Post-fire forest loss evaluation method and device based on unmanned aerial vehicle and microwave radar

Abstract

The invention discloses a post-fire forest loss evaluation method and device based on an unmanned aerial vehicle and a microwave radar. The method comprises the following steps: S1, acquiring post-fire environment information and forest biophysical parameters of a predetermined area; s2, selecting a forest loss evaluation model according to the post-fire field environment information and the forest biophysical parameters; and S3, inputting the post-fire field environment information and the forest biophysical parameters of the target area into the forest loss evaluation model for post-fire forest loss evaluation. By adopting the technical scheme of the invention, the forest loss caused by the fire can be quantitatively evaluated.

Description

Technical Field

[0001] The present invention belongs to the field of computer technology, and in particular relates to a method and device for evaluating forest loss after fire based on unmanned aerial vehicles and microwave radar. Background Art

[0002] With the impact of global climate change and human activities, the risk of forest fires is increasing, causing serious damage to forest ecosystems. Forest fires are extremely destructive natural disasters that not only damage forest resources, destroy the ecological environment, and affect economic development and social stability, but also pose a huge threat to people's lives, property and safety. At present, it is impossible to quantitatively evaluate the forest losses caused by fires. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method and device for evaluating forest loss after fire based on unmanned aerial vehicles and microwave radar, which can quantitatively evaluate the forest loss caused by fire.

[0004] To achieve the above object, the present invention adopts the following technical solution:

[0005] A method for evaluating forest loss after fire based on UAV and microwave radar, including:

[0006] Step S1, obtaining post-fire environmental information and forest biophysical parameters of a predetermined area;

[0007] Step S2, selecting a forest loss assessment model based on post-fire site environmental information and forest biophysical parameters;

[0008] Step S3: input the post-fire on-site environmental information and forest biophysical parameters of the target area into the forest loss assessment model to conduct post-fire forest loss assessment.

[0009] Preferably, in step S1, the post-fire site environmental information of the predetermined area and the airborne L-band fully polarized synthetic aperture radar data (Polarimetric Synthetic Aperture Radar, PolSAR) are preprocessed, and the post-fire site environmental information (forest fire intensity and burned area) and forest biophysical parameters are extracted and calculated. The forest biophysical parameters are obtained by polarization decomposition of the airborne L-band fully polarized synthetic aperture radar data.

[0010] Preferably, in step S2, the PolSAR data and UAV remote sensing images of the predetermined area are used as independent variables, the measured biomass of the predetermined area is used as the dependent variable, and a forest loss assessment model for the predetermined area is constructed using random forest.

[0011] Preferably, in step S3, the processed PolSAR data and UAV remote sensing data of the target area are input into the optimal forest loss assessment model to calculate the forest loss amount in the predetermined area, and then graded calculations are performed according to the forest fire intensity and the burned area to obtain the changes in forest loss amount under different forest fire intensities and burned areas.

[0012] The present invention also provides a post-fire forest loss assessment device based on an unmanned aerial vehicle and a microwave radar, comprising:

[0013] An acquisition module, which obtains post-fire environmental information and forest biophysical parameters of a predetermined area;

[0014] The processing module determines the intensity and burned area of ​​forest fires based on the post-fire on-site environmental information and forest biophysical parameters, and constructs a forest loss assessment model;

[0015] The evaluation module inputs the post-fire on-site environmental information and forest biophysical parameters of the target area into the forest loss evaluation model to conduct post-fire forest loss evaluation.

[0016] The acquisition modules include:

[0017] The first processing unit is used to pre-process the post-fire site environmental information and airborne L-band fully polarized synthetic aperture radar data of the predetermined area, wherein all UAV images are subjected to radiometric calibration, atmospheric correction, geometric correction and image cropping, and microwave radar data are subjected to radiometric calibration, multi-viewing, filtering, terrain radiometric correction, geocoding and geometric correction;

[0018] The second processing unit is used to extract and calculate post-fire site environmental information and forest biophysical parameters. The forest biophysical parameters are obtained by polarization decomposition of airborne L-band fully polarized synthetic aperture radar data.

[0019] The processing module takes the PolSAR data and UAV remote sensing images of the predetermined area as independent variables, and the measured biomass of the predetermined area as the dependent variable, and uses random forest to construct a forest loss assessment model for the predetermined area.

[0020] The evaluation modules include:

[0021] The first calculation unit evaluates the forest fire intensity and the burned area and performs graded calculations to obtain the burned area at different forest fire intensities on a fine scale;

[0022] The second calculation unit inputs the processed PolSAR data and UAV remote sensing data of the target area into the forest loss assessment model to calculate the forest loss amount of the predetermined area.

[0023] The present invention obtains post-fire environmental information and forest biophysical parameters of a predetermined area; selects a forest loss assessment model based on the post-fire on-site environmental information and forest biophysical parameters; and inputs the post-fire on-site environmental information and forest biophysical parameters of a target area into the forest loss assessment model to perform post-fire forest loss assessment. The technical solution of the present invention can quantitatively assess forest loss caused by fire. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0025] Figure 1 The present invention is a flowchart of a method for evaluating forest loss after fire based on UAV and microwave radar according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0028] Embodiment 1:

[0029] like Figure 1 As shown, an embodiment of the present invention provides a method for evaluating forest loss after fire based on a drone and a microwave radar, comprising:

[0030] Step S1, obtaining post-fire environmental information and forest biophysical parameters of a predetermined area;

[0031] Step S2: constructing a forest loss assessment model based on post-fire site environmental information and forest biophysical parameters;

[0032] Step S3: input the post-fire on-site environmental information and forest biophysical parameters of the target area into the forest loss assessment model to conduct post-fire forest loss assessment.

[0033] As an implementation method of an embodiment of the present invention, step S1 includes preprocessing the post-fire site environment information and airborne L-band fully polarized synthetic aperture radar data (Polarimetric Synthetic Aperture Radar, PolSAR) of a predetermined area, and extracting and calculating the post-fire site environment information and forest biophysical parameters. The post-fire site environment information includes forest fire intensity and burned area, and the forest biophysical parameters are obtained by polarization decomposition of airborne L-band fully polarized synthetic aperture radar data.

[0034] First, the UAV remote sensing images and PolSAR data are preprocessed. All UAV remote sensing images are subjected to radiometric calibration, atmospheric correction, geometric correction and image cropping. The PolSAR data preprocessing includes radiometric calibration, multi-viewing, filtering, terrain radiometric correction, geocoding and geometric correction.

[0035] Secondly, the processed PolSAR data are used to extract polarization decomposition parameters and construct polarization scattering ratio parameters.

[0036] Polarization decomposition parameters:

[0037] The PolSAR data were subjected to incoherent target polarization decomposition. A total of 9 polarization decomposition methods were used, including Freeman3 decomposition, Freeman2 decomposition, Yamaguchi3 decomposition, An&Yang3 decomposition, An&Yang4 decomposition, Yamaguchi4_Y4O decomposition, Yamaguchi4_Y4R decomposition, Singh-Yamaguchi6 decomposition, and Singh-Yamaguchi7 decomposition, and 36 polarization decomposition parameters were obtained (Table 1).

[0038] Table 1

[0039]

[0040] Polarization scattering ratio parameter construction:

[0041] Decomposition of AnYang4 (R 1 )、AnYang3 decomposition (R 2 )、Freeman2 decomposition(R 3 )、Freeman3 decomposition(R 4 ), Yamaguchi3 decomposition (R 5 ) to obtain the polarization decomposition parameters and perform ratio calculation to obtain five polarization scattering ratio parameters. The calculation formula is as follows:

[0042]

[0043] Among them, P AY4_V , PAY4_D , P AY4_O , P AY4_H , P AY4_G They are respectively the surface scattering, even-order scattering, surface scattering, helical scattering and surface scattering components of each polarization decomposition method.

[0044] Finally, the fire intensity was graded and the burned area was calculated. A regression equation was established based on the composite burn index (CBI) and post-fire NBR obtained from field surveys. The NBR thresholds corresponding to different levels of fire intensity were determined based on the regression equation, and the fire intensity and burned area were then determined.

[0045] Obtaining CBI through field investigation:

[0046] In a 20m×20m square plot, four layers were investigated according to the vertical structure of the forest: (A) surface fuel and soil layer; (B) herbaceous, low shrubs and small trees <1m; (C) tall shrubs and 1-5m trees; (D) 5-20m canopy layer (Table 2). Several predictor variables were set in each layer and visually estimated, with the variable value range from 0 to 3 (0 represents no fire; 1 represents light fire; 2 represents moderate fire; 3 represents moderate fire).

[0047] Table 2

[0048]

[0049]

[0050] The estimated values ​​of each layer are combined to obtain the CBI value of the entire plot. The calculation formula is as follows:

[0051]

[0052] Among them, X ij is the CBI score of the jth survey factor in the i-th layer; n is the total number of survey factors in the i-th layer.

[0053] The total CBI value calculation formula of the sample plot is:

[0054]

[0055] Among them, CBI is the comprehensive fire index of the sample site; CBI i is the CBI value of the i-th layer; k is the number of strata in the sample plot.

[0056] NBR calculation:

[0057] According to the NBR calculation formula, NBR is extracted based on the remote sensing image after atmospheric correction. The theoretical value range of NBR is [-1,1]. For the convenience of calculation and analysis, the original NBR value is multiplied by 1000 and converted into an integer. The NBR calculation formula is as follows:

[0058]

[0059] Among them, NIR is the near infrared band in remote sensing images, and SWIR is the short infrared band in remote sensing images.

[0060] As an implementation method of an embodiment of the present invention, in step S2, the PolSAR data and UAV remote sensing images of the predetermined area are used as independent variables, and the measured biomass of the predetermined area is used as the dependent variable, and a forest loss assessment model for the predetermined area is constructed using Random Forest (RF); 75% of the sample data is selected as a training set, and the remaining 25% of the data is used as a test set, and the root mean square error, relative root mean square error and determination coefficient are compared to measure the model accuracy.

[0061] Sample plot data survey and forest loss calculation:

[0062] The biomass survey of the sample plot includes the tree layer and the shrub layer. According to the fire intensity (unburned area, mild, moderate and severe), sample plots are set up in the predetermined area for investigation. 3-5 sample plots are set up for each scenario, and 3-5 20m×20m sample plots are set up in each sample plot. Each tree (DBH≥2.5cm) in each sample plot is measured, and its tree species, diameter at breast height (DBH) and tree height (H) are recorded; for shrubs, 5 2m×2 small sample plots are set up in the shape of plum blossoms in the large sample plot of 20m×20m. The dominant shrub species (including young trees with a measured diameter of <2.5cm) are investigated in the shrub sample plot, and the name, coverage, number of plants (height <1m is not included), and average height of the dominant shrubs are recorded.

[0063] The biomass of the tree layer was calculated based on the measured diameter at breast height and tree height using the allometric growth equation for different tree species. The biomass of the shrub layer was calculated using the full harvest method. The samples taken from each plot were dried in the laboratory for 24-48 hours (75°C) until constant weight was reached, and the dry weight was measured to obtain the biomass of the shrub layer. The sum of the biomass of the tree layer and the shrub layer was converted to the biomass per unit area of ​​the forest (t / hm2).

[0064] Random Forest:

[0065] Construct a post - fire forest loss model using the random forest method. The process of the random forest mainly includes: (1) Input the sample set; (2) Randomly select the training set; (3) Select m feature variables from M feature variables (m < M); (4) Based on the above steps, construct n decision trees and calculate the average value to obtain the final prediction result.

[0066] Model accuracy evaluation parameters:

[0067] In the embodiments of the present invention, the root mean squared error (RMSE), relative root mean squared error (rRMSE), and coefficient of determination (R - Square, R 2 ) can be further used to evaluate the training and test results of the model, and obtain the accuracy of the forest loss evaluation model.

[0068]

[0069] Among them, RMSE is the root mean squared error; y i is the true value, is the predicted value; is the average of the true values; N is the sample size. Among them, the smaller the RMSE value, the better the model prediction effect; the smaller the rRMSE, the better the model prediction effect; the larger the R 2 value, the higher the correlation between the independent variable and the dependent variable.

[0070] As an implementation manner of the embodiments of the present invention, in step S3, the processed PolSAR data and UAV remote sensing data of the target area are input into the forest loss evaluation model to calculate the forest loss amount of the predetermined area.

[0071] Embodiment 2:

[0072] The embodiments of the present invention also provide a post - fire forest loss evaluation device based on UAV and microwave radar, including:

[0073] An acquisition module, which acquires the post - fire environmental information and forest biophysical parameters of the predetermined area;

[0074] A processing module, which selects a forest loss evaluation model according to the post - fire on - site environmental information and forest biophysical parameters;

[0075] An evaluation module, which inputs the post - fire on - site environmental information and forest biophysical parameters of the target area into the forest loss evaluation model for post - fire forest loss evaluation.

[0076] As an implementation manner of the embodiments of the present invention, the acquisition module includes:

[0077] The first processing unit is used to pre-process the post-fire site environmental information and airborne L-band fully polarized synthetic aperture radar data of the predetermined area, wherein all UAV images are subjected to radiometric calibration, atmospheric correction, geometric correction and image cropping, and microwave radar data are subjected to radiometric calibration, multi-viewing, filtering, terrain radiometric correction, geocoding and geometric correction;

[0078] The second processing unit is used to extract and calculate post-fire site environmental information and forest biophysical parameters. The forest biophysical parameters are obtained by polarization decomposition of airborne L-band fully polarized synthetic aperture radar data.

[0079] As an implementation method of an embodiment of the present invention, the processing module takes the PolSAR data and UAV remote sensing images of a predetermined area as independent variables, and the measured biomass of the predetermined area as the dependent variable, and uses random forest to construct a forest loss assessment model for the predetermined area; 75% of the sample data is selected as a training set, and the remaining 25% of the data is used as a test set, and the root mean square error, relative root mean square error and determination coefficient are compared to measure the model accuracy.

[0080] As an implementation of the embodiment of the present invention, the evaluation module includes:

[0081] The first calculation unit is used to input the processed PolSAR data and the UAV remote sensing data of the target area into the optimal forest loss assessment model to calculate the forest loss amount of the predetermined area;

[0082] The second calculation unit is used to perform graded calculations according to the forest fire intensity and the burned area, and obtain the changes in forest loss under different forest fire intensities and burned areas.

[0083] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for evaluating forest loss after fire based on drones and microwave radar, characterized in that: include: Step S1, obtaining post-fire environmental information and forest biophysical parameters of a predetermined area; Step S2, selecting a forest loss assessment model based on post-fire site environmental information and forest biophysical parameters; Step S3: input the post-fire on-site environmental information and forest biophysical parameters of the target area into the forest loss assessment model to conduct post-fire forest loss assessment.

2. The method for evaluating forest loss after fire based on drone and microwave radar as claimed in claim 1, characterized in that: In step S1, the post-fire site environmental information and airborne L-band fully polarized synthetic aperture radar data of the predetermined area are preprocessed, and the post-fire site environmental information and forest biophysical parameters are extracted and calculated. The forest biophysical parameters are obtained by polarization decomposition of the airborne L-band fully polarized synthetic aperture radar data.

3. The method for evaluating forest loss after fire based on drone and microwave radar as claimed in claim 2, characterized in that: In step S2, the PolSAR data and UAV remote sensing images of the predetermined area are used as independent variables, the measured biomass of the predetermined area is used as the dependent variable, and a forest loss assessment model for the predetermined area is constructed using random forest.

4. The method for evaluating forest loss after fire based on unmanned aerial vehicle and microwave radar as claimed in claim 3, characterized in that: In step S3, the processed PolSAR data and UAV remote sensing data of the target area are input into the optimal forest loss assessment model to calculate the forest loss amount in the predetermined area, and then graded calculations are performed according to the forest fire intensity and the burned area to obtain the changes in forest loss amount under different forest fire intensities and burned areas.

5. A post-fire forest loss assessment device based on drones and microwave radar, characterized in that: include: An acquisition module, which obtains post-fire environmental information and forest biophysical parameters of a predetermined area; The processing module selects the forest loss assessment model based on the post-fire site environmental information and forest biophysical parameters; The evaluation module inputs the post-fire on-site environmental information and forest biophysical parameters of the target area into the forest loss evaluation model to conduct post-fire forest loss evaluation.

6. The post-fire forest loss assessment device based on unmanned aerial vehicle and microwave radar as claimed in claim 5, characterized in that: The acquisition modules include: The first processing unit is used to pre-process the post-fire site environmental information and airborne L-band fully polarized synthetic aperture radar data of the predetermined area, wherein all UAV images are subjected to radiometric calibration, atmospheric correction, geometric correction and image cropping, and microwave radar data are subjected to radiometric calibration, multi-viewing, filtering, terrain radiometric correction, geocoding and geometric correction; The second processing unit is used to extract and calculate post-fire site environmental information and forest biophysical parameters. The forest biophysical parameters are obtained by polarization decomposition of airborne L-band fully polarized synthetic aperture radar data.

7. The post-fire forest loss assessment device based on unmanned aerial vehicle and microwave radar as claimed in claim 6, characterized in that: The processing module takes the PolSAR data and UAV remote sensing images of the predetermined area as independent variables, and the measured biomass of the predetermined area as the dependent variable, and uses random forest to construct a forest loss assessment model for the predetermined area.

8. The post-fire forest loss assessment device based on unmanned aerial vehicle and microwave radar as claimed in claim 7, characterized in that: The evaluation modules include: The first calculation unit is used to input the processed PolSAR data and the UAV remote sensing data of the target area into the optimal forest loss assessment model to calculate the forest loss amount of the predetermined area; The second calculation unit is used to perform graded calculations according to the forest fire intensity and the burned area, and obtain the changes in forest loss under different forest fire intensities and burned areas.

Images