A method, device, equipment and storage medium for assessing forestry loss value

Abstract

The present invention relates to the field of remote sensing technology, and discloses a method, device, equipment and storage medium for evaluating the value of forestry losses. The present invention combines a second time series image data set with a first fitting model to calculate a first difference image, which can intuitively and quantitatively show the changes in vegetation in cloudless areas of remote sensing images after a typhoon disaster relative to a non-disaster state. Furthermore, by training the model with a set of environmental influencing factors and the first difference image, it is possible to dig out the deep-seated relationship rules of the impact of typhoon disasters on forests hidden behind complex data. Furthermore, by using the constructed model to predict the difference in normalized vegetation index in the cloud-covered area to generate a second difference image, it is possible to improve the spatial representation information of the disaster situation in the entire typhoon-affected area. Finally, by combining the actual forestry loss value, the first difference image and the second difference image, a rapid evaluation of the forestry loss value in the entire typhoon-affected area can be achieved.

Description

Technical Field

[0001] The invention relates to the field of remote sensing technology, and in particular to a forestry loss value assessment method, device, equipment and storage medium. Background Art

[0002] Rapidly identifying the extent of forest damage in typhoon-affected areas and assessing the value of forestry losses caused by disasters are of great significance for post-typhoon loss verification, damage compensation, emergency deployment, etc. At present, optical satellite remote sensing data is an important data source and means for investigating and assessing forestry disaster losses in typhoon-affected areas. However, tropical regions have high frequency and long duration of cloud coverage, making it difficult to obtain high-quality optical remote sensing data with low cloud coverage, which poses a challenge to quickly assessing the extent of forestry damage in the entire disaster area. Summary of the invention

[0003] In view of this, the present invention provides a forestry loss value assessment method, device, equipment and storage medium to solve the problem that the cloud coverage frequency in tropical areas is high and the time is long, and it is difficult to obtain high-quality optical remote sensing data with low cloud coverage, which leads to the inability to quickly assess the extent of forestry damage in the entire disaster area.

[0004] In a first aspect, the present invention provides a method for assessing forestry loss value, the method comprising:

[0005] A set of environmental influencing factors, a first fitting model, a second time series image data set at a preset time after a typhoon disaster, and the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster are obtained. The set of environmental influencing factors is used to characterize the impact of typhoon disasters on forests. The first fitting model is used to characterize the periodic change of the normalized vegetation index in the cloud-free area of ​​the remote sensing image before the typhoon disaster over time; based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated; the set of environmental influencing factors and the first difference image are trained with a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image; the machine learning model is used to simulate and predict the difference of the normalized vegetation index in the cloud-covered area of ​​the remote sensing image after the typhoon disaster to obtain a second difference image; based on the actual forestry loss value, the first difference image and the second difference image, the forestry loss value in the typhoon-affected area is calculated.

[0006] The forestry loss value assessment method provided by the present invention calculates the first difference image by combining the second time series image data set with the first fitting model, which can intuitively and quantitatively show the changes in vegetation in the cloudless area of ​​the remote sensing image after the typhoon disaster relative to the unaffected state. Furthermore, by taking the set of environmental influencing factors and the first difference image as input and training with a preset machine learning algorithm, it is possible to dig out the deep-seated relationship rules of the impact of typhoon disasters on forests hidden behind the complex data and construct a corresponding machine learning model for the cloudless area of ​​the remote sensing image. Furthermore, the trained machine learning model can be used to reasonably infer the normalized vegetation index difference of the cloud-covered area to generate a second difference image. Furthermore, by combining the second difference image with the first difference image, the spatial representation of the disaster situation in the entire typhoon-affected area can be improved. Finally, by combining the actual forestry loss value with the first difference image reflecting the degree of vegetation damage in the cloud-free area and the second difference image of the degree of vegetation damage in the cloud-covered area obtained through simulation prediction, a rapid assessment of the forestry loss value in the entire typhoon-affected area can be achieved, which is beneficial to provide a basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., and improve the efficiency of post-disaster rescue and relief work.

[0007] In an optional implementation, obtaining a first fitting model includes:

[0008] The first time series image dataset before the typhoon disaster is obtained; based on the first time series image dataset, a first fitting model is constructed through iterative weighted least squares processing.

[0009] The forestry loss value assessment method provided by the present invention utilizes the first time series image data set before the typhoon disaster, and constructs a first fitting model through iterative weighted least squares processing. The first fitting model can more accurately fit the periodic change of the normalized vegetation index in the cloud-free area over time, thereby improving the fitting accuracy of the model to the data, thereby better capturing the changing trend of the normalized vegetation index, and providing a reliable model basis for the subsequent accurate prediction of the simulated normalized vegetation index value in the cloud-free area after the typhoon disaster, thereby providing more accurate data support for calculating the difference image and assessing the forestry loss value.

[0010] In an optional implementation, obtaining the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster includes:

[0011] In the cloudless area of ​​the remote sensing image after the typhoon, multiple standard sites are determined according to the forest type; based on the multiple standard sites, the forest resource damage caused by the typhoon is investigated to obtain the damaged forest stock, the number of damaged trees and the degree of damage; based on the damaged forest stock, the number of damaged trees and the degree of damage, the actual forestry loss value in the cloudless area of ​​the remote sensing image after the typhoon is calculated.

[0012] The forestry loss value assessment method provided by the present invention conducts surveys based on multiple standards, can obtain accurate actual loss data in cloudless areas, and provide reliable dependent variable data for constructing a fitting model of the difference between forestry loss value and normalized vegetation index, so that the model can more accurately reflect the relationship between forestry loss and vegetation index changes.

[0013] In an optional implementation, based on the second time series image data set, after being processed by the first fitting model, a first difference image of a cloud-free area of ​​a remote sensing image after a typhoon disaster is calculated, including:

[0014] The first fitting model is used to predict the simulated normalized vegetation index at a preset time after the typhoon disaster to obtain the simulated normalized vegetation index value; based on the simulated normalized vegetation index value and the second time series image data set, the first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated.

[0015] The forestry loss value assessment method provided by the present invention uses the first fitting model to predict the simulated normalized vegetation index at a preset time after the typhoon disaster, which can intuitively reflect the changes in vegetation in the cloudless area before and after the typhoon disaster. Furthermore, by comparing the difference between the simulated normalized vegetation index value and the second time series image data set actually observed, the degree of vegetation damage in the cloudless area caused by the typhoon disaster can be accurately identified, providing detailed spatial distribution information for further analysis of the forestry loss in the entire disaster area.

[0016] In an optional implementation, based on the simulated normalized vegetation index value and the second time series image data set, calculating a first difference image of a cloud-free area of ​​a remote sensing image after a typhoon disaster includes:

[0017] The second time series image dataset is processed by a mask method to obtain a third time series image dataset of cloud-free areas of remote sensing images after the typhoon disaster; the first difference image is calculated using the third time series image dataset and the simulated normalized vegetation index value.

[0018] The forestry loss value assessment method provided by the present invention can accurately exclude cloud-covered areas through mask processing, and only retain the image information of cloud-free areas, thereby generating a third time series image data set that more accurately reflects the actual disaster situation in the cloud-free areas, thereby improving data quality and reliability. Furthermore, the first difference image is calculated using the third time series image data set and the simulated normalized vegetation index value, which can intuitively and quantitatively show the difference in the degree of vegetation affected by the typhoon at various locations in the cloud-free area.

[0019] In an optional implementation, the forestry loss value in the typhoon-affected area is calculated based on the real forestry loss value, the first difference image and the second difference image, including:

[0020] A second fitting model is constructed based on the actual forestry loss value and the first difference image; a target difference image is determined based on the first difference image and the second difference image; and the target difference image is input into the second fitting model to obtain the forestry loss value in the typhoon-affected area.

[0021] The forestry loss value assessment method provided by the present invention combines the actual forestry loss value with the first difference image, can dig out the intrinsic quantitative relationship between the forestry loss value and the normalized vegetation index difference and construct a corresponding second fitting model. Furthermore, the first difference image and the second difference image are integrated and the target difference image is determined, so that the vegetation damage difference information of the entire typhoon-affected area (covering cloudless areas and cloud-covered areas) can be unified into one image. The degree of deviation of vegetation in different areas (regardless of whether there is cloud cover) from the normal growth state after being damaged by the typhoon is fully presented. Finally, the target difference image that integrates the vegetation damage difference information of the entire typhoon-affected area is input into the second fitting model, which can realize the quantitative assessment of the forestry loss value in the entire typhoon-affected area. Therefore, by implementing the present invention, full use is made of the known actual loss value of cloud-free areas and the difference image of cloud-covered areas obtained through model prediction, making the assessment results more scientific and accurate, providing a reliable basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., helping to improve the efficiency of post-disaster rescue and relief work, and better meeting the emergency needs of rapid assessment of forestry loss value after typhoon disasters in tropical cloudy areas.

[0022] In a second aspect, the present invention provides a forestry loss value assessment device, the device comprising:

[0023] The acquisition module is used to acquire an environmental influencing factor set, a first fitting model, a second time series image data set at a preset time after a typhoon disaster, and the actual forestry loss value of a cloud-free area in a remote sensing image after a typhoon disaster. The environmental influencing factor set is used to characterize the impact of typhoon disasters on forests, and the first fitting model is used to characterize the periodic change of the normalized vegetation index in the cloud-free area of ​​the remote sensing image before the typhoon disaster over time; the processing module is used to calculate a first difference image of the cloud-free area in the remote sensing image after the typhoon disaster based on the second time series image data set after processing by the first fitting model; the training module is used to train the environmental influencing factor set and the first difference image through a preset machine learning algorithm to obtain a machine learning model of the cloud-free area in the remote sensing image; the prediction module is used to simulate and predict the normalized vegetation index difference in the cloud-covered area of ​​the remote sensing image after the typhoon disaster using the machine learning model to obtain a second difference image; the calculation module is used to calculate the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image, and the second difference image.

[0024] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the forestry loss value assessment method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0025] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the forestry loss value assessment method of the first aspect or any corresponding embodiment thereof.

[0026] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the forestry loss value assessment method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 is a flow chart of a method for assessing forestry loss value according to an embodiment of the present invention;

[0029] Figure 2 is a flow chart of another forestry loss value assessment method according to an embodiment of the present invention;

[0030] Figure 3 is a flow chart of another forestry loss value assessment method according to an embodiment of the present invention;

[0031] Figure 4 is a structural block diagram of a forestry loss value assessment device according to an embodiment of the present invention;

[0032] Figure 5 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0034] An embodiment of the present invention provides a method for assessing the value of forestry losses, which combines a second time series image data set with a first fitting model to calculate a first difference image, constructs a machine learning model with a set of environmental influencing factors and the first difference image training, and then uses the model to infer and generate a second difference image, and finally combines the actual forestry loss value, the first difference image and the second difference image to achieve the effect of quickly assessing the value of forestry losses in the entire typhoon-affected area.

[0035] According to an embodiment of the present invention, an embodiment of a forestry loss value assessment method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] In this embodiment, a forestry loss value assessment method is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 1 is a flow chart of a method for assessing forestry loss value according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0037] Step S101, obtaining a set of environmental influencing factors, a first fitting model, a second time series image data set at a preset time after the typhoon disaster, and the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster.

[0038] Among them, the set of environmental influencing factors is used to characterize the impact of typhoon disasters on forests, which may include wind speed, rainfall, terrain, canopy height, pre-disaster normalized difference vegetation index (NDVI), etc.

[0039] Furthermore, the first fitting model is used to characterize the periodic changes of the normalized vegetation index in the cloud-free area of ​​the remote sensing image before the typhoon disaster. Through the first fitting model, the normal growth and change patterns of the vegetation in the area at different time stages when not disturbed by the typhoon can be clearly presented, such as knowing the NDVI fluctuation trend of vegetation in different seasons, which can provide a reliable reference standard for comparing the changes in vegetation index after the disaster.

[0040] Furthermore, the second time series image dataset at a preset time after the typhoon disaster represents a set of actual image data about the disaster-stricken area collected within a preset specific time period after the typhoon, which can intuitively reflect the actual vegetation state change process in the disaster-stricken area after the typhoon disaster occurs.

[0041] Furthermore, the actual forestry loss value in the cloudless areas of remote sensing images after the typhoon is represented by the actual quantitative value of forestry loss obtained by conducting field surveys on the cloudless areas in remote sensing images after the typhoon, which can accurately measure the losses caused to forestry by the typhoon in the cloudless areas.

[0042] Step S102, based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated.

[0043] Among them, the first difference image can visually show the changes in vegetation in the cloudless area of ​​the remote sensing image after the typhoon compared with the unaffected state. At the same time, the difference value of each pixel in the first difference image (corresponding to a certain area on the ground) reflects the degree to which the vegetation in this area deviates from the normal growth state due to the influence of the typhoon.

[0044] For example, if the difference of a certain pixel is positive, it means that the actual situation of vegetation in the area after the disaster is better than that predicted according to normal periodic changes. It may be that there are local protection factors that make the vegetation less affected by the typhoon. If the difference is negative and the absolute value is large, it means that the vegetation in the area has been severely damaged, such as trees may fall down, branches and leaves fall off in large numbers, etc., resulting in the actual vegetation coverage, growth vitality and other aspects being far less than the expected state before the disaster.

[0045] Specifically, the first fitting model can be used to predict the vegetation index state that the cloudless area should present at the corresponding time after the disaster in the absence of the typhoon disaster. Furthermore, by combining with the second time series image dataset, the first difference image of the cloudless area of ​​the remote sensing image after the typhoon disaster can be calculated.

[0046] In an optional implementation, the first fitting model is first used to predict the simulated normalized difference vegetation index of a cloud-free area at a preset time point after the typhoon disaster, to obtain a corresponding simulated NDVI value.

[0047] Secondly, for the second time series image dataset at a preset time after the typhoon disaster, the actual normalized vegetation index value of the cloud-free area at the same preset time point is extracted.

[0048] Among them, professional remote sensing image processing software can be used to perform corresponding band analysis, calculation and other operations on the image data (according to the NDVI calculation formula " ",in Represents the near-infrared band of remote sensing images. Represents the red band of the remote sensing image), thereby obtaining the actual NDVI value of the cloud-free area of ​​the remote sensing image.

[0049] Finally, the difference between the actual NDVI value of the cloud-free area at each corresponding position (for example, the actual ground area corresponding to pixels) and the simulated NDVI value predicted by the first fitting model is calculated, that is, the simulated NDVI value is subtracted from the actual NDVI value to obtain the difference data for each position, and then these difference data are sorted according to the corresponding spatial position relationship to form a new image, that is, the first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster.

[0050] Step S103, training the set of environmental influencing factors and the first difference image through a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image.

[0051] Among them, the preset machine learning algorithm can be a linear regression algorithm, a decision tree algorithm, a random forest algorithm, a support vector machine (SVM) algorithm, a gradient boosting tree algorithm, a neural network algorithm, etc.

[0052] Specifically, the data in the first difference image is used as the dependent variable, and the influencing factors in the environmental influencing factor set are used as independent variables, and are input into a preset machine learning algorithm for training to obtain a machine learning model for the cloud-free area of ​​the remote sensing image.

[0053] Among them, during the training process, the algorithm can automatically find the potential relationship pattern between environmental influencing factors and NDVI differences, and by adjusting the parameters of the model, enable the model to predict the corresponding NDVI differences as accurately as possible based on the input environmental influencing factors.

[0054] Furthermore, the accuracy of the trained model can be verified and evaluated by the following methods:

[0055] (1) Divide the data into training and test sets: Divide the prepared data into training and test sets, for example, according to a certain ratio (such as 80% of the data as training set and 20% of the data as test set) to ensure that the data distribution of the training and test sets is representative.

[0056] (2) Evaluation metric selection: According to the model type and task objectives, select appropriate evaluation metrics, such as mean square error (MSE), mean absolute error (MAE), coefficient of determination (R²), etc. For regression tasks, MSE and MAE can measure the error between the predicted value and the true value, and R² can reflect the degree to which the model explains the variance of the dependent variable.

[0057] (3) Use the test set to evaluate the trained model. Input the independent variable data in the test set into the model to obtain the predicted NDVI difference, which is then compared with the actual NDVI difference in the test set to calculate the value of the evaluation index. Further, the performance of the model is judged based on the value of the evaluation index. If the evaluation index does not meet expectations, consider adjusting the algorithm parameters or replacing the algorithm and retraining until satisfactory model accuracy is obtained and the corresponding machine learning model for cloud-free areas of remote sensing images is determined.

[0058] Step S104, using a machine learning model to simulate and predict the difference in normalized vegetation index of the cloud coverage area of ​​the remote sensing image after the typhoon disaster, to obtain a second difference image.

[0059] In tropical regions, cloud cover is frequent and lasts for a long time, which makes it impossible to directly obtain the actual vegetation information in the cloud-covered area from remote sensing images, and it is difficult to calculate the actual normalized vegetation index difference in the area.

[0060] Specifically, according to the description of step S103, the machine learning model has learned the relationship between environmental influencing factors and NDVI differences. Therefore, when the prepared environmental influencing factor data of the cloud coverage area is input into the trained machine learning model, the machine learning model can make predictions based on the learned patterns.

[0061] For example, if a decision tree model is used, it can be based on the input environmental influencing factor data, follow the decision rules of the tree, start from the root node, make judgments based on the characteristic conditions of each node, gradually reach the leaf nodes, and finally output the predicted NDVI difference; if it is a neural network model, the data will be transmitted between multiple layers of the network, and after nonlinear transformation of the activation function and weighted summation of the weights, the predicted NDVI difference will be obtained; other models such as random forests, support vector machines, etc. can also process the input data according to their respective algorithm principles to obtain the predicted NDVI difference.

[0062] Furthermore, since the input environmental impact factor data has spatial location information, each data point corresponds to a geographical unit (such as a grid unit) in the cloud coverage area, and the NDVI difference predicted by the model will also have spatial location information accordingly.

[0063] Finally, the predicted NDVI difference values ​​can be arranged and organized according to their corresponding spatial position relationships to generate a second difference image.

[0064] Furthermore, the grayscale value or color value of each pixel (representing a certain geographical range) in the second difference image can represent the magnitude of the predicted NDVI difference at that location.

[0065] Step S105, calculating the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image and the second difference image.

[0066] Specifically, by combining the actual forestry loss value with the first difference image reflecting the degree of vegetation damage in the cloud-free area and the second difference image of the degree of vegetation damage in the cloud-covered area obtained through simulation prediction, a rapid assessment of the forestry loss value in the entire typhoon-stricken area can be achieved, which is beneficial to provide a basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., and improve the efficiency of post-disaster rescue and relief work.

[0067] In an optional embodiment, first, the NDVI difference data of the cloud-free area obtained from the first difference image and the NDVI difference data of the cloud-covered area obtained from the second difference image are merged and integrated according to a unified geographic coordinate system to form a complete comprehensive difference data set covering the entire typhoon-stricken area (including cloud-free areas and cloud-covered areas).

[0068] Secondly, the real forestry loss value and the comprehensive difference data set were analyzed to observe the correlation characteristics between the two in terms of numerical changes. For example, by drawing scatter plots, we can see whether the real forestry loss value shows a corresponding increase or decrease trend as the NDVI difference increases or decreases, whether it is a linear relationship or a nonlinear relationship, etc., to preliminarily explore the inherent quantitative connection pattern between them.

[0069] Furthermore, according to the correlation characteristics obtained from the analysis, a suitable mathematical method can be selected to construct a model for calculating the value of forestry losses. For example, if it is found that the two show a relatively obvious linear relationship, a linear regression method can be used to construct the model.

[0070] Finally, the NDVI difference corresponding to each pixel in the integrated regional comprehensive difference data set can be substituted into the constructed calculation model in turn to obtain the predicted value of forestry loss corresponding to each pixel position. Then, according to the scope of the entire disaster-stricken area, the predicted values ​​of forestry loss corresponding to all pixels are summed or other appropriate statistical summary methods are used to finally obtain the total value of forestry loss in the entire typhoon-affected area, and complete the calculation and evaluation of the forestry loss value of the entire disaster-stricken area.

[0071] The forestry loss value assessment method provided in this embodiment calculates the first difference image by combining the second time series image data set with the first fitting model, which can intuitively and quantitatively show the changes in vegetation in the cloudless area of ​​the remote sensing image after the typhoon disaster relative to the unaffected state. Furthermore, by taking the set of environmental influencing factors and the first difference image as input and training with the preset machine learning algorithm, it is possible to dig out the deep-seated relationship between the impact of typhoon disasters on forests hidden behind the complex data and construct a corresponding machine learning model for the cloudless area of ​​the remote sensing image. Furthermore, the trained machine learning model can be used to reasonably infer the difference in the normalized vegetation index of the cloud-covered area to generate a second difference image. Furthermore, by combining the second difference image with the first difference image, the spatial representation of the disaster situation in the entire typhoon-affected area can be improved. Finally, by combining the actual forestry loss value with the first difference image reflecting the degree of vegetation damage in the cloud-free area and the second difference image of the degree of vegetation damage in the cloud-covered area obtained through simulation prediction, a rapid assessment of the forestry loss value in the entire typhoon-affected area can be achieved, which is beneficial to provide a basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., and improve the efficiency of post-disaster rescue and relief work.

[0072] In this embodiment, a forestry loss value assessment method is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 2 is a flow chart of a method for assessing forestry loss value according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0073] Step S201, obtaining a set of environmental influencing factors, a first fitting model, a second time series image dataset of a preset time after the typhoon disaster, and the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster.

[0074] In an optional implementation, the first fitting model may be obtained by following the steps below:

[0075] Step a1, obtaining the first time series image dataset before the typhoon disaster.

[0076] Step a2: constructing a first fitting model based on the first time series image data set through iterative weighted least squares processing.

[0077] Among them, iterative weighted least squares method represents an improved least squares method, which can deal with problems such as heteroscedasticity in the data by assigning different weights to different data points, so that the model can better fit the data.

[0078] Specifically, there may be some cloud-covered areas in the first time series image data obtained, and clouds will block ground vegetation information, affecting the accurate calculation and model construction of vegetation index. Therefore, some cloud detection algorithms can be used, such as cloud detection methods based on spectral features, texture features, etc., to mark and remove the cloud-covered parts of the image data, leaving only the cloud-free image parts.

[0079] Furthermore, for the selected cloud-free image areas, the NDVI value of each cloud-free area can be calculated pixel by pixel (corresponding to a certain geographical unit on the ground) according to the NDVI calculation formula using the corresponding band information in the image data, thereby obtaining the NDVI data of the cloud-free image area at each time phase. Through this operation, the original image data can be converted into NDVI data that can reflect the growth status and coverage of vegetation, which is convenient for analyzing the changes of vegetation over time.

[0080] Furthermore, the NDVI data of the cloud-free image area at different phases were used as the dependent variable, and the corresponding time variable (such as time series in units of year, month, day, etc.) was used as the independent variable. The iterative weighted least squares method was used to fit the model and construct a fitting model for the periodic changes of the normalized vegetation index NDVI of the cloud-free image before the disaster, which is the first fitting model.

[0081] In an optional implementation, the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster can be obtained by the following steps:

[0082] Step b1, in the cloud-free area of ​​the remote sensing image after the typhoon disaster, determine multiple standard sites according to the forest type.

[0083] Step b2, based on multiple standard areas, investigate the damage to forest resources caused by the typhoon to obtain the damaged tree stock, the number of damaged trees and the degree of damage.

[0084] Step b3, based on the damaged tree stock, the number of damaged trees and the degree of damage, calculate the actual forestry loss value in the cloudless area of ​​the remote sensing image after the typhoon disaster.

[0085] Specifically, standard plots may be set up for each forest type in accordance with established specifications (generally 20 m × 20 m) in areas affected by the typhoon and where there is no cloud coverage in remote sensing images.

[0086] Among them, in order to ensure the accuracy and representativeness of the survey results, no less than 10 standard plots are set up for each forest type.

[0087] For example, if there are two main forest types, coniferous forest and broad-leaved forest, in a typhoon-stricken area and the cloud-free area of ​​the remote sensing image, then for the coniferous forest type, at least 10 plots of 20 meters × 20 meters can be evenly selected within its distribution range as standard sites; the same operation can be performed for the broad-leaved forest. The selection of these standard sites should cover the different growth environments, topographical conditions and other conditions of the forest type as much as possible, so that they can fully represent the overall situation of this type of forest in the entire cloud-free area.

[0088] Furthermore, the damage to forest resources caused by typhoons can be investigated on site within the standard area, and the damaged tree stock, number of damaged trees and degree of damage can be recorded.

[0089] Finally, the value of forestry loss in each standard area can be evaluated through "timber loss (yuan) = damaged forest stock (cubic meter) × number of trees × degree of damage (%) × forest market price (yuan / cubic meter)" and ultimately the actual forestry loss value of the cloud-free area in the remote sensing image after the typhoon disaster can be obtained.

[0090] Step S202, based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated.

[0091] Specifically, the above step S202 includes:

[0092] Step S2021, using the first fitting model to predict the simulated normalized difference vegetation index at a preset time after the typhoon disaster, to obtain a simulated normalized difference vegetation index value.

[0093] The preset time is the same date as the preset time in step S101.

[0094] Specifically, according to the description of step S101, the first fitting model is a periodic variation fitting model of the cloud-free image normalized difference vegetation index NDVI constructed by using time series image data and iterative weighted least squares algorithm before the disaster.

[0095] Therefore, the preset time can be substituted into the first fitting model, and the simulated normalized vegetation index value corresponding to the preset time can be calculated according to the periodic variation law of the vegetation index reflected in the model.

[0096] Step S2022, based on the simulated normalized difference vegetation index value and the second time series image data set, calculate a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster.

[0097] Specifically, by comparing the differences between the simulated normalized vegetation index values ​​and the actually observed second time series image dataset, the extent of vegetation damage caused by typhoon disasters in cloudless areas can be accurately identified, providing detailed spatial distribution information for further analysis of forestry losses in the entire disaster-stricken area.

[0098] In some optional implementations, the above step S2022 includes:

[0099] Step c1, processing the second time series image dataset by a masking method to obtain a third time series image dataset of cloud-free areas of remote sensing images after the typhoon disaster.

[0100] Step c2, calculating a first difference image using the third time series image data set and the simulated normalized difference vegetation index value.

[0101] Specifically, the masking method is usually based on a cloud detection algorithm to identify cloud-covered areas in the image, and then remove these cloud-covered areas (or mark them as invalid values), retaining only the image data of the cloud-free areas, thereby obtaining a third time series image dataset of cloud-free areas in remote sensing images after the typhoon disaster.

[0102] Furthermore, the actual NDVI value of the cloud-free area at each time point can be calculated using the third time series image dataset according to the NDVI calculation formula.

[0103] Furthermore, the difference between the actual NDVI value of the cloud-free area and the simulated normalized vegetation index value is calculated pixel by pixel (corresponding to a certain area on the ground).

[0104] For example, for a pixel in a cloud-free area, the true NDVI value is 0.6 and the simulated NDVI value is 0.8, then the difference value of the pixel is 0.6-0.8=-0.2.

[0105] Finally, after performing difference calculation on all pixels in the cloud-free area, the calculated differences can be arranged according to the positions of the pixels to form a first difference image.

[0106] Step S203: The environmental influencing factor set and the first difference image are trained with a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0107] Step S204, using a machine learning model to simulate and predict the difference in the normalized vegetation index of the cloud-covered area of ​​the remote sensing image after the typhoon, to obtain a second difference image. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0108] Step S205, based on the actual forestry loss value, the first difference image and the second difference image, calculate the forestry loss value in the typhoon-affected area. Figure 1 Step S105 of the illustrated embodiment will not be described in detail here.

[0109] The forestry loss value assessment method provided in this embodiment uses the first time series image data set before the typhoon disaster, and constructs the first fitting model through iterative weighted least squares processing, which can more accurately fit the periodic change of the normalized vegetation index in the cloud-free area over time, improve the fitting accuracy of the model to the data, and thus better capture the change trend of the normalized vegetation index. At the same time, based on multiple standards, the survey can obtain accurate actual loss data in the cloud-free area. Further, the first fitting model is used to predict the simulated normalized vegetation index at a preset time after the typhoon disaster, which can intuitively reflect the changes in vegetation in the cloud-free area before and after the typhoon disaster. Further, through the mask method processing, the cloud-covered area can be accurately excluded, and only the image information of the cloud-free area is retained, thereby generating a third time series image data set that more accurately reflects the actual disaster situation in the cloud-free area, and improving the data quality and reliability. Further, the third time series image data set and the simulated normalized vegetation index value are used to calculate the first difference image, which can intuitively and quantitatively show the difference in the degree of vegetation affected by the typhoon at various locations in the cloud-free area.

[0110] In this embodiment, a forestry loss value assessment method is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 3 is a flow chart of a method for assessing forestry loss value according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0111] Step S301, obtain a set of environmental influencing factors, a first fitting model, a second time series image dataset of a preset time after the typhoon disaster, and the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0112] Step S302, based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated. Figure 2 Step S202 of the illustrated embodiment will not be described in detail here.

[0113] Step S303: The environmental influencing factor set and the first difference image are trained with a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0114] Step S304: Use a machine learning model to simulate and predict the difference in the normalized vegetation index of the cloud-covered area of ​​the remote sensing image after the typhoon, and obtain a second difference image. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0115] Step S305, calculating the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image and the second difference image.

[0116] Specifically, the above step S305 includes:

[0117] Step S3051, constructing a second fitting model based on the actual forestry loss value and the first difference image.

[0118] Specifically, the actual forestry loss value is taken as the dependent variable, and the NDVI difference data in the first difference image is taken as the independent variable, and a fitting model between the two is constructed using appropriate mathematical methods, such as linear regression, polynomial regression or other machine learning algorithms.

[0119] Step S3052: determine a target difference image based on the first difference image and the second difference image.

[0120] Specifically, the first difference image (NDVI difference in cloud-free areas) and the second difference image (NDVI difference in cloud-covered areas) can be merged according to the spatial coordinate correspondence of the images to obtain a target difference image covering the entire typhoon-affected area (including cloud-free areas and cloud-covered areas), and each pixel in the image has corresponding NDVI difference data.

[0121] Step S3053, inputting the target difference image into the second fitting model to obtain the forestry loss value in the typhoon-affected area.

[0122] Specifically, the NDVI difference data corresponding to each pixel in the target difference image may be input into the second fitting model as an independent variable.

[0123] Furthermore, the predicted value of forestry loss value corresponding to each pixel position can be calculated based on the functional relationship of the second fitting model.

[0124] Furthermore, the predicted values ​​of forestry loss values ​​corresponding to all areas covered by the entire target difference image (ie, the entire typhoon-affected area) may be summarized, for example, by summing up, to obtain the total value of forestry loss in the entire typhoon-affected area.

[0125] Furthermore, based on the target difference image and the calculated forestry loss value, the location of the forest disaster after the typhoon (determined by the corresponding ground position of the difference image pixel position), the type of disaster (which can be judged by combining the characteristics of the NDVI difference and related knowledge) and the loss value can be determined.

[0126] The forestry loss value assessment method provided in this embodiment combines the actual forestry loss value with the first difference image, and can dig out the intrinsic quantitative relationship between the forestry loss value and the normalized vegetation index difference and construct a corresponding second fitting model. Furthermore, by integrating the first difference image and the second difference image and determining the target difference image, the vegetation damage difference information of the entire typhoon-affected area (covering cloudless areas and cloud-covered areas) can be unified into one image. It fully presents the degree of deviation of vegetation in different areas (regardless of whether there is cloud cover or not) from the normal growth state after being damaged by the typhoon. Finally, the target difference image that integrates the vegetation damage difference information of the entire typhoon-affected area is input into the second fitting model, which can realize the quantitative assessment of the forestry loss value in the entire typhoon-affected area. Therefore, by implementing the present invention, full use is made of the known actual loss value of cloud-free areas and the difference image of cloud-covered areas obtained through model prediction, making the assessment results more scientific and accurate, providing a reliable basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., helping to improve the efficiency of post-disaster rescue and relief work, and better meeting the emergency needs of rapid assessment of forestry loss value after typhoon disasters in tropical cloudy areas.

[0127] In one example, a method for assessing the value of forestry losses after a typhoon disaster in a tropical cloudy area is provided, including:

[0128] Step 1: Using time series image data and iterative weighted least squares algorithm, a fitting model of the periodic variation of the normalized difference vegetation index (NDVI) of cloud-free images before the disaster was constructed. Based on this, the undisturbed simulated NDVI value on a certain day after the typhoon’s passing date was predicted;

[0129] Among them, the calculation formula of normalized difference vegetation index (NDVI) is:

[0130]

[0131] Where: Represents the near-infrared band of remote sensing images. Represents the red band of remote sensing imagery.

[0132] Step 2: After the disaster area actually suffered a typhoon disaster, obtain the remote sensing image of the same date in step 1, filter out the cloud-covered area with a mask, and calculate the true NDVI value of the cloud-free area of ​​the remote sensing image on that date;

[0133] Step 3: Calculate the difference image between the actual NDVI value of the cloud-free area in step 2 and the simulated NDVI value predicted in step 1;

[0134] Step 4: Select environmental factors that affect forests caused by typhoon disasters, including wind speed, rainfall, topography, canopy height, pre-disaster NDVI, etc., and process and analyze the data of environmental factors;

[0135] Step 5: Select a machine learning algorithm, take the NDVI difference as the dependent variable, and the environmental influencing factors in step 4 as the independent variable, train the machine learning model of the cloud-free area of ​​the remote sensing image in step 2, and perform model accuracy verification and evaluation;

[0136] Step 6: Use the above machine learning model to simulate and predict the NDVI difference of the cloud coverage area of ​​the remote sensing image after the typhoon disaster;

[0137] Step 7: In the typhoon-affected areas and in the areas without cloud coverage in remote sensing images, select standard plots according to the forest types. The general specifications are 20 meters × 20 meters, and no less than 10 standard plots are set for each forest type;

[0138] Step 8: Conduct an on-site survey of forest resource damage caused by the typhoon in the standard area, record the damaged tree stock, number of trees and degree of damage, and evaluate the forestry loss value in the standard area according to "wood loss (yuan) = damaged tree stock (cubic meters) × number of trees × degree of damage (%) × market price of trees (yuan / cubic meter)";

[0139] Step 9: Based on the forestry loss value data from the standard site survey and the corresponding NDVI difference data, a fitting model is constructed for the two;

[0140] Step 10: Apply the fitting model of step 9 to all typhoon-affected areas, and calculate the forestry loss value in the typhoon-affected areas based on the full-area NDVI difference image obtained from steps 1 to 6, including information such as the location of forest damage after the typhoon, the type of damage, and the loss value. The full-area NDVI difference image is composed of the NDVI difference image of the cloud-free area in step 3 and the NDVI difference image of the cloud-covered area in step 6.

[0141] The method for assessing the value of forestry losses after typhoon disasters in tropical cloudy areas provided in this example is aimed at the emergency demand for rapid assessment of the value of forestry losses after typhoon disasters in tropical cloudy areas. The cloudless areas of optical remote sensing images are used in combination with environmental influencing factor data to construct a prediction model for NDVI changes in cloudless areas of remote sensing images of disaster areas. The post-disaster NDVI difference image of the cloud-covered area is obtained by applying the prediction model, and then the forestry loss value of the standard land and the NDVI difference are fitted to quickly assess the value of forestry losses in the entire typhoon-affected area. The present invention combines satellite remote sensing data, typhoon disaster mechanism analysis, forestry standard land surveys and other methods to propose a rapid assessment method for the value of forestry losses after typhoon disasters in tropical cloudy areas, which is beneficial to provide a basis for post-disaster loss verification for local forestry management departments, emergency departments, etc., and improve the efficiency of post-disaster rescue and relief.

[0142] In this embodiment, a forestry loss value assessment device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0143] This embodiment provides a forestry loss value assessment device, such as Figure 4 As shown, the device comprises:

[0144] The acquisition module 401 is used to obtain a set of environmental influencing factors, a first fitting model, a second time series image data set at a preset time after the typhoon disaster, and the actual forestry loss value of the cloud-free area of ​​the remote sensing image after the typhoon disaster. The set of environmental influencing factors is used to characterize the impact of typhoon disasters on forests, and the first fitting model is used to characterize the periodic changes of the normalized vegetation index in the cloud-free area of ​​the remote sensing image before the typhoon disaster over time.

[0145] The processing module 402 is used to calculate a first difference image of a cloud-free area of ​​a remote sensing image after a typhoon disaster based on the second time series image data set after being processed by a first fitting model.

[0146] The training module 403 is used to train the set of environmental influencing factors and the first difference image through a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image.

[0147] The prediction module 404 is used to simulate and predict the normalized vegetation index difference of the cloud coverage area of ​​the remote sensing image after the typhoon disaster by using a machine learning model to obtain a second difference image.

[0148] The calculation module 405 is used to calculate the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image and the second difference image.

[0149] In some optional implementations, obtaining the first fitting model in the obtaining module 401 includes:

[0150] The first time series image dataset before the typhoon disaster is obtained; based on the first time series image dataset, a first fitting model is constructed through iterative weighted least squares processing.

[0151] In some optional implementations, the acquisition module 401 acquires the real forestry loss value of the cloudless area of ​​the remote sensing image after the typhoon disaster, including:

[0152] In the cloudless area of ​​the remote sensing image after the typhoon, multiple standard sites are determined according to the forest type; based on the multiple standard sites, the forest resource damage caused by the typhoon is investigated to obtain the damaged forest stock, the number of damaged trees and the degree of damage; based on the damaged forest stock, the number of damaged trees and the degree of damage, the actual forestry loss value in the cloudless area of ​​the remote sensing image after the typhoon is calculated.

[0153] In some optional implementations, the processing module 402 includes:

[0154] The prediction submodule is used to predict the simulated normalized vegetation index at a preset time after the typhoon disaster using the first fitting model to obtain a simulated normalized vegetation index value.

[0155] The calculation submodule is used to calculate the first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster based on the simulated normalized vegetation index value and the second time series image data set.

[0156] In some optional implementations, the calculation submodule includes:

[0157] The processing unit is used to process the second time series image data set by a mask method to obtain a third time series image data set of cloud-free areas of the remote sensing image after the typhoon disaster.

[0158] The calculation unit is used to calculate a first difference image using the third time series image data set and the simulated normalized vegetation index value.

[0159] In some optional implementations, the calculation module 405 includes:

[0160] A submodule is constructed to construct a second fitting model based on the actual forestry loss value and the first difference image.

[0161] The determination submodule is used to determine a target difference image based on the first difference image and the second difference image.

[0162] The input submodule is used to input the target difference image into the second fitting model to obtain the forestry loss value in the typhoon-affected area.

[0163] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0164] The forestry loss value assessment device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0165] The embodiment of the present invention also provides a computer device having the above Figure 4 The forestry loss value assessment device shown.

[0166] See also Figure 5 , Figure 5 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 5 As shown, the computer device comprises: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other, and can be installed on a common mainboard or installed in other ways as required. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.

[0167] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0168] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0169] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0170] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0171] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0172] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0173] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0174] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for assessing forestry loss value, characterized in that: The method comprises: Obtaining a set of environmental influencing factors, a first fitting model, a second time series image data set at a preset time after a typhoon disaster, and a real forestry loss value of a cloud-free area in a remote sensing image after a typhoon disaster, wherein the set of environmental influencing factors is used to characterize the impact of a typhoon disaster on forests, and the first fitting model is used to characterize the periodic change of a normalized vegetation index over time in a cloud-free area in a remote sensing image before a typhoon disaster; Based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated; The set of environmental influencing factors and the first difference image are trained with a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image; Using the machine learning model to simulate and predict the difference of the normalized vegetation index of the cloud coverage area of ​​the remote sensing image after the typhoon disaster, to obtain a second difference image; Based on the actual forestry loss value, the first difference image and the second difference image, the forestry loss value in the typhoon-affected area is calculated.

2. The method according to claim 1, characterized in that: Get the first fitted model, including: Obtain the first time series image dataset before the typhoon disaster; Based on the first time series image data set, the first fitting model is constructed through iterative weighted least squares processing.

3. The method according to claim 1, characterized in that Obtain the true forestry loss value of cloud-free areas in remote sensing images after typhoon disasters, including: Determine a plurality of standard sites according to forest types in the cloud-free area of ​​the remote sensing image after the typhoon disaster; Based on the multiple standard sites, the damage to forest resources caused by the typhoon is investigated to obtain the damaged forest stock, the number of damaged trees and the degree of damage; The actual forestry loss value of the cloudless area in the remote sensing image after the typhoon disaster is calculated based on the damaged tree stock, the number of damaged trees and the degree of damage.

4. The method according to claim 1, characterized in that: Based on the second time series image data set, after being processed by the first fitting model, a first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated, including: Using the first fitting model to predict the simulated normalized difference vegetation index at the preset time after the typhoon disaster, to obtain a simulated normalized difference vegetation index value; Based on the simulated normalized difference vegetation index value and the second time series image data set, the first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster is calculated.

5. The method according to claim 4, characterized in that Calculating the first difference image of the cloud-free area of ​​the remote sensing image after the typhoon disaster based on the simulated normalized difference vegetation index value and the second time series image data set includes: Processing the second time series image dataset by a mask method to obtain a third time series image dataset of the cloud-free area of ​​the remote sensing image after the typhoon disaster; The first difference image is calculated using the third time series image data set and the simulated normalized difference vegetation index value.

6. The method according to claim 1, characterized in that Calculating the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image, and the second difference image, including: constructing a second fitting model based on the actual forestry loss value and the first difference image; determining a target difference image based on the first difference image and the second difference image; The target difference image is input into the second fitting model to obtain the forestry loss value in the typhoon-affected area.

7. A forestry loss value assessment device, characterized in that: The device comprises: An acquisition module is used to acquire a set of environmental influencing factors, a first fitting model, a second time series image data set at a preset time after the typhoon disaster, and a real forestry loss value of a cloud-free area in a remote sensing image after the typhoon disaster, wherein the set of environmental influencing factors is used to characterize the impact of typhoon disasters on forests, and the first fitting model is used to characterize the periodic change of the normalized vegetation index over time in a cloud-free area in a remote sensing image before the typhoon disaster; A processing module, configured to calculate a first difference image of a cloud-free area of ​​the remote sensing image after the typhoon disaster based on the second time series image data set and processed by the first fitting model; A training module, used for training the set of environmental influencing factors and the first difference image through a preset machine learning algorithm to obtain a machine learning model of the cloud-free area of ​​the remote sensing image; A prediction module, used to simulate and predict the normalized vegetation index difference of the cloud coverage area of ​​the remote sensing image after the typhoon disaster using the machine learning model to obtain a second difference image; A calculation module is used to calculate the forestry loss value in the typhoon-affected area based on the actual forestry loss value, the first difference image and the second difference image.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the forestry loss value assessment method according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the forestry loss value assessment method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the forestry loss value assessment method according to any one of claims 1 to 6.

Images