A highway side slope deformation monitoring method and system

Abstract

The application provides a highway slope deformation monitoring method and system, and the method comprises the following steps: based on the monitoring image of the real-time obtained highway slope deformation area, the image displacement value of each grid is calculated; for the grid marked with the center of the infrared light source, the image displacement value of the grid is calculated based on the image coordinates of the corresponding infrared light source center in the continuous monitoring image; for the grid which is not marked with the center of the infrared light source but is marked with the natural ground feature point: when the average brightness of the monitoring image is greater than the preset threshold, the image displacement value is calculated based on the image coordinates of the natural ground feature point in the continuous monitoring image; when the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement of the grid is valued as the average value of the displacement values of the adjacent grids; based on the image displacement value of each grid, the displacement of the corresponding slope area is calculated. The application can extract the deformation range of the highway slope with high precision, low cost and all-weather.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of highway slope deformation monitoring, and particularly relates to a highway slope deformation monitoring method and system. BACKGROUND

[0002] Highway slope landslide disaster is a major threat to highway traffic safety. In order to effectively prevent and respond to landslide disasters, highway slope deformation monitoring and early warning technology is particularly important.

[0003] Highway slope deformation monitoring technology is a research focus in the field of highway engineering. Its main goal is to monitor the time series deformation and sliding scale of the slope, timely warn of slope landslide risks, assess the scale and destructive nature of the landslide, provide data support for slope maintenance, reinforcement, and disaster warning, and ensure the safety and smoothness of highway traffic. With the development of surveying and mapping remote sensing technology, highway slope deformation monitoring technology has developed rapidly, forming a variety of technical systems, mainly including the following several common monitoring methods:

[0004] Ground measurement technology: The horizontal and vertical displacement changes of the slope monitoring points are measured by devices such as level and total station. This technology has high precision and strong adaptability. However, ground measurement technology is greatly limited by terrain and visibility conditions, and manual observation frequency is low, making it impossible to achieve all-weather, real-time monitoring.

[0005] Satellite navigation positioning technology: The three-dimensional coordinates of the monitoring points are obtained in real time by receiving satellite signals, and the displacement changes of the slope are calculated. This technology has the advantages of all-weather, real-time, high precision, and can realize remote automatic monitoring, and is not limited by visibility conditions, and is suitable for large-scale monitoring. However, in areas such as mountainous areas where signal shielding is serious, the precision is affected, and the equipment cost is high.

[0006] Satellite remote sensing monitoring technology: Synthetic Aperture Radar Interferometry (InSAR) technology is used to obtain the small deformation information of the slope surface by interferometric processing of radar images obtained at different times. This technology can monitor large-scale slope deformation with millimeter-level precision and is not limited by weather and lighting conditions. However, data processing is complex, and in areas with complex terrain and dense vegetation, the precision is low, and the data purchase cost is high.

[0007] Ground laser scanning technology: The slope is scanned by a laser scanner to obtain point cloud data, and then the deformation and range of the slope are calculated. This technology can quickly obtain three-dimensional information of the slope with high measurement precision and is suitable for comprehensive reflection of deformation conditions. However, the equipment cost is high, the data processing amount is large, and the scanning range is limited, and the adaptability to complex terrain is poor.

[0008] Unmanned aerial vehicle monitoring technology: Unmanned aerial vehicles equipped with optical cameras, laser radars and other sensors can quickly obtain high-resolution images and three-dimensional point cloud data for slope monitoring. Unmanned aerial vehicle monitoring has high flexibility and low cost, and can quickly respond to monitoring needs. However, it is limited by weather, flight distance and endurance time, and data processing also requires certain technical support.

[0009] Although the above technologies have made certain progress in slope deformation monitoring, they still face the following challenges:

[0010] Difficulty in all-weather monitoring: Except for satellite navigation positioning technology, which can achieve all-weather monitoring, other technologies are generally limited by environmental conditions such as light and weather, and cannot work stably at night or in low-light conditions.

[0011] High monitoring cost: Satellite navigation positioning technology and ground laser scanning technology require expensive equipment, and satellite remote sensing technology faces high data purchase costs, making monitoring costs high and difficult to be widely applied.

[0012] Data processing complexity: The data processing complexity of most technologies is high, especially for remote sensing technology and laser scanning technology, which require a lot of post-processing, increasing the difficulty of operation and maintenance.

[0013] Monitoring object limitation: Except for satellite remote sensing technology, most methods can only monitor a small number of discrete points, making it difficult to extract and locate the deformation area of the slope, resulting in difficulty in assessing the scope and danger of landslides.

[0014] Therefore, existing technologies cannot simultaneously meet the requirements of low cost, all-weather, and convenient data processing, especially in the monitoring of slope deformation at night or in complex environments. In order to overcome these limitations, there is an urgent need for a high-precision, low-cost, all-weather, and deformation range extraction method for highway slope deformation monitoring. SUMMARY

[0015] The purpose of the present application is to solve the problems in the above background art, and to provide a high-precision, low-cost, all-weather, and deformation range extraction method for highway slope deformation monitoring.

[0016] The technical solution adopted by the present application is: a highway slope deformation monitoring method, wherein the deformation area of the highway slope is provided with an infrared light source; comprising the following steps:

[0017] Grid division is performed on the initial image of the highway slope deformation area, and natural feature points and infrared light source centers are extracted, and the corresponding grid and image coordinates are marked;

[0018] Based on the real-time acquired monitoring image of the highway slope deformation area, the image displacement value of each grid is calculated;

[0019] For the grid marked with the center of the infrared light source, the image displacement value of the grid is calculated based on the image coordinates of the corresponding center of the infrared light source in the continuous monitoring images;

[0020] For the grid not marked with the center of the infrared light source but marked with the natural ground feature point:

[0021] When the average brightness of the monitoring image is greater than the preset threshold, the image displacement value is calculated based on the image coordinates of the natural ground feature point in the continuous monitoring images;

[0022] When the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids;

[0023] For the grid without the infrared light source monitoring point and the natural ground feature point, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids;

[0024] Based on the image displacement value of each grid, the displacement of the corresponding slope area is calculated.

[0025] In the technical scheme, a plurality of monitoring piles are embedded along the line from top to bottom in the deformation area of the highway slope; an infrared light source is arranged on each monitoring pile; a reference pile is embedded in the stable area outside the deformation area of the highway slope, and a monitoring camera is installed on the reference pile to obtain an initial image and a monitoring image.

[0026] In the technical scheme, the process of calculating the image displacement value of the grid marked with the center of the infrared light source includes: for each grid marked with the center of the infrared light source, the absolute difference value of the image longitudinal coordinates of the center of the infrared light source in the continuous two monitoring images is calculated as the image displacement value of the grid at the corresponding monitoring time; for all monitoring times in a specified time period, the cumulative sum of all image displacement values of the grid in the time period is calculated as the image displacement value of the grid in the specified time period.

[0027] In the technical scheme, the process of calculating the image displacement value based on the image coordinates of the natural ground feature point in the continuous monitoring images includes: for any grid marked with the natural ground feature point, the average value of the difference values of the image longitudinal coordinates of all natural ground feature points in the grid in the continuous two monitoring images is calculated as the image displacement value of the grid at the corresponding monitoring time; for all monitoring times in a specific time period, the cumulative sum of all image displacement values of the grid in the time period is calculated as the image displacement value of the grid in the specific time period; the specific time period is the time period between the specified monitoring time and the monitoring time corresponding to the last monitoring image with brightness greater than the preset threshold corresponding to the specified monitoring time.

[0028] In the technical solution, the process of obtaining the natural feature point of each monitoring image in the specific time period except the designated monitoring moment comprises: constructing a search space based on the average pixel displacement of the monitoring images in the specific time period and in combination with the natural feature point matched at the designated monitoring moment; and performing natural feature point matching on each monitoring image based on the search space.

[0029] In the technical solution, for the monitoring image p i of the designated moment, the natural feature point pt i_q_k is obtained by searching the search space.

[0030] Area i_r_q_k

[0031] ={x i_q_k -1<x<x i_q_k +1,y i_q_k -2*Avg_shift i_r <y<y i_q_k +1}

[0032] wherein the image coordinates of the natural feature point pt i_q_k are (x i_q_k , y i_q_k ), and Avg_shift i_r represents the average pixel displacement of the monitoring images in the specific time period.

[0033] In the technical solution, for all monitoring moments in the specific time period, the cumulative sum of the average pixel displacement of the monitoring images in the time period is calculated as the average pixel displacement of the monitoring images in the time period; and for any monitoring image, the average value of the absolute difference between the monitoring image and the vertical coordinates of the center of all infrared light sources in the previous monitoring image is calculated as the average pixel displacement of the monitoring image.

[0034] In the technical solution, at least one grid marked with the center of the infrared light source exists in the neighborhood of each grid.

[0035] In the technical solution, the following steps are further included: connecting the grids with the image displacement values greater than the set threshold value in the specific time period to form a closed area, marking the slope area corresponding to the closed area as the slope landslide range, and calculating the area of the slope area.

[0036] The application further provides a highway slope deformation monitoring system for implementing the highway slope deformation monitoring method.

[0037] The beneficial effects of the present application are: the present application can accurately demarcate the monitoring area by grid division of the deformation area of the highway slope and extraction of natural feature points and infrared light source centers. In this way, the deformation data of each grid can be accurately calculated, thereby improving the monitoring accuracy. The present application can select different processing methods for images under different conditions (such as brightness greater than or less than a preset threshold), ensuring that effective monitoring can continue even in insufficient light conditions (such as at night), overcoming the limitations of traditional methods under low light conditions. The present application can monitor both small deformation and large displacement of the slope by combining infrared light sources with natural feature points, and can reflect the overall deformation trend of the slope in real time.

[0038] Further, the present application can ensure the stability and accuracy of the monitoring system by burying multiple monitoring piles in the deformation area of the slope and installing reference piles in the stable area. The reference piles provide a fixed reference point, so that each monitoring image can be compared with the initial image to ensure data accuracy; setting multiple monitoring piles and reference piles ensures comprehensive coverage of the deformation area and the stable area, which not only detects the deformation area, but also monitors the stable area in real time, improving the comprehensiveness of the entire highway slope deformation monitoring.

[0039] Further, the present application can accurately obtain the displacement value of each grid by calculating the image ordinate difference of the infrared light source center. Using continuous monitoring images for comparison can clearly depict the displacement trajectory and range of the deformation area; calculating the cumulative sum of image displacement values within a certain time period can reflect the overall deformation trend within the entire time period, thereby providing early warning of slope landslide and other disasters.

[0040] Further, the present application uses natural feature points to enhance monitoring accuracy: calculating displacement values based on image coordinates of natural feature points in continuous monitoring images can provide more feature point data, making displacement calculation more accurate and comprehensive. This method combines image comparison to improve the recognition ability of the deformation area; by adjusting the displacement calculation method according to the brightness of the monitoring image, the mean value of the neighborhood displacement value can be filled in when the light condition is insufficient, ensuring that the monitoring data will not be lost due to insufficient light.

[0041] Further, the present application can improve the matching accuracy of image feature points by calculating the average pixel displacement within a certain time period and combining the natural feature points at the specified monitoring time to construct a search space. This ensures more accurate feature point matching between monitoring images, reducing errors caused by image distortion or changes in camera angle; in each monitoring image, feature point matching is performed based on the search space, which can reduce unnecessary computational load, thereby improving the real-time performance and response speed of the monitoring system.

[0042] Further, the present application defines the search space of natural feature points in the monitoring image at a specific moment, which can effectively limit the search range and improve the accuracy and efficiency of feature point matching. It avoids the waste of computing resources caused by a large search space; through a more accurate search space definition, it can more accurately match the displacement of natural feature points, which helps to more finely monitor the micro-deformation of the slope.

[0043] Further, the present application calculates the average pixel displacement accumulation of all monitoring images in a specific time period, which can dynamically adjust the calculation result of image displacement, so that the monitoring result in the whole time period is more accurate; for each monitoring image, the average pixel displacement is calculated by the difference of the infrared light source center longitudinal coordinate with the previous image, which helps to discover deformation in time, avoid lag feedback, and early landslide warning.

[0044] Further, the present application ensures that there is at least one grid marked with the infrared light source center in the neighborhood of each grid, which can ensure the stability of the monitoring system and avoid the monitoring blind area caused by the lack of data source in a certain grid; ensuring that each grid has corresponding monitoring data source is conducive to providing more comprehensive and accurate slope deformation monitoring.

[0045] Further, the present application can accurately identify the landslide range of the slope by calculating the grid whose image displacement value is greater than the set threshold and connecting these grids to form a closed region. This method provides data support for landslide disaster assessment, which can effectively demarcate the landslide influence range and perform subsequent processing; by calculating the area of the landslide region, it can provide a reference for the impact assessment of landslide disaster, and further provide data support for disaster warning and rescue work. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 The present application provides a method flowchart. DETAILED DESCRIPTION

[0047] The present application will be further described in detail below in combination with the drawings and specific embodiments, which facilitate a clear understanding of the present application, but they do not constitute a limitation of the present application.

[0048] Example 1

[0049] As shown in the drawings, the present application provides a highway slope deformation monitoring method, and the deformation area of the highway slope is provided with an infrared light source; comprising the following steps: Figure 1

[0050] The initial image of the highway slope deformation area is divided into grids, and the natural feature points and the infrared light source center are extracted, and the corresponding grid and image coordinates are marked;

[0051] ​calculating the image displacement value of each grid based on the monitoring image of the real-time obtained highway slope deformation area;

[0052] For the grid marked with the center of the infrared light source, the image displacement value of the grid is calculated based on the image coordinates of the corresponding center of the infrared light source in the continuous monitoring image;

[0053] For the grid which is not marked with the center of the infrared light source but marked with the natural feature points of the ground object:

[0054] When the average brightness of the monitoring image is greater than the preset threshold, the image displacement value is calculated based on the image coordinates of the natural feature points of the ground object in the continuous monitoring image;

[0055] When the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids;

[0056] For the grid which is neither marked with the center of the infrared light source nor marked with the natural feature points of the ground object, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids;

[0057] Based on the image displacement value of each grid, the displacement of the corresponding slope area is calculated.

[0058] The principles of the present application will be further illustrated below with specific embodiments.

[0059] The embodiment specifically includes the following steps:

[0060] Step 1: The monitoring camera and the infrared light source monitoring point are laid out, and the grid of the initial image is divided.

[0061] The monitoring piles are buried along the monitoring slope from top to bottom, and the infrared light sources are stably installed on the monitoring piles. The number of monitoring piles for the same monitoring slope is not less than 2, and the number of monitoring piles is recorded as a. The reference piles are buried in the stable area outside the slope deformation area, and the monitoring camera is installed on the reference pile. The focal length of the monitoring camera is f (unit: meter), and the pixel size of the monitoring camera is μ (unit: meter). The shooting angle of the monitoring camera is adjusted to ensure that the monitoring camera is directed to the infrared light source. The distance D = d1 d2…d a from the monitoring camera to the infrared light source is measured (unit: meter), wherein a≥2.

[0062] Before starting to execute the monitoring process, the initial image of the highway slope deformation area is obtained by the monitoring camera, the initial image of the highway slope deformation area is divided into grids, and the natural feature points and the center of the infrared light source are extracted, and the corresponding grid and image coordinates are marked. It should be noted that after starting to execute the monitoring process, the current highway slope deformation area is obtained as the monitoring image at each monitoring time, and the monitoring image corresponds to the monitoring time one by one.

[0063] The initial image p0 is divided into a grid, and the length and width of the grid are both 128 pixels. The grid is numbered as G 0_q The slope area corresponding to a single grid is n = 1, 2,..., a.

[0064] If the average brightness value L i of the monitoring image p i is less than or equal to 40, the image is taken in the night or low-light conditions, the image is dark, and the natural feature point extraction is not performed.

[0065] If the average brightness value L i of the monitoring image p i is greater than 40, the initial image is taken in sufficient light conditions, and the average brightness is greater than 40. The SIFT algorithm is used for image feature point extraction for each grid of the initial monitoring image p0, and the natural feature point set PT 0_q = {pt 0_q_1 , pt 0_q_2 ,..., pt 0_q_k} is obtained. The image coordinates of the natural feature point pt 0_q_k are (x 0_q_k , y 0_q_k ).

[0066] The initial value of the slope displacement value corresponding to each grid in the initial image is 0.

[0067] Step 2: Time series monitoring image acquisition and light condition determination

[0068] The monitoring camera automatically takes time series monitoring images of the slope at a fixed time interval, which can be set to 15 minutes, 30 minutes, or 1 hour. The time series monitoring image dataset is P = {p1 p2…p i}, and i is the monitoring image serial number. p i represents the latest monitoring image.

[0069] After taking each monitoring image, the average brightness value L where s is the total number of pixels in the monitoring image, and g j is the gray value of each pixel in the monitoring image.

[0070] If L i ≤ 40, it is determined that the shooting time of the current image p i is at night or in insufficient light conditions.

[0071] If L i > 40, it is determined that the current image p i is taken in sufficient light conditions.

[0072] Step 3: Obtain the time sequence displacement value of the infrared light source monitoring point

[0073] The image coordinates of the center of the infrared light source in the monitoring image are extracted by fitting the center of the ellipse. For each monitoring image, the image coordinates of the center of the nth infrared light source in the monitoring image p i are recorded as O i_n = (x i_n , y i_n ) (in pixels), where n = (1, 2, …, a).

[0074] The pixel displacement of the center of the infrared light source in the monitoring image p i relative to the previous monitoring image p i-1 is represented as shift i_i-1_n = |y i_n -y i-1_n |.

[0075] For all monitoring times in a specified time period (i.e. the time period between taking the monitoring image p r and taking the monitoring image p i ), the cumulative sum of all image displacement values of the grid in the time period is calculated as the image displacement value of the grid in the specified time period. Based on the image displacement value of each grid, the displacement of the corresponding slope area is calculated.

[0076] Preferably, in the time period between taking the monitoring image p i-1 and taking the monitoring image p i , the deformation of the slope area corresponding to the monitoring point where the infrared light source is located is

[0077] The cumulative deformation of the monitoring point where the infrared light source is located between the monitoring image p i and any previous monitoring image p r is r≤j≤i,1≤r≤i-1.

[0078] Based on the above calculation method, the cumulative deformation of each monitoring point where the infrared light source is located can be obtained.

[0079] Step 4: Image grid division and natural feature point extraction

[0080] Divide the monitoring image p i into grids, and the length and width of the grid are both 128 pixels. The grid is numbered as G i_q . The slope area corresponding to a single grid is

[0081] If the monitoring image pi Average brightness value L i If the value is ≤40, the image was taken at night or in low light conditions, resulting in a dark image, and no natural feature points will be extracted.

[0082] If the monitored image p i Average brightness value L i If the value is greater than 40, the image was taken under sufficient lighting conditions. This applies to the monitoring image p. i Each grid cell is used to extract image feature points using the SIFT algorithm to obtain the natural feature point set PT. i_q =

[0083] pt i_q_1 pt i_q_2 , ..., pt i_q_k}. Record the feature points pt of natural land features. i_q_k The image coordinates are (x i_q_k y i_q_k ).

[0084] Step 5: Matching natural feature points and obtaining temporal displacement values ​​from time-series monitoring images

[0085] Pick

[0086] If the monitored image p i Average brightness value L i >40, select monitoring image p i The previous monitoring image with an average brightness value greater than 40 is denoted as p. r .

[0087] Matching monitoring image p i and p r The corresponding natural feature points in the grid. To speed up the matching process and narrow down the range of candidate matching points, based on the monitoring image p i Relative to the monitoring image p r The average pixel shift is Avg_shift i_r Construct the search space.

[0088] The monitoring image p is calculated using the pixel displacement of a infrared light sources installed on the slope. i

[0089] Compared to the previous monitoring image p i-1 The average pixel displacement is

[0090] Calculate the captured monitoring image p r To capture monitoring images p i The cumulative sum of the average pixel shifts of the monitored images within the specified time period is taken as the average pixel shift Avg_shift of the monitored images within that time period.i_r .

[0091] For monitoring image p i Natural feature points pt i_q_k Its monitoring image p r

[0092] The search space for candidate matching points is defined as follows:

[0093] Area i_r_q_k

[0094] ={x i_q_k -1 <x<x i_q_k +1,y i_q_k -2*Avg_shift i_r <y<y i_q_k +1

[0095] For monitoring image p i Natural feature points pt i_q_k Select monitoring image p i-1 Located in the search space Area i_r_q_k The SIFT feature matching algorithm was used to match the feature points of natural land features within the area.

[0096] Record monitoring images p i With monitoring image p i-1 In grid G i_q The coordinates of the matching points within the grid. Using grid G. i_q The average of the differences in the Y coordinates of all matching points within the image is used as the monitoring image p. i Relative to the monitoring image p i-1 In grid G i_q The average pixel displacement within the monitored image p i Image displacement value shift corresponding to the monitoring time i_i-1_G .

[0097] For capturing monitoring images p r To capture monitoring images p i For all monitoring times within a given time period, calculate the cumulative sum of all image displacement values ​​for that grid within that time period, and use this sum as the image displacement value for that grid within that specific time period. Based on the image displacement value of that grid, calculate the displacement of the corresponding slope area.

[0098] Preferably, when capturing monitoring images p i With monitoring image p i-1 The time period between, grid G i_q The corresponding slope displacement is

[0099] Then, when capturing monitoring images pi between any monitoring image p r and the grid G i_q The cumulative displacement of the corresponding slope region is wherein r≤j≤i, 1≤r≤i-1.

[0100] Step 6: extraction of the slope landslide range, taking the following calculation strategy:

[0101] (1) For the grid with infrared light source monitoring points, the displacement value calculated by the infrared light source monitoring points according to the method described in step 3 is taken as the corresponding slope displacement value of the grid;

[0102] (2) For the grid without marked infrared light source center but with natural feature points:

[0103] When the average brightness of the monitoring image is greater than the preset threshold, the displacement value calculated by the natural feature points according to the method described in step 5 is taken as the corresponding slope displacement value of the grid;

[0104] When the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement value is assigned as the average value of the eight-neighbor image displacement values, and then the corresponding slope region displacement value is calculated.

[0105] (3) For the grid without infrared light source monitoring points and without extracted natural feature points, the image displacement value is assigned as the average value of the eight-neighbor image displacement values, and then the corresponding slope region displacement value is calculated.

[0106] Specifically, the process of assigning the image displacement value of the grid as the average value of the eight-neighbor grid displacement values includes: counting the set of grids with non-zero displacement values in the eight-neighbor grids of the grid as H, the number of elements in H is z, and taking the average displacement value of the grid set H as the corresponding slope displacement value of the grid

[0107] Based on the above strategy, the time-series cumulative displacement value of the corresponding slope of each grid divided by the monitoring image can be obtained.

[0108] For the time period between the shooting monitoring image p i and any monitoring image p r , the grid G i_q The cumulative displacement of the corresponding slope region is SumV i_r_G , and the displacement threshold Thre is set.

[0109] If SumV i_r_G > Thre, it is determined that the grid G i_q The corresponding slope between the shooting monitoring image p i and the monitoring image p r has a large cumulative displacement;

[0110] If SumV i_r_G < Thre, then it is determined that the grid G i_q The corresponding slope in the captured monitoring image p i and the monitoring image p r have a smaller cumulative displacement between them.

[0111] Preferably, the monitoring image and the divided grid are displayed through visualization software, and the grids where SumV i_r_G > Thre are connected in the visualization interface to form a closed area, which is the slope landslide range.

[0112] Preferably, the area of the grids where SumV i_r_G > Thre is accumulated through a background program, which is the total landslide area of the slope and is displayed.

[0113] Compared with the prior art, the present invention has the following remarkable advantages and effects:

[0114] (1) Low monitoring cost. For a single highway slope, the present invention only needs to deploy 1 camera and several infrared light source targets, and the cost is greatly reduced compared with satellite navigation positioning technology, satellite remote sensing monitoring technology, etc.;

[0115] (2) All-weather and all-time uninterrupted monitoring. The present invention uses infrared light source targets to replace conventional targets, enabling high-precision identification and positioning of monitoring targets at night or under low-light conditions. Under sufficient sunlight conditions, both infrared light source targets and natural feature targets are monitored simultaneously, and the displacement amount of the infrared light source target is used as the initial displacement value of the natural feature target to assist in the displacement calculation of the natural feature target, realizing all-weather and all-time uninterrupted monitoring;

[0116] (3) Simple data processing process. Using the displacement value of the infrared light source target as the initial value of the natural feature point displacement value greatly reduces the matching search range of natural feature points, improving the matching accuracy and calculation efficiency;

[0117] (4) Realize slope surface monitoring. Existing monitoring technologies all monitor the displacement of a small number of discrete points on the slope. The present invention uses the method of setting infrared light source monitoring points and extracting natural feature monitoring points to achieve uniform distribution of slope monitoring points, thereby realizing slope surface monitoring and facilitating the evaluation of the scale and destructiveness of slope landslide bodies.

[0118] Embodiment 2

[0119] The present invention also provides a highway slope deformation monitoring system, which is used to implement the highway slope deformation monitoring method described in the above technical solution, including:

[0120] The grid division module is configured to divide an initial image of the highway slope deformation area into grids, and extract natural feature points and infrared light source centers, and mark corresponding grids and image coordinates.

[0121] The displacement calculation module is configured to calculate an image displacement value of each grid based on a real-time acquired monitoring image of the highway slope deformation area.

[0122] For the grid marked with the infrared light source center, the image displacement value of the grid is calculated based on image coordinates of the corresponding infrared light source center in continuous monitoring images.

[0123] For the grid not marked with the infrared light source center but marked with the natural feature point:

[0124] When the average brightness of the monitoring image is greater than a preset threshold, the image displacement value is calculated based on image coordinates of the natural feature point in continuous monitoring images.

[0125] When the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement of the grid is assigned as an average value of displacement values of neighboring grids.

[0126] For the grid neither having the infrared light source monitoring point nor having the natural feature point, the image displacement of the grid is assigned as an average value of displacement values of neighboring grids.

[0127] Based on the image displacement value of each grid, a displacement of a corresponding slope area is calculated.

[0128] Embodiment 3

[0129] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the highway slope deformation monitoring method in the above technical solution.

[0130] Embodiment 4

[0131] The application further provides an electronic device, which comprises a memory and a processor, the memory and the processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to execute the highway slope deformation monitoring method in the above technical solution.

[0132] Those skilled in the art will understand that embodiments of the application can be provided as methods, systems, or computer program products. Therefore, the application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the application can be in the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0133] The embodiments of methods, apparatuses (systems) and computer program products according to the present application can be described in reference to flowchart and / or block diagram illustrations of methods, apparatuses (systems) and computer program products according to embodiments of the present application. It will be understood that each block of the flowchart and / or block diagrams and combinations of blocks in the flowchart and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 Figure 1

[0134] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart and / or block diagram block or blocks. Figure 1 Figure 1

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 Figure 1

[0136] The embodiments of the present application described above with reference to the drawings are merely illustrative, and not restrictive, and many modifications can be made by those skilled in the art without departing from the spirit and scope of the present application, and these modifications are also intended to be within the scope of the present application.

[0137] The contents not described in detail in the specification are the prior art known to those skilled in the art.​​​​​​

Claims

1. A method for monitoring deformation of a highway slope, characterized by: The deformation zone of the highway slope is provided with an infrared light source; the method comprises the following steps: Grid division is performed on the initial image of the deformation zone of the highway slope, and natural feature points and the center of the infrared light source are extracted, and the corresponding grid and image coordinates are marked; Based on the real-time monitoring image of the deformation zone of the highway slope, the image displacement value of each grid is calculated; For the grid marked with the center of the infrared light source, the image displacement value of the grid is calculated based on the image coordinates of the corresponding center of the infrared light source in the continuous monitoring image; For the grid not marked with the center of the infrared light source but marked with the natural feature points: When the average brightness of the monitoring image is greater than a preset threshold, the image displacement value is calculated based on the image coordinates of the natural feature points in the continuous monitoring image; When the average brightness of the monitoring image is less than or equal to the preset threshold, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids; For the grid without the monitoring point of the infrared light source and the natural feature points, the image displacement of the grid is assigned as the average value of the displacement values of the neighboring grids; Based on the image displacement value of each grid, the displacement of the corresponding slope region is calculated.

2. The method of claim 1, wherein: The deformation zone of the highway slope is provided with a plurality of monitoring piles buried along the line from top to bottom; each monitoring pile is provided with an infrared light source; a reference pile is buried in the stable region outside the deformation zone of the highway slope, and a monitoring camera is installed on the reference pile, which is used to acquire the initial image and the monitoring image.

3. The method of claim 1, wherein: The process of calculating the image displacement value of the grid marked with the center of the infrared light source comprises: for each grid marked with the center of the infrared light source, the absolute difference value of the image vertical coordinates of the center of the infrared light source in the continuous two monitoring images is calculated as the image displacement value of the grid at the corresponding monitoring time; for all monitoring times within a specified time period, the cumulative sum of all image displacement values of the grid within the time period is calculated as the image displacement value of the grid within the specified time period.

4. The method of claim 1, wherein: The process of calculating the image displacement value based on the image coordinates of the natural feature points in the continuous monitoring image comprises: for any grid marked with the natural feature points, the average value of the difference values of the image vertical coordinates of all natural feature points in the grid in the continuous two monitoring images is calculated as the image displacement value of the grid at the corresponding monitoring time; for all monitoring times within a specific time period, the cumulative sum of all image displacement values of the grid within the time period is calculated as the image displacement value of the grid within the specific time period; the specific time period is the time period between the specified monitoring time and the monitoring time corresponding to the last monitoring image with brightness greater than the preset threshold.

5. The method of claim 4, wherein: The process of acquiring the natural feature points of each monitoring image within the specific time period except the specified monitoring time comprises: based on the average pixel displacement of the monitoring images within the specific time period, the natural feature points of each monitoring image are matched in combination with the natural feature points matched at the specified monitoring time.

6. The method of claim 4, wherein: The monitoring image p at a specified time i Natural terrain feature point pt i_q_k The definition of the search space is as follows: Area i_r_q_k = {x i_q_k -1 < x < x i_q_k +1, y i_q_k -2 * Avg_shift i_r < y < y i_q_k +1} where the image coordinates of the natural terrain feature point pt i_q_k are (x i_q_k , y i_q_k ), and Avg_shift i_r represents the average pixel displacement of the monitoring image within a specified time period.

7. The method of claim 6, wherein: For all monitoring instants within a certain time period, the cumulative sum of the average pixel displacement of the monitoring images within the time period is calculated as the average pixel displacement of the monitoring images within the time period; for any monitoring image, the average value of the absolute difference between its and the vertical coordinates of all infrared light source centers in the previous monitoring image is calculated as the average pixel displacement of the monitoring image.

8. The method of claim 1, wherein: At least one grid marked with the center of the infrared light source exists in the neighborhood of each grid.

9. The method of claim 4, wherein: The method further comprises the following steps: connecting the grids with image displacement values greater than a set threshold within a certain time period to form a closed region, marking the slope region corresponding to the closed region as the slope landslide range, and calculating the area of the slope region.

10. A highway slope deformation monitoring system, characterized by: The system is used to implement the highway slope deformation monitoring method of any one of claims 1-9.

Images