Method and device for judging seismic destructive influence field based on multi-source information fusion

Abstract

The invention discloses a multi-source information fusion earthquake destructive influence field determination method and device, and relates to the field of earthquake influence field determination, and the method comprises the steps: obtaining multi-source information data; the multi-source information data comprises video stream data, WIFI state data and geographic information data; performing feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data comprises a video intensity index and a WIFI intensity index; performing information fusion on the feature extraction data to obtain fusion intensity information; performing optimization processing on the fusion intensity information by adopting space-time Kalman filtering to obtain a dynamic intensity field; projecting the dynamic intensity field to a GIS map to obtain a dynamic intensity contour line; the dynamic intensity isoline is used for displaying an earthquake influence range and intensity distribution so as to judge an earthquake influence field; the GIS map is determined based on the geographic information data. According to the method, the earthquake influence field can be quickly and accurately judged.

Description

Technical Field

[0001] The present application relates to the field of earthquake impact field determination, and particularly to a method and device for determining earthquake destructive impact field by fusing multi-source information. Background Art

[0002] The suddenness and destructiveness of earthquake disasters determine that quickly and accurately grasping the disaster situation information after an earthquake is crucial for emergency response. Drawing the earthquake destructive impact field and conducting disaster intensity assessment are the basis for post-earthquake emergency decision-making and resource allocation. At present, the earthquake field mainly relies on the following technical means for rapid intensity assessment:

[0003] ① Instrumental intensity rapid reporting based on strong motion network; ② Disaster loss assessment based on remote sensing images; ③ Empirical intensity calculation based on ground motion prediction models.

[0004] However, these traditional methods still have limitations in practical applications, such as sparse distribution of strong motion stations, long acquisition and processing cycles of remote sensing data, and limited accuracy of empirical models.

[0005] Therefore, how to quickly and accurately determine the earthquake impact field is of great importance. Summary of the Invention

[0006] The purpose of the present application is to provide a method and device for determining earthquake destructive impact field by fusing multi-source information, which can quickly and accurately determine the earthquake impact field.

[0007] To achieve the above purpose, the present application provides the following solutions:

[0008] In the first aspect, the present application provides a method for determining earthquake destructive impact field by fusing multi-source information, including:

[0009] Obtaining multi-source information data; the multi-source information data includes: video stream data, WIFI status data, and geographic information data;

[0010] Performing feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index;

[0011] Performing information fusion on the feature extraction data to obtain fused intensity information;

[0012] Using spatio-temporal Kalman filtering to optimize the fused intensity information to obtain a dynamic intensity field;

[0013] Project the dynamic intensity field onto the GIS map to obtain dynamic intensity isolines; the dynamic intensity isolines are used to display the earthquake influence range and intensity distribution for determining the earthquake influence field; the GIS map is determined based on the geographic information data.

[0014] Optionally, perform feature extraction processing on the multi-source information data to obtain feature extraction data, specifically including:

[0015] Perform macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index;

[0016] Calculate the interruption rate of the WIFI status data to obtain a WIFI intensity index.

[0017] Optionally, perform macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index, including:

[0018] Perform denoising and enhancement processing on the video stream data to obtain processed video stream data;

[0019] Segment the processed video stream data frame by frame to obtain multiple frames of images, perform macroblock partitioning on each frame of image, and use the block matching algorithm to calculate the motion vectors of macroblocks between adjacent frames; the motion vectors represent the displacement amount and direction of objects in the image between frames;

[0020] Based on a set motion threshold, determine the number of moving macroblocks in each frame of image according to the motion vectors;

[0021] Determine the average moving pixel ratio based on the number of all moving macroblocks;

[0022] Determine the average moving pixel ratio as the video intensity index.

[0023] Optionally, the expression of the average moving pixel ratio is:

[0024]

[0025] where P is the average moving pixel ratio; N is the total number of video frames corresponding to the video stream data; M i is the number of moving macroblocks in the i-th frame of image; S macroblock is the macroblock area; S frame is the image frame area.

[0026] Optionally, calculate the interruption rate of the WIFI status data to obtain a WIFI intensity index, including:

[0027] Calculate the interruption rate of the WIFI status data to obtain the WIFI network interruption rate;

[0028] Based on the WIFI interruption intensity calculation model, determine the WIFI intensity index according to the WIFI network interruption rate; the WIFI interruption intensity calculation model is a mathematical model determined based on regional WIFI hotspot detection data and post-earthquake WIFI interruption statistical data.

[0029] Optionally, the expression of the WIFI network interruption rate is:

[0030]

[0031] The expression of the WIFI intensity index is:

[0032]

[0033] where R is the WIFI network interruption rate; N before is the number of WIFI hotspots in the area before the earthquake; N after is the number of WIFI hotspots in the area after the earthquake; I w is the WIFI intensity index; a is the lower limit of intensity; b is the upper limit of intensity; k1 is the sensitivity coefficient of the WIFI interruption intensity calculation model; R0 is the inflection point interruption rate of the WIFI interruption intensity calculation model.

[0034] Optionally, the expression of the fused intensity information is:

[0035] I f = w v ×I v + w w ×I w ;

[0036] where, I f is the fused intensity information; w v is the weight coefficient of the video intensity index; w w is the weight coefficient of the WIFI intensity index; I v is the video intensity index; I w is the WIFI intensity index.

[0037] Optionally, use spatio-temporal Kalman filtering to optimize the fused intensity information to obtain a dynamic intensity field, specifically including:

[0038] For the fused intensity information, use the neighborhood-based outlier detection method to detect and remove outliers;

[0039] Use spatial interpolation method to interpolate and convert the discrete intensity value points after removal processing into a continuous intensity scene to obtain a dynamic intensity field;

[0040] Among them, when the spatial interpolation method uses inverse distance weighted interpolation for interpolation conversion, the corresponding expression is:

[0041]

[0042] Among them, I(x0, y0) is the intensity value of the point to be estimated (x0, y0); λ k is the weight coefficient; I k is the intensity value of the k-th known intensity value point; n is the total number of known intensity value points; d((x0, y0), (x k , y k )) is the distance between the point to be estimated (x0, y0) and the k-th known intensity value point (x k , y k ); p is the distance attenuation exponent.

[0043] Optionally, the method for determining the seismic destructive impact field by multi-source information fusion further includes:

[0044] Determine the velocity of the pixels in each frame of image according to the motion vector; the expression corresponding to the velocity of the pixels is:

[0045]

[0046] Among them, v i (x, y) is the velocity of the pixel (x, y); MV ix is the component of the motion vector of the pixel (x, y) in the x direction in the i-th frame of image; MV iy is the component of the motion vector of the pixel (x, y) in the y direction in the i-th frame of image; Δd i (x, y) is the true motion distance corresponding to the motion vector; Δt is the time interval; F is the video frame rate; D i (x + MV ix , y + MV iy ) is the pixel depth value corresponding to the motion vector of the pixel (x, y) in the i-th frame of image; D i (x, y) is the pixel depth value corresponding to the pixel (x, y).

[0047] In a second aspect, the present application provides a device for determining a seismic destructive impact field by multi-source information fusion, including:

[0048] A multi-source information data acquisition module, configured to acquire multi-source information data; the multi-source information data includes: video stream data, WIFI status data, and geographic information data;

[0049] A feature extraction module, configured to perform feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index;

[0050] An information fusion module, configured to perform information fusion on the feature extraction data to obtain fused intensity information;

[0051] A dynamic intensity field determination module, configured to perform optimization processing on the fused intensity information by using spatio-temporal Kalman filtering to obtain a dynamic intensity field;

[0052] A dynamic intensity isoline determination module, configured to project the dynamic intensity field onto a GIS map to obtain dynamic intensity isolines; the dynamic intensity isolines are used to display the earthquake influence range and intensity distribution for determining the earthquake influence field; the GIS map is determined based on the geographic information data.

[0053] According to the specific embodiments provided by the present application, the present application discloses the following technical effects:

[0054] The present application provides a method and device for determining a seismic destructive influence field through multi-source information fusion. The method includes: acquiring multi-source information data; the multi-source information data includes: video stream data, WIFI status data, and geographic information data; performing feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index; performing information fusion on the feature extraction data to obtain fused intensity information; performing optimization processing on the fused intensity information by using spatio-temporal Kalman filtering to obtain a dynamic intensity field; projecting the dynamic intensity field onto a GIS map to obtain dynamic intensity isolines; the dynamic intensity isolines are used to display the earthquake influence range and intensity distribution for determining the earthquake influence field; the GIS map is determined based on the geographic information data. By combining the video stream data with the WIFI status data, through feature extraction and information fusion, the present application can make the WIFI status data and the video stream data complement each other effectively, and can comprehensively and objectively reflect the factors corresponding to the seismic destructive influence. Then, by performing optimization processing on the fused intensity information by using spatio-temporal Kalman filtering, the accuracy of the seismic influence field determination can be improved. Finally, the earthquake influence range and intensity distribution are displayed through the dynamic intensity isolines, which can more intuitively display the earthquake influence for determining the earthquake influence field. Thus, the present application can quickly and accurately determine the earthquake influence field. Description of the Drawings

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the following described drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0056] Figure 1Flow chart of the method for determining the seismic destructive influence field by multi-source information fusion;

[0057] Figure 2 It is the technical flow chart corresponding to the method for determining the seismic destructive influence field by multi-source information fusion. Specific implementation manner

[0058] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present application.

[0059] In recent years, with the rapid improvement of the social informatization level, the distribution of WIFI has become increasingly popular. WIFI is a common method for connecting mobile phones indoors. When the seismic intensity reaches a certain level, communication and power supply generally interrupt. Judging the damage degree of houses by seismic intensity is a very important indicator. In addition, the publicly available network video surveillance system has achieved wide coverage. These massive non-private public surveillance video data provide a new data source and technical idea for earthquake disaster assessment. Social videos can truly and intuitively record the ground vibration and damage scenes during the earthquake, and contain rich information on the spatial distribution of the disaster situation. The present application innovatively proposes to fuse social video surveillance data and WIFI network interruption information to construct a set of rapid determination technical methods for the seismic destructive influence field by multi-source information fusion. The purpose of this method is to make up for the deficiencies of traditional earthquake disaster assessment methods, fully explore and utilize new social perception data, provide faster, more refined and comprehensive disaster situation information support for earthquake emergency response, and improve the scientificity and timeliness of post-earthquake emergency decision-making.

[0060] The present application integrates macroscopic and microscopic perspectives. It can not only intuitively analyze the seismic surface damage from the video, but also reflect the damage of more extensive urban infrastructure and social operations through WIFI interruption. Relevant research has been carried out, for example: Xia Min (2008) estimated ground motion parameters based on surveillance videos; Tian Guowei (2015) measured structural vibrations using video image processing technology; Hashimoto et al. (2013) analyzed social media videos and the degree of earthquake damage. These studies have laid a theoretical foundation and practical experience for the research and development of the technology of the present application.

[0061] In addition to video surveillance, as an important part of the modern social infrastructure, the operating status of the WIFI network can indirectly reflect the degree of earthquake impact. When a destructive earthquake occurs, strong ground shaking may cause physical damage, power interruption, line damage or network congestion to WIFI devices (such as routers and optical network terminals), resulting in WIFI signal interruption. Especially when the earthquake intensity reaches VIII degrees or above, power and communication facilities are extremely vulnerable to damage, and the phenomenon of WIFI interruption will be more common and serious. This application incorporates the concept of a relationship model between WIFI interruption and intensity. Based on the assumption that the greater the earthquake intensity, the higher the probability and wider the range of WIFI interruption, by analyzing the characteristics of the number, duration, and spatial distribution of regional WIFI interruptions after the earthquake, the earthquake impact intensity information can be inverted. WIFI data has the advantages of low data acquisition cost, wide coverage, and good real-time performance, which effectively complements video surveillance data and can more comprehensively and objectively reflect the earthquake disaster situation.

[0062] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] In an exemplary embodiment, as Figure 1 shown, a method for determining the destructive earthquake impact field by multi-source information fusion is provided, including:

[0064] Step 100: Obtain multi-source information data. The multi-source information data includes: video stream data, WIFI status data, and geographical information data.

[0065] Step 200: Perform feature extraction processing on the multi-source information data to obtain feature extraction data. The feature extraction data includes: video intensity index and WIFI intensity index.

[0066] Among them, performing feature extraction processing on the multi-source information data to obtain feature extraction data specifically includes:

[0067] Performing macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index; calculating the interruption rate of the WIFI status data to obtain a WIFI intensity index.

[0068] Specifically, performing macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index includes:

[0069] Performing denoising and enhancement processing on the video stream data to obtain processed video stream data.

[0070] The processed video stream data is segmented frame by frame to obtain multiple frames of images. Each frame of image is divided into macroblocks, and the motion vectors of the macroblocks between adjacent frames are calculated using the block matching algorithm. The motion vector represents the displacement amount and direction of an object in the image between frames.

[0071] Based on a set motion threshold, the number of moving macroblocks in each frame of image is determined according to the motion vector. Based on the number of all moving macroblocks, the average moving pixel ratio is determined. The average moving pixel ratio is determined as the video intensity index.

[0072] The expression for the average moving pixel ratio is:

[0073]

[0074] Where P is the average moving pixel ratio; N is the total number of video frames corresponding to the video stream data; M i is the number of moving macroblocks in the i-th frame of image; S macroblock is the macroblock area; S frame is the image frame area.

[0075] In one embodiment, the interruption rate of the WIFI status data is calculated to obtain the WIFI intensity index, including:

[0076] The interruption rate of the WIFI network is calculated according to the WIFI status data. Based on the WIFI interruption intensity inference model, the WIFI intensity index is determined according to the WIFI network interruption rate. The WIFI interruption intensity inference model is a mathematical model determined based on the regional WIFI hotspot detection data and the post-earthquake WIFI interruption statistical data.

[0077] The expression for the WIFI network interruption rate is:

[0078]

[0079] The expression for the WIFI intensity index is:

[0080]

[0081] Where R is the WIFI network interruption rate; N before is the number of WIFI hotspots in the area before the earthquake; N after is the number of WIFI hotspots in the area after the earthquake; I w is the WIFI intensity index; a is the lower limit of intensity; b is the upper limit of intensity; k1 is the sensitivity coefficient of the WIFI interruption intensity inference model; R0 is the inflection point interruption rate of the WIFI interruption intensity inference model.

[0082] Step 300: Perform information fusion on the feature extraction data to obtain the fused intensity information.

[0083] The expression for the fused intensity information is as follows:

[0084] I f = w v × I v + w w × I w .

[0085] Where, I f is the fused intensity information; w v is the weight coefficient of the video intensity index; w w is the weight coefficient of the WIFI intensity index; I v is the video intensity index; I w is the WIFI intensity index.

[0086] Step 400: Optimize the fused intensity information using spatio-temporal Kalman filtering to obtain a dynamic intensity field.

[0087] In one embodiment, optimizing the fused intensity information using spatio-temporal Kalman filtering to obtain a dynamic intensity field specifically includes:

[0088] For the fused intensity information, use the neighborhood-based outlier detection method to detect and remove outliers.

[0089] Use spatial interpolation method to interpolate and convert the discrete intensity value points after the removal process into a continuous intensity scene surface to obtain a dynamic intensity field.

[0090] Where, when the inverse distance weighted interpolation is used for the spatial interpolation method for interpolation conversion, the corresponding expression is:

[0091]

[0092] Where, I(x0,y0) is the intensity value of the point to be estimated (x0,y0); λ k is the weight coefficient; I k is the intensity value of the k-th known intensity value point; n is the total number of known intensity value points; d((x0,y0),(x k ,y k )) is the distance between the point to be estimated (x0,y0) and the k-th known intensity value point (x k ,y k ); p is the distance decay exponent.

[0093] Step 500: Project the dynamic intensity field onto the GIS map to obtain dynamic intensity isograms. The dynamic intensity isograms are used to display the earthquake influence range and intensity distribution for the determination of the earthquake influence field; the GIS map is determined based on geographic information data.

[0094] As an alternative implementation, the method for determining the seismic destructive impact field through multi-source information fusion further includes:

[0095] Determine the velocity of pixels in each frame of the image according to the motion vector; the expression corresponding to the velocity of the pixels is:

[0096]

[0097] where v i (x,y) is the velocity of the pixel (x,y); MV ix is the component of the motion vector of the pixel (x,y) in the x direction in the i-th frame of the image; MV iy is the component of the motion vector of the pixel (x,y) in the y direction in the i-th frame of the image; Δd i (x,y) is the actual motion distance corresponding to the motion vector; Δt is the time interval; F is the video frame rate; D i (x + MV ix , y + MV iy ) is the pixel depth value corresponding to the motion vector of the pixel (x,y) in the i-th frame of the image; D i (x,y) is the pixel depth value corresponding to the pixel (x,y).

[0098] The core of the technical solution mentioned in this application lies in fusing video surveillance motion feature analysis and WIFI network interruption statistics to construct a rapid determination model for the seismic destructive impact field. The overall technical flow chart is as Figure 2 shown.

[0099] First, access video stream data, WIFI status data, and geographic information data as multi-source inputs. Then, in the feature extraction layer, process the video (video stream data) and WIFI status data in parallel: perform macro-block motion analysis and depth estimation on the video stream data to extract the video intensity index; at the same time, calculate the interruption rate of the WIFI status data to obtain the WIFI intensity index. Next, in the model fusion core layer, fuse the video intensity index and the WIFI intensity index to integrate the advantages of the two types of data. The fused intensity information, that is, the fused intensity information, is optimized through spatio-temporal Kalman filtering to generate a dynamic intensity field, reducing noise and improving accuracy. Finally, project the dynamic intensity field onto the GIS map and visually display the earthquake impact range and intensity distribution in the form of dynamic intensity isolines, providing visual support for post-earthquake emergency decision-making. The entire process efficiently integrates multi-source information and aims to quickly and accurately determine the seismic impact field.

[0100] Estimation of seismic intensity based on video surveillance (video stream data).

[0101] Based on the standardized public video surveillance information source (video stream data), the macroblock motion estimation method is used to extract the ground motion characteristics when an earthquake occurs from the surveillance video, and then infer the earthquake impact intensity.

[0102] Seismic video analysis mainly includes the following steps:

[0103] Video preprocessing: Perform preprocessing operations such as denoising and enhancement on video stream data to improve video quality.

[0104] Frame-by-frame macroblock motion estimation: Seismic video (obtained video stream data) is divided into frames, and each frame is divided into macroblocks. Block matching algorithms, such as block matching algorithms based on the mean absolute difference (MAD) algorithm, are used to calculate the motion vector (MV) of macroblocks between adjacent frames. The motion vector reflects the displacement and direction of objects in the image between frames.

[0105] Motion pixel statistics: Count the number of motion macroblocks in each frame. Set a motion threshold. When the motion vector amplitude of a macroblock exceeds the threshold, it is determined to be a motion macroblock. Accumulate the number of motion macroblocks M in each frame.

[0106] Video intensity index calculation: Calculate the average motion pixel ratio P of all image frames in the earthquake video clip (video stream data) and use it as the video intensity index.

[0107] The expression of average motion pixel ratio is:

[0108]

[0109] Where P is the average motion pixel ratio; N is the total number of video frames corresponding to the video stream data; M i is the number of motion macroblocks in the i-th frame image; S macroblock is the macroblock area; S frame is the image frame area.

[0110] Earthquake velocity and acceleration estimation.

[0111] In order to further infer earthquake ground motion parameters (such as peak ground velocity PGV and peak ground acceleration PGA) from the video intensity index, the present application combines a single image depth estimation method.

[0112] Since the installation height and angle of the surveillance camera are unknown, it is difficult to directly determine the correspondence between image pixels and real-world distances. Through single-image depth estimation algorithms (such as deep learning-based depth estimation algorithms like DeepLabv3+ and MiDaS), the depth value D(x, y) of each pixel point in the scene can be approximately estimated.

[0113] In a seismic video clip, the velocity of the pixel (x, y) in the i-th frame image, that is, the velocity v i (x, y) can be estimated from its motion vector MV i (x, y) corresponding to the true motion distance Δd i (x, y) divided by the time interval Δt (i.e., the reciprocal of the video frame rate F). Δd i (x, y) is approximately equal to the difference between the depth values corresponding to the starting pixel and the ending pixel of the motion vector.

[0114] The expression corresponding to the velocity of the pixel is:

[0115]

[0116] where v i (x, y) is the velocity of the pixel (x, y); MV ix is the component of the motion vector of the pixel (x, y) in the x direction in the i-th frame image; MV iy is the component of the motion vector of the pixel (x, y) in the y direction in the i-th frame image; Δd i (x, y) is the true motion distance corresponding to the motion vector; Δt is the time interval; F is the video frame rate; D i (x + MV ix , y + MV iy ) is the depth value of the pixel point corresponding to the motion vector of the pixel (x, y) in the i-th frame image; D i (x, y) is the depth value of the pixel point corresponding to the pixel (x, y).

[0117] Furthermore, the average peak pixel velocity V peak in the seismic video clip can be calculated as a reference index for seismic ground velocity. For example: take the average of the maximum values of all pixel velocities in all frames. Similarly, the average peak pixel acceleration A peak can be estimated.

[0118] Estimation of seismic intensity based on WIFI interruption.

[0119] This application innovatively proposes a method for estimating seismic intensity based on regional WIFI network interruption data.

[0120] Based on the theory that "seismic intensity is positively correlated with the degree of WIFI interruption", a WIFI interruption intensity calculation model is constructed. The data sources in the process of constructing the WIFI interruption intensity calculation model include: regional WIFI hotspot detection data (such as WIFI hotspot information obtained based on operator or third-party WIFI map service APIs) and post-earthquake WIFI interruption statistical data (such as the reduction ratio of the number of WIFI hotspots in the region or the interruption duration within a certain time after the earthquake).

[0121] That is, based on the WIFI interruption intensity calculation model, the WIFI intensity index is determined according to the WIFI network interruption rate; the WIFI interruption intensity calculation model is a mathematical model determined based on regional WIFI hotspot detection data and post-earthquake WIFI interruption statistical data.

[0122]

[0123] Among them, I w is the WIFI intensity index; a is the lower limit of intensity; b is the upper limit of intensity; k1 is the sensitivity coefficient of the WIFI interruption intensity calculation model; R0 is the inflection point interruption rate of the WIFI interruption intensity calculation model. R is the WIFI network interruption rate, which represents the reduction ratio of the number of WIFI hotspots after the earthquake.

[0124] a, b, k, and R0 can be calibrated through historical earthquake data and WIFI interruption data.

[0125] The expression of the WIFI network interruption rate is:

[0126]

[0127] Among them, R is the WIFI network interruption rate; N before is the number of WIFI hotspots in the region before the earthquake; N after is the number of WIFI hotspots in the region after the earthquake.

[0128] Determination of the intensity field based on multi-source information fusion.

[0129] This application uses multi-source information fusion technology to fuse the video intensity index I v and the WIFI intensity index I w to obtain the fused intensity information I f , and finally an earthquake destructive impact field can be generated.

[0130] I f = w v × I v + w w × I w .

[0131] Among them, I fis the integrated intensity information; w v is the weight coefficient of the video intensity index; w w is the weight coefficient of the WIFI intensity index; I v is the video intensity index; I w is the WIFI intensity index.

[0132] w v + w w = 1. The weight coefficients of the video intensity index and the WIFI intensity index can be determined according to factors such as data quality, reliability, and data source type.

[0133] Intensity field generation and visualization.

[0134] Video intensity point and WIFI intensity point identification: Based on video surveillance and WIFI interruption data, that is, video stream data and WIFI status data, a series of video intensity index points (I v points) and WIFI intensity index points (I w points) are calculated respectively.

[0135] Interference point and outlier deletion: Outlier detection and elimination are performed on the video intensity index points (I v points) and WIFI intensity index points (I w points). An outlier detection method based on the neighborhood is adopted. For example, for each intensity value point (video intensity index point and WIFI intensity index point), calculate the average intensity value of its k adjacent points around it. If the intensity value of this point deviates from the neighborhood average value by more than a set threshold (such as 2 standard deviations), it is determined as a suspected interference point or outlier. (Improve the elimination criterion, for example, combine the neighborhood point quantity threshold to avoid misdeleting real high-intensity points). For example, adopt the following criterion:

[0136] Interference point elimination criterion.

[0137] If the value I i of the intensity value point satisfies:

[0138]

[0139] and among the 10 adjacent points around it, the number of adjacent points that satisfy the above conditions is less than 3, then delete this intensity value point.

[0140] Among them, is the average intensity value of the 10 adjacent points around intensity point i, σ neighbor(i) is the standard deviation of the intensity values of the 10 adjacent points around it, and θ is the threshold coefficient (for example, θ = 2).

[0141] Intensity field interpolation: Using spatial interpolation methods such as Inverse Distance Weighted (IDW) or Kriging, the discrete intensity value points are interpolated into a continuous intensity field surface.

[0142] The corresponding expression for the Inverse Distance Weighted (IDW) is:

[0143]

[0144] Where, I(x0,y0) is the intensity value of the point to be estimated (x0,y0); λ k is the weight coefficient; I k is the intensity value of the k-th known intensity value point; n is the total number of known intensity value points.

[0145]

[0146] d((x0,y0),(x k ,y k )) is the distance between the point to be estimated (x0,y0) and the k-th known intensity value point (x k ,y k ); p is the distance attenuation exponent, usually taking p = 2.

[0147] Based on the same inventive concept, the embodiments of the present application also provide a multi-source information fusion earthquake destructive impact field determination device for implementing the multi-source information fusion earthquake destructive impact field determination method involved above. The implementation solutions provided by this device to solve problems are similar to the implementation solutions recorded in the above method. Therefore, the specific limitations in one or more embodiments of the multi-source information fusion earthquake destructive impact field determination device provided below can refer to the limitations on the multi-source information fusion earthquake destructive impact field determination method in the above text, and will not be elaborated here.

[0148] In an exemplary embodiment, a multi-source information fusion earthquake destructive impact field determination device is provided, including:

[0149] A multi-source information data acquisition module for acquiring multi-source information data; the multi-source information data includes: video stream data, WIFI status data, and geographical information data.

[0150] A feature extraction module for performing feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index.

[0151] An information fusion module for performing information fusion on the feature extraction data to obtain fusion intensity information.

[0152] The dynamic intensity field determination module is used to optimize the fused intensity information by using spatio-temporal Kalman filtering to obtain a dynamic intensity field.

[0153] The dynamic intensity isogram determination module is used to project the dynamic intensity field onto a GIS map to obtain dynamic intensity isograms. The dynamic intensity isograms are used to display the earthquake influence range and intensity distribution for determining the earthquake influence field; the GIS map is determined based on the geographical information data.

[0154] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0155] In this article, specific examples are used to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for determining the seismic destructive influence field by multi-source information fusion, characterized in that, Including: Obtain multi-source information data; The multi-source information data includes: video stream data, WIFI status data, and geographical information data; Perform feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index; Perform information fusion on the feature extraction data to obtain fusion intensity information; Use spatio-temporal Kalman filtering to optimize the fusion intensity information to obtain a dynamic intensity field; Project the dynamic intensity field onto a GIS map to obtain dynamic intensity isolines; the dynamic intensity isolines are used to display the earthquake influence range and intensity distribution for determining the earthquake influence field; the GIS map is determined based on the geographical information data.

2. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 1, characterized in that, Performing feature extraction processing on the multi-source information data to obtain feature extraction data specifically includes: Perform macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index; Calculate the interruption rate of the WIFI status data to obtain a WIFI intensity index.

3. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 2, characterized in that Performing macroblock motion analysis and depth estimation on the video stream data to obtain a video intensity index includes: Perform denoising and enhancement processing on the video stream data to obtain processed video stream data; Segment the processed video stream data frame by frame to obtain multiple frames of images, perform macroblock division on each frame of image, and use a block matching algorithm to calculate the motion vectors of macroblocks between adjacent frames; the motion vectors represent the displacement amount and direction of objects in the image between frames; Based on a set motion threshold, determine the number of moving macroblocks in each frame of image according to the motion vectors; Based on the number of all moving macroblocks, determine the average moving pixel ratio; Determine the average moving pixel ratio as the video intensity index.

4. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 3, wherein The expression of the average moving pixel ratio is: Wherein, P is the average motion pixel ratio; N is the total number of video frames corresponding to the video stream data; M i is the number of motion macroblocks in the i-th frame image; S macroblock is the macroblock area; S frame is the image frame area.

5. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 2, wherein Calculating the interruption rate of the WIFI status data to obtain a WIFI intensity index includes: Calculate the WIFI network interruption rate according to the WIFI status data; Based on the WIFI interruption intensity inference model, determine the WIFI intensity index according to the WIFI network interruption rate; the WIFI interruption intensity inference model is a mathematical model determined based on regional WIFI hotspot detection data and post-earthquake WIFI interruption statistics data.

6. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 5, characterized in that, The expression of the WIFI network interruption rate is: The expression of the WIFI intensity index is: Among them, R is the WIFI network interruption rate; N before is the number of WIFI hotspots in the area before the earthquake; N after is the number of WIFI hotspots in the area after the earthquake; I w is the WIFI intensity index; a is the lower limit of intensity; b is the upper limit of intensity; k1 is the sensitivity coefficient of the WIFI interruption intensity calculation model; R0 is the inflection point interruption rate of the WIFI interruption intensity calculation model.

7. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 1, characterized in that, The expression of the fusion intensity information is: I f = w v × I v + w w × I w ; Among them, I f is the fused intensity information; w v is the weight coefficient of the video intensity index; w w is the weight coefficient of the WIFI intensity index; I v is the video intensity index; I w is the WIFI intensity index.

8. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 1, characterized in that, Using spatio-temporal Kalman filtering to optimize the fusion intensity information to obtain a dynamic intensity field specifically includes: For the fusion intensity information, use an outlier detection method based on neighborhoods to detect and remove outliers; Use a spatial interpolation method to interpolate and convert the discrete intensity value points after the removal process into a continuous intensity scene to obtain a dynamic intensity field; Among them, when the spatial interpolation method uses inverse distance weighted interpolation for interpolation conversion, the corresponding expression is: Among them, I(x0, y0) is the intensity value of the point (x0, y0) to be estimated; λ k is the weight coefficient; I k is the intensity value of the k-th known intensity value point; n is the total number of known intensity value points; d((x0, y0), (x k , y k )) is the distance between the point (x0, y0) to be estimated and the k-th known intensity value point (x k , y k ); p is the distance attenuation exponent.

9. The method for determining the seismic destructive influence field by multi-source information fusion according to claim 3, characterized in that, The method for determining the earthquake destructive influence field by multi-source information fusion further includes: Determine the speed of pixels in each frame of image according to the motion vectors; the expression corresponding to the speed of pixels is: where v i (x, y) is the velocity of pixel (x, y); MV ix is the x-component of the motion vector of pixel (x, y) in the i-th frame image; MV iy is the y-component of the motion vector of pixel (x, y) in the i-th frame image; Δd i (x, y) is the actual motion distance corresponding to the motion vector; Δt is the time interval; F is the video frame rate; D i (x + MV ix , y + MV iy ) is the pixel depth value corresponding to the motion vector of pixel (x, y) in the i-th frame image; D i (x, y) is the pixel depth value corresponding to pixel (x, y).

10. A device for determining the seismic destructive impact field with multi-source information fusion, characterized in that, The earthquake destructive impact field determination device for multi-source information fusion includes: A multi-source information data acquisition module, which is used to acquire multi-source information data; the multi-source information data includes: video stream data, WIFI status data, and geographical information data; A feature extraction module, which is used to perform feature extraction processing on the multi-source information data to obtain feature extraction data; the feature extraction data includes: video intensity index and WIFI intensity index; An information fusion module, which is used to perform information fusion on the feature extraction data to obtain fusion intensity information; A dynamic intensity field determination module, which is used to perform optimization processing on the fusion intensity information by using spatio-temporal Kalman filtering to obtain a dynamic intensity field; A dynamic intensity isoline determination module, which is used to project the dynamic intensity field onto a GIS map to obtain dynamic intensity isolines; the dynamic intensity isolines are used to display the earthquake influence range and intensity distribution for the determination of the earthquake impact field; the GIS map is determined based on the geographical information data.

Images