Landslide deformation image detection and early warning method and system

Abstract

The invention discloses a landslide deformation image detection and early warning method and system, and relates to the technical field of image processing, and the method comprises the steps: analyzing the influence of the position of a potential sliding surface of a side slope on the side slope to determine the non-part position of the side slope, analyzing the influence of the potential sliding surface on the side slope to determine the area position of the affected outer surface of the side slope, and carrying out the early warning. And the reliability of slope landslide phenomenon analysis is improved. A slope image feature sequence on each slope position is constructed, and a reliable basis is provided for subsequent integration of slope deformation content and slope positions. Integrating slope deformation contents to obtain a deformation description set on each slope position; the deformation description sets on all the slope positions are integrated according to the association among different slope positions to perform early warning and prediction of the landslide phenomenon, and the deformation association among different slope positions is considered to perform integral detection and early warning on the slope landslide phenomenon, so that the precision and adaptability of slope landslide detection and early warning are improved; and the slope landslide phenomenon and landslide protection are effectively monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a landslide deformation image detection and early warning method and system. Background Art

[0002] As a common geological disaster, a landslide is often accompanied by the deformation of a mountain body or slope, posing a serious threat to the safety of people's lives and property. With the rapid development of fields such as remote sensing technology, image processing technology, and artificial intelligence, the detection and early warning of landslide deformation images have become possible. By means of satellite remote sensing images, unmanned aerial vehicle (UAV) aerial photography, etc., high-resolution images of the landslide area are obtained, and image analysis algorithms are used to automatically calculate the displacement information of the landslide body, realizing the real-time monitoring of landslide deformation. At the same time, by combining multi-source information such as meteorological data and geological conditions, a landslide prediction model is established to improve the accuracy and reliability of early warning. The implementation of the landslide deformation image detection and early warning scheme is of great significance for timely discovering landslide hazards, giving early warnings, and taking measures to reduce disaster losses.

[0003] In the prior art, the possibility of a landslide is analyzed only based on the deformation conditions in the images of certain parts of the slope, but all parts of the slope cannot be effectively combined, resulting in poor accuracy and low adaptability of slope landslide detection and early warning, and the natural safety of the slope cannot be effectively guaranteed.

[0004] Therefore, how to improve the accuracy and adaptability of slope landslide detection and early warning is a technical problem to be solved at present. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of poor accuracy and low adaptability in slope landslide detection and early warning in the prior art, and a landslide deformation image detection and early warning method is proposed. The method includes:

[0006] Obtain slope structure information and the position of the potential sliding surface of the slope, define the position of the slope part, analyze the influence of the position of the potential sliding surface of the slope on the slope to determine the non-slope part position, and set image monitoring points according to the slope part position and the non-slope part position, so as to collect the slope image content at different positions of the slope;

[0007] Obtain slope images at different positions of the slope within a period of time, extract the image features of the slope images at each slope position, and construct an image feature sequence of the slope images at each slope position;

[0008] Analyze the image feature sequence of the slope images at each slope position to determine the slope deformation content at each slope position, and integrate the slope deformation content to obtain a deformation description set at each slope position;

[0009] Integrate the deformation description sets at all slope positions based on the associations between different slope positions to issue early warnings and predictions for the landslide phenomenon of the slope.

[0010] In some embodiments of the present application, analyze the influence of the potential slip surface position of the slope on the slope to determine the non-site positions of the slope, including,

[0011] Collect slope geological data, and identify the potential slip surface positions and slip data on the slope through the slope geological data. The slip data includes the slip direction, slip speed, and slip distance;

[0012] Input the slip direction, slip speed, slip distance, and potential slip surface position into numerical simulation software to simulate the sliding process of the slip surface and obtain the simulation results;

[0013] The simulation results include the slip path and regional stress-strain related data. Statistically analyze the slip path to form a slip influence zone, and split the slip influence zone according to the geological soil categories to obtain multiple small slip influence zones. Determine the potential slip surfaces affected by each small slip influence zone and the safety factor of the potential slip surface, and match the regional stress-strain related data to each small slip influence zone according to the position;

[0014] Evaluate the stability index of the small slip influence zone according to the safety factor of the slip surface and the regional stress-strain related data of the small slip influence zone;

[0015]

[0016] Among them, is the stability index of the i1-th small slip influence zone, is the stable conversion coefficient, n1 is the category data of the regional stress-strain related data of the i1-th small slip influence zone, is the combined weight of the regional stress-strain related data of the i2-th category data, is the average value of the regional stress-strain related data of the i2-th category data, n2 is the number of potential slip surfaces affected by the i1-th small slip influence zone, is the influence weight of the i3-th potential slip surface affected by the i1-th small slip influence zone, is the safety factor of the i3-th potential slip surface affected by the i1-th small slip influence zone, and k1 is a preset constant;

[0017] Superimpose all the small slip influence zones according to the stability index to obtain the unstable outer surface area of the slope, and use the unstable outer surface area of the slope as the non-site position of the slope.

[0018] In some embodiments of the present application, set image monitoring points according to the slope site positions and slope non-site positions, including,

[0019] Both the slope part position and the non-slope part position are position intervals of a certain part on the slope. Calculate the respective area sizes according to the position intervals of the slope part position and the non-slope part position, deploy image monitoring points for the slope part position through a preset first layout strategy, and deploy image monitoring points for the non-slope part position through a preset second layout strategy.

[0020] In some embodiments of the present application, extract the image features of the slope images at each slope position, and construct the slope image feature sequences at each slope position, including,

[0021] The image features of the slope images include color features, texture features, shape features, and other features. Confirm the target objects in the slope images and the combination of image feature types required for the target objects, and combine a preset target object recognition algorithm to identify the contour of the target objects in the slope images;

[0022] Integrate the color features, texture features, shape features, and other features according to the contour of the target objects, and construct the slope image feature sequences of each target object contour at each slope position over time.

[0023] In some embodiments of the present application, analyze the slope image feature sequences at each slope position to determine the slope deformation content at each slope position, including,

[0024] The slope deformation content includes the contour deformation content of the target objects and the non-contour deformation content of the target objects;

[0025] Calculate the geometric features of each target object contour on the slope image feature sequences at each slope position, and convert the geometric features into shape descriptors, and calculate the contour deformation content of each target object contour at each slope position through the shape descriptors;

[0026] Calculate the change in the image features inside each target object contour on the slope image feature sequences at each slope position, and generate the non-contour deformation content of each target object at each slope position.

[0027] In some embodiments of the present application, integrate the slope deformation content to obtain the deformation description set at each slope position, including,

[0028] Comprehensively combine the contour deformation content and the non-contour deformation content of the target objects under each target object to generate a comprehensive deformation degree index, and thus determine the comprehensive deformation degree index of each target object;

[0029] Construct the deformation description set at each slope position according to the contour deformation content of the target objects, the non-contour deformation content of the target objects, and the comprehensive deformation degree index.

[0030] In some embodiments of the present application, the method further includes determining the association between different slope positions, including,

[0031] The slope position includes the slope part position and the slope non - part position. The slope part position includes the slope top, the slope surface, and the slope toe. The slope non - part position is the unstable outer surface area of the slope;

[0032] Analyze the deformation transfer paths between the unstable outer surface area of the slope, the slope top, the slope surface, and the slope toe, and quantify the influence degree of each slope position on other slope positions in the deformation transfer path.

[0033] In some embodiments of the present application, according to the association between different slope positions, integrate the deformation description sets at all slope positions to warn and predict the landslide phenomenon of the slope, including,

[0034] Determine the landslide warning value according to the comprehensive deformation degree index and the influence degree, and rely on the landslide warning value to realize the warning and prediction of the landslide phenomenon of the slope;

[0035]

[0036] Among them, Ew is the landslide warning value, m is the number of slope positions, γ j is the deformation influence coefficient determined by the influence degree of the j - th slope position, B j is the comprehensive deformation degree index of the j - th slope position, max(γ j B j ) is the maximum value in γ j B j , and k2 is a preset constant.

[0037] Correspondingly, the present application also provides a landslide deformation image detection and warning system, including,

[0038] The first module is used to obtain the slope structure information and the potential sliding surface position of the slope, define the slope part position, analyze the influence of the potential sliding surface position of the slope on the slope to determine the slope non - part position, and set image monitoring points according to the slope part position and the slope non - part position, so as to collect the slope image content of different positions of the slope;

[0039] The second module is used to obtain the slope images at different positions of the slope within a period of time, extract the image features of the slope images at each slope position, and construct the image feature sequence of the slope images at each slope position;

[0040] The third module is used to analyze the image feature sequence of the slope images at each slope position to determine the slope deformation content at each slope position, and integrate the slope deformation content to obtain the deformation description set at each slope position;

[0041] The fourth module is used to integrate the deformation description sets at all slope positions according to the associations between different slope positions, so as to give early warnings and predictions for the landslide phenomena of slopes.

[0042] By applying the above technical solutions, analyze the influence of the potential sliding surface position of the slope on the slope to determine the non-part positions of the slope, analyze the influence of the potential sliding surface on the slope to determine the regional positions of the outer surface of the slope affected, and improve the reliability of slope landslide phenomenon analysis. Set image monitoring points according to the slope part positions and non-part positions of the slope to ensure that the image conditions at different positions of the slope can be captured in time. Construct the slope image feature sequences at each slope position, so as to accurately capture the changes at different positions of the slope, and provide a reliable basis for the subsequent integration of slope deformation content and slope positions. Integrate the slope deformation content to obtain the deformation description sets at each slope position; integrate the deformation description sets at all slope positions according to the associations between different slope positions to give early warnings and predictions for the landslide phenomena of slopes, take into account the deformation associations between different slope positions to conduct overall detection and early warning of slope landslide phenomena, improve the accuracy and adaptability of slope landslide detection and early warning, and effectively monitor slope landslide phenomena and landslide protection. Description of the Drawings

[0043] Figure 1 It is a schematic flow chart of the landslide deformation image detection and early warning method proposed by the present invention;

[0044] Figure 2 It is a schematic structural diagram of the landslide deformation image detection and early warning system proposed by the present invention. Detailed Embodiments

[0045] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0046] Refer to Figure 1 , the landslide deformation image detection and early warning method includes the following steps:

[0047] Step S101, obtain the slope structure information and the potential sliding surface position of the slope, define the slope part positions, analyze the influence of the potential sliding surface position of the slope on the slope to determine the non-part positions of the slope, and set image monitoring points according to the slope part positions and non-part positions of the slope, so as to collect the slope image content at different positions of the slope.

[0048] In this embodiment, for potential sliding surface identification and geomechanical analysis, based on geological exploration data and using geomechanical theory, analyze the stability of the slope and identify potential sliding surfaces. The slope part positions include the slope top, slope surface, slope toe, etc. Specifically,

[0049] I. Top of the slope

[0050] Importance: The top of the slope is the starting area of landslide occurrence and the place where the landslide body initially shows deformation and movement.

[0051] Monitoring content: Attention should be focused on signs such as cracks, settlements, and exposed land at the top of the slope, which may be early warning signals of an impending landslide.

[0052] II. Slope surface

[0053] Importance: The slope surface is the main path for the landslide body to slide and the area where the deformation is most obvious during the landslide process.

[0054] Monitoring content: It is necessary to monitor the crack expansion, soil loosening, rock fragment falling, etc. on the slope surface. These signs indicate that the landslide body is gradually losing stability.

[0055] III. Bottom of the slope (toe of the slope)

[0056] Importance: The bottom of the slope is the place where the landslide body finally accumulates and is the main area where the landslide causes harm to the surrounding environment and facilities.

[0057] Monitoring content: Attention should be paid to the accumulation situation at the toe of the slope, signs such as ground uplift or settlement, which may indicate that a landslide has occurred or is about to occur.

[0058] IV. Potential landslide surface (unstable outer surface area of the slope affected by the potential landslide surface)

[0059] Importance: The potential landslide surface refers to the weak surface or soft interlayer in the slope where sliding may occur and is the key area for landslide occurrence.

[0060] Monitoring content: Attention should be focused on the deformation of the potential landslide surface, such as crack expansion, soil creep, etc. These signs indicate that the landslide body is gradually sliding along the potential landslide surface. Since the potential landslide surface is located inside the slope and the image inside the slope cannot be directly obtained, an unstable outer surface area of the slope affected by the potential landslide surface is defined here, where the image can be directly collected.

[0061] In some embodiments of the present application, to determine the non - critical position of the slope by analyzing the influence of the potential landslide surface position on the slope, including,

[0062] Collecting slope geological data, identifying the potential landslide surface position and sliding data on the slope through the slope geological data. The sliding data includes sliding direction, sliding speed, and sliding distance;

[0063] Input the sliding direction, sliding speed, sliding distance, and the position of the potential sliding surface into numerical simulation software to simulate the sliding process of the sliding surface and obtain the simulation results;

[0064] The simulation results include the sliding path and regional stress-strain related data. The sliding path is statistically analyzed to form the sliding influence zone, and the sliding influence zone is split according to the geological soil category to obtain multiple small sliding influence zones. Determine the potential sliding surface affected by each small sliding influence zone and the safety factor of the potential sliding surface, and match the regional stress-strain related data to each small sliding influence zone according to the position;

[0065] Evaluate the stability index of the small sliding influence zone based on the safety factor of the sliding surface and the regional stress-strain related data of the small sliding influence zone;

[0066]

[0067] Among them, is the stability index of the i1-th small sliding influence zone, is the stable conversion coefficient, n1 is the category data of the regional stress-strain related data of the i1-th small sliding influence zone, is the combination weight of the regional stress-strain related data of the i2-th category data, is the average value of the regional stress-strain related data of the i2-th category data, n2 is the number of potential sliding surfaces affected by the i1-th small sliding influence zone, is the influence weight of the i3-th potential sliding surface affected by the i1-th small sliding influence zone, is the safety factor of the i3-th potential sliding surface affected by the i1-th small sliding influence zone, and k1 is a preset constant;

[0068] Superimpose all small sliding influence zones according to the stability index to obtain the unstable slope outer surface area, and use the unstable slope outer surface area as the non-site position of the slope.

[0069] In this embodiment, parameters such as the sliding direction, sliding speed, and sliding distance of the sliding surface are analyzed. Using a Geographic Information System (GIS) or numerical simulation software, according to the characteristics of the sliding surface and the sliding parameters, the sliding process of the sliding surface is simulated. Analyze the slope areas that may be affected during the sliding process of the sliding surface, including the sliding path, sliding influence zone, etc. Extract the sliding path of the sliding surface, and analyze the characteristics such as the trend, length, and curvature of the sliding path to understand the movement trend of the sliding surface. Determine the range of the sliding influence zone. The sliding influence zone usually includes the area around the sliding path, and its width and length depend on the scale and sliding speed of the sliding surface. The regional stress-strain related data includes stress-strain data, other related force data, etc. The safety factor of the potential sliding surface is the ratio of the anti-sliding force to the sliding force of the sliding surface, and the small sliding influence zone may be affected by multiple potential sliding surfaces.

[0070] In this embodiment, It represents the correction of the stable influence amount of the relative average value of the safety factor of the sliding surface affected by the small sliding influence zone. k1 is used to control the size of the correction function. For the area affected by multiple sliding surfaces, the superposition principle is used to determine the possible positions of the unstable slope outer surface areas, and the changes of these unstable slope outer surface areas are captured through images to detect and warn of landslide situations.

[0071] In some embodiments of the present application, image monitoring points are set according to the slope part position and the non-slope part position, including,

[0072] Both the slope part position and the non-slope part position are position intervals of a certain part on the slope. According to the position intervals of the slope part position and the non-slope part position, the regional sizes of each are calculated, and image monitoring points are deployed for the slope part position through a preset first deployment strategy, and image monitoring points are deployed for the non-slope part position through a preset second deployment strategy.

[0073] In this embodiment, the image monitoring point is the specific target position for arranging the image capture device. The first deployment strategy is to set the image monitoring points for the slope part position by taking into account the two principles of multi-angle and uniformity. The second deployment strategy is to, on the basis of taking into account the two principles of multi-angle and uniformity, add the principle of density to set the image monitoring points for the non-slope part position of the slope. Because the non-slope part position of the slope is the area of the slope outer surface affected by the potential sliding surface of the slope, it is necessary to collect images frequently and densely to ensure that changes can be captured in a timely manner.

[0074] Step S102, obtain slope images at different positions of the slope within a period of time, extract the image features of the slope images at each slope position, and construct an image feature sequence of the slope images at each slope position.

[0075] In this embodiment, the slope image feature sequence at each slope position is the slope image feature sequence of each target object in the image, where the target object is the target situation that needs to be identified in the slope image, and the target objects include cracks, landslide bodies, slope surfaces, ground vegetation, etc.

[0076] In slope image monitoring, the content that can be used as an identification target is quite rich, and these targets are usually closely related to the stability and safety of the slope. The following is a detailed description of the identifiable targets in the slope image:

[0077] 1. Cracks

[0078] Description: Cracks are the cracking phenomena that appear on the surface or inside of the slope, and they are important precursors of slope instability.

[0079] Identification features: Cracks usually appear as linear or irregular shapes, and their colors may be different from the surrounding rock and soil. Sometimes there are displacement or opening phenomena.

[0080] Significance of identification: By identifying the location, length, width, and trend of cracks, the degree of slope rupture and potential sliding risk can be evaluated.

[0081] 2. Landslide body

[0082] Description: The landslide body refers to the rock and soil mass that slides as a whole or partially on the slope due to the action of gravity.

[0083] Identification features: The landslide body usually has obvious sliding traces, such as sliding surfaces and sliding zones, and its shape, color, and texture may be different from the surrounding rock and soil mass.

[0084] Significance of identification: Identifying the landslide body can determine the sliding range and degree of the slope, providing an important basis for landslide early warning and prevention.

[0085] 3. Changes in slope surface morphology

[0086] Description: Changes in slope surface morphology refer to the morphological changes that occur on the slope surface due to factors such as weathering, erosion, and landslides.

[0087] Identification features: Changes in slope surface morphology may be manifested as uneven slope surfaces, toe erosion, crest collapse, etc.

[0088] Significance of identification: By identifying the changes in slope surface morphology, the evolution process of the slope can be understood, and its stability status can be evaluated.

[0089] 4. Changes in vegetation cover

[0090] Description: Changes in vegetation cover refer to the changes in the distribution, species, and density of vegetation on the slope.

[0091] Recognition features: Vegetation cover changes may be manifested as sparse, withered, dead, or newly grown vegetation, etc.

[0092] Recognition significance: Vegetation cover changes may be related to the moisture condition, soil fertility, and stability of the slope. By identifying these changes, the ecological environment and stability of the slope can be indirectly evaluated.

[0093] In some embodiments of the present application, the image features of the slope images at each slope position are extracted, and a sequence of slope image features at each slope position is constructed, including,

[0094] The image features of the slope image include color features, texture features, shape features, and other features. The target object within the slope image and the combination of image feature types required for the target object are confirmed, and the contour of the target object within the slope image is identified by combining a preset target object recognition algorithm;

[0095] The color features, texture features, shape features, and other features are integrated according to the target object contour, and a sequence of slope image features of each target object contour at each slope position is constructed over time.

[0096] In this embodiment, the combination of image feature types required for the target object means that different target objects require different combinations of image feature types for recognition. For example, for landslide body recognition, the contour and boundary of the landslide body are recognized through shape features and texture features. For vegetation cover recognition, color features and texture features are used to distinguish vegetation and non-vegetation areas. The target object recognition algorithms include Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), etc., and are combined with edge detection algorithms for classification and recognition of the target. Integrating the color features, texture features, shape features, and other features according to the target object contour means integrating all the image features near and within the target object contour according to the position of the target object contour.

[0097] Step S103: Analyze the sequence of slope image features at each slope position to determine the slope deformation content at each slope position, and integrate the slope deformation content to obtain a deformation description set at each slope position.

[0098] In this embodiment, the slope deformation content includes target object contour deformation content and target object non-contour deformation content. The target object contour deformation content is the deformation content related to the contour, and the target object non-contour deformation content is the deformation content related to the interior of the contour. The target object contour deformation content, target object non-contour deformation content, and target object are integrated to obtain a deformation description set at each slope position.

[0099] In some embodiments of the present application, analyzing the sequence of slope image features at each slope position to determine the slope deformation content at each slope position includes,

[0100] The slope deformation content includes the contour deformation content and the non - contour deformation content of the target object;

[0101] Calculate the geometric features of each target object contour on the slope image feature sequence at each slope position, convert the geometric features into shape descriptors, and calculate the contour deformation content of each target object at each slope position through the shape descriptors;

[0102] Calculate the change of image features inside each target object contour on the slope image feature sequence at each slope position, and generate the non - contour deformation content of each target object at each slope position.

[0103] In this embodiment, a contour extraction algorithm (such as Canny edge detection) can be used to obtain its contour information. Then, by comparing the shape and position changes of the contours in images at different time points, the target object is tracked. Align the contours at different time points to accurately compare the changes between them. This may require the use of image registration techniques to ensure that the contours are compared in the same spatial reference system. Calculate the geometric features of the contour, such as perimeter, area, aspect ratio, curvature, etc. These features can help describe the overall shape and size of the contour. Use shape descriptors (such as Fourier descriptors, Hough transform, etc.) to capture the detailed shape information of the contour. These descriptors are very useful for describing the subtle changes of the contour. The contour deformation content of the target object includes the amount of deformation, deformation direction, deformation speed, etc. The contour deformation content of the target object can reflect the deformation situation of the target object. For example, for cracks, pay attention to the changes in their length, width and depth, as well as the appearance of new cracks or the expansion of cracks. For landslide bodies, pay attention to the changes in their volume, shape and displacement, as well as the stability and sliding trend of the landslide body. For the slope morphology and vegetation cover, pay attention to their overall changes and possible erosion, settlement or vegetation degradation phenomena.

[0104] It should be noted that the Fourier descriptor extracts shape features by converting the coordinate data of the contour into a frequency - domain representation. This conversion enables the shape information of the contour to be represented as a combination of a series of Fourier coefficients, which reflect the shape changes of the contour at different frequencies. Advantages: Fourier descriptors are invariant to translation, rotation and scale changes of the contour, so they can effectively capture the subtle changes of the contour, even if these changes may not be easily detected in the time domain. Applications: In slope image analysis, Fourier descriptors can be used to detect the small expansion of cracks or the shape changes of landslide bodies, thus providing important information about slope stability.

[0105] In this embodiment, the image feature changes inside the contour of the target object. For example, the color feature describes the distribution and changes of colors in an image or an image region. For the target object in the slope image, such as the vegetation-covered area, the change of the color feature can reflect the growth condition or health condition of the vegetation, thereby indirectly reflecting the stability of the slope. The texture feature describes the spatial distribution pattern of grayscale or colors in an image or an image region. For the target object in the slope image, such as the landslide body or the crack area, the change of the texture feature can reflect the deformation or damage of its internal structure.

[0106] In some embodiments of the present application, the deformation content of the slope is integrated to obtain a deformation description set at each slope position, including,

[0107] Integrate the contour deformation content and the non-contour deformation content of the target object under each target object to generate a comprehensive deformation degree index, and thereby determine the comprehensive deformation degree index of each target object;

[0108] Construct a deformation description set at each slope position according to the target object contour deformation content, the target object non-contour deformation content, and the comprehensive deformation degree index.

[0109] In this embodiment, for the comprehensive deformation degree index, for the contour deformation, a comprehensive index of the deformation amount, deformation speed, and deformation direction can be calculated. For the non-contour deformation, the change degree of the image feature can be extracted, such as the crack length, the color change degree, etc. The indexes of the contour deformation and the non-contour deformation are weighted and summed to obtain the comprehensive deformation degree index of the target object. Here, the target object contour deformation content and the target object non-contour deformation content are integrated. The construction of the deformation description set at each slope position is to integrate each target object, so as to generate the specific deformation description situation and the comprehensive deformation degree index at each slope position.

[0110] It can be understood that the deformation description set includes the specific deformation content of the target object contour deformation content and the target object non-contour deformation content and a comprehensive deformation degree index (integrating the target object contour deformation content, the target object non-contour deformation content, and all target objects).

[0111] In some embodiments of the present application, the method further includes determining the association between different slope positions, including,

[0112] The slope position includes the slope part position and the slope non-part position. The slope part position includes the slope top, the slope surface, and the slope foot. The slope non-part position is the unstable outer surface area of the slope;

[0113] Analyze the deformation transfer paths among the unstable outer surface area of the slope, the slope top, the slope surface, and the slope foot, and quantify the influence degree of each slope position on other slope positions in the deformation transfer path.

[0114] In this embodiment, according to the geological structure and slope characteristics, a mechanical model is established, considering the slope stability, stress distribution and deformation mechanism. The model should be able to reflect the interaction and influence between different positions. By analyzing the simulation results and determining the degree and range of the mutual influence between the deformations at different positions, the deformation transfer coefficient between each slope part can be obtained, thereby quantifying the influence degree between the deformation transfer paths.

[0115] The deformation transfer paths between slope positions are as follows. Only some possible examples are given.

[0116] From the slope top to the slope surface:

[0117] Transfer mechanism: The cracks and settlements at the top may gradually expand downward, affecting the stability of the slope surface. The change in load at the top (such as the increased weight due to rainfall) may also exacerbate the deformation of the slope surface.

[0118] From the slope surface to the slope bottom:

[0119] Transfer mechanism: The deformation and movement on the slope surface will be directly transmitted to the slope toe. The sliding speed and direction of the slope surface determine the degree and range of the influence on the slope toe.

[0120] The role of the potential landslide surface (the unstable outer surface area of the slope):

[0121] Transfer mechanism: The deformation of the potential landslide surface is a precursor to slope instability, and its deformation may cause deformation and movement at the top, slope surface and slope toe simultaneously or successively.

[0122] Step S104, according to the association between different slope positions, integrate the deformation description sets at all slope positions to give early warning and prediction of the landslide phenomenon of the slope.

[0123] In some embodiments of the present application, according to the association between different slope positions, integrate the deformation description sets at all slope positions to give early warning and prediction of the landslide phenomenon of the slope, including

[0124] Determine the landslide early warning value according to the comprehensive deformation degree index and influence degree, and rely on the landslide early warning value to realize the early warning and prediction of the landslide phenomenon of the slope;

[0125]

[0126] Among them, Ew is the landslide early warning value, m is the number of slope positions, γ j is the deformation influence coefficient determined by the influence degree of the jth slope position, B j is the comprehensive deformation degree index of the jth slope position, max(γ j B j ) is γ j Bj The maximum value in

[0127] In this embodiment, the larger the landslide warning value here, the higher the possibility of slope landslide. Moreover, the deformation description mainly focuses on the description of the deformation trend, and the landslide can be predicted by combining the deformation trend. The influence degree is the influence degree of a certain slope position on other slope positions or on the overall slope, which represents the correction of the sum of the deformation degrees of all slope positions by the slope position with the greatest influence. k2 is used to balance and ensure the magnitude of the correction function.

[0128] Correspondingly, the present application also provides a landslide deformation image detection and warning system, as Figure 2 shown, including,

[0129] The first module is used to obtain slope structure information and the position of the potential sliding surface of the slope, define the position of the slope part, analyze the influence of the position of the potential sliding surface of the slope on the slope to determine the non-part position of the slope, and set image monitoring points according to the position of the slope part and the non-part position of the slope, so as to collect the slope image content at different positions of the slope;

[0130] The second module is used to obtain slope images at different positions of the slope within a period of time, extract the image features of the slope images at each slope position, and construct an image feature sequence of the slope images at each slope position;

[0131] The third module is used to analyze the image feature sequence of the slope images at each slope position to determine the slope deformation content at each slope position, and integrate the slope deformation content to obtain a deformation description set at each slope position;

[0132] The fourth module is used to integrate the deformation description sets at all slope positions according to the association between different slope positions to warn and predict the landslide phenomenon of the slope.

[0133] By applying the above technical solutions, analyze the influence of the position of the potential sliding surface of the slope on the slope to determine the non-site position of the slope, analyze the influence of the potential sliding surface on the slope, and determine the regional position of the outer surface of the slope affected, so as to improve the reliability of slope landslide phenomenon analysis. Set up image monitoring points according to the slope site position and the slope non-site position to ensure that the images of different positions of the slope can be captured in time. Construct the slope image feature sequence at each slope position, so as to accurately capture the changes at different positions of the slope, and provide a reliable basis for the subsequent integration of slope deformation content and slope position. Integrate the slope deformation content to obtain the deformation description set at each slope position; according to the correlation between different slope positions, integrate the deformation description sets at all slope positions to warn and predict the landslide phenomenon of the slope, taking into account the deformation correlation between different slope positions to conduct overall detection and warning of the slope landslide phenomenon, improving the accuracy and adaptability of slope landslide detection and warning, and effectively monitoring the slope landslide phenomenon and landslide protection.

[0134] Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by hardware or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of the present invention.

[0135] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the drawings are not necessarily essential for implementing the present invention.

[0136] Those skilled in the art can understand that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the description of the implementation scenario, or can be correspondingly changed and located in one or more systems different from the present implementation scenario. The modules in the above implementation scenario can be combined into one module, or further split into multiple sub-modules.

[0137] The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. Landslide deformation image detection and early warning method, characterized in that including, obtaining slope structure information and the position of the potential slip surface of the slope, defining the position of the slope part, analyzing the influence of the position of the potential slip surface of the slope on the slope to determine the non-part position of the slope, and setting image monitoring points according to the position of the slope part and the non-part position of the slope, so as to collect slope image content at different positions of the slope; obtaining slope images at different positions of the slope within a period of time, extracting the image features of the slope images at each slope position, and constructing an image feature sequence of the slope images at each slope position; analyzing the image feature sequence of each slope position to determine the slope deformation content at each slope position, and integrating the slope deformation content to obtain a deformation description set at each slope position; warning and predicting the landslide phenomenon of the slope by integrating the deformation description sets at all slope positions according to the association between different slope positions.

2. The landslide deformation image detection and early warning method according to claim 1, characterized in that analyzing the influence of the position of the potential slip surface of the slope on the slope to determine the non-part position of the slope, including, collecting slope geological data, and identifying the position of the potential slip surface and the sliding data on the slope through the slope geological data, where the sliding data includes the sliding direction, sliding speed, and sliding distance; inputting the sliding direction, sliding speed, sliding distance, and the position of the potential slip surface into numerical simulation software to simulate the sliding process of the slip surface and obtain the simulation results; the simulation results include the sliding path and region stress-strain related data, statistically analyzing the sliding path to form a sliding influence zone, splitting the sliding influence zone according to the geological soil category to obtain multiple small sliding influence zones, determining the potential slip surface affected by each small sliding influence zone and the safety factor of the potential slip surface, and matching the region stress-strain related data to each small sliding influence zone according to the position; evaluating the stability index of the small sliding influence zone according to the safety factor of the slip surface and the region stress-strain related data of the small sliding influence zone; Among them, is the stability index of the i1-th small sliding influence zone, is the stable conversion coefficient, n1 is the category data of the regional stress and strain related data of the i1-th small sliding influence zone, is the combination weight of the regional stress and strain related data of the i2-th category data, is the average value of the regional stress and strain related data of the i2-th category data, n2 is the number of potential sliding surfaces affected by the i1-th small sliding influence zone, is the influence weight of the i3-th potential sliding surface affected by the i1-th small sliding influence zone, is the safety factor of the i3-th potential sliding surface affected by the i1-th small sliding influence zone, k1 is a preset constant; superimposing all the small sliding influence zones according to the stability index to obtain the unstable outer surface area of the slope, and taking the unstable outer surface area of the slope as the non-part position of the slope.

3. The landslide deformation image detection and warning method according to claim 1, characterized in that setting image monitoring points according to the position of the slope part and the non-part position of the slope, including, both the position of the slope part and the non-part position of the slope are the position intervals of a certain part on the slope. Calculate the regional size of each according to the position intervals of the position of the slope part and the non-part position of the slope, deploy image monitoring points for the position of the slope part through a preset first deployment strategy, and deploy image monitoring points for the non-part position of the slope through a preset second deployment strategy.

4. The landslide deformation image detection and warning method according to claim 1, characterized in that extracting the image features of the slope images at each slope position and constructing an image feature sequence of the slope images at each slope position, including, the image features of the slope image include color features, texture features, shape features, and other features. Confirm the target object in the slope image and the combination of image feature types required by the target object, and identify the contour of the target object in the slope image by combining a preset target object recognition algorithm; integrating color features, texture features, shape features, and other features according to the contour of the target object, and constructing an image feature sequence of the slope images of each target object contour at each slope position over time.

5. The landslide deformation image detection and early warning method according to claim 4, wherein, Analyze the sequence of slope image features at each slope position to determine the slope deformation content at each slope position, including The slope deformation content includes the contour deformation content of the target object and the non - contour deformation content of the target object; Calculate the geometric features of each target object contour on the slope image feature sequence at each slope position, convert the geometric features into shape descriptors, and calculate the contour deformation content of each target object at each slope position through the shape descriptors; Calculate the change in image features inside each target object contour on the slope image feature sequence at each slope position, and generate the non - contour deformation content of each target object at each slope position.

6. The landslide deformation image detection and early warning method according to claim 5, characterized in that Integrate the slope deformation content to obtain the deformation description set at each slope position, including Integrate the contour deformation content and non - contour deformation content of the target object under each target object to generate a comprehensive deformation degree index, and thus determine the comprehensive deformation degree index of each target object; Construct the deformation description set at each slope position according to the contour deformation content of the target object, the non - contour deformation content of the target object, and the comprehensive deformation degree index.

7. The landslide deformation image detection and early warning method according to claim 2, characterized in that The method further includes determining the association between different slope positions, including The slope position includes the slope part position and the non - slope part position. The slope part position includes the slope top, the slope surface, and the slope toe. The non - slope part position is the unstable outer surface area of the slope; Analyze the deformation transfer paths between the unstable outer surface area of the slope, the slope top, the slope surface, and the slope toe, and quantify the influence degree of each slope position on other slope positions in the deformation transfer path.

8. The landslide deformation image detection and early warning method according to claim 6 or 7, characterized in that, Integrate the deformation description sets at all slope positions according to the association between different slope positions to warn and predict the landslide phenomenon of the slope, including Determine the landslide warning value according to the comprehensive deformation degree index and the influence degree, and use the landslide warning value to realize the warning and prediction of the landslide phenomenon of the slope; Among them, Ew is the landslide warning value, m is the number of slope positions, and γ j is the deformation influence coefficient determined by the influence degree of the j-th slope position, B j is the comprehensive deformation degree index of the j-th slope position, max(γ j B j ) is the maximum value in γ j B j , and k2 is a preset constant.

9. Landslide deformation image detection and early warning system, characterized in that, including The first module is used to obtain the slope structure information and the position of the potential slip surface of the slope, define the slope part position, analyze the influence of the potential slip surface position of the slope on the slope to determine the non - slope part position, and set image monitoring points according to the slope part position and the non - slope part position, so as to collect the slope image content at different positions of the slope; The second module is used to obtain the slope images at different positions of the slope within a period of time, extract the image features of the slope images at each slope position, and construct the slope image feature sequence at each slope position; The third module is used to analyze the slope image feature sequence at each slope position to determine the slope deformation content at each slope position, and integrate the slope deformation content to obtain the deformation description set at each slope position; The fourth module is used to integrate the deformation description sets at all slope positions according to the association between different slope positions to warn and predict the landslide phenomenon of the slope.

Images