​​the control area is calculated, the actual coverage area of ​​the high-confidence relapse vaccine is statistically analyzed, and the ratio of the total area of ​​the control area to the actual coverage area of ​​the high-confidence relapse vaccine is calculated to obtain the relapse vaccine coverage rate data. Then, the control area relapse comprehensive assessment index is calculated using the control area relapse comprehensive assessment algorithm. The calculation formula for the comprehensive assessment algorithm for recurrence in the treatment area is as follows: ,in, To eliminate the recurrence comprehensive assessment index in the treatment area, The total area of ​​high-confidence relapsed vaccines. To exclude the total area of ​​the autonomous region, Confidence level for identifying relapsed vaccines. The near-infrared reflectance reference value for Spartina alterniflora. The average reflectance of native vegetation in the near-infrared band.

[0018] Furthermore, in the prevention and control assessment module, based on a gradient of relapsed seedling coverage rates of 0%, 3%, 5%, 10%, 15%, 20%, 30%, and 40%, combined with... Classification of treatment effectiveness levels: When At that time, the treatment effect was excellent; when At that time, the treatment effect was good; when At that time, the treatment effect was generally poor; when At that time, the treatment effect was poor; and data including relapse vaccine coverage rate and a comprehensive relapse assessment index for the treated area were generated. The assessment conclusions regarding the treatment effectiveness level and corresponding control recommendations.

[0019] Furthermore, in the results display module, the thematic map of relapsed vaccine distribution is marked with differentiated colors according to the level of treatment effectiveness and the coverage rate of relapsed vaccines. 0% coverage and... The area is blue, with 3%-10% coverage and Yellow, 10%-20% coverage and Orange, >20% coverage and It is red.

[0020] Compared with existing technologies, this monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography has the following advantages:

[0021] I. This invention fully utilizes the unique spectral characteristics of Spartina alterniflora—low absorption in the red band and high reflectivity in the near-infrared band—and the ability to invert photosynthetic pigment content from the spectral derivative in the red-edge region, through a spectral identification algorithm for Spartina alterniflora and a confidence verification algorithm for recurrent seedling identification. By extracting complete spectral data of each pixel in the image region by region and calculating the Spartina alterniflora-specific spectral identification index, combined with the confidence verification algorithm for recurrent seedling identification, high-precision differentiation of suspected Spartina alterniflora recurrent seedlings is achieved. This innovation effectively solves the identification problem caused by the similarity of spectral characteristics between Spartina alterniflora recurrent seedlings and native vegetation in the early stages of growth, improving the accuracy and reliability of identification and providing technical support for the control of Spartina alterniflora.

[0022] Second, this invention constructs a quantitative evaluation system based on spatial calibration data through a comprehensive recurrence assessment algorithm for the treated area. It not only calculates the total area of ​​the treated area and the actual coverage area of ​​high-confidence recurring seedlings, but also obtains recurring seedling coverage data. Combining multi-dimensional information such as the confidence level of recurring seedling identification, the near-infrared reflectance benchmark values ​​of Spartina alterniflora and native vegetation, it calculates a comprehensive recurrence assessment index for the treated area. Based on the recurring seedling coverage gradient, it classifies the treatment effectiveness levels and generates evaluation conclusions. This innovation enables a comprehensive and scientific evaluation of the Spartina alterniflora control effect, providing a strong basis for optimizing and adjusting control strategies, and improving control efficiency and ecological protection effectiveness.

[0023] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0025] Figure 1 A flowchart of a monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography;

[0026] Figure 2 This is a framework diagram of a monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography.

[0027] Figure 3 This is a framework diagram of the seedling identification module in a monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography. Detailed Implementation

[0028] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0029] Example 1:

[0030] In a monitoring scenario for the recurrence of Spartina alterniflora in spring-harvested areas of coastal mudflats, monitoring was conducted during the critical period of seedling germination after spring recurrence. A hexacopoeia drone was used as the carrier, equipped with a 12-channel multispectral imaging device and a complete set of auxiliary sensors. The operation followed a preset flight path: the flight altitude was set at 100m, the flight speed was controlled at 4m / s, and the image forward overlap rate was maintained at 88%, and the lateral overlap rate was maintained at 82%. The multispectral imaging device collected data in the 500-600nm green band, the 600-700nm red band, and the 700-1300nm near-field band. Multi-band spectral data in the infrared band is simultaneously captured by a 12-channel spectral sensor for multi-band imaging. Meanwhile, auxiliary sensors record environmental data throughout the process: a tide level sensor acquires real-time tide levels in the monitored area; a light intensity sensor acquires light intensity during the operation period; a wind sensor acquires wind speed data; a temperature and humidity sensor records air temperature and relative humidity; a three-axis attitude sensor tracks the UAV's pitch, roll, and yaw angles; a barometric altimeter records flight altitude; and a speed sensor acquires flight speed data. This data ultimately forms complete raw baseline data, such as... Figure 1 As shown.

[0031] For the original basic data, noise filtering was first performed to remove invalid spectral data caused by slight drone jitter and environmental interference. Then, radiometric correction was carried out to eliminate the influence of changes in light intensity on the spectral image. Next, geometric correction was performed, combining drone flight attitude data and tidal flat topographic reference map to correct the geometric distortion of the image. Finally, data verification was completed to check the integrity and consistency of spectral data and environmental data in each band, and standardized spectral image data was output to lay the data foundation for subsequent seedling identification.

[0032] Based on standardized spectral image data, and relying on the Spartina alterniflora spectral recognition algorithm and the recurrence seedling identification confidence verification algorithm, recurrence seedlings were distinguished and spatially calibrated. First, taking advantage of the characteristics of Spartina alterniflora, such as low red light absorption, high near-infrared reflectance, and the ability to invert photosynthetic pigment content from the spectral derivative in the red edge region, complete spectral data of each pixel in the image were extracted region by region. Then, the Spartina alterniflora-specific spectral recognition index was calculated using the Spartina alterniflora spectral recognition algorithm. The calculation formula for the Spartina alterniflora spectral recognition algorithm is as follows: ,in, A spectral identification index specific to Spartina alterniflora, The first derivative of the spectrum at 723 nm in the red-edge region is obtained from the analysis of the acquired multi-band spectral data. The average reflectance in the near-infrared band is 700-1300nm. The average reflectance in the 600-700nm red light band is... The single-point reflectivity is at 723nm. The single-point reflectivity is 703nm. The pixels were marked as suspected recurrent Spartina alterniflora seedlings; then, the recurrent seedling identification confidence score was calculated for the suspected recurrent seedlings target pixels using a recurrent seedling identification confidence score verification algorithm. The formula for calculating the confidence level of the relapse vaccine identification algorithm is as follows: ,in, Confidence level for identifying relapsed vaccines. A spectral identification index specific to Spartina alterniflora, The average reflectance in the near-infrared band is 700-1300nm. To monitor the average red light reflectance of native vegetation in the region, The average reflectance in the 600-700nm red light band is... The red light band reflectance of Spartina alterniflora is used as the reference value; The pixels were identified as high-confidence candidate pixels for recurrent seedlings. For these high-confidence candidate pixels, spectral reflectance data in the 723nm and 703nm red-edge bands were extracted. Red-edge difference feature parameters and red-edge normalized feature parameters were constructed to invert the chlorophyll a and chlorophyll b photosynthetic pigment content of the corresponding plants. At the same time, the fluctuation amplitude of spectral data in the 723nm red-edge region was combined to accurately distinguish between recurrent Spartina alterniflora seedlings and native vegetation such as Reed. Finally, the spatial location of the high-confidence candidate pixels for recurrent seedlings was determined, clarifying that the recurrent seedlings in this treatment area are mainly concentrated in the area on both sides of the tidal flats in the eastern part of the mudflats.

[0033] Based on the completed spatial calibration data, the total area of ​​the treatment area was first calculated to be 1500 mu, and then the actual coverage area of ​​the high-confidence relapse seedlings was calculated to be 45 mu, resulting in a relapse seedling coverage rate of 3%. Subsequently, the comprehensive relapse assessment index of the treatment area was calculated using the comprehensive relapse assessment algorithm. The calculation formula for the comprehensive assessment algorithm for recurrence in the treatment area is as follows: ,in, To eliminate the recurrence comprehensive assessment index in the treatment area, The total area of ​​high-confidence relapsed vaccines. To exclude the total area of ​​the autonomous region, Confidence level for identifying relapsed vaccines. The near-infrared reflectance reference value for Spartina alterniflora. The average near-infrared reflectance of native vegetation; combined with a 3% gradient of regenerated seedling coverage and Numerical values ​​are used to classify the treatment effect level: when At that time, the treatment effect was excellent; when At that time, the treatment effect was good; when At that time, the treatment effect was generally poor; when At that time, the treatment effect was poor; after calculation of the area The score was 0.78, indicating a good treatment effect. The final evaluation conclusion included a 3% relapse rate and a score of 0.78. The numerical values, the good treatment effect level, and the control recommendations for this level, namely, "strengthening routine patrols of the areas on both sides of the tidal ditch and carrying out targeted treatment in early summer".

[0034] Based on the assessment results, a thematic map of recurrent seedling distribution was generated for the treated tidal flat area. The thematic map of recurrent seedling distribution uses differentiated color markers according to the treatment effectiveness level and the recurrent seedling coverage rate. Specifically, 0% coverage and... The area is blue, with 3%-10% coverage and Yellow, 10%-20% coverage and Orange, >20% coverage and The area is marked in red to visually represent the specific location, distribution range, and coverage intensity of recurrent seedlings. The recurrent seedling coverage rate in this area is 3% and the RCEI is 0.78, corresponding to the color yellow. The specific location, distribution range, and coverage intensity of recurrent seedlings on both sides of the tidal channels in the central and eastern parts of the area are also marked on the map. A standardized monitoring report is generated simultaneously, which integrates the monitoring operation parameters, data processing procedures, seedling identification results, control assessment conclusions, and management recommendations, providing complete data support for the local Spartina alterniflora control work.

[0035] In summary, for the spring control of Spartina alterniflora in coastal mudflats, a hexacopoeia equipped with a 12-channel multispectral imaging device and a complete set of auxiliary sensors was used to collect multi-band spectral and environmental data along a preset flight path. After data preprocessing, standardized spectral image data was output. Through Spartina alterniflora spectral recognition algorithms and recurrence seedling identification confidence verification algorithms, recurrence seedlings were distinguished and spatially calibrated to identify their core distribution areas. The recurrence seedling coverage rate was then calculated. Through a comprehensive recurrence assessment algorithm in the control area, the control effect was determined and control recommendations were generated. Finally, a thematic map of recurrence seedling distribution and a standardized monitoring report were output, providing data support for the regional Spartina alterniflora control work.

[0036] Example 2:

[0037] In the autumn control and monitoring of Spartina alterniflora in estuarine wetlands, monitoring was conducted at the end of the autumn vegetation growth period to distinguish between newly replanted seedlings and residual plants from previous years. An octocopter drone was used as the carrier, equipped with a 12-channel multispectral imaging device and a complete set of auxiliary sensors. The drone operated along a pre-set flight path: the flight altitude was set at 80m, the flight speed was controlled at 3m / s, the forward overlap rate was increased to 90%, and the lateral overlap rate was maintained at 85%, ensuring the accuracy of image acquisition under the complex terrain of the intertidal zone. The multispectral imaging device acquired 500-600nm green light and 600-700nm red light... Multi-band spectral data in the 700-1300nm near-infrared band, with a 12-channel spectral sensor simultaneously completing multi-band imaging; auxiliary sensors record operational environment data: a tide monitoring sensor acquires real-time tide levels; a light intensity sensor acquires light intensity; a wind sensor acquires wind speed; a temperature and humidity sensor acquires air temperature and relative humidity; a three-axis attitude sensor acquires the UAV's pitch, roll, and yaw angles to obtain UAV flight attitude stability; a barometric altimeter confirms flight altitude; and a speed sensor acquires flight speed data, ultimately forming the original basic data covering the entire review area.

[0038] A preprocessing workflow is initiated for the raw baseline data. First, noise filtering is performed to remove spectral noise caused by intertidal water vapor interference. Next, radiometric correction is implemented to correct spectral deviations caused by changes in autumn sunlight angle. Then, geometric correction is carried out, combining high-precision topographic data of the estuary wetlands to eliminate geometric errors in the images generated by low-altitude UAV operations. Finally, data verification is completed to ensure that there are no missing or abnormal spectral and environmental data in each band, outputting standardized spectral image data that meets the requirements for seedling identification, such as... Figure 2 As shown.

[0039] Based on standardized spectral image data, a Spartina alterniflora spectral recognition algorithm and a recurrence seedling identification confidence verification algorithm were used for recurrence seedling identification and spatial calibration. The first step involved extracting complete spectral data from each pixel of the image. Based on the spectral characteristics of Spartina alterniflora, the Spartina alterniflora spectral recognition algorithm was used to calculate... The calculation formula for the Spartina alterniflora spectral recognition algorithm is as follows: ;Will The first step is to mark the pixels as suspected relapse target pixels; the second step is to use a relapse identification confidence verification algorithm to calculate the confidence level of the suspected relapse target pixels. The formula for calculating the confidence level of the relapse vaccine identification algorithm is as follows: ; Filter out High-confidence candidate pixels for recurrent seedlings were selected; subsequently, reflectance data of the 723nm and 703nm red-edge bands of the high-confidence candidate pixels for recurrent seedlings were extracted, and red-edge difference feature parameters and red-edge normalized feature parameters were constructed to invert photosynthetic pigment content. Combined with the fluctuation amplitude of spectral data in the 723nm red-edge region, the recurrent seedlings of *Spartina alterniflora* were successfully distinguished from native wetland vegetation such as *Suaeda salsa* and *Tamarix chinensis*, completing spatial location calibration. Figure 3 As shown, it was ultimately found that the recurrent seedlings in the re-examination area were mainly distributed in the shallow water area in the south of the wetland, and were scattered in a point-like pattern.

[0040] Based on spatial calibration data, the total area of ​​the re-examination area was calculated to be 2200 mu, and the area with high-confidence relapse seedling coverage was 110 mu, resulting in a relapse seedling coverage rate of 5%. The relapse comprehensive assessment algorithm for the treated area was used to calculate... The calculation formula for the comprehensive assessment algorithm for recurrence in the treatment area is as follows: Combined with a 5% relapse vaccination coverage gradient and Numerical values ​​are used to classify the treatment effect level: when At that time, the treatment effect was excellent; when At that time, the treatment effect was good; when At that time, the treatment effect was generally poor; when At that time, the treatment effect was poor; after calculation of the area The score was 0.62, classifying its treatment effectiveness as moderate; an evaluation conclusion was then generated, including a 5% relapse rate and a score of 0.62. The report included general treatment effectiveness levels and control recommendations such as "conducting comprehensive remedial treatment in the southern shallow water area before the germination period in the following spring, while setting up monitoring sampling points to strengthen dynamic tracking."

[0041] Based on the assessment results, a thematic map of relapse vaccine distribution was generated. This map uses differentiated color markers based on the cure effectiveness level and relapse vaccine coverage rate. Specifically, 0% coverage and... The area is blue, with 3%-10% coverage and Yellow, 10%-20% coverage and Orange, >20% coverage and The red color visually indicates the specific location, distribution range, and coverage intensity of recurrent seedlings. The recurrent seedling coverage rate in this area is 5% and the RCEI is 0.62, corresponding to the orange color. The map clearly shows the point distribution location, coverage range, and intensity of recurrent seedlings in the southern shallow area. At the same time, a standardized monitoring report is generated, which summarizes the entire process information of this autumn review and provides a monitoring basis for optimizing the subsequent Spartina alterniflora control plan in this estuary wetland.

[0042] In summary, for the autumn control and follow-up survey area of ​​Spartina alterniflora in the estuary wetland, an octocopter drone equipped with a matching 12-channel multispectral imaging device and auxiliary sensors was used to complete the collection of multi-band spectral and operational environment raw data according to a preset flight path. After preprocessing to obtain standardized spectral image data, and relying on the Spartina alterniflora spectral recognition algorithm and the recurrence seedling identification confidence verification algorithm, the recurrence seedlings were identified and spatially calibrated to determine their distribution pattern. Then, the recurrence seedling coverage rate was calculated. Through the comprehensive recurrence assessment algorithm of the control area, the control effect was judged and supplementary treatment suggestions were given. Thematic maps and monitoring reports were generated simultaneously, providing accurate data basis for subsequent control plan optimization.

[0043] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on multispectral aerial photography by unmanned aerial vehicles, characterized in that, The system includes: Multispectral acquisition module: Using a multi-rotor UAV as a carrier, equipped with multispectral imaging equipment and auxiliary sensors, it collects multi-band spectral data of the Spartina alterniflora control area according to a preset route, and records the operating environment data simultaneously, thereby forming the original basic data; Data preprocessing module: Based on the original basic data, it performs noise filtering, radiometric correction, geometric correction and data verification operations on the spectral image, and outputs standardized spectral image data; Seedling identification module: Based on standardized spectral image data, the module uses a Spartina alterniflora spectral identification algorithm and a recurring seedling identification confidence verification algorithm to distinguish and spatially calibrate recurring seedlings; Prevention and control assessment module: Based on spatial calibration data, the module uses a comprehensive recurrence assessment algorithm in the control area, combined with the recurrence seedling coverage rate, to classify the control effectiveness level and generate assessment conclusions. Results Display Module: Based on the evaluation conclusions, generate a thematic map of relapsed vaccine distribution and a standardized monitoring report.

2. The monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, The multispectral acquisition module acquires multi-band spectral data in the 500-600nm green band, 600-700nm red band, and 700-1300nm near-infrared band using a multispectral imaging device, and is equipped with a 12-channel spectral sensor for simultaneous multi-band imaging.

3. The monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, The multispectral acquisition module records operational environment data through auxiliary sensors, including a tide level monitoring sensor, a light intensity sensor, a wind force sensor, a temperature and humidity sensor, a three-axis attitude sensor, a barometric altimeter, and a speed sensor. Specifically, the tide level monitoring sensor collects real-time tide levels in the monitored area; the light intensity sensor collects light intensity data; the wind force sensor collects wind speed data; the temperature and humidity sensor collects air temperature and humidity data; the three-axis attitude sensor collects pitch, roll, and yaw angle flight attitude data during UAV aerial photography; the barometric altimeter collects flight altitude data; and the speed sensor collects flight speed data. The preset flight path is a flight altitude of 80-120m, a flight speed ≤5m / s, an image forward overlap rate ≥85%, and a lateral overlap rate ≥80%.

4. The monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, In the seedling identification module, based on standardized spectral image data and considering the characteristics of Spartina alterniflora—low absorption in the red band, high reflectivity in the near-infrared band, and the ability to invert photosynthetic pigment content from the spectral derivative in the red-edge region—re-seedling identification is performed. Complete spectral data for each pixel in the standardized spectral image is extracted region by region, and a specific spectral identification index for Spartina alterniflora is calculated using a Spartina alterniflora spectral identification algorithm. ,when The target pixel is marked as a suspected recurrence of Spartina alterniflora seedlings.

5. A monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography, as described in claim 4, is characterized in that... In the seedling identification module, the calculation formula for the Spartina alterniflora spectral identification algorithm is as follows: ,in, A spectral identification index specific to Spartina alterniflora, The first derivative of the spectrum at 723 nm in the red-edge region is obtained from the analysis of the acquired multi-band spectral data. The average reflectance in the near-infrared band is 700-1300nm. The average reflectance in the 600-700nm red light band is... The single-point reflectivity is at 723nm. The reflectivity at a single point in the 703nm band is [value missing].

6. The monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, In the seedling identification module, the confidence level of recurrent seedling identification is calculated using a recurrent seedling identification confidence verification algorithm. The formula for calculating the confidence level of the relapse vaccine identification algorithm is as follows: ,in, Confidence level for identifying relapsed vaccines. A spectral identification index specific to Spartina alterniflora, The average reflectance in the near-infrared band is 700-1300nm. To monitor the average red light reflectance of native vegetation in the region, The average reflectance in the 600-700nm red light band is... This is the baseline value for the red light band reflectance of Spartina alterniflora.

7. A monitoring and evaluation system for identifying recurrent Spartina alterniflora seedlings based on UAV multispectral aerial photography according to claim 6, characterized in that, In the seedling identification module, when At that time, pixels were identified as high-confidence candidate pixels for recurrent seedlings. For these high-confidence candidate pixels, spectral reflectance data of the 723nm and 703nm red-edge bands were extracted from standardized spectral images. Two types of red-edge feature parameters were constructed: one type was the red-edge difference feature parameter obtained by differentiating the reflectance of the 723nm and 703nm bands, and the other type was the red-edge normalized feature parameter obtained by calculating the difference and sum of the reflectance of the 723nm and 703nm bands and then performing ratio processing on the difference and sum. The chlorophyll a and chlorophyll b photosynthetic pigment content of the plants corresponding to the target pixels were inverted using the red-edge difference feature parameter and the red-edge normalized feature parameter. At the same time, the fluctuation range of the spectral data in the 723nm red-edge region was combined to distinguish the vegetation species corresponding to the high-confidence candidate pixels for recurrent seedlings, thus completing the distinction between the recurrent Spartina alterniflora seedlings and native vegetation. Finally, the spatial location of the high-confidence candidate pixels for recurrent seedlings was determined to clarify the specific spatial location and distribution pattern of the recurrent seedlings in the treated area.

8. The monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, In the prevention and control assessment module, based on spatial calibration data, the total area of ​​the control area is calculated, the actual coverage area of ​​the high-confidence relapsed vaccine is statistically analyzed, and the ratio of the total area of ​​the control area to the actual coverage area of ​​the high-confidence relapsed vaccine is calculated to obtain the relapsed vaccine coverage rate data. Then, the control area relapse comprehensive assessment index is calculated using the control area relapse comprehensive assessment algorithm. The calculation formula for the comprehensive assessment algorithm for recurrence in the treatment area is as follows: ,in, To eliminate the recurrence comprehensive assessment index in the treatment area, The total area of ​​high-confidence relapsed vaccines. To exclude the total area of ​​the autonomous region, Confidence level for identifying relapsed vaccines. The near-infrared reflectance reference value for Spartina alterniflora. The average reflectance of native vegetation in the near-infrared band.

9. A monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 8, characterized in that, In the prevention and control assessment module, based on the recurrence seedling coverage rate gradient of 0%, 3%, 5%, 10%, 15%, 20%, 30%, and 40%, combined with... Classification of treatment effectiveness levels: When At that time, the treatment effect was excellent; when At that time, the treatment effect was good; when At that time, the treatment effect was generally poor; when At that time, the treatment effect was poor; and data including relapse vaccine coverage rate and a comprehensive relapse assessment index for the treated area were generated. The assessment conclusions regarding the treatment effectiveness level and corresponding control recommendations.

10. A monitoring and evaluation system for identifying recurrent seedlings of Spartina alterniflora based on UAV multispectral aerial photography according to claim 1, characterized in that, In the results display module, the thematic map of relapsed vaccine distribution is marked with differentiated colors according to the level of treatment effectiveness and the coverage rate of relapsed vaccines. 0% coverage and... The area is blue, with 3%-10% coverage and Yellow, 10%-20% coverage and Orange, >20% coverage and It is red.

Images