Mountain debris flow disaster intelligent early warning system based on remote sensing network

Abstract

The invention discloses an intelligent early warning system for debris flow disasters in a mountainous area based on a remote sensing network, and particularly relates to the technical field of intelligent early warning of debris flow disasters. Calculating a brightness change coefficient and a color difference coefficient of the low-resolution remote sensing image at the same pixel point from the perspective of high time resolution, quantifying a brightness change fluctuation condition and a color difference condition of the low-resolution remote sensing image at the same pixel point, and calculating a first early warning coefficient; obtaining high-resolution remote sensing images at different moments, calculating boundary similarity, verifying debris flow disaster precursor signals captured in the low-resolution remote sensing images in the high-resolution remote sensing images according to the boundary similarity, and calculating a second early warning coefficient to construct a debris flow precursor comprehensive early warning model; and outputting a debris flow precursor comprehensive early warning index, and comprehensively evaluating the precursor risk of the debris flow disaster, thereby realizing efficient and accurate debris flow disaster precursor early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent early warning of debris flow disasters. More specifically, the present invention relates to an intelligent early warning system for mountain debris flow disasters based on a remote sensing network. Background Art

[0002] Debris flow disasters have the characteristics of suddenness and strong non-linearity, and usually occur in mountainous areas and other areas vulnerable to natural disasters. The precursor signals of debris flows, including changes in soil moisture, the appearance of ground cracks, and changes in vegetation conditions, often show weak local changes. Due to their complexity and locality, these precursor signals are usually difficult to capture timely and accurately through traditional ground observation means. Therefore, remote sensing technology, especially remote sensing image data, has become a key data source in the intelligent early warning system for mountain debris flow disasters.

[0003] High-resolution remote sensing images have high spatial accuracy and can display the minute changes of natural elements such as terrain, vegetation, and soil in detail. However, due to the low acquisition frequency, they often cannot provide real-time monitoring data, especially in the stage of frequently changing disaster precursors. Relatively speaking, although low-resolution remote sensing images can provide more frequent monitoring, their spatial resolution is not sufficient to capture the subtle changes of debris flow precursors. Therefore, how to effectively capture these weak disaster precursor signals in low-resolution images and verify them in high-resolution images has become an important technical challenge faced by the current debris flow disaster early warning system. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an intelligent early warning system for mountain debris flow disasters based on a remote sensing network to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent early warning system for mountain debris flow disasters based on a remote sensing network includes a low-resolution early warning module, a precursor area division module, a high-resolution early warning module, and a comprehensive early warning module;

[0007] The low-resolution early warning module is used to obtain low-resolution remote sensing images at different times, establish a time series of low-resolution remote sensing images in chronological order, and obtain a first early warning coefficient according to the time series of low-resolution remote sensing images;

[0008] The precursor area division module is used to compare the first early warning coefficient with a preset first early warning coefficient threshold, and mark the debris flow precursor areas in the low-resolution remote sensing images and the high-resolution remote sensing images;

[0009] A high-resolution early warning module, which is used to obtain high-resolution remote sensing images at different times, calculate the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in the high-resolution remote sensing image at each time, and obtain a second early warning coefficient according to the boundary similarity between the debris flow precursor area and the non-debris flow precursor area;

[0010] A comprehensive early warning module, which is used to construct a debris flow precursor comprehensive early warning model according to the first early warning coefficient and the second early warning coefficient, output a debris flow precursor comprehensive early warning index, and conduct a comprehensive early warning on the precursor risk of debris flow disasters.

[0011] In a preferred embodiment, obtaining the first early warning coefficient according to the low-resolution remote sensing image time series specifically includes the following:

[0012] Mark the low-resolution remote sensing image time series as Dt = {dtx i} = {dtx1, dtx2,..., dtx I}, where dtx i represents the low-resolution remote sensing image at the i-th moment, i ∈ {1, 2,..., I}, and I is a positive integer;

[0013] Calculate the brightness change coefficient and color difference coefficient of the low-resolution remote sensing images at different times at the same pixel point;

[0014] The acquisition logic of the brightness change coefficient is as follows:

[0015] Calculate the brightness value Ld of the pixel point of the low-resolution remote sensing image at the i-th moment i , and the calculation expression is as follows: Ld i = 0.299 * R i + 0.587 * G i + 0.114B i , where R i , G i , B i respectively represent the red, green, and blue components of the pixel point at the i-th moment;

[0016] Calculate the brightness change rate ΔLd between adjacent moments i , and the calculation expression is as follows: where Ld i+1 represents the brightness value of the pixel point at the i + 1-th moment, and ε is a very small positive value used to avoid the denominator being zero;

[0017] Calculate the average value μLd of the brightness change rate, and the expression is as follows:

[0018] Calculate the standard deviation σLd of the brightness change rate, and the calculation expression is as follows:

[0019] Calculate the brightness change coefficient Ldx, and the calculation expression is as follows:

[0020] The acquisition logic of the color difference coefficient is as follows:

[0021] Mark the RGB value combination of the pixel point as a color vector

[0022] Calculate the color difference value ΔSc between adjacent moments i , and the calculation expression is as follows: where represents the color vector at the (i + 1)-th moment, represents the color vector at the i-th moment, represents the Euclidean norm of the color vector at the i-th moment;

[0023] Calculate the average value μSc of the color difference value, and the expression is as follows:

[0024] Calculate the standard deviation σSc of the color difference value, and the calculation expression is as follows:

[0025] Calculate the color difference coefficient Scy, and the calculation expression is as follows:

[0026] Calculate the first warning coefficient YJ1 according to the brightness change coefficient and the color difference coefficient, and the calculation formula is as follows:

[0027] YJ1 = α * Ldx + β * Scy, where α and β respectively represent the preset proportionality coefficients of the brightness change coefficient and the color difference coefficient, and both α and β are greater than 0.

[0028] In a preferred embodiment, compare the first warning coefficient with the preset first warning coefficient threshold, and mark the debris flow precursor area in the low-resolution remote sensing image and the high-resolution remote sensing image, specifically as follows:

[0029] Traverse the pixel points in the low-resolution remote sensing image, compare the first warning coefficient of each pixel point with the preset first warning coefficient threshold. If the first warning coefficient is greater than the first warning coefficient threshold, mark the corresponding pixel point as a debris flow precursor pixel point, mark the area represented by all debris flow precursor pixel points as the debris flow precursor area, and at the same time mark the same area in the high-resolution remote sensing image as the debris flow precursor area; if the first warning coefficient is less than or equal to the first warning coefficient threshold, mark the corresponding pixel point as a non-debris flow precursor pixel point, mark the area represented by all non-debris flow precursor pixel points as the non-debris flow precursor area, and at the same time mark the same area in the high-resolution remote sensing image as the non-debris flow precursor area.

[0030] In a preferred embodiment, high-resolution remote sensing images at different times are obtained, and the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in each high-resolution remote sensing image is calculated as follows:

[0031] Take the boundary line between the debris flow precursor area and the non-debris flow precursor area as the boundary line, mark the pixel points on both sides of the boundary line as boundary pixel points. If the boundary pixel point belongs to the debris flow precursor area, add the boundary point to the set of boundary pixel points of the debris flow precursor area; if the boundary pixel point belongs to the non-debris flow precursor area, add the boundary point to the set of boundary pixel points of the non-debris flow precursor area.

[0032] Represent the set of boundary pixel points of the debris flow precursor area and the set of boundary pixel points of the non-debris flow precursor area as boundary pixel point vectors respectively: Where represents the boundary pixel point vector of the debris flow precursor area, Xsn j represents the boundary pixel points in the set of boundary pixel points of the debris flow precursor area, j ∈ {1, 2,..., J}, and J is a positive integer; Where represents the boundary pixel point vector of the non-debris flow precursor area, Xsf j represents the boundary pixel points in the set of boundary pixel points of the non-debris flow precursor area, j ∈ {1, 2,..., J}, and J is a positive integer;

[0033] Calculate the boundary similarity BJX according to the boundary pixel point vector of the debris flow precursor area and the boundary pixel point vector of the non-debris flow precursor area. The calculation expression is as follows: Where represents the dot product operation of the boundary pixel point vector of the debris flow precursor area and the boundary pixel point vector of the non-debris flow precursor area, represents the Euclidean norm of the boundary pixel point vector of the debris flow precursor area, Represents the Euclidean norm of the vector of boundary pixel points in the non-debris flow precursor area.

[0034] In a preferred embodiment, the boundary similarity is compared with a preset boundary similarity threshold. If the boundary similarity is greater than the boundary similarity threshold, the high-resolution remote sensing image at this moment is marked as a non-debris flow precursor image; if the boundary similarity is less than or equal to the boundary similarity threshold, the high-resolution remote sensing image at this moment is marked as a high-resolution debris flow precursor image.

[0035] Calculate the second warning coefficient YJ2 based on the high-resolution debris flow precursor image and the non-debris flow precursor image. The calculation expression is as follows: Where TX1 is the number of high-resolution debris flow precursor images, and TX2 is the number of non-debris flow precursor images.

[0036] In a preferred embodiment, a comprehensive debris flow precursor warning model is constructed based on the first warning coefficient and the second warning coefficient, and the comprehensive debris flow precursor warning index Nsl is output. The formula on which the model is based is as follows In the formula, T1 represents the acquisition time period of the low-resolution remote sensing image, T2 represents the acquisition time period of the high-resolution remote sensing image, YJ1 m Represents the first warning coefficient of the m-th pixel point in the low-resolution remote sensing image, YJ2 represents the second warning coefficient, a1 and a2 respectively represent the preset proportionality coefficients of the first warning coefficient and the second warning coefficient, and both a1 and a2 are greater than 0.

[0037] In a preferred embodiment, the comprehensive debris flow precursor warning index is compared with a preset comprehensive debris flow precursor warning index threshold to conduct a comprehensive warning of the precursor risk of debris flow disasters, specifically as follows:

[0038] If the comprehensive debris flow precursor warning index is greater than the comprehensive debris flow precursor warning index threshold, a comprehensive debris flow precursor warning signal is generated; if the comprehensive debris flow precursor warning index is less than or equal to the comprehensive debris flow precursor warning index threshold, there is no need to generate a comprehensive debris flow precursor warning signal.

[0039] The technical effects and advantages of the present invention:

[0040] The present invention obtains low-resolution remote sensing images at different times, analyzes the time series of low-resolution remote sensing images, calculates the brightness change coefficient and color difference coefficient of the low-resolution remote sensing images at the same pixel points from the perspective of high time resolution, quantifies the brightness change fluctuation and color difference of the low-resolution remote sensing images at the same pixel points, and calculates the first warning coefficient to provide an early signal for the precursor risk of debris flow disasters; and marks the debris flow precursor areas in the low-resolution remote sensing images and high-resolution remote sensing images according to the first warning coefficient, obtains high-resolution remote sensing images at different times, calculates the boundary similarity between the debris flow precursor areas and non-debris flow precursor areas in the high-resolution remote sensing images at each time, verifies the debris flow disaster precursor signals captured in the low-resolution remote sensing images in the high-resolution remote sensing images according to the boundary similarity, and simultaneously calculates the second warning coefficient. Combining the first warning coefficient, the acquisition time period of the remote sensing images, a debris flow precursor comprehensive warning model is constructed, and a debris flow precursor comprehensive warning index is output to comprehensively evaluate the precursor risk of debris flow disasters, thereby realizing efficient and accurate early warning of debris flow disaster precursors, improving the prediction and early warning capabilities of mountain debris flow disasters, and being of great significance for reducing disaster losses and ensuring the safety of people's lives and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings;

[0042] Figure 1 It is a flowchart of the system according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0044] Embodiment: The present invention provides a Figure 1 shown intelligent early warning system for mountain debris flow disasters based on a remote sensing network, including a low-resolution warning module, a precursor area division module, a high-resolution warning module, and a comprehensive warning module;

[0045] The low-resolution warning module is used to obtain low-resolution remote sensing images at different times, establish a time series of low-resolution remote sensing images in chronological order, and obtain the first warning coefficient according to the time series of low-resolution remote sensing images;

[0046] The precursor area division module is used to compare the first warning coefficient with a preset first warning coefficient threshold, and mark the debris flow precursor area in the low-resolution remote sensing image and the high-resolution remote sensing image;

[0047] The high-resolution warning module is used to obtain high-resolution remote sensing images at different times, calculate the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in each high-resolution remote sensing image, and obtain the second warning coefficient according to the boundary similarity between the debris flow precursor area and the non-debris flow precursor area;

[0048] The comprehensive warning module is used to construct a debris flow precursor comprehensive warning model according to the first warning coefficient and the second warning coefficient, output a debris flow precursor comprehensive warning index, and conduct a comprehensive warning on the precursor risk of debris flow disasters;

[0049] The low-resolution warning module is used to obtain low-resolution remote sensing images at different times, establish a low-resolution remote sensing image time series in chronological order of the low-resolution remote sensing images, and obtain the first warning coefficient according to the low-resolution remote sensing image time series;

[0050] It should be noted that the low-resolution remote sensing image is a data type with a large coverage area, high acquisition frequency, and suitable for macro monitoring. Although its spatial accuracy is low, its high temporal resolution makes it a key data source in the debris flow disaster warning system, suitable for dynamic change monitoring and large-scale risk screening. The spatial resolution of the low-resolution remote sensing image is usually between dozens of meters and hundreds of meters, indicating that the ground area covered by each pixel is relatively large;

[0051] Obtaining the first warning coefficient according to the low-resolution remote sensing image time series specifically includes the following:

[0052] Mark the low-resolution remote sensing image time series as Dt = {dtx i} = {dtx1, dtx2,..., dtx I}, where dtx i represents the low-resolution remote sensing image at the i-th moment, i ∈ {1, 2,..., I}, and I is a positive integer;

[0053] Calculate the brightness change coefficient and color difference coefficient of the low-resolution remote sensing images at different times at the same pixel point;

[0054] The brightness change coefficient and the color difference coefficient measure the brightness change fluctuation and color difference of low-resolution remote sensing images at the same pixel point from the perspective of high temporal resolution. When the brightness change fluctuation and color difference of low-resolution remote sensing images at the same pixel point are relatively severe, it may indicate subtle changes in debris flow precursors; the brightness change coefficient measures the brightness (light intensity) fluctuation of the same pixel point at different times and reflects the dynamic changes in local areas of optical images. A large brightness fluctuation may mean a significant change in the state of surface materials (such as soil or vegetation); the color difference coefficient reflects the color change of the same pixel point at different times, including the comprehensive change of hue, saturation, and brightness, and is especially suitable for capturing spectral property changes caused by changes in surface materials or vegetation cover.

[0055] The acquisition logic of the brightness change coefficient is as follows:

[0056] Calculate the brightness value Ld of the pixel point of the low-resolution remote sensing image at the i-th moment i , and the calculation expression is as follows: Ld i = 0.299 * R i + 0.587 * G i + 0.114B i , where R i , G i , B i respectively represent the red, green, and blue components of the pixel point at the i-th moment;

[0057] Calculate the brightness change rate ΔLd between adjacent moments i , and the calculation expression is as follows: where Ld i+1 represents the brightness value of the pixel point at the (i + 1)-th moment, and ε is a very small positive value used to avoid the denominator being zero;

[0058] Calculate the average value μLd of the brightness change rate, and the expression is as follows:

[0059] Calculate the standard deviation σLd of the brightness change rate, and the calculation expression is as follows:

[0060] Calculate the brightness change coefficient Ldx, and the calculation expression is as follows:

[0061] The acquisition logic of the color difference coefficient is as follows:

[0062] Combine the RGB values of the pixel point and label it as a color vector

[0063] Calculate the color difference value ΔSc between adjacent moments i , and the calculation expression is as follows: where represents the color vector at the (i + 1)-th moment, represents the color vector at the i-th moment, represents the Euclidean norm of the color vector at the i-th moment;

[0064] Calculate the average value μSc of the color difference values, and the expression is as follows:

[0065] Calculate the standard deviation σSc of the color difference values, and the calculation expression is as follows:

[0066] Calculate the color difference coefficient Scy, and the calculation expression is as follows:

[0067] Calculate the first warning coefficient YJ1 according to the brightness change coefficient and the color difference coefficient, and the calculation formula is as follows:

[0068] YJ1 = α * Ldx + β * Scy, where α and β respectively represent the preset proportional coefficients of the brightness change coefficient and the color difference coefficient, and both α and β are greater than 0;

[0069] It should be noted that the first warning coefficient calculated above refers to the first warning coefficient of a pixel point in the low-resolution remote sensing image, that is, each pixel point in the low-resolution remote sensing image corresponds to a first warning coefficient;

[0070] It should be noted that before calculating the first warning coefficient, it is necessary to ensure that both the brightness change coefficient and the color difference coefficient have been normalized; α and β are set according to the actual situation. For example, the expert weighting method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of each index through professional opinion surveys and comprehensive evaluations;

[0071] The brightness change coefficient and the color difference coefficient complement each other at high temporal resolution, and respectively measure the dynamic changes of the low-resolution remote sensing image from the perspectives of brightness and color, providing a reliable basis for debris flow precursor detection. When the fluctuations of both are severe, it indicates that geological activities may have entered the initial stage. The first warning coefficient combining these characteristics can provide early signals for the warning system and help reduce disaster losses.

[0072] The precursor area division module is used to compare the first warning coefficient with the preset first warning coefficient threshold, and mark the debris flow precursor area in the low-resolution remote sensing image and the high-resolution remote sensing image;

[0073] Compare the first warning coefficient with the preset first warning coefficient threshold, and mark the debris flow precursor area in the low-resolution remote sensing image and the high-resolution remote sensing image, specifically as follows:

[0074] Traverse the pixel points in the low-resolution remote sensing image, compare the first warning coefficient of each pixel point with the preset first warning coefficient threshold. If the first warning coefficient is greater than the first warning coefficient threshold, it indicates that the fluctuation difference of the low-resolution remote sensing image in terms of brightness and color is relatively serious, suggesting that the geological activity may have entered the initial stage. Mark the corresponding pixel points as debris flow precursor pixel points, mark the area represented by all debris flow precursor pixel points as the debris flow precursor area, and at the same time mark the same area in the high-resolution remote sensing image as the debris flow precursor area. If the first warning coefficient is less than or equal to the first warning coefficient threshold, it indicates that the fluctuation difference of the low-resolution remote sensing image in terms of brightness and color does not reach the preset significance standard, and the geological activity may be in a relatively stable state without showing obvious characteristics of debris flow precursors. Mark the corresponding pixel points as non-debris flow precursor pixel points, mark the area represented by all non-debris flow precursor pixel points as the non-debris flow precursor area, and at the same time mark the same area in the high-resolution remote sensing image as the non-debris flow precursor area;

[0075] A high-resolution warning module for obtaining high-resolution remote sensing images at different times, calculating the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in each high-resolution remote sensing image, and obtaining a second warning coefficient based on the boundary similarity between the debris flow precursor area and the non-debris flow precursor area;

[0076] Obtain high-resolution remote sensing images at different times, and calculate the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in each high-resolution remote sensing image, specifically as follows:

[0077] Take the boundary line between the debris flow precursor area and the non-debris flow precursor area as the boundary line, mark the pixel points on both sides of the boundary line as boundary pixel points. If the boundary pixel point belongs to the debris flow precursor area, add the boundary point to the set of boundary pixel points of the debris flow precursor area; if the boundary pixel point belongs to the non-debris flow precursor area, add the boundary point to the set of boundary pixel points of the non-debris flow precursor area;

[0078] Represent the set of boundary pixel points of the debris flow precursor area and the set of boundary pixel points of the non-debris flow precursor area as boundary pixel point vectors respectively: Where represents the boundary pixel point vector of the debris flow precursor area, Xsn j represents the boundary pixel points in the set of boundary pixel points of the debris flow precursor area, j ∈ {1, 2,..., J}, and J is a positive integer; Where represents the boundary pixel point vector of the non-debris flow precursor area, Xsf jDenote the boundary pixel points in the set of boundary pixel points of the non-debris flow precursor area, where \(j\in\{1,2,\cdots,J\}\) and \(J\) is a positive integer;

[0079] Calculate the boundary similarity \(BJX\) based on the boundary pixel point vectors of the debris flow precursor area and the non-debris flow precursor area. The calculation expression is as follows: where denotes the dot product operation of the boundary pixel point vector of the debris flow precursor area and the boundary pixel point vector of the non-debris flow precursor area, denotes the Euclidean norm of the boundary pixel point vector of the debris flow precursor area, denotes the Euclidean norm of the boundary pixel point vector of the non-debris flow precursor area;

[0080] Compare the boundary similarity with the preset boundary similarity threshold. If the boundary similarity is greater than the boundary similarity threshold, it indicates that the difference between the debris flow precursor area and the non-debris flow precursor area in the high-resolution remote sensing image is smaller, and mark the high-resolution remote sensing image at this moment as a non-debris flow precursor image, effectively avoiding false alarms; if the boundary similarity is less than or equal to the boundary similarity threshold, it indicates that the difference between the debris flow precursor area and the non-debris flow precursor area in the high-resolution remote sensing image is more obvious, that is, better verify the debris flow disaster precursor signals captured in the low-resolution remote sensing image in the high-resolution remote sensing image, and mark the high-resolution remote sensing image at this moment as a high-resolution debris flow precursor image;

[0081] Calculate the second warning coefficient \(YJ2\) based on the high-resolution debris flow precursor image and the non-debris flow precursor image. The calculation expression is as follows: where \(TX1\) is the number of high-resolution debris flow precursor images, and \(TX2\) is the number of non-debris flow precursor images;

[0082] The comprehensive warning module is used to construct a comprehensive debris flow precursor warning model based on the first warning coefficient and the second warning coefficient, output the comprehensive debris flow precursor warning index, and conduct a comprehensive warning on the precursor risk of debris flow disasters;

[0083] Construct a comprehensive debris flow precursor warning model based on the first warning coefficient and the second warning coefficient, and output the comprehensive debris flow precursor warning index \(Nsl\). The formula on which the model is based is as follows In the formula, \(T1\) represents the acquisition time period of the low-resolution remote sensing image, \(T2\) represents the acquisition time period of the high-resolution remote sensing image, \(YJ1\) m represents the first warning coefficient of the \(m\)th pixel point in the low-resolution remote sensing image, \(YJ2\) represents the second warning coefficient, \(a1\) and \(a2\) respectively represent the preset proportional coefficients of the first warning coefficient and the second warning coefficient, and both \(a1\) and \(a2\) are greater than 0;

[0084] It should be noted that before constructing the comprehensive early warning model for debris flow precursors, it is necessary to ensure that both the first early warning coefficient and the second early warning coefficient have been normalized; a1 and a2 are set according to the actual situation. For example, the expert weighting method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of each index through professional opinion surveys and comprehensive evaluations.

[0085] From the above calculation expressions, it can be seen that the larger the acquisition time period of the low-resolution remote sensing image, the larger the acquisition time period of the high-resolution remote sensing image, the larger the first early warning coefficient, and the larger the second early warning coefficient, the larger the comprehensive early warning index for debris flow precursors, which means the higher the precursor risk of debris flow disasters, and more urgent and effective prevention and control measures need to be taken to reduce the possible losses caused by the disasters. On the contrary, the smaller the acquisition time period of the low-resolution remote sensing image, the smaller the acquisition time period of the high-resolution remote sensing image, the smaller the first early warning coefficient, and the smaller the second early warning coefficient, the smaller the comprehensive early warning index for debris flow precursors, which means the lower the precursor risk of debris flow disasters, the smaller the assessment of the occurrence probability and urgency of debris flow, and the lower the possibility of disaster occurrence. Conventional monitoring measures can be taken without immediately implementing large-scale emergency plans.

[0086] The comprehensive early warning model for debris flow precursors constructed based on the first early warning coefficient and the second early warning coefficient effectively improves the accuracy and reliability of debris flow precursor identification by integrating low-resolution and high-resolution remote sensing image data. By calculating the first early warning coefficient and the second early warning coefficient and combining boundary similarity and spatio-temporal dynamic analysis, this model can accurately assess the risk of debris flow precursor areas and issue early warning signals in a timely manner. At the same time, it reduces the false alarm rate and missed alarm rate, provides a scientific basis for decision-makers, optimizes resource allocation, and thus reduces the losses caused by debris flow disasters and improves the disaster prevention and control ability.

[0087] Compare the comprehensive early warning index for debris flow precursors with the preset threshold of the comprehensive early warning index for debris flow precursors to conduct a comprehensive early warning of the precursor risk of debris flow disasters, as follows:

[0088] If the comprehensive early warning index for debris flow precursors is greater than the threshold of the comprehensive early warning index for debris flow precursors, a comprehensive early warning signal for debris flow precursors is generated, indicating that the precursor risk of debris flow disasters is extremely high, and more urgent and effective prevention and control measures need to be taken to reduce the possible losses caused by the disasters; if the comprehensive early warning index for debris flow precursors is less than or equal to the threshold of the comprehensive early warning index for debris flow precursors, there is no need to generate a comprehensive early warning signal for debris flow precursors, indicating that the precursor risk of debris flow disasters is extremely low, the assessment of the occurrence probability and urgency of debris flow is small, and the possibility of disaster occurrence is low. Conventional monitoring measures can be taken without immediately implementing large-scale emergency plans.

[0089] The present invention obtains low-resolution remote sensing images at different times, analyzes the time series of low-resolution remote sensing images, calculates the brightness change coefficient and color difference coefficient of the low-resolution remote sensing images at the same pixel points from the perspective of high temporal resolution, quantifies the brightness change fluctuation and color difference of the low-resolution remote sensing images at the same pixel points, and calculates the first warning coefficient to provide early signals for the precursor risks of debris flow disasters; and marks the debris flow precursor areas in the low-resolution remote sensing images and high-resolution remote sensing images according to the first warning coefficient, obtains high-resolution remote sensing images at different times, calculates the boundary similarity between the debris flow precursor areas and non-debris flow precursor areas in the high-resolution remote sensing images at each time, verifies the debris flow disaster precursor signals captured in the low-resolution remote sensing images in the high-resolution remote sensing images according to the boundary similarity, and at the same time calculates the second warning coefficient. Combining the first warning coefficient, the acquisition time period of the remote sensing images, a comprehensive early warning model for debris flow precursors is constructed, and a comprehensive early warning index for debris flow precursors is output to comprehensively evaluate the precursor risks of debris flow disasters, so as to achieve efficient and accurate early warning of debris flow disaster precursors, improve the prediction and early warning capabilities of mountain debris flow disasters, and is of great significance for reducing disaster losses and ensuring the safety of people's lives and property.

[0090] The above formulas are all dimensionless and take their numerical values for calculation. The formula is a formula obtained by collecting a large amount of data for software simulation to get the closest real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0091] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on the computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0092] It should be understood that in various embodiments of the present application, the magnitudes of the sequence numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0093] If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0094] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network is characterized by: It includes low-resolution early warning module, precursor area division module, high-resolution early warning module and comprehensive early warning module; A low-resolution early warning module is used to obtain low-resolution remote sensing images at different times, establish a low-resolution remote sensing image time series according to the chronological order of the low-resolution remote sensing images, and obtain a first early warning coefficient according to the low-resolution remote sensing image time series; A precursor region division module is used to compare the first warning coefficient with a preset first warning coefficient threshold value, and mark the debris flow precursor region in the low-resolution remote sensing image and the high-resolution remote sensing image; A high-resolution early warning module is used to obtain high-resolution remote sensing images at different times, and calculate the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in the high-resolution remote sensing image at each time, and obtain a second early warning coefficient according to the boundary similarity between the debris flow precursor area and the non-debris flow precursor area; The comprehensive warning module is used to construct a comprehensive warning model for debris flow precursors based on the first warning coefficient and the second warning coefficient, output a comprehensive warning index for debris flow precursors, and provide a comprehensive warning for the precursor risks of debris flow disasters.

2. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 1 is characterized by: The first warning coefficient is obtained according to the time series of low-resolution remote sensing images, which specifically includes the following: The low-resolution remote sensing image time series is marked as Dt = {dtx i }={dtx1,dtx2,...,dtx I }, where dtx i represents the low-resolution remote sensing image at the i-th moment, i∈{1,2,...,I}, I is a positive integer; Calculate the brightness variation coefficient and color difference coefficient of the same pixel point of low-resolution remote sensing images at different times; The logic for obtaining the brightness change coefficient is as follows: Calculate the brightness value Ld of the pixel of the low-resolution remote sensing image at the i-th moment i , the calculation expression is as follows: Ld i =0.299*R i +0.587*G i +0.114B i , where R i , G i , B i Respectively represent the red, green, and blue components of the pixel at the i-th moment; Calculate the brightness change rate ΔLd between adjacent moments i , the calculation expression is as follows: Where Ld i+1 Represents the brightness value of the pixel at time i+1, ε is a very small positive value to avoid the denominator being zero; Calculate the average value of the brightness change rate μLd, the expression is as follows: Calculate the standard deviation of the brightness change rate σLd, the calculation expression is as follows: Calculate the brightness change coefficient Ldx, the calculation expression is as follows: The logic for obtaining the color difference coefficient is as follows: Mark the RGB value combination of the pixel as a color vector Calculate the color difference value ΔSc at adjacent moments i , the calculation expression is as follows: in represents the color vector at the i+1th moment, represents the color vector at the i-th moment, represents the Euclidean norm of the color vector at the i-th moment; Calculate the average value of the color difference value μSc, the expression is as follows: Calculate the standard deviation of the color difference value σSc, the calculation expression is as follows: Calculate the color difference coefficient Scy, the calculation expression is as follows: The first warning coefficient YJ1 is calculated based on the brightness change coefficient and the color difference coefficient. The calculation formula is as follows: YJ1=α*Ldx+β*Scy, wherein α and β represent preset proportional coefficients of brightness variation coefficient and color difference coefficient respectively, and both α and β are greater than 0.

3. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 2 is characterized by: The first warning coefficient is compared with the preset first warning coefficient threshold, and the debris flow precursor area is marked in the low-resolution remote sensing image and the high-resolution remote sensing image, as follows: The pixel points in the low-resolution remote sensing image are traversed, and the first warning coefficient of each pixel point is compared with the preset first warning coefficient threshold value. If the first warning coefficient is greater than the first warning coefficient threshold value, the corresponding pixel point is marked as a debris flow precursor pixel point, and the area represented by all debris flow precursor pixels is marked as a debris flow precursor area, and the same area in the high-resolution remote sensing image is marked as a debris flow precursor area; If the first warning coefficient is less than or equal to the first warning coefficient threshold, the corresponding pixel point will be marked as a non-debris flow precursor pixel point, and the area represented by all non-debris flow precursor pixels will be marked as a non-debris flow precursor area. At the same time, the same area in the high-resolution remote sensing image will be marked as a non-debris flow precursor area.

4. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 3 is characterized by: Obtain high-resolution remote sensing images at different times, and calculate the boundary similarity between the debris flow precursor area and the non-debris flow precursor area in the high-resolution remote sensing image at each moment, as follows: The boundary line between the debris flow precursor area and the non-debris flow precursor area is taken as the boundary line, and the pixels on both sides of the boundary line are marked as boundary pixels. If the boundary pixel belongs to the debris flow precursor area, the boundary point is added to the boundary pixel point set of the debris flow precursor area; if the boundary pixel belongs to the non-debris flow precursor area, the boundary point is added to the boundary pixel point set of the non-debris flow precursor area; The boundary pixel set of the debris flow precursor area and the boundary pixel set of the non-debris flow precursor area are represented by boundary pixel vectors: in Represents the pixel vector of the boundary of the debris flow precursor area, Xsn j represents the boundary pixel point in the boundary pixel point set of debris flow precursor area, j∈{1,2,...,J}, J is a positive integer; in Indicates the pixel vector of the boundary of the non-debris flow precursor area, Xsf j represents the boundary pixel point in the boundary pixel point set of the non-debris flow precursor area, j∈{1,2,...,J}, J is a positive integer; The boundary similarity BJX is calculated based on the boundary pixel vectors of the debris flow precursor area and the boundary pixel vectors of the non-debris flow precursor area. The calculation expression is as follows: in It represents the dot product operation between the boundary pixel vector of debris flow precursor area and the boundary pixel vector of non-debris flow precursor area. The Euclidean norm of the pixel vector at the boundary of the debris flow precursor area, Represents the Euclidean norm of the pixel vector at the boundary of the non-debris flow precursor area.

5. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 4 is characterized by: The boundary similarity is compared with a preset boundary similarity threshold. If the boundary similarity is greater than the boundary similarity threshold, the high-resolution remote sensing image at that moment is marked as a non-debris flow precursor image; if the boundary similarity is less than or equal to the boundary similarity threshold, the high-resolution remote sensing image at that moment is marked as a high-resolution debris flow precursor image; The second warning coefficient YJ2 is calculated based on the high-resolution debris flow precursor image and the non-debris flow precursor image. The calculation expression is as follows: TX1 is the number of high-resolution debris flow precursor images, and TX2 is the number of non-debris flow precursor images.

6. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 5 is characterized by: According to the first warning coefficient and the second warning coefficient, a debris flow precursor comprehensive warning model is constructed to output the debris flow precursor comprehensive warning index Nsl. The formula based on the model is as follows Where, T1 represents the acquisition time period of low-resolution remote sensing images, T2 represents the acquisition time period of high-resolution remote sensing images, and YJ1 m represents the first warning coefficient of the mth pixel in the low-resolution remote sensing image, YJ2 represents the second warning coefficient, a1 and a2 represent the preset proportional coefficients of the first warning coefficient and the second warning coefficient respectively, and a1 and a2 are both greater than 0.

7. The intelligent early warning system for debris flow disasters in mountainous areas based on remote sensing network according to claim 6 is characterized by: The debris flow precursor comprehensive warning index is compared with the preset debris flow precursor comprehensive warning index threshold to conduct a comprehensive warning of the precursor risk of debris flow disasters, as follows: If the comprehensive warning index of debris flow precursors is greater than the comprehensive warning index threshold of debris flow precursors, a comprehensive warning signal of debris flow precursors is generated; if the comprehensive warning index of debris flow precursors is less than or equal to the comprehensive warning index threshold of debris flow precursors, there is no need to generate a comprehensive warning signal of debris flow precursors.

Images