A landslide prediction method and system based on image processing

Abstract

The application provides a landslide prediction method and system based on image processing, remote sensing images of a mountain area are shot and analyzed to obtain surface external characteristic information and soil internal characteristic information of the mountain area, and a risk sub-area existing in the mountain area is identified, so that a large range of surface and internal soil state identification of the mountain area can be performed, and a potential risk area of the mountain area caused by vegetation and rainfall can be accurately determined; cruise detection is further performed on the risk sub-area to obtain surface dynamic images and internal dynamic information of the risk sub-area, and relative displacement change data between the surface and the internal layer of the risk sub-area is obtained by analysis, so that the soil layer movement of the risk sub-area can be accurately identified; and finally, based on the relative displacement change data, a covered area and a landslide running direction corresponding to a landslide event of the risk sub-area are predicted, full-range and early prediction of the landslide area are realized, and the accuracy and reliability of the mountain landslide prediction are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geological monitoring, in particular to a landslide prediction method and system based on image processing. BACKGROUND

[0002] Mountain landslide is a serious geological disaster, which is affected by the geological conditions of the mountain itself and the external weather conditions. The existing monitoring of mountain landslide usually sets contact type sensing devices such as fiber grating sensors in the mountain area to detect the soil of the mountain in real time. The above-mentioned method can accurately detect the movement of the soil layer of the mountain, and can respond when the soil layer of the mountain is loose or slides. However, the detection range of the above-mentioned method for mountain landslide is determined by the distribution range of the fiber grating sensor, so that it is impossible to detect mountain landslide in a large area, and the above-mentioned method cannot provide early warning of landslide, which reduces the coverage and predictability of mountain landslide detection, and cannot improve the accuracy and reliability of mountain landslide prediction. SUMMARY

[0003] The purpose of the present application is to provide a landslide prediction method and system based on image processing, which photographs and analyzes remote sensing images of a mountain area to obtain surface external feature information and soil internal feature information of the mountain area, thereby identifying risk sub-regions existing in the mountain area. This can identify the surface and internal soil state of the mountain area in a large range, accurately determine the potential risk area of the mountain area caused by vegetation and rainfall, and further cruise and detect the risk sub-regions to obtain surface dynamic images and internal dynamic information of the risk sub-regions, thereby analyzing the relative displacement change data between the surface and the internal layer of the risk sub-regions to accurately identify the soil layer movement of the risk sub-regions. Based on the relative displacement change data, the coverage area and the landslide movement direction corresponding to the landslide event of the risk sub-regions are predicted, the full range and early prediction of the landslide area are realized, and the accuracy and reliability of the mountain landslide prediction are improved.

[0004] The present application is achieved by the following technical solutions:

[0005] A landslide prediction method based on image processing, comprising:

[0006] Remote sensing of a mountain area is performed to obtain remote sensing images of the mountain area; the remote sensing images of the mountain area are analyzed to obtain surface external feature information and soil internal feature information of the mountain area; and based on the surface external feature information and the soil internal feature information, risk sub-regions existing in the mountain area are identified;

[0007] cruising detection is performed on the risk sub-region to obtain surface dynamic images and inner layer dynamic information of the risk sub-region; analysis is performed on the surface dynamic images and the inner layer dynamic information to obtain relative displacement change data of the surface and the inner layer of the risk sub-region;

[0008] Based on the relative displacement change data, the coverage area and the landslide movement direction corresponding to the risk sub-region are predicted; and a corresponding notification message is generated and sent based on the coverage area and the landslide movement direction.

[0009] Optionally, remote sensing is performed on the mountain area to obtain a mountain area remote sensing image; analysis is performed on the mountain area remote sensing image to obtain surface external feature information and soil internal feature information of the mountain area; and based on the surface external feature information and the soil internal feature information, a risk sub-region existing in the mountain area is identified, including:

[0010] Based on the region boundary position information of the mountain area, directional remote sensing scanning is performed on the mountain area to obtain a mountain area remote sensing image;

[0011] Visible spectrum analysis is performed on the mountain area remote sensing image to obtain surface vegetation distribution feature information of the mountain area; and far-infrared spectrum analysis is performed on the mountain area remote sensing image to obtain soil water content feature information of the mountain area;

[0012] Based on the surface vegetation distribution feature information, vegetation distribution sparsity density feature information of the entire surface range of the mountain area is determined; and based on the soil water content feature information, distribution position information of all soil sub-regions in a water content saturation state in the entire soil range of the mountain area is determined;

[0013] Based on the vegetation distribution sparsity density feature information and the distribution position information of all soil sub-regions in a water content saturation state, all risk sub-regions existing in the mountain area are identified.

[0014] Optionally, cruising detection is performed on the risk sub-region to obtain surface dynamic images and inner layer dynamic information of the risk sub-region; analysis is performed on the surface dynamic images and the inner layer dynamic information to obtain relative displacement change data of the surface and the inner layer of the risk sub-region, including:

[0015] Based on the region boundary position information of the risk sub-region, full-range cruising scanning and scanning radar detection are performed on the risk sub-region to obtain surface soil dynamic images and inner layer soil movement state information of the risk sub-region, respectively;

[0016] The dynamic image of the surface layer soil is subjected to frame contour recognition to obtain contour position change characteristic information of the surface layer soil of the risk sub-region; based on the contour position change characteristic information, movement speed and movement direction information of the surface layer soil in a unit time are determined;

[0017] The movement state information of the inner layer soil is analyzed to obtain movement speed and movement direction information of the inner layer soil of the risk sub-region in a unit time;

[0018] The movement speed and movement direction information of the surface layer soil and the inner layer soil in a unit time are compared to obtain movement speed difference and movement direction included angle between the surface layer soil and the inner layer soil in a unit time, which are used as the relative displacement change data.

[0019] Optionally, based on the relative displacement change data, a covered area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-region are predicted; based on the covered area and the landslide movement direction, a corresponding notification message is generated and sent, including:

[0020] The movement speed difference is compared with a preset speed difference threshold value and the movement direction included angle is compared with a preset included angle threshold value; if the movement speed difference is greater than the preset speed difference threshold value or the movement direction included angle is greater than the preset included angle threshold value, the risk sub-region is determined as a whole as a covered area corresponding to a landslide event, and an angle range corresponding to the movement direction included angle is determined as a landslide movement direction range; otherwise, the risk sub-region is not determined as a covered area corresponding to a landslide event.

[0021] Based on the location range of the risk sub-region where a landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal.

[0022] A landslide prediction system based on image processing, comprising:

[0023] A remote sensing shooting and analysis module is configured to remotely sense a mountain area to obtain a mountain area remote sensing image, and analyze the mountain area remote sensing image to obtain surface external characteristic information and soil internal characteristic information of the mountain area;

[0024] A risk sub-region identification module is configured to identify a risk sub-region existing in the mountain area based on the surface external characteristic information and the soil internal characteristic information;

[0025] A cruise detection and analysis module is configured to cruise and detect the risk sub-region to obtain surface dynamic image and inner layer dynamic information of the risk sub-region;

[0026] a displacement change data determination module configured to analyze the surface layer dynamic image and the inner layer dynamic information to obtain relative displacement change data of the surface layer and the inner layer of the risk sub-region;

[0027] a landslide event prediction module configured to predict, based on the relative displacement change data, a coverage area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-region;

[0028] a notification execution module configured to generate and send a corresponding notification message based on the coverage area and the landslide movement direction.

[0029] Optionally, the remote sensing photographing and analyzing module is configured to perform remote sensing photographing on the mountain region to obtain a mountain region remote sensing image; and analyze the mountain region remote sensing image to obtain surface external characteristic information and soil internal characteristic information of the mountain region, including:

[0030] perform visible spectrum analysis on the mountain region remote sensing image to obtain surface vegetation distribution characteristic information of the mountain region; and perform far-infrared spectrum analysis on the mountain region remote sensing image to obtain soil water content characteristic information of the mountain region;

[0031] the risk sub-region identification module is configured to identify, based on the surface external characteristic information and the soil internal characteristic information, a risk sub-region existing in the mountain region, including:

[0032] determine vegetation distribution sparsity density characteristic information of the entire surface range of the mountain region based on the surface vegetation distribution characteristic information; and determine distribution position information of all soil sub-regions in a water content saturation state within the entire soil range of the mountain region based on the soil water content characteristic information;

[0033] identify all risk sub-regions existing in the mountain region based on the vegetation distribution sparsity density characteristic information and the distribution position information of all soil sub-regions in a water content saturation state.

[0034] Optionally, the cruise detection and analysis module is configured to perform cruise detection on the risk sub-region to obtain surface layer dynamic image and inner layer dynamic information of the risk sub-region, including:

[0035] perform full-range cruise scanning photographing and scanning radar detection on the risk sub-region based on the region boundary position information of the risk sub-region to respectively obtain surface layer soil dynamic image and inner layer soil movement state information of the risk sub-region;

[0036] The contour position change characteristic information of the surface soil of the risk sub-area is obtained by performing picture contour recognition on the surface soil dynamic image.

[0037] The moving speed and moving direction information of the inner layer soil of the risk sub-area in a unit time are obtained by analyzing the inner layer soil movement state information.

[0038] The displacement change data determination module is configured to analyze the surface layer dynamic image and the inner layer dynamic information to obtain relative displacement change data of the surface layer and the inner layer of the risk sub-area, including:

[0039] The moving speed difference and the moving direction angle between the surface layer soil and the inner layer soil in a unit time are obtained by comparing the moving speed and moving direction information of the surface layer soil and the inner layer soil in a unit time, which are used as the relative displacement change data.

[0040] Optionally, the landslide event prediction module is configured to predict the coverage area and the landslide movement direction corresponding to the risk sub-area based on the relative displacement change data, including:

[0041] The moving speed difference is compared with a preset speed difference threshold value, and the moving direction angle is compared with a preset angle threshold value, if the moving speed difference is greater than the preset speed difference threshold value or the moving direction angle is greater than the preset angle threshold value, the risk sub-area is determined as a whole to be the coverage area corresponding to the landslide event, and the angle range corresponding to the moving direction angle is determined as the landslide movement direction range; otherwise, the risk sub-area is not determined as the coverage area corresponding to the landslide event.

[0042] The notification execution module is configured to generate and send a corresponding notification message based on the coverage area and the landslide movement direction, including:

[0043] Based on the location range of the risk sub-area where the landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal.

[0044] Compared with the prior art, the present application has the following advantages:

[0045] The landslide prediction method and system based on image processing provided in the application shoot and analyze remote sensing images of mountain areas, obtain surface external characteristic information and soil internal characteristic information of the mountain areas, identify risk sub-areas existing in the mountain areas, thereby enabling wide-range surface and internal soil state identification of the mountain areas, and accurately determining potential risk areas of the mountain areas caused by vegetation and rainfall; the risk sub-areas are also subjected to cruise detection, surface dynamic images and internal dynamic information of the risk sub-areas are obtained, relative displacement change data between the surface and the internal layer of the risk sub-areas are analyzed and obtained, and the soil layer movement of the risk sub-areas is accurately identified; based on the relative displacement change data, the covered area and the landslide movement direction corresponding to the landslide event of the risk sub-areas are predicted, full-range and early prediction of the landslide area is realized, and the accuracy and reliability of the mountain landslide prediction are improved. BRIEF DESCRIPTION OF DRAWINGS

[0046] 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 needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor. Among them:

[0047] Figure 1 The flowchart of the landslide prediction method based on image processing provided in the present application.

[0048] Figure 2 The structure diagram of the landslide prediction system based on image processing provided in the present application. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings. It can be understood that the specific embodiments described here are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, in order to facilitate the description, only the parts related to the present application are shown in the drawings, but not all the structures. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0050] The terms "comprising" and "having" and any variations thereof in the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units not listed, or optionally also includes other steps or units inherent to the process, method, product or device.

[0051] Reference to“an embodiment” herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase“in one embodiment” in various places in the specification are not necessarily referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. As will be understood by those skilled in the art, an embodiment described herein can be incorporated into other embodiments.

[0052] Referring to Figure 1 An embodiment of the application provides a landslide prediction method based on image processing. The landslide prediction method based on image processing comprises:

[0053] Remote sensing is performed on a mountain area to obtain a remote sensing image of the mountain area; the remote sensing image of the mountain area is analyzed to obtain surface external characteristic information and soil internal characteristic information of the mountain area; and a risk sub-area existing in the mountain area is identified based on the surface external characteristic information and the soil internal characteristic information.

[0054] Cruise detection is performed on the risk sub-area to obtain surface layer dynamic images and internal layer dynamic information of the risk sub-area; and relative displacement change data of the surface layer and the internal layer of the risk sub-area are obtained by analyzing the surface layer dynamic images and the internal layer dynamic information.

[0055] Based on the relative displacement change data, a covered area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-area are predicted; and a corresponding notification message is generated and sent based on the covered area and the landslide movement direction.

[0056] The landslide prediction method based on image processing can obtain surface external characteristic information and soil internal characteristic information of a mountain area by performing remote sensing and analysis on a remote sensing image of the mountain area, and identify a risk sub-area existing in the mountain area based on the information, so that the surface and internal soil state of the mountain area can be identified in a large range, and a potential risk area of the mountain area caused by vegetation and rainfall can be accurately determined. Cruise detection is performed on the risk sub-area to obtain surface layer dynamic images and internal layer dynamic information of the risk sub-area, and relative displacement change data between the surface layer and the internal layer of the risk sub-area are obtained by analyzing the surface layer dynamic images and the internal layer dynamic information, so that the soil layer movement of the risk sub-area can be accurately identified. Based on the relative displacement change data, a covered area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-area are predicted, so that full-range and advance prediction of the landslide area can be realized, and the accuracy and reliability of the mountain landslide prediction can be improved.

[0057] In another embodiment, a mountain area is remotely sensed to obtain a mountain area remote sensing image; the mountain area remote sensing image is analyzed to obtain surface external characteristic information and soil internal characteristic information of the mountain area; based on the surface external characteristic information and the soil internal characteristic information, a risk sub-area existing in the mountain area is identified, including:

[0058] Based on the region boundary position information of the mountain area, the mountain area is directionally remotely sensed to obtain a mountain area remote sensing image;

[0059] The mountain area remote sensing image is analyzed in a visible spectrum to obtain surface vegetation distribution characteristic information of the mountain area; the mountain area remote sensing image is analyzed in a far infrared spectrum to obtain soil water content characteristic information of the mountain area;

[0060] Based on the surface vegetation distribution characteristic information, vegetation distribution sparse density characteristic information of the entire surface range of the mountain area is determined; based on the soil water content characteristic information, distribution position information of all soil sub-areas in a water content saturation state in the entire soil range of the mountain area is determined;

[0061] Based on the vegetation distribution sparse density characteristic information and the distribution position information of all soil sub-areas in a water content saturation state, all risk sub-areas existing in the mountain area are identified.

[0062] The beneficial effects of the above embodiments are that, in actual monitoring, the size and position of the mountain area region are determined based on the position of the region boundary of the mountain area region, and the mountain area region is remotely sensed in the visible spectrum band and the far infrared spectrum band to obtain a mountain area region remote sensing image with double spectrum band characteristics, which can accurately visually identify the mountain area region in the visible spectrum band and the far infrared spectrum band. The stability of the soil layer of the mountain area region is closely related to the vegetation coverage and soil water content of the mountain area region. The more dense the vegetation coverage of the mountain area region, the more solid the soil layer. The higher the soil water content of the mountain area region, the more loose the soil, and the less solid the soil layer. Identifying the vegetation coverage and soil water content of the mountain area region can effectively determine whether the soil layer of the mountain area region has a potential risk of landslide. At the same time, the degree of ground vegetation coverage of the mountain area region can be analyzed through the visible spectrum band in the remote sensing image. The more dense the ground vegetation coverage, the greater the intensity of the green band component in the visible spectrum band in the remote sensing image, and the wider the distribution. In addition, water has a strong absorption rate of far infrared light. When the soil water content of the mountain area region is large, the soil absorbs far infrared light with greater intensity. By analyzing the intensity and distribution of the far infrared spectrum band in the remote sensing image, the soil water content state of the mountain area region can be accurately identified. Therefore, the visible spectrum analysis is performed on the mountain area region remote sensing image to obtain the ground vegetation distribution characteristic information of the mountain area region. The far infrared spectrum analysis is performed on the mountain area region remote sensing image to obtain the soil water content characteristic information of the mountain area region. Based on the ground vegetation distribution characteristic information, the vegetation distribution density characteristic information of the entire ground surface range of the mountain area region is determined. Based on the soil water content characteristic information, the distribution position information of all soil sub-regions in the saturated state of the soil water content in the entire soil range of the mountain area region is determined, so as to quantitatively analyze the vegetation coverage and soil water content of the mountain area region. The soil sub-region with a vegetation distribution density less than a preset density threshold or in a saturated limit state of soil water content is determined as a corresponding risk sub-region, and the risk sub-region is positioned and identified, which facilitates subsequent cruise detection of the risk sub-region by the unmanned aerial vehicle for accurate orientation, thereby improving the efficiency of cruise detection.

[0063] In another embodiment, the risk sub-region is cruise detected to obtain surface layer dynamic images and inner layer dynamic information of the risk sub-region; the surface layer dynamic images and the inner layer dynamic information are analyzed to obtain relative displacement change data of the surface layer and the inner layer of the risk sub-region, including:

[0064] Based on the region boundary position information of the risk sub-region, the risk sub-region is full-range cruise scanned and scanned radar detected to obtain surface layer soil dynamic images and inner layer soil movement state information of the risk sub-region;

[0065] The contour position change characteristic information of the surface soil of the risk sub-region is obtained by performing frame contour recognition on the surface soil dynamic image; and the moving speed and moving direction information of the surface soil in a unit time are determined based on the contour position change characteristic information;

[0066] The moving speed and moving direction information of the inner layer soil in a unit time are obtained by analyzing the inner layer soil movement state information of the risk sub-region;

[0067] The moving speed difference and moving direction angle between the surface soil and the inner layer soil in a unit time are obtained by comparing the moving speed and moving direction information of the surface soil and the inner layer soil in a unit time, which are used as the relative displacement change data.

[0068] The beneficial effects of the above embodiments are that, based on the region boundary position of the risk sub-region, the flight equipment such as a drone is used to perform cruise detection on the risk sub-region, which can include but is not limited to full-range cruise scanning and scanning radar detection of the risk sub-region, and the soil state of the risk sub-region is identified from two different aspects. Specifically, by performing full-range cruise scanning and scanning radar detection on the risk sub-region, the surface soil dynamic image and the inner layer soil movement state information of the risk sub-region are obtained, respectively, wherein the scanning and photographing can identify the surface soil state of the risk sub-region, and the scanning radar detection can identify the movement state of the inner layer soil of the risk sub-region. The contour position change characteristic information of the surface soil of the risk sub-region is obtained by performing frame contour recognition on the surface soil dynamic image, and the moving speed and moving direction information of the surface soil in a unit time are determined based on the contour position change characteristic information; and the moving speed and moving direction information of the inner layer soil in a unit time are obtained by analyzing the inner layer soil movement state information of the risk sub-region, which can distinguish and identify the moving state of the surface soil and the inner layer soil of the risk sub-region. Then, the moving speed and moving direction information of the surface soil and the inner layer soil in a unit time are compared, and the moving speed difference and moving direction angle between the surface soil and the inner layer soil in a unit time are obtained, which can quantitatively identify the movement difference of the soil layers at different depths in the risk sub-region, and provide a reliable basis for subsequent determination of whether a landslide event occurs in the risk sub-region.

[0069] In another embodiment, based on the relative displacement change data, the coverage area and the landslide movement direction corresponding to the occurrence of a landslide event in the risk sub-region are predicted; and based on the coverage area and the landslide movement direction, a corresponding notification message is generated and sent, including:

[0070] comparing the moving speed difference with a preset speed difference threshold and comparing the moving direction angle with a preset angle threshold, if the moving speed difference is greater than the preset speed difference threshold or the moving direction angle is greater than the preset angle threshold, the risk sub-region is determined as a whole as a coverage region corresponding to a landslide event, and an angle range corresponding to the moving direction angle is determined as a landslide movement direction range; otherwise, the risk sub-region is not determined as the coverage region corresponding to the landslide event;

[0071] Based on the location range of the risk sub-region where the landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal.

[0072] The above embodiments have the beneficial effects that the moving speed difference is compared with the preset speed difference threshold and the moving direction angle is compared with the preset angle threshold, if the moving speed difference is greater than the preset speed difference threshold or the moving direction angle is greater than the preset angle threshold, the risk sub-region is determined as a whole as a coverage region corresponding to a landslide event, and an angle range corresponding to the moving direction angle is determined as a landslide movement direction range, which can identify the influence range and movement coverage direction range of the landslide event occurring in the risk sub-region. Based on the location range of the risk sub-region where the landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal, wherein the notification message can include but is not limited to the location range of the risk sub-region where the landslide event occurs and the landslide movement direction range, and the user terminal can be but is not limited to a mobile terminal such as a smart phone held by a resident located in the influence range of the landslide event, improving the timeliness and accuracy of landslide warning.

[0073] Please refer to Figure 2 An embodiment of the present application provides a landslide prediction system based on image processing. The landslide prediction system based on image processing comprises:

[0074] A remote sensing shooting and analyzing module is configured to remotely sense a mountain area to obtain a remote sensing image of the mountain area, and analyze the remote sensing image of the mountain area to obtain surface external characteristic information and soil internal characteristic information of the mountain area;

[0075] A risk sub-region identifying module is configured to identify a risk sub-region existing in the mountain area based on the surface external characteristic information and the soil internal characteristic information;

[0076] A cruise detection and analysis module is configured to cruise and detect the risk sub-region to obtain surface dynamic image and internal dynamic information of the risk sub-region;

[0077] a displacement change data determination module configured to analyze the surface layer dynamic image and the inner layer dynamic information to obtain relative displacement change data of the surface layer and the inner layer of the risk sub-region;

[0078] a landslide event prediction module configured to predict, based on the relative displacement change data, a coverage area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-region;

[0079] a notification execution module configured to generate and send a corresponding notification message based on the coverage area and the landslide movement direction.

[0080] The landslide prediction system based on image processing has the advantages that the system captures and analyzes remote sensing images of a mountain region to obtain surface external characteristic information and soil internal characteristic information of the mountain region, and identifies risk sub-regions existing in the mountain region based on the information, so that the system can identify the surface and internal soil state of the mountain region on a large scale, and accurately determine potential risk regions of the mountain region caused by vegetation and rainfall. The system also performs cruise detection on the risk sub-regions to obtain surface layer dynamic images and inner layer dynamic information of the risk sub-regions, and analyzes the relative displacement change data between the surface layer and the inner layer of the risk sub-regions based on the information, so as to accurately identify the soil layer movement of the risk sub-regions. Furthermore, the system predicts, based on the relative displacement change data, a coverage area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-regions, so as to realize full-range and early prediction of the landslide region, and improve the accuracy and reliability of the mountain landslide prediction.

[0081] In another embodiment, the remote sensing capturing and analyzing module is configured to capture remote sensing images of a mountain region to obtain mountain region remote sensing images, and analyze the mountain region remote sensing images to obtain surface external characteristic information and soil internal characteristic information of the mountain region, including:

[0082] performing visible spectrum analysis on the mountain region remote sensing images to obtain surface vegetation distribution characteristic information of the mountain region, and performing far-infrared spectrum analysis on the mountain region remote sensing images to obtain soil water content characteristic information of the mountain region;

[0083] The risk sub-region identification module is configured to identify risk sub-regions existing in the mountain region based on the surface external characteristic information and the soil internal characteristic information, including:

[0084] determining vegetation distribution sparse density characteristic information of the entire surface of the mountain region based on the surface vegetation distribution characteristic information, and determining distribution position information of all soil sub-regions in a water content saturation state within the entire soil of the mountain region based on the soil water content characteristic information;

[0085] Based on the vegetation distribution sparse density feature information and the distribution position information of all soil sub-regions in the water content saturation state, all risk sub-regions existing in the mountain region are identified.

[0086] The beneficial effects of the above embodiments are that, in actual monitoring, the size and position of the region range of the mountain region are determined based on the position of the region boundary of the mountain region, and the mountain region is shot by remote sensing in the visible spectrum band and the far infrared spectrum band to obtain a mountain region remote sensing image with double spectrum band characteristics, so that the mountain region can be accurately visually identified in the visible spectrum band and the far infrared spectrum band. Whether the soil layer of the mountain region is stable is closely related to the vegetation coverage and soil water content of the mountain region. When the vegetation coverage of the mountain region is more dense, the corresponding soil layer is more solid; when the soil water content of the mountain region is higher, the soil becomes more loose, and the corresponding soil layer is less solid. Therefore, identifying the vegetation coverage and soil water content of the mountain region can effectively determine whether the soil layer of the mountain region has a potential risk of landslide. At the same time, the degree of surface vegetation coverage of the mountain region can be analyzed in the visible spectrum band of the remote sensing image. When the surface vegetation coverage is more dense, the composition intensity of the green band in the visible spectrum band of the remote sensing image is greater and the distribution is wider. In addition, water has a strong absorption rate for far infrared light. When the soil water content of the mountain region is large, the corresponding soil absorption intensity of far infrared light is also greater. By analyzing the intensity and distribution of the far infrared spectrum band in the remote sensing image, the soil water content state of the mountain region can be accurately identified. Therefore, the visible spectrum analysis is performed on the mountain region remote sensing image to obtain the surface vegetation distribution feature information of the mountain region; the far infrared spectrum analysis is performed on the mountain region remote sensing image to obtain the soil water content feature information of the mountain region; then, based on the surface vegetation distribution feature information, the vegetation distribution sparse density feature information of the entire surface range of the mountain region is determined; based on the soil water content feature information, the distribution position information of all soil sub-regions in the water content saturation state in the entire soil range of the mountain region is determined, so as to quantitatively analyze the vegetation coverage and soil water content of the mountain region. Then, the soil sub-region with a vegetation distribution density less than a preset density threshold or in a saturated limit state of soil water content is determined as a corresponding risk sub-region, and the risk sub-region is positioned and identified, so as to facilitate subsequent cruise detection of the risk sub-region by the unmanned aerial vehicle for accurate orientation, thereby improving the efficiency of cruise detection.

[0087] In another embodiment, the cruise detection analysis module is configured to perform cruise detection on the risk sub-region to obtain surface layer dynamic images and inner layer dynamic information of the risk sub-region, including:

[0088] Based on the risk sub-area boundary position information, the risk sub-area is subjected to full-range cruise scanning shooting and scanning radar detection, and surface soil dynamic image and inner soil movement state information of the risk sub-area are obtained respectively;

[0089] The surface soil dynamic image is subjected to picture contour recognition, and contour position change characteristic information of the surface soil of the risk sub-area is obtained; based on the contour position change characteristic information, moving speed and moving direction information of the surface soil within a unit time are determined;

[0090] The inner soil movement state information is subjected to analysis, and moving speed and moving direction information of the inner soil of the risk sub-area within a unit time are obtained;

[0091] The displacement change data determination module is used for analyzing the surface dynamic image and the inner dynamic information, and obtaining relative displacement change data of the surface and the inner layer of the risk sub-area, including:

[0092] The moving speed and moving direction information of the surface soil and the inner soil within a unit time are compared, and moving speed difference and moving direction included angle between the surface soil and the inner soil within a unit time are obtained, which are used as the relative displacement change data

[0093] The above embodiment has the beneficial effect that, based on the position of the area boundary of the risk sub-region, the flight device such as a UAV is used to perform cruise detection on the risk sub-region. The cruise detection can include, but is not limited to, full-range cruise scanning and shooting and scanning radar detection on the risk sub-region, to identify the soil state of the risk sub-region from two different aspects. Specifically, by performing full-range cruise scanning and shooting and scanning radar detection on the risk sub-region, the surface soil dynamic image and the inner soil movement state information of the risk sub-region are obtained, respectively. The scanning and shooting can identify the surface soil state of the risk sub-region, and the scanning radar detection can identify the movement state of the inner soil of the risk sub-region. By performing picture contour identification on the surface soil dynamic image, the contour position change feature information of the surface soil of the risk sub-region is obtained, to determine the moving speed and moving direction information of the surface soil within a unit time; and by analyzing the inner soil movement state information, the moving speed and moving direction information of the inner soil of the risk sub-region within a unit time is obtained, so that the moving states of the surface soil and the inner soil of the risk sub-region can be distinguished and identified. Then, the moving speed difference and the moving direction angle between the surface soil and the inner soil within a unit time are obtained by comparing the moving speed and moving direction information of the surface soil and the inner soil within a unit time, so that the movement difference of the soil layers at different depths in the risk sub-region can be quantitatively identified, to provide a reliable basis for subsequent determination of whether a landslide event occurs in the risk sub-region.

[0094] In another embodiment, the landslide event prediction module is configured to predict, based on the relative displacement change data, a coverage area and a landslide movement direction corresponding to a landslide event occurring in the risk sub-region, including:

[0095] The moving speed difference is compared with a preset speed difference threshold value, and the moving direction angle is compared with a preset angle threshold value. If the moving speed difference is greater than the preset speed difference threshold value or the moving direction angle is greater than the preset angle threshold value, the risk sub-region is determined as a whole as the coverage area corresponding to the landslide event, and an angle range corresponding to the moving direction angle is determined as the landslide movement direction range. Otherwise, the risk sub-region is not determined as the coverage area corresponding to the landslide event.

[0096] The notification execution module is configured to generate and send a corresponding notification message based on the coverage area and the landslide movement direction, including:

[0097] Based on the position range of the risk sub-region where the landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal.

[0098] The above embodiment has the beneficial effect that the movement speed difference is compared with a preset speed difference threshold value, and the movement direction angle is compared with a preset angle threshold value, if the movement speed difference is greater than the preset speed difference threshold value or the movement direction angle is greater than the preset angle threshold value, the risk sub-region is determined as a whole as a landslide event corresponding coverage area, and the angle range corresponding to the movement direction angle is determined as a landslide movement direction range, which can identify the landslide event influence range and the movement coverage direction range of the risk sub-region. Based on the location range of the risk sub-region where the landslide event occurs and the landslide movement direction range, a corresponding notification message is generated and sent to a corresponding user terminal, wherein the notification message can include but is not limited to the location range of the risk sub-region where the landslide event occurs and the landslide movement direction range, and the user terminal can be but is not limited to a mobile terminal such as a smart phone held by a resident located in the landslide event influence range, improving the timeliness and accuracy of landslide warning.

[0099] Overall, the landslide prediction method and system based on image processing shoot and analyze remote sensing images of mountainous regions to obtain surface external feature information and soil internal feature information of the mountainous regions, so as to identify risk sub-regions existing in the mountainous regions, which can identify the surface and internal soil state of the mountainous regions on a large scale, accurately determine the potential risk area of the mountainous region caused by vegetation and rainfall; the surface layer dynamic image and the internal layer dynamic information of the risk sub-region are obtained by cruise detection, and the relative displacement change data between the surface layer and the internal layer of the risk sub-region is obtained by analysis, so as to accurately identify the soil layer movement of the risk sub-region; based on the relative displacement change data, the coverage area and the landslide movement direction corresponding to the landslide event of the risk sub-region are predicted, the full range and early prediction of the landslide area are realized, and the accuracy and reliability of the mountain landslide prediction are improved.

[0100] The above is only one specific embodiment of the present application, and any improvement made on the basis of the concept of the present application is considered to be within the scope of protection of the present application.

Claims

1. A landslide prediction method based on image processing, characterized in that, include: Step S1: Take remote sensing images of the mountain area to obtain remote sensing images of the mountain area; analyze the remote sensing images of the mountain area to obtain the external surface features and internal soil features of the mountain area. Based on the external surface features and the internal soil features, risk sub-regions within the mountainous area are identified. Step S2: Conduct cruise detection on the risk sub-region to obtain surface dynamic images and inner dynamic information of the risk sub-region; analyze the surface dynamic images and inner dynamic information to obtain relative displacement change data of the surface and inner layers of the risk sub-region; Step S3: Based on the relative displacement change data, predict the coverage area and landslide movement direction corresponding to the landslide event in the risk sub-region; Based on the coverage area and the direction of landslide movement, generate and send corresponding notification messages; Step S1 includes: Based on the location information of the mountain area's boundary, directional remote sensing scanning and imaging are performed on the mountain area to obtain remote sensing images of the mountain area. Visible spectral analysis was performed on the remote sensing images of the mountain area to obtain the surface vegetation distribution characteristics of the mountain area; far-infrared spectral analysis was performed on the remote sensing images of the mountain area to obtain the soil moisture content characteristics of the mountain area. Based on the surface vegetation distribution characteristics, the vegetation density characteristics of the entire surface area of ​​the mountain region are determined; based on the soil moisture content characteristics, the distribution location information of all soil sub-regions in the soil saturation state within the entire soil area of ​​the mountain region is determined. Based on the vegetation distribution density characteristics and the location information of all soil sub-regions in a water-saturated state, all risk sub-regions in the mountain area are identified. Step S2 includes: Based on the regional boundary location information of the risk sub-region, a full-range cruise scan and scanning radar detection are performed on the risk sub-region to obtain dynamic images of the surface soil and information on the movement state of the inner soil in the risk sub-region, respectively. The dynamic image of the surface soil is subjected to image contour recognition to obtain the contour position change feature information of the surface soil in the risk sub-region; based on the contour position change feature information, the moving speed and moving direction information of the surface soil per unit time are determined. The movement state information of the inner soil layer is analyzed to obtain the movement speed and direction information of the inner soil layer in the risk sub-region per unit time. The moving speed and direction information of the surface soil and the inner soil are compared per unit time to obtain the difference in moving speed and the angle of moving direction between the surface soil and the inner soil per unit time, which are used as the relative displacement change data. Step S3 includes: Step 31: Compare the moving speed difference with a preset speed difference threshold and the moving direction angle with a preset angle threshold. If the moving speed difference is greater than the preset speed difference threshold or the moving direction angle is greater than the preset angle threshold, then the entire risk sub-region is determined as the coverage area corresponding to the landslide event, and the angle range corresponding to the moving direction angle is determined as the landslide movement direction range; otherwise, the risk sub-region is not determined as the coverage area corresponding to the landslide event. Step 32: Based on the location range of the risk sub-region where the landslide event occurred and the range of the landslide movement direction, generate a corresponding notification message and send the notification message to the corresponding user terminal.

2. A landslide prediction system based on image processing, employing the landslide prediction method based on image processing as described in claim 1, characterized in that, include: The remote sensing imaging and analysis module is used to perform remote sensing imaging of the mountain area to obtain remote sensing images of the mountain area; and to analyze the remote sensing images of the mountain area to obtain external surface feature information and internal soil feature information of the mountain area. The risk sub-region identification module is used to identify risk sub-regions existing in the mountain area based on the external surface feature information and the internal soil feature information; The cruise detection and analysis module is used to conduct cruise detection on the risk sub-region to obtain surface dynamic images and inner dynamic information of the risk sub-region. The displacement change data determination module is used to analyze the surface dynamic image and the inner layer dynamic information to obtain the relative displacement change data of the surface and inner layer of the risk sub-region. The landslide event prediction module is used to predict the coverage area and landslide movement direction corresponding to the landslide event in the risk sub-region based on the relative displacement change data. The notification execution module is used to generate and send corresponding notification messages based on the coverage area and the direction of landslide movement; The remote sensing imaging and analysis module is used to perform remote sensing imaging of the mountain area to obtain remote sensing images of the mountain area; and to analyze the remote sensing images of the mountain area to obtain external surface feature information and internal soil feature information of the mountain area, including: Visible spectral analysis was performed on the remote sensing images of the mountain area to obtain the surface vegetation distribution characteristics of the mountain area; far-infrared spectral analysis was performed on the remote sensing images of the mountain area to obtain the soil moisture content characteristics of the mountain area. The risk sub-region identification module is used to identify risk sub-regions existing in the mountain area based on the external surface feature information and the internal soil feature information, including: Based on the surface vegetation distribution characteristics, the vegetation density characteristics of the entire surface area of ​​the mountain region are determined; based on the soil moisture content characteristics, the distribution location information of all soil sub-regions in the soil saturation state within the entire soil area of ​​the mountain region is determined. Based on the vegetation distribution density characteristics and the location information of all soil sub-regions in a water-saturated state, all risk sub-regions in the mountain area are identified. The cruise detection and analysis module is used to conduct cruise detection on the risk sub-region to obtain surface dynamic images and inner dynamic information of the risk sub-region, including: Based on the regional boundary location information of the risk sub-region, a full-range cruise scan and scanning radar detection are performed on the risk sub-region to obtain dynamic images of the surface soil and information on the movement state of the inner soil in the risk sub-region, respectively. The dynamic image of the surface soil is subjected to image contour recognition to obtain the contour position change feature information of the surface soil in the risk sub-region; based on the contour position change feature information, the moving speed and moving direction information of the surface soil per unit time are determined. The movement state information of the inner soil layer is analyzed to obtain the movement speed and direction information of the inner soil layer in the risk sub-region per unit time. The displacement change data determination module is used to analyze the surface dynamic image and the inner layer dynamic information to obtain the relative displacement change data of the surface and inner layer of the risk sub-region, including: The moving speed and direction information of the surface soil and the inner soil are compared per unit time to obtain the difference in moving speed and the angle of moving direction between the surface soil and the inner soil per unit time, which are used as the relative displacement change data. The landslide event prediction module is used to predict the coverage area and landslide movement direction corresponding to a landslide event in the risk sub-region based on the relative displacement change data, including: The moving speed difference is compared with a preset speed difference threshold, and the moving direction angle is compared with a preset angle threshold. If the moving speed difference is greater than the preset speed difference threshold or the moving direction angle is greater than the preset angle threshold, then the entire risk sub-region is determined as the coverage area corresponding to the landslide event, and the angle range corresponding to the moving direction angle is determined as the landslide movement direction range; otherwise, the risk sub-region is not determined as the coverage area corresponding to the landslide event. The notification execution module is used to generate and send corresponding notification messages based on the coverage area and the landslide movement direction, including: Based on the location range of the risk sub-region where the landslide event occurred and the range of the landslide movement direction, a corresponding notification message is generated and sent to the corresponding user terminal.

Images