Land ecological bearing capacity assessment method based on image analysis

Abstract

The invention relates to the technical field of ecological assessment, in particular to a land ecological bearing capacity assessment method based on image analysis, which comprises the following steps of: acquiring multi-source remote sensing image data including optical remote sensing images, synthetic aperture radar images and unmanned aerial vehicle images, preprocessing the acquired data, synchronizing the multi-source data, and acquiring the multi-source remote sensing image data; generating comprehensive remote sensing image data; and based on the comprehensive remote sensing image data, extracting earth surface vegetation, soil moisture and landform information to obtain ecological element information. The temporary ecological bearing capacity is calculated by fusing ecological element information and real-time meteorological data, dynamic and accurate evaluation is provided for land health conditions, and response of an ecological system to environmental change can be reflected in real time. And through a detailed task allocation table, the transparency and the executive force of cross-department collaboration are enhanced. In addition, a random forest model is utilized to analyze ecological performance, so that performance evaluation not only depends on a single algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ecological assessment, and particularly relates to a land ecological carrying capacity assessment method based on image analysis. BACKGROUND

[0002] Ecological assessment is a key technical field in environmental science and ecology, focusing on assessing and analyzing the health of ecosystems, biodiversity, and their response to human activities. Land ecological carrying capacity assessment is an important topic in the field of ecological assessment, aiming to determine the maximum extent of ecological activities that a piece of land can support without suffering irreversible damage.

[0003] However, existing technologies fail to effectively integrate ecological assessment with subsequent ecological management and performance monitoring, lacking a continuous management framework to support the entire process from assessment to implementation. Secondly, the efficiency of department coordination and task allocation is low, as there is a lack of a clear, data-based task execution monitoring, making it difficult to track the implementation status of various measures, which may lead to resource waste and delay of ecological restoration work. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art, and a land ecological carrying capacity assessment method based on image analysis is proposed.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solution, a land ecological carrying capacity assessment method based on image analysis, comprising the following steps:

[0006] Collecting multi-source remote sensing image data, including optical remote sensing images, synthetic aperture radar images and unmanned aerial vehicle images, preprocessing the collected data, synchronizing multi-source data, and generating comprehensive remote sensing image data; based on the comprehensive remote sensing image data, extracting surface vegetation, soil moisture and topographic information to obtain ecological element information;

[0007] Using the ecological element information, calculating ecological indicators, and simultaneously fusing real-time weather data to obtain temporary ecological carrying capacity; performing spatial analysis on the temporary ecological carrying capacity to determine the land health status of the region, and generating regional ecological health results;

[0008] According to the regional ecological health results, setting ecological restoration targets, decomposing the ecological restoration targets into department and project management tasks, and establishing an ecological target task allocation table; based on the ecological target task allocation table, monitoring the task execution progress to obtain task execution monitoring results;

[0009] Using the task execution monitoring results, assessing the ecological performance of each department and project, identifying performance problems, and obtaining performance analysis results.

[0010] Preferably, the step of obtaining comprehensive remote sensing image data is:

[0011] Collecting image data from three sources of optical remote sensing, synthetic aperture radar and unmanned aerial vehicle, recording the type and timestamp of image data from each source, obtaining a preliminary image data set;

[0012] Based on the preliminary image data set, noise removal and geometric correction are performed on each type of image data, radiation correction is performed on optical remote sensing image, despeckling is performed on synthetic aperture radar image, and image resolution and format are unified, obtaining a pre-processed image data set;

[0013] Based on the pre-processed image data set, time synchronization is used to correct data time deviation, spatial alignment is applied, and comprehensive remote sensing image data is generated.

[0014] Preferably, the step of obtaining ecological element information is:

[0015] Analyzing the comprehensive remote sensing image data, using spectral recognition and classification of ground vegetation, distinguishing vegetation types by analyzing reflection and absorption characteristics of different wavelengths, judging growth conditions, obtaining ground vegetation information;

[0016] Based on the ground vegetation information, combining microwave radar data and vegetation cover data, calculating the soil moisture level of each region, obtaining soil moisture information;

[0017] Based on the soil moisture information, calculating slope, slope direction and surface roughness, obtaining ecological element information.

[0018] Preferably, the step of obtaining temporary ecological carrying capacity is:

[0019] Based on the ecological element information, determining the net primary productivity of various plants, obtaining preliminary NPP data;

[0020] Based on the preliminary NPP data, introducing real-time meteorological data, calculating the temporary ecological carrying capacity, the formula is:

[0021]

[0022] Where, NPP initial is the preliminary NPP data value, P is the real-time precipitation, P0 is the average precipitation, k and q are adjustment coefficients, T is the real-time air temperature, T0 is the average air temperature, A is the adjustment coefficient of air temperature influence, M is the soil moisture, M max is the maximum soil moisture, and EC is the temporary ecological carrying capacity.

[0023] Preferably, the step of obtaining regional ecological health results is:

[0024] Based on the temporary ecological carrying capacity, spatial analysis is performed to map the carrying capacity of each region and form a spatial distribution map;

[0025] Based on the spatial distribution map, the land health index of the region is calculated, and the calculation formula is:

[0026]

[0027] Wherein, H is the land health index, Z is the total number of analysis regions, EC i is the ecological carrying capacity of the i-th region, EC opt is the ideal ecological carrying capacity, σ i is the coefficient of variation of the ecological carrying capacity of the i-th region;

[0028] Based on the land health index, the land health status is evaluated, and the regional ecological health results are generated according to the environmental sensitivity and adaptability of the region.

[0029] Preferably, the obtaining step of the ecological target task allocation table is:

[0030] Based on the regional ecological health results, vegetation coverage problems and water resource consumption problems are identified, and an ecological restoration target list is formed;

[0031] Based on the ecological restoration target list, ecological restoration measures are formulated, including afforestation and improvement of water conservancy, to obtain an ecological restoration measure plan;

[0032] Based on the ecological restoration measure plan, responsibility allocation and time schedule are determined, and an ecological target task allocation table is established.

[0033] Preferably, the obtaining step of the task execution monitoring result is:

[0034] Using the ecological target task allocation table as the basis, task monitoring is deployed to track the task completion of each department and team in real time, and task execution monitoring data is obtained;

[0035] Based on the task execution monitoring data, the progress of each task and the compliance of the predetermined target are evaluated, and delays and deviations are identified, to obtain a task progress analysis result;

[0036] Based on the task progress analysis result, the execution of all tasks is integrated and evaluated, adjustment measures or task priority are formulated, and a task execution monitoring result is obtained.

[0037] Preferably, the obtaining step of the performance analysis result is:

[0038] Based on the task execution monitoring result, the time point and feedback of task completion are recorded, and the original data set of ecological performance is obtained;

[0039] Based on the original data set of ecological performance, a random forest model is applied and an ecological performance index is calculated, with the formula adjusted as follows:

[0040]

[0041] Wherein, EPI is the ecological performance index, n is the number of data points, f i (X) is the predicted value of the i-th data point calculated by the random forest model, W i is the weight associated with the data point i;

[0042] Based on the ecological performance index, the performance level of each department and project is evaluated, and task delays or quality inconsistencies are identified, to obtain performance analysis results.

[0043] Compared with the prior art, the advantages and positive effects of the present application are as follows:

[0044] The present application calculates the temporary ecological carrying capacity by fusing ecological element information and real-time meteorological data, providing dynamic and accurate assessment of land health status, and can reflect the response of the ecological system to environmental changes in real time. And through the detailed task allocation table, the transparency and execution of cross-department collaboration are enhanced. In addition, the ecological performance is analyzed by using the random forest model, so that the performance evaluation not only depends on a single algorithm, but also enhances the accuracy of evaluation through a multi-element model. This method optimizes resource allocation, improves the efficiency of project management and the adaptability of ecological restoration measures, and promotes the sustainability of ecological system management. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 The present application is a schematic diagram of the steps. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0047] Please refer to Figure 1 The present application provides a technical scheme, a land ecological carrying capacity evaluation method based on image analysis, comprising the following steps:

[0048] Collecting multi-source remote sensing image data, including optical remote sensing image, synthetic aperture radar image and unmanned aerial vehicle image, pre-processing the collected data, synchronizing multi-source data, and generating comprehensive remote sensing image data; based on the comprehensive remote sensing image data, extracting ground vegetation, soil moisture and topographic information to obtain ecological element information;

[0049] The ecological index is calculated by using the ecological element information, and the temporary ecological carrying capacity is obtained by fusing real-time meteorological data; the temporary ecological carrying capacity is subjected to spatial analysis, the land health condition of the region is judged, and the regional ecological health result is generated;

[0050] According to the regional ecological health result, the ecological restoration target is set, the ecological restoration target is decomposed into department and project management tasks, and an ecological target task allocation table is established; based on the ecological target task allocation table, the task execution progress is monitored, and a task execution monitoring result is obtained;

[0051] The ecological performance of each department and project is evaluated using the task execution monitoring result, the performance problem is identified, and a performance analysis result is obtained.

[0052] The acquisition steps of the comprehensive remote sensing image data are as follows:

[0053] Image data is collected from three sources of optical remote sensing, synthetic aperture radar and unmanned aerial vehicle, the type and timestamp of image data of each source are recorded, and a preliminary image data set is obtained;

[0054] Based on the preliminary image data set, noise removal and geometric correction are performed on each type of image data, radiation correction is performed on optical remote sensing image, and despeckling processing is performed on synthetic aperture radar image, and the image resolution and format are unified, to obtain a pre-processed image data set;

[0055] Based on the pre-processed image data set, time synchronization is used to correct the data time deviation, and spatial alignment is applied to generate comprehensive remote sensing image data.

[0056] Specifically, based on the type and time identification content listed in the optical remote sensing data, synthetic aperture radar data and unmanned aerial vehicle image data, first of all, the resolution, band number, shooting time and geographic reference coordinates of the optical remote sensing data need to be recorded, the pulse repetition frequency, viewing angle, polarization mode and observation time of the synthetic aperture radar data are recorded, and the flight height, lens parameters, shooting range and shooting time of the unmanned aerial vehicle image data are recorded, then the above records are compared with the pre-prepared sampling conditions, for example, the flight height corresponds to the interval of 50 meters to 200 meters, the pulse repetition frequency corresponds to the interval of 0 hertz to 10 kilohertz, the lens focal length corresponds to the interval of 10 millimeters to 100 millimeters, and the shooting time is limited to 24 hours within a day and is configured according to the shooting plan of the day, the determination method of these numerical ranges can be summarized and counted according to the historical experience data collected in the same region in the past, thereby forming a preliminary scheme of multiple interval upper and lower limits, if any single value of the current record breaks the established interval, the collection method needs to be adjusted or it is determined that the data is temporarily excluded from subsequent processing, all optical remote sensing, synthetic aperture radar and unmanned aerial vehicle image records are verified and summarized one by one through the above method, and finally the preliminary image data set is obtained.

[0057] Based on the preliminary image data set obtained in the foregoing, first, the brightness deviation existing in the optical remote sensing image is processed. The brightness value is compared with the pre-defined possible range, for example, the brightness value is limited within the interval of 0 to 255, if a pixel exceeds the range, it is marked as an abnormal pixel and is corrected by using an interpolation method, at the same time, the shooting angle and time information are referred to complete the geometric correction. In the geometric correction link, the shooting coordinate system and the target geographic coordinate system need to be compared, and the deviation is corrected in sections, for example, first, the longitudinal coordinate is compared with the interval of 0 to 1000 meters, and then the transverse coordinate is compared with the interval of 0 to 1000 meters, if the coordinate drift occurs, the additional measurement data of the geographic reference point can be combined to further locate and compensate, then the synthetic aperture radar image is processed to remove the speckle, the echo intensity is compared with the normalized range of 0 to 1, if there are continuous noise points higher than 1 in a region, it is judged as a speckle and is denoised by using a filtering method, and the color space of the unmanned aerial vehicle image data is uniformly converted. After the above operations are completed, the resolution of the three types of images is adjusted again, so that the row and column number is within the pre-defined size range, for example, the height direction is not more than 3000 pixels, and the width direction is not more than 4000 pixels, finally, all the corrected images are uniformly formatted and the necessary metadata is recorded, and finally the pre-processed image data set is obtained.

[0058] Based on the pre-processed image data set obtained in the foregoing, first, the time stamps of the optical remote sensing, synthetic aperture radar and unmanned aerial vehicle images are compared synchronously, if it is found that the shooting time recorded by some of them exceeds the pre-set allowed deviation range, for example, the alignment of the minute level is allowed to be ± 1 minute, and the alignment of the second level is allowed to be ± 5 seconds, then the reference time needs to be re-compared and the missing frames need to be filled in if necessary, after the time synchronization is completed, the spatial alignment is performed, the offset is checked by comparing the four corner points of each image in the geographic coordinate system, for example, the longitude is compared with the interval of -180° to 180°, and the latitude is compared with the interval of -90° to 90°, if there is a situation of exceeding these ranges, the grid comparison can be performed by using the reference map, and any coordinate offset is gradually corrected, when the multi-source data meets the alignment conditions in the time and space dimensions, the three types of images are spliced according to the unified projection mode, and the pixel-level fusion is performed on the overlapping area, and the flight height and observation viewing angle recorded are used to further calibrate the fusion details, after all the steps are completed, the comprehensive remote sensing image data is finally generated.

[0059] The steps for obtaining ecological element information are:

[0060] The comprehensive remote sensing image data is analyzed, the ground vegetation is recognized and classified by using spectrum, the vegetation types are distinguished by analyzing the reflection and absorption characteristics of different wavelengths, the growth conditions are judged, and the ground vegetation information is obtained;

[0061] Based on the ground vegetation information, combined with microwave radar data and vegetation cover data, the soil moisture level of each region is calculated to obtain the soil moisture information;

[0062] Based on the soil moisture information, the slope, slope direction and ground roughness are calculated to obtain the ecological element information.

[0063] Specifically, based on the comprehensive remote sensing image data obtained in the foregoing, first, the corresponding multi-band information is read and a reflection intensity table for each band is established. In the table, an upper limit and a lower limit determined by experience statistics are set for visible light and near-infrared bands, for example, the reflection intensity upper limit can be 1.0, and the lower limit can be 0.0. If it is found that the reflection intensity of any band is lower than 0.0 or exceeds 1.0, it is judged that there may be noise points and marked accordingly. Then, after confirming that the band range is valid, spectral recognition is performed. Here, the actual reflectivity of each band can be compared with the pre-collected typical vegetation reflection curve. If the matching degree exceeds a certain preset threshold, it is determined that there is a similar spectral feature. The preset threshold can be calculated according to the average value of the reflectivity distribution measured in the same region in previous years, for example, when the average matching degree obtained in the sample of each main vegetation type is 0.75, it is set as the spectral matching threshold. Then, based on the determined vegetation type data, the growth condition is judged. Here, the vegetation density and leaf area index of each region can be calculated, and the vegetation density can be used as a reference to determine the growth condition. For example, if the vegetation density is greater than 0.5, it is determined that the growth condition is good; if the vegetation density is between 0.3 and 0.5, it is determined that the growth condition is normal; and if the vegetation density is less than 0.3, it is determined that the growth condition is poor. The above process obtains the classification label of the growth condition to obtain the ground vegetation information. As a measure, NDVI greater than 0.3 is considered to be in good condition, 0.2 to 0.3 is considered to be normal, and less than 0.2 is considered to be poor. The interval division can refer to the average threshold given by the local vegetation growth record and be corrected in combination with seasonal changes. Through the above process, the classification label of the growth condition is obtained to obtain the ground vegetation information.

[0064] Based on the ground vegetation information obtained in the foregoing, first, the vegetation type and its distribution proportion in each region are read and associated with the echo information in the microwave radar data. The waveform intensity corresponding to the spatial coordinates of the region is found in the echo information. Here, the intensity can be limited between 0.0 and 1.0. If it exceeds this range, it is marked as an abnormal echo and the data source is checked. Then, the existing vegetation cover data is introduced. The cover data includes parameters such as vegetation density and leaf area index of each region. Then, the vegetation density and echo information are combined to calculate the soil moisture level. Here, an empirical coefficient established according to the soil reflection characteristics can be used to correct the estimated value. For example, the unit of soil moisture is defined as volumetric water content, and the reference interval is set to 0% to 50%. If the calculated value exceeds this interval, the leaf area index in the vegetation cover data will be compared again to confirm whether there is an abnormality. When the calculated values of all regions fall within a reasonable range, the numerical distribution of soil moisture can be output to obtain the soil moisture information.

[0065] Based on the soil moisture information obtained, the water content of each region is first compared and overlaid with the pre-prepared digital elevation model. In this process, the slope is calculated based on the change in elevation. The slope value can be set in a floating interval of 0° to 90°. If the slope value exceeds this interval, it is necessary to check again whether there is data missing in the elevation model. After the slope calculation is completed, the aspect is determined based on the geographic orientation information. The azimuth angle is limited to 0° to 360°. If a negative value or more than 360° appears during calculation, the projection reference needs to be checked. Then, the surface roughness is determined by sampling the surface relief in a grid manner on the same elevation model. For example, the height difference distribution between adjacent grids is taken as a reference. If the height difference falls between 0 meters and 0.5 meters, it is recorded as low roughness. If it exceeds 0.5 meters, it can be considered as relatively higher roughness. This 0.5 meter is the average critical value selected from the historical regional topographic change records. After completing the above series of calculations and recording the slope, aspect, and surface relief information for each sampling point, the ecological factor information is obtained.

[0066] The temporary ecological carrying capacity acquisition step is:

[0067] Based on the ecological factor information, the net primary productivity of various plants is determined to obtain preliminary NPP data.

[0068] Based on the preliminary NPP data, real-time weather data is introduced to calculate the temporary ecological carrying capacity. The formula is:

[0069]

[0070] where NPP initial is the preliminary NPP data value, P is the real-time precipitation, P0 is the average precipitation, k and q are adjustment coefficients, T is the real-time air temperature, T0 is the average air temperature, A is the adjustment coefficient of air temperature influence, M is the soil moisture, M max is the maximum soil moisture, and EC is the temporary ecological carrying capacity.

[0071] Specifically, based on the ecological element information obtained in the foregoing, first, the recorded vegetation type and vegetation density in each region are confirmed, and the temperature and sunshine duration data of the region are combined, and the information is matched with the biomass parameters actually observed in the plant growth cycle, for example, the observed dry matter quality is compared with the interval of 0 grams to 1000 grams, if the dry matter quality of a certain region exceeds the interval, the sampling process needs to be reviewed and the on-site conditions of the soil and vegetation are further verified, when the recorded dry matter quality is consistent with the vegetation coverage and soil moisture content in the foregoing, the value is included in the subsequent calculation, and the historical biomass growth record in the region is referred to to determine the growth rate of the plant in a certain period of time, for example, the biomass of a batch of vegetation is recorded to increase from 100 grams to 150 grams in 7 days, and the daily average growth is about 7.14 grams, the growth rate is multiplied by the vegetation coverage, and the unit area increment of all similar vegetation is summed up to obtain the local net primary production contribution value of this type of plant, and the unit area net primary production contribution value of other types of plants is obtained by repeating the process, and the actual distribution area of each plant is weighted, and a plurality of classification intervals can be set in the process to identify whether there is a situation that the data deviates significantly from the regular growth rule in the region, for example, if the unit area growth rate is 50% higher than the average, it is marked as abnormal and the parameters such as ground temperature, soil moisture content or sampling period are compared again, when it is confirmed that all sampling values are normal, the net primary production contribution values of various plants are combined and summed up to obtain the preliminary NPP data.

[0072] The formula has the beneficial effect of comprehensively considering the dynamic changes of multiple parameters such as the preliminary NPP data, the precipitation level, the temperature state and the soil moisture;

[0073] NPP initial The acquisition step is to accumulate the 7-day measured biomass growth value by the preliminary NPP data obtained in the foregoing, for example, the unit area net primary production increment is recorded to be 35 grams in 7 days, and NPP initial = 35 grams / unit area, which matches the biomass actually observed by different vegetation communities, obtained from the growth cycle monitoring in the region;

[0074] The acquisition step of P is to record the precipitation of a day by the meteorological monitoring device and compare it with the interval of 0 mm to 300 mm, if it is not in the interval, the device reading is reviewed and the monitoring point is increased if necessary, when the record is 8 mm, P = 8 mm, which is measured on the same day;

[0075] The acquisition step of P0 is to sum up the precipitation of the same region in the past 30 days and then average it, for example, the total precipitation in the past 30 days is 210 mm, and P0 = 210 / 30 = 7 mm is calculated;

[0076] The obtaining step of k is extracting a precipitation adjustment coefficient from the data of the influence of multi-year regional precipitation on plant growth, segmenting and analyzing the regression relationship between NPP and precipitation in each period, and recording the value obtained by cumulative observation in the region for 4 years, and k=0.02 is selected as the index amplification rate of the influence of precipitation;

[0077] The obtaining step of q is estimating the slope and offset of the curve fitting in the negative feedback section of the same regression relationship, and q=1.5 is used to correct the temporary bearing capacity reduction amplitude when there is excessive precipitation;

[0078] The obtaining step of T is taking the current average temperature from the real-time record of the thermometer, and the recorded temperature is 25 degrees Celsius, so T=25, and the temperature is confirmed to be within the range of 0-45 degrees Celsius;

[0079] The obtaining step of T0 is selecting the average temperature in the same month within 10 years, and calculating T0=22 degrees Celsius;

[0080] The obtaining step of A is collecting high-temperature influence data on vegetation respiration intensity and transpiration rate in the field for multiple rounds of statistics, and combining the plant respiration rate change curve in different temperature zones to obtain the adjustment coefficient of temperature influence, which is currently defined as A=0.6;

[0081] The obtaining step of M is referring to the soil moisture information obtained before, and if the volume moisture content is 0.35 when measured, then M=0.35;

[0082] M max The obtaining step of M is sampling and recording the saturated moisture content in the same region at multiple points to obtain the most prominent value of about 0.5, so M max =0.5;

[0083] Calculation process:

[0084] Calculate the numerator

[0085]

[0086] Calculate the denominator

[0087]

[0088] Calculate

[0089]

[0090] Add the above results to get the value in the parentheses:

[0091]

[0092] Calculation

[0093]

[0094] cos(2.19911)≈-0.58850

[0095] |cos(2.19911)|=0.58850

[0096] Multiply the result of the above step by 16.10728 to get the final EC value:

[0097] EC=16.10728×0.58850≈9.49

[0098] The result shows that under the conditions of about 8 mm of precipitation, soil volumetric water content of about 0.35, and temperature of 25 degrees Celsius, the temporary ecological carrying capacity is about 9.49. If subsequent monitoring shows that EC is higher than 10, it indicates that it can support higher vegetation net primary production. If EC is lower than 5, it indicates that the local growth conditions are relatively limited. Therefore, the calculated result of 9.49 can be regarded as the temporary ecological carrying capacity under the current climate and soil conditions.

[0099] The steps to obtain the regional ecological health results are as follows:

[0100] Based on the temporary ecological carrying capacity, perform spatial analysis to map the carrying capacity of each region and form a spatial distribution map;

[0101] Based on the spatial distribution map, calculate the regional land health index, and the calculation formula is:

[0102]

[0103] Where H is the land health index, Z is the total number of analysis regions, EC i is the ecological carrying capacity of the i-th region, EC opt is the ideal ecological carrying capacity, and σ i is the coefficient of variation of the ecological carrying capacity of the i-th region.

[0104] Based on the land health index, assess the land health status, and generate regional ecological health results according to the environmental sensitivity and adaptability of the region.

[0105] Specifically, based on the temporary ecological carrying capacity value obtained in the foregoing, first, the carrying capacity is compared and marked with the corresponding coordinates in combination with the geographical boundary information of each region. In this process, a plurality of classification thresholds can be set to distinguish different carrying capacity levels. For example, a carrying capacity value less than 2.0 is regarded as a lower level, a carrying capacity value from 2.0 to 5.0 is regarded as a general level, a carrying capacity value from 5.0 to 10.0 is regarded as a relatively higher level, and a carrying capacity value greater than 10.0 is separately classified as an extremely high level. The determination of the classification threshold can refer to the field observation data of the vegetation growth conditions in a similar geographical environment for many years. The different carrying capacity intervals are compared with the corresponding plant net primary productivity and water conditions. After the grading standards are established, the specific carrying capacity values in the region are compared with the thresholds one by one. If the carrying capacity of a region is in the interval from 2.0 to 5.0, it is recorded as a general level. If the carrying capacity is greater than 10.0, it is marked as an extremely high level. Then, the carrying capacity grading operation of each region is completed block by block. The corresponding blocks are combined with the spatial coordinates. The carrying capacity levels of these blocks are marked in the spatial coordinate system. Finally, a mapping relationship containing the carrying capacity levels of all blocks is generated. The mapping relationship is superimposed with the underlying geographical boundary to form a spatial distribution map.

[0106] The formula has the beneficial effect of comprehensively considering the difference between the actual ecological carrying capacity and the ideal value of each region and measuring the fluctuation of the ecological carrying capacity among different regions by the coefficient of variation;

[0107] The obtaining step of H is that the ecological carrying capacity, the ideal ecological carrying capacity and the coefficient of variation of each region are sequentially collected and brought into the formula to obtain an average value.

[0108] The obtaining step of Z is that the number of all divided regions in the entire analysis range is counted. The specific value of Z can be determined through the region division data record.

[0109] EC i The obtaining step of EC is that the temporary ecological carrying capacity value obtained in the foregoing is used to summarize the corresponding measurement and calculation results of each region, and EC is recorded. i ;

[0110] EC opt The obtaining step of EC is that, in combination with historical data and field observation for many years, a carrying capacity value that is relatively stable and can reflect the best growth conditions of plants in the region is selected. Thus, the ideal ecological carrying capacity value is confirmed, and EC is recorded. opt ;

[0111] σ i The obtaining step of σ is that the distribution difference of the ecological carrying capacity is measured and collected in a region for many times. The standard deviation or variance result of the carrying capacity of all measurement points is recorded and mapped as the coefficient of variation, and σ is recorded. i ;

[0112] Calculation process:

[0113] Specify the number of regions Z. For example, if 5 sub-regions are defined, then Z = 5.

[0114] The carrying capacity EC1, EC2, EC3, EC4, and EC5 of each area are collected, for example, 9.49, 6.20, 11.10, 4.80, and 7.90 respectively;

[0115] Set ideal ecological carrying capacity EC opt =8.0, this value is based on the historical average level of similar vegetation distribution areas at the same latitude;

[0116] Obtain the coefficients of variation σ1 to σ5 of each region, for example, 0.45, 0.35, 0.55, 0.40, and 0.42 are obtained through measurement and recording, respectively;

[0117] Calculate for each region And accumulate, taking area 1 as an example:

[0118]

[0119] Calculate the other four regions in the same way and add them together, for example, the total is S≈3.762;

[0120] Final calculation

[0121] The results show that in the five selected areas, the overall land health index is approximately 0.752. If the subsequent calculated result is greater than 1.0, it means that the carrying capacity of most areas is close to the ideal value and the variation is relatively low. If the result is less than 0.5, it means that there is a large gap between the carrying capacity and the ideal value in some areas. This value is not only a measure of the land health status within the current statistical scope, but can also serve as the data basis for subsequent adjustment strategies.

[0122] Based on the land health index value obtained in the foregoing, the corresponding environmental sensitivity record of each region is called, which includes specific quantitative data of environmental change factor sensitivity and land restoration difficulty, for example, the water resource shortage degree of a region is expressed in a 0.0 to 1.0 grading table, if the value is close to 1.0, it means that the water resource is relatively scarce, at the same time, the soil erosion rate or the underground water decline trend in the region can be referred to for supplementary judgment, compare these quantitative data with the health index calculated in the foregoing, if a higher environmental sensitivity is observed in a region with a lower health index, mark it as an urgent focus, then combine the adaptability parameter of the region, which is integrated by the soil bearing range and the biodiversity record, specifically, the biodiversity monitoring data in the past 3 to 5 years can be collected and matched with the soil analysis results to determine whether each region has the self-regulating ability to cope with climate or vegetation changes, after combining the environmental sensitivity and adaptability data, a priority list of regions with low health index and high sensitivity is output, finally, compare the list with the actual geographical distribution, arrange the tracking monitoring order for each key region, and generate the regional ecological health result.

[0123] The obtaining step of the ecological target task allocation table is:

[0124] Based on the regional ecological health result, identify vegetation coverage problems and water resource consumption problems, and form an ecological restoration target list;

[0125] Based on the ecological restoration target list, develop ecological restoration measures, including afforestation and water improvement, and obtain an ecological restoration measure plan;

[0126] Based on the ecological restoration measure plan, determine the responsibility allocation and the time schedule, and establish the ecological target task allocation table.

[0127] Specifically, based on the previously obtained regional ecological health results, the coverage and water resource utilization of each region are first called as basic data, where the coverage can be obtained from the previously recorded vegetation distribution ratio. If the vegetation coverage of a certain area is less than 30%, it is marked as a lower coverage area, and if the coverage is higher than 70%, it is marked as a higher coverage area. These two thresholds can be determined by referring to local historical vegetation observations and combining the results of surface image comparison. At the same time, the water resource consumption curve of each place is called and compared with the water intake in the hydrological monitoring record. For example, the daily average water intake is compared with the 0 cubic meter to 500 cubic meter interval. If the daily water intake of a certain area exceeds 300 cubic meters for several consecutive days, it is recorded as a high consumption section. This 300 cubic meter limit can be derived from long-term statistical data of local population and industrial water demand. When it is found that the coverage is too low or the water consumption is long-term higher than the above-mentioned limit, it is determined that the area has the corresponding problem. All areas marked with vegetation coverage or water resource problems are summarized, and the key attention areas are identified. The key attention areas can be ranked according to the deviation of coverage and water consumption. For example, if the coverage deviation is more than 10 percentage points from the 30% threshold and the daily water intake is long-term close to or higher than 500 cubic meters, it is determined as the first ranked area. After completing the screening and recording of all areas, an induction table covering vegetation coverage problems and water resource consumption problems is integrated to form an ecological restoration target list.

[0128] Based on the previously obtained ecological restoration target list, the appropriate green increase content is first specified for each area with low vegetation coverage. For example, the dense planting scheme of selecting specific trees and shrubs on open grassland is selected. The tree species to be planted are selected in combination with existing soil nutrients and climate conditions. If the air temperature in this area is between 15°C and 25°C and the soil moisture content is maintained between 15% and 25%, then warm-loving tree species are preferred. When the soil pH reaches 6.5 to 7.5, plants with similar adaptability are matched. At the same time, the areas with prominent water resource consumption problems are formulated with corresponding engineering content, such as setting up small water storage tanks or renovating irrigation channels. The daily water consumption is compared with the established water allocation interval. If the daily water consumption is higher than the allowed amount for 5 consecutive days, it is determined that the area needs to improve water supply. This allocation interval can be calculated from the population size and crop water requirement. After completing the plan of tree planting and water conservancy improvement, the urgency and actual operation difficulty of each area are ranked. If soil improvement is required in a certain area, it is also included in the green increase measures. Finally, a document covering time schedule, construction focus, and detailed requirements of tree planting and water conservancy is summarized to obtain the ecological restoration measures plan.

[0129] Based on the ecological restoration measures plan obtained in the foregoing, a corresponding person in charge is first designated in each participating department or team and the management scope thereof is recorded, for example, the afforestation group is responsible for two types of areas, i.e., low-coverage grassland and abandoned forest land, and the water conservancy group is responsible for two types of areas, i.e., high-water-consumption agricultural irrigation areas and urban water supply pipeline reconstruction areas. Then, the plan nodes of tree planting or water conservancy reconstruction are compared with the local seasonal regularity. If the tree planting group plans to carry out green increase operation in winter, it is necessary to compare whether the average temperature in this stage is between 5°C and 15°C. If the water conservancy group plans to carry out channel construction in the rainy season, it is necessary to regard whether the weekly rainfall is within the range of 0 mm to 50 mm as a feasible condition. These thresholds can be determined by referring to statistical data of the local climate cycle in the past. After all the division of work and schedules are completed, the priorities of various tasks are matched, for example, areas that urgently need to increase vegetation coverage and whose water conservancy facilities are in disrepair are listed as priorities, and areas that have medium coverage and whose water consumption is not continuously over-standard are arranged in the subsequent arrangement. According to these priorities and the executable manpower of each department, a segmented progress plan is arranged, and the targets and persons in charge of each period are recorded accordingly. After all the tasks and schedules are matched and completed, an ecological target task allocation table is established.

[0130] The task execution monitoring result obtaining step is:

[0131] Using the ecological target task allocation table as the basis, task monitoring is deployed to track the completion of tasks by each department and team in real time, and task execution monitoring data is obtained.

[0132] Based on the task execution monitoring data, the progress of each task and the compliance of the predetermined target are evaluated, delays and deviations are identified, and task progress analysis results are obtained.

[0133] Based on the task progress analysis results, the execution of all tasks is integrated and evaluated, adjustment measures or updated task priorities are formulated, and task execution monitoring results are obtained.

[0134] Specifically, using the ecological target task allocation table as the basis, first read the work items and time nodes corresponding to each department and team, and correspond these items with the previously established ecological restoration measures plan and responsibility allocation information. In the corresponding process, it is necessary to verify whether the start and end dates confirmed by each department fall within the time period set in advance, for example, compare the execution time of the tree planting group with the interval of March 1 to April 30, if it exceeds the interval, mark it as having a progress problem, compare the execution time of the water conservancy group with the interval of May 1 to June 30, if it falls within this interval, mark it as normal. These intervals can be determined according to local temperature and precipitation rules. When the correspondence is completed, combine it with the percentage of completion of the stage target, for example, compare the tree planting completion percentage with the interval of 0% to 100%, if it is less than 20%, mark it as progress lag, when it exceeds 80%, mark it as close to completion. For departments whose progress exceeds the preset threshold or deviates significantly from the expected schedule, the threshold can be set by statistical analysis of the average time of similar tasks in the past, for example, less than 2% daily progress increment for two weeks is considered lagging, more than 5% increment in a short period is considered rapid local progress. Through daily or weekly comparison process, record the execution of each department and classify, finally form task execution monitoring data.

[0135] Based on the task execution monitoring data obtained above, first analyze the completion time and completion quality indicators of each task, and compare them one by one with the planned target given in advance. If it is found that the time consumption of a task exceeds 30% of the original time for many consecutive days, it is determined to be a serious delay. The value of 30% can be selected by reviewing the delay distribution proportion of past projects, and at the same time, referring to the statistics of manpower and equipment in each stage, the delayed link is refined into a problem of resource scheduling or personnel management, for example, in the tree planting task, if the daily tree planting number is significantly insufficient compared with the interval of 0 to 50, it is determined that the tree planting progress does not meet the established quantity target, in the water conservancy reconstruction, if the daily pipeline replacement number is less than 2, it is considered as progress slowing down. These judgment thresholds can be set based on the specific equipment efficiency and daily productivity statistics of manual labor. After comparing all task completion degrees with the predetermined target, if multiple tasks have delays or deviations, record their positions and reasons in the task list, if individual task progress meets the predetermined target, mark it as normal. After completing all comparisons and marking, classify and summarize, and extract task progress analysis results from them.

[0136] Based on the task progress analysis results obtained, all departments and team execution are arranged and compared on the same timeline, and the project entries that do not meet the schedule requirements on time are found and marked with priority, the priority can be determined according to the influence range and delay degree, for example, the project that has not entered the scheduled process for more than 5 days or the task completion degree is less than 50%, the 5-day threshold can be determined by referring to the fluctuation of the average milestone progress of the existing projects, the resource allocation and the number of professional personnel required for the project are checked at the same time, if the personnel need to be increased, the existing personnel reserve is combined to dispatch, the updated plan schedule is compared with the existing work arrangement again to determine whether the new schedule can be satisfied within the preset interval, if there is still conflict, continue to adjust the division of work, for the project with normal progress analysis results, maintain the existing allocation and enter the next cycle of observation, all the tasks adjusted or rearranged are listed in the tracking entries of the next stage after the record is completed, and finally the task execution monitoring result is obtained.

[0137] The performance analysis result obtaining step is:

[0138] Based on the task execution monitoring result, the time point and feedback of task completion are recorded to obtain the original data set of ecological performance;

[0139] Based on the original data set of ecological performance, a random forest model is applied and the ecological performance index is calculated, and the formula is adjusted as:

[0140]

[0141] Wherein, EPI is the ecological performance index, n is the number of data points, f i (X) is the predicted value of the ith data point calculated by the random forest model, W i is the weight associated with the data point i;

[0142] Based on the ecological performance index, the performance level of each department and project is evaluated, the task delay or quality inconsistency is identified, and the performance analysis result is obtained.

[0143] Specifically, based on the task execution monitoring results obtained earlier, first extract the time node recorded by each task in the completion link, compare the time node with the original progress plan, if there is a deviation of more than two days, mark it as obvious delay, the value of two days can be set in combination with the average delay period of similar type work in the past, at the same time, extract the completion quality information from the team feedback, for example, for tree planting work, the survival rate record of unit area can be read and compared with 0% to 100% interval, if the survival rate is continuously lower than 60%, it is marked as quality problem, the threshold of 60% can be referred to the average survival rate of similar environment in recent years, for water conservancy transformation, the number of pipeline replacement per day can be recorded and compared with 0 to 10 execution interval, if the average number of replacement per day is less than 3, it is marked as progress lag, this value of 3 can be set according to the historical construction efficiency of local construction team, after obtaining the completion time and related feedback of all tasks, check whether it meets the pre-set completion date and quality index in turn, if it is found that the completion time of multiple tasks exceeds the maximum allowable delay specified in the foregoing, further analyze the causes to facilitate subsequent aggregation, after all the records are sorted out, a detailed data containing completion time and feedback results is formed, and finally the original data set of ecological performance is obtained.

[0144] The advantage of the formula is to comprehensively consider the normalized relationship between the prediction value output by the random forest model and the weight of each data point, and reflect the distribution characteristics of the overall ecological performance by squaring and taking square root;

[0145] The EPI acquisition step is to first aggregate the ecological performance data points of all projects and departments and input them into the random forest model one by one, calculate the prediction value after obtaining the prediction value, and calculate it combined with the corresponding weight;

[0146] The acquisition step of n is to count all the collected monitoring points or project items one by one to get the total number of data points, for example, there are 8 monitoring points in a evaluation, then n = 8;

[0147] f i The acquisition step of (X) is to input the input features of the i th data point into the random forest model, and the vegetation restoration amount, water conservancy transformation progress and other indicators are used as features in the training process. The prediction value output by the model for the i th data point is obtained through multiple rounds of iterative training, for example, the completion degree and survival rate of tree planting of a department are input as features, the output prediction value can be recorded as f i (X), the value range can be calibrated according to the historical project evaluation records obtained earlier;

[0148] W iThe acquisition step is to assign values according to the importance or influence range of each data point, for example, to quantify the influence coefficient of the index by expert scoring or historical data, if a monitoring point involves a larger area of vegetation coverage, the weight can be relatively high, and its range is set to 0.1 to 1.0, if the final record of the point weight is 0.8, then W i = 0.8;

[0149] Calculation process:

[0150] Get the prediction value of a single data point and the corresponding weight, for example, the first data point prediction value f1(X) = 0.75, weight W1 = 0.8, and so on, record all f i (X) and W i ;

[0151] Divide the prediction value of each data point by the corresponding weight and square it, for example, for the first data point:

[0152]

[0153] Add up the results corresponding to all data points and take the square root, for example, assuming that the operation result of 8 data points is accumulated to 5.64, then:

[0154] The result shows that when the random forest model predicts the data of each monitoring point, the ecological performance index calculated by the above weighted sum of squares and square root is about 2.375, if the subsequent observation EPI value is greater than 3.0, it means that the overall ecological performance score tends to be higher, if it is lower than 1.0, it means that most monitoring points are not ideal, therefore 2.375 can be regarded as a quantitative measure of ecological performance under the current comprehensive conditions.

[0155] Based on the ecological performance index obtained, the actual completion time, quality index and corresponding EPI performance of each department or project are first compared. In the comparison process, several classification intervals can be set, for example, departments with EPI less than 1.0 are considered as lower performance, 1.0 to 2.0 as general performance, 2.0 to 3.0 as higher performance, and more than 3.0 as extremely high performance. These intervals can be determined in combination with the distribution statistics of previous ecological project evaluations. Then the performance points of each department are summarized and the distribution in the above intervals is viewed. If a department is in the lower performance interval for two consecutive times, it is marked as possibly having serious delay or quality deficiency. Further confirmation is made in combination with the delay days or quality feedback in the task execution monitoring results. For departments that have reached the extremely high performance interval for many times, the corresponding completion time and resource input indicators are recorded to check whether there is obvious waste of manpower or material resources. After the performance of all departments is completed, the results are integrated into a list, and the daily progress or quality differences of each project are analyzed. Finally, a comprehensive performance comparison information covering all departments and projects is obtained, forming the performance analysis result.

[0156] The above is only a preferred embodiment of the present application, and does not limit the present application in other forms. Any skilled person in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application technical solution.

Claims

1. A land ecological carrying capacity evaluation method based on image analysis, characterized in that, The following steps are involved: Collect multi-source remote sensing image data, including optical remote sensing images, synthetic aperture radar images, and drone images, pre-process the collected data, synchronize multi-source data, and generate comprehensive remote sensing image data; based on the comprehensive remote sensing image data, extract surface vegetation, soil moisture, and topography information to obtain ecological element information; Utilizing the ecological factor information, calculating ecological indicators and integrating real-time meteorological data to obtain a temporary ecological carrying capacity; performing spatial analysis on the temporary ecological carrying capacity to determine the health status of the land in the region and generate a regional ecological health result; According to the ecological health results of the region, set ecological restoration goals, break down the ecological restoration goals into departmental and project management tasks, and establish an ecological goal task allocation table; Based on the ecological target task allocation table, monitor the task execution progress and obtain the task execution monitoring results; Use the task execution monitoring results to evaluate the ecological performance of each department and project, identify performance problems, and obtain performance analysis results.

2. The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The steps for obtaining the comprehensive remote sensing image data are as follows: Collect image data from three sources: optical remote sensing, synthetic aperture radar, and drones. Record the image data type and timestamp of each source to obtain a preliminary image dataset. Based on the preliminary image dataset, performing noise removal and geometric correction on each image data, performing radiation correction on the optical remote sensing image, performing despeckle processing on the synthetic aperture radar image, unifying the image resolution and format, and obtaining a preprocessed image dataset; Based on the preprocessed image dataset, time synchronization is used to correct the data time deviation, and spatial alignment is applied to generate comprehensive remote sensing image data. 3.The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The steps for obtaining the ecological element information are as follows: Analyzing the comprehensive remote sensing image data, using spectrum to identify and classify surface vegetation, distinguishing vegetation types and determining growth conditions by analyzing reflection and absorption characteristics of different wavelengths, and obtaining surface vegetation information; Based on the surface vegetation information, combined with microwave radar data and vegetation cover data, soil moisture levels in each area are calculated to obtain soil moisture information; Based on the soil moisture information, the slope, aspect and surface roughness are calculated to obtain ecological element information.

4. The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The steps for obtaining the temporary ecological carrying capacity are: Based on the ecological factor information, the net primary production of various plants is determined to obtain preliminary NPP data; Based on the preliminary NPP data, real-time meteorological data was introduced to calculate the temporary ecological carrying capacity. The formula is: where NPP initial is the preliminary NPP data value, P is the real-time precipitation, P0 is the average precipitation, k and q are adjustment coefficients, T is the real-time temperature, T0 is the average temperature, A is the adjustment coefficient of temperature influence, M is the soil moisture, M max is the maximum soil moisture, and EC is the temporary ecological carrying capacity.

5. The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The steps for obtaining the regional ecological health results are as follows: Based on the temporary ecological carrying capacity, perform spatial analysis to map the carrying capacity of each area and form a spatial distribution map; Based on the spatial distribution map, the regional land health index is calculated using the following formula: Wherein, H is the land health index, Z is the total number of analysis regions, EC i is the ecological carrying capacity of the ith region, EC opt is the ideal ecological carrying capacity, σ i is the variation coefficient of the ecological carrying capacity of the ith region; Based on the land health index, the land health status is assessed and regional ecological health results are generated according to the environmental sensitivity and adaptability of the region.

6. The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The steps for obtaining the ecological target task allocation table are as follows: Based on the ecological health results of the region, identify vegetation coverage issues and water resource consumption issues and form a list of ecological restoration targets; Based on the list of ecological restoration targets, develop ecological restoration measures, including afforestation and water improvement, to obtain an ecological restoration measure plan; Based on the ecological restoration measure plan, determine responsibility allocation and time schedule, and establish an ecological target task allocation table.

7. The land ecological carrying capacity evaluation method based on image analysis according to claim 1, characterized in that, The task execution monitoring result obtaining step is: Using the ecological target task allocation table as a basis, deploy task monitoring to track the completion of tasks by each department and team in real time to obtain task execution monitoring data; Based on the task execution monitoring data, evaluate the progress of each task and the degree of compliance with the predetermined target, identify delays and deviations, and obtain a task progress analysis result; Based on the task progress analysis result, integrate and evaluate the execution of all tasks, develop adjustment measures or update task priorities, and obtain a task execution monitoring result. 8.The land ecological carrying capacity evaluation method based on image analysis of claim 1, wherein, The performance analysis result obtaining step is: Based on the task execution monitoring result, record the time point of task completion and feedback to obtain an original data set of ecological performance; Based on the original data set of ecological performance, apply a random forest model and calculate an ecological performance index, with the formula adjusted as follows: where EPI is the ecological performance index, n is the number of data points, f i (X) is the predicted value of the ith data point calculated by the random forest model, W i is the weight associated with data point i; Based on the ecological performance index, evaluate the performance level of each department and project, identify task delays or quality inconsistencies, and obtain a performance analysis result.

Images