A monitoring method and monitoring system for a slope

Abstract

The present invention relates to the technical field of highway slope monitoring, and specifically to a monitoring method and a monitoring system for slopes. The monitoring method includes the following steps: arranging a plurality of targets on the slope; establishing a coordinate system on the target surface of the target to obtain the world coordinates of the target; acquiring an image containing each target surface and establishing a pixel coordinate system on the image; obtaining the displacement of the target in the world coordinates through the change of the target in the pixel coordinates; triggering slope early warning through the displacement of the target. The problems of insufficient accuracy and high cost in existing slope monitoring are solved. It is possible to simultaneously monitor multiple points on the slope, flexibly arrange the positions of the monitoring points according to the actual situation, and achieve large-scale monitoring of the slope at a relatively low cost. The obtained monitoring data can be directly associated with the displacement of the slope. When dealing with an ongoing early warning, the true state of the slope can also be directly observed based on the collected image.

Description

Technical Field

[0001] The present invention relates to the technical field of highway slope monitoring, and more specifically, to a monitoring method and a monitoring system for slopes. Background Art

[0002] In recent years, with the continuous leap of intelligent technology and informatization level, many new perspectives and new ideas have emerged in the field of highway slope monitoring and disaster warning. These innovative technologies not only improve the efficiency and accuracy of slope monitoring, but also provide a solid technical guarantee for the safe operation of highways. In terms of slope monitoring technology, the adaptability to complex environments is particularly crucial. Highway slopes are often in natural environments with complex terrains and variable climates. Therefore, slope monitoring technology must have strong environmental adaptability to ensure accurate and stable operation under various conditions. In addition, the quantitative analysis of slope deformation has also become a research hotspot. By accurately measuring and quantitatively analyzing key parameters such as slope displacement and deformation rate, the deformation law of slopes can be understood more deeply, providing a more scientific basis for disaster warning.

[0003] Through high-definition cameras and image processing technology, real-time video of the slope can be obtained, and the features such as the shape, deformation, and cracks of the slope can be intuitively observed and analyzed. This not only improves the intuitiveness and accuracy of monitoring, but also provides richer information for disaster warning. However, although image analysis technology plays an important role in slope monitoring, its accuracy problem cannot be ignored. Since the deformation of the slope is a slow and continuous process, errors often occur in the conversion and processing of existing monitoring images, resulting in the failure to detect the displacement and deformation features caused by some obvious image changes in a timely manner. This not only affects the accuracy of monitoring, but also may bring potential risks to the safe operation of highways. To solve this problem, the existing technology reduces the error in the image conversion process by optimizing the image processing algorithm; or uses more advanced cameras and sensors to improve the quality and resolution of images. However, these methods often require increasing the number and complexity of monitoring devices. For example, commonly, by the way of corresponding the acquisition device to the monitoring point, in large-scale monitoring, for each additional monitoring point, one more set of acquisition devices is required, resulting in an increase in monitoring costs. Summary of the Invention

[0004] The present invention aims to overcome at least one defect of the above-mentioned prior art, and provides a monitoring method and a monitoring system for slopes, which are used to solve the problems of insufficient accuracy and high cost in existing slope monitoring.

[0005] The technical solution adopted by the present invention is a monitoring method for a slope, which includes the following steps: setting a plurality of targets on the slope; establishing a coordinate system on the target surface of the target to obtain the world coordinates of the target; obtaining an image containing each target surface and establishing a pixel coordinate system on the image; obtaining the displacement of the target in the world coordinates through the change of the target in the pixel coordinates; triggering the slope warning through the displacement of the target.

[0006] The step of "obtaining an image containing each target surface and establishing a pixel coordinate system on the image" is specifically: performing a rigid body transformation on the world coordinates of any point on the target surface to convert it into acquisition coordinates; performing a perspective projection on the acquisition coordinates to convert them into image coordinates; performing a secondary transformation on the image coordinates to obtain pixel coordinates.

[0007] Let be the rotation matrix, be the translation matrix, be the 0 matrix, and the point in the world coordinates is converted into a point on the acquisition coordinates through the transformation matrix ;

[0008] Let be the focal length, and the point on the acquisition coordinates is converted into a point on the image coordinates through the transformation matrix ;

[0009] Let , be the physical size of the pixel point in, and the point on the pixel coordinates is converted into a point on the pixel coordinates through the transformation matrix ;

[0010] The relational formula between the world coordinates and the pixel coordinates is: , where the coordinates of the pixel are ; the coordinate values in the world coordinates are , is the scale factor, is the center of the pixel coordinate system.

[0011] After the step of "obtaining the displacement of the target in the world coordinates", the following steps are further included: setting an anchor target within the visual range of the image; obtaining a correction parameter through the anchor target; correcting the displacement data through the correction parameter.

[0012] The step of "triggering the slope warning through the displacement of the target" is specifically: continuously obtaining images within the time ;

[0013] Obtaining the displacement change of the target in the world coordinates from images; the displacement change includes: the change in the axis and the change in the The change of the axis is , and that of the axis of the target is ; Obtain the amplitude of the = , the amplitude of the , ; Obtain the range of the amplitudes , ; Set the warning threshold . When and / or , and ;

[0014] Obtain the local slope , , where is or , is the time;

[0015] If , give a warning.

[0016] Furthermore, a monitoring system for a slope is provided, including a monocular image collector, a plurality of monitoring targets, a control host and a storage; the monitoring targets are dispersedly arranged on the slope to be monitored, and monitoring points are provided on the target surface of the monitoring targets; the monocular image collector is fixed outside the slope to be monitored and frames all the monitoring targets within its field of view for obtaining digital images; the control host establishes the pixel coordinates of each monitoring point in the digital image and obtains the displacement change of the monitoring target according to the displacement change of the monitoring point in the pixel coordinates to realize the monitoring and warning of the slope.

[0017] The target surface of the monitoring target includes: a first unit and a second unit; the first unit and the second unit adopt a pair of contrast colors; the first unit and the second unit are irregularly distributed on the target surface so that the monitoring points in the pixel coordinates change with any displacement.

[0018] The monitoring system further includes an anchor target; the anchor target is arranged outside the slope and between the monocular image collector and the monitoring target; the anchor target is within the field of view of the monocular image collector; the monocular image collector is directly opposite the anchor target; the distance between the monocular image collector and the anchor target is less than the distance between the monocular image collector and the monitoring target.

[0019] The monitoring system also includes a fill light device; the fill light device uses an array infrared light source with a wavelength of 800nm ​​to 900nm, and the array infrared light source is in the same direction as the monocular image collector.

[0020] The monitoring target includes a target plate, a target rod and a fixed foundation, wherein the fixed foundation is made of concrete and a weed control cloth is arranged around the fixed foundation.

[0021] Compared with the prior art, the beneficial effects of the present invention are: it can monitor multiple points on the slope at the same time, flexibly arrange the positions of the monitoring points according to the actual situation, and realize large-scale monitoring of the slope at a relatively low cost. The acquired monitoring data can be directly associated with the displacement of the slope, and when responding to the warning that is occurring, the real state of the slope can be directly observed according to the collected image. Different from indirect monitoring data, there is no need to conduct a secondary on-site investigation, which effectively reduces the operation and maintenance cost and labor cost of slope monitoring. By adding specific supplementary lighting equipment, the problem of insufficient illumination and inability to accurately identify the target in large-scale monitoring is effectively solved. And by designing an irregular target surface and adding a weeding cloth, the problems that the current monocular visual monitoring system encounters in the actual engineering implementation process are optimized. In order to further improve the accuracy of the data, the monitoring interference correction caused by the shaking of the monocular image collector is innovatively included in the calculation, and the uncontrollable shaking problem is solved based on the synchronous correction of the anchor target. In addition, combined with the actual engineering warning effect, a warning algorithm based on time, amplitude and local slope is provided, which greatly improves the accuracy of the warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The flowchart of the monitoring method in the present invention.

[0023] Figure 2 Schematic diagram of the monitoring system of the present invention.

[0024] Figure 3 It is a schematic diagram of the image in the present invention.

[0025] Figure 4 This is a comparison chart of data changes before and after fill light in the present invention.

[0026] Figure 5 It is a comparison diagram of the data changes before and after correction by the correction parameters in the present invention.

[0027] Figure 6 Schematic diagram of the target in the present invention.

[0028] Figure 7 It is a schematic diagram of the target surface in the present invention.

[0029] Figure 8 This is an example diagram of the calibration ruler in the present invention.

[0030] Description of reference numerals: Highway 100, monitoring target 101, target board 111, first unit 121, second unit 122, target rod 112, fixed foundation 113, weed control fabric 114, anchored target 102, monocular image collector 103, slope 200. Specific implementation mode

[0031] The attached drawings of the present invention are only for illustrative purposes and should not be construed as a limitation of the present invention. To better illustrate the following embodiments, some components in the attached drawings will be omitted, enlarged or reduced, which do not represent the dimensions of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the attached drawings may be omitted.

[0032] Embodiment 1

[0033] This embodiment is a monitoring method for slope 200, including the following steps:

[0034] Set a number of targets on slope 200;

[0035] Establish a coordinate system on the target surface of the target and obtain the world coordinates of the target;

[0036] Obtain the images containing each target surface and establish a pixel coordinate system on the images;

[0037] Obtain the displacement of the target in the world coordinates through the change of the target in the pixel coordinates;

[0038] Trigger the slope 200 early warning through the displacement of the target.

[0039] Through the association between the world coordinates and the pixel coordinates in the image, the simultaneous monitoring of multiple targets on a slope 200 is realized. Reflecting the change of the world coordinates by the change of the pixel coordinates can accurately reflect the position offset of the target through the image, so as to realize the precise monitoring of a large-range slope 200, meet the requirements of high-precision large-range monitoring, and adapt to complex environments to master the displacement changes of the surface layer and slope body of slope 200 in real time and effectively. This solution obtains the pictures of the targets formed on the slope 200 to be monitored, calculates the target displacement, so as to reflect the state of slope 200. Due to the wide monitoring range, multi-point monitoring should be adopted to ensure the stable monitoring of each level of slope 200.

[0040] The step of "obtain the images containing each target surface and establish a pixel coordinate system on the images" is specifically:

[0041] Perform a rigid body transformation on the world coordinates of any point on the target surface to convert them into acquisition coordinates;

[0042] Perform a perspective projection on the acquisition coordinates to convert them into image coordinates;

[0043] Perform a quadratic transformation on the image coordinates to obtain pixel coordinates.

[0044] The world coordinates are transformed through rigid body transformation, perspective projection, and quadratic transformation to form pixel coordinates, establishing the correlation between the world coordinates and the pixel coordinates. Thus, the world coordinates of each point on the target board 111 can be converted into pixel coordinates. The change of the target is accurately obtained through the change of the pixel coordinates. By providing a conversion step between the world coordinates and the image coordinates, the correlation between the world coordinates and the pixel coordinates is established in a concise and effective manner.

[0045] Let be the rotation matrix, be the translation matrix, be the zero matrix, and the points in the world coordinates are transformed through the transformation matrix into points on the acquisition coordinates;

[0046] Let be the focal length, and the points on the acquisition coordinates are transformed through the transformation matrix into points on the image coordinates;

[0047] Let , be the physical size of the pixel point at, and the points on the pixel coordinates are transformed through the transformation matrix into points on the pixel coordinates;

[0048] The relational expression between the world coordinates and the pixel coordinates is: , where the coordinates of the pixel are ; the coordinate values in the world coordinates are , is the scale factor, is the center of the pixel coordinate system.

[0049] After the step of "obtaining the displacement of the target in the world coordinates", the following steps are further included:

[0050] Set the anchor target 102 within the visual range of the image;

[0051] Obtain the correction parameters through the anchor target 102;

[0052] Correct the displacement data through the correction parameters.

[0053] The monitoring data deviation caused by shaking is mainly due to the monocular camera. This deviation is reflected in that the amplitude change in the world coordinates is larger than the change value of the target itself, seriously interfering with the monitoring results and judgments. By increasing the setting of the anchoring target 102 to synchronously obtain the shaking caused by environmental force majeure, and then through the correction parameters expressing the shaking, the shaking of the image acquisition device caused by force majeure can be effectively eliminated, thereby correcting the deviation in the monitoring results, improving the monitoring accuracy of displacement, and improving the monitoring data problems caused by the instability of the image acquisition device.

[0054] The step of "triggering the warning of the slope 200 through the displacement of the target" is specifically as follows:

[0055] Within the time continuously obtain images;

[0056] From images, obtain the displacement change of the target in the world coordinates;

[0057] The displacement change includes: the change of the axis and

[0058] the axis change of the target is , the axis change of the target is ;

[0059] Obtain the amplitude of the axis , axis amplitude , ;

[0060] Obtain the range of the amplitude, ;

[0061] Set the warning threshold , when and / or , and ;

[0062] Obtain the local slope , , where is or , is the time;

[0063] If , issue a warning.

[0064] The change of the target world coordinates can be accurately used for the long-term warning of the slope 200 through the above calculations. By combining accurate quantitative judgment and time, the current state of the slope 200 can be judged in time, and phenomena such as landslides and even cracks can be quickly grasped, with a high accuracy rate.

[0065] Embodiment 2

[0066] This embodiment is a monitoring system for the slope 200, including a monocular image collector 103, a plurality of monitoring targets 101, a control host and a storage; the monitoring targets 101 are dispersedly arranged on the slope 200 to be monitored, and monitoring points are arranged on the target surfaces of the monitoring targets 101; the monocular image collector 103 is fixed outside the slope 200 to be monitored and frames all the monitoring targets 101 within its field of view, and is used to obtain digital images; the control host establishes the pixel coordinates of each monitoring point in the digital image, and obtains the displacement change of the monitoring target 101 according to the displacement change of the monitoring point in the pixel coordinates, so as to realize the monitoring and warning of the slope 200.

[0067] By means of a monocular image collector 103 cooperating with a plurality of monitoring targets 101, which is different from the combination of the existing image collector and the target where the acquisition device and the monitoring points correspond one by one, the construction cost of the monitoring system is greatly reduced. By judging the true displacement of the monitoring target 101 through the pixel coordinates, while realizing the simultaneous monitoring and mutual comparison of multiple targets, the requirements for monitoring accuracy and sensitivity are met.

[0068] The target surface of the monitoring target 101 includes: a first unit 121 and a second unit 122; the first unit 121 and the second unit 122 adopt a pair of contrasting colors; the first unit 121 and the second unit 122 are irregularly distributed on the target surface, so that the monitoring points in the pixel coordinates change with any displacement.

[0069] Through the irregular distribution of two units with contrasting colors, it is prevented that the colors or arrangements of the pixels before and after displacement are exactly the same during any displacement of the pixels constituting the monitoring points, avoiding the wrong judgment caused by the limitation of pixel recognition after displacement, which helps to correct the lack of displacement monitoring through pixel coordinates and improve the accuracy and precision of monitoring.

[0070] The monitoring system further includes an anchoring target 102; the anchoring target 102 is arranged outside the slope 200 and between the monocular image collector 103 and the monitoring target 101; the anchoring target 102 is within the field of view of the monocular image collector 103; the monocular image collector 103 faces the anchoring target 102 directly; the distance between the monocular image collector 103 and the anchoring target 102 is less than the distance between the monocular image collector 103 and the monitoring target 101. Further, the position where the anchoring target 102 is located is lower than the position where the monitoring target 101 is located.

[0071] The setting of the anchoring target 102 is further defined by position and distance, which helps to fit the area where the monocular image collector 103 is located more closely, withstand environmental impacts more similar to those of the monocular image collector 103, so as to obtain the pixel displacement of the monocular image collector 103 caused by the environment. Through the monitoring and data acquisition of the anchoring target 102, it can be used to correct the fluctuations caused by the interference of the monocular image collector 103 itself to the monitoring target 101, thereby improving the monitoring accuracy and precision.

[0072] The monitoring system further includes a supplementary lighting device; the supplementary lighting device uses an array infrared light source with a wavelength of 800 nm to 900 nm, and the array infrared light source is in the same direction as the orientation of the monocular image collector 103.

[0073] In the combination of one monocular image collector 103 and multiple monitoring targets 101 in the monitoring system, the distance between the monocular image collector 103 and the monitoring targets 101 is far, which is to frame each monitoring target 101 within the same visual range. Inevitably, it is necessary to increase the illumination of the monitoring targets 101 to ensure that the monitoring targets 101 can be clearly identified. The coverage range of the supplementary lights of the existing targets is small and the distance is short, which cannot meet the good supplementary lighting of each monitoring target 101 at night. By using an infrared light supplementary lighting device with a specific wavelength band for supplementary lighting, not only the supplementary lighting range is increased, but also the supplementary lighting requirements of the monitoring targets 101 in all directions are met through an array manner. Preferably, the wavelength is 850 nm. Through the supplementary lighting device, the monitoring system can meet the all-day and uninterrupted monitoring of the slope 200. The array is a 3×5 matrix array. As Figure 4 shown, the upper curve is the displacement change of the X-axis in the world coordinate, and the lower curve is the displacement change of the Y-axis in the world coordinate. The ordinate of both is displacement, and the abscissa is the corresponding time. To the left of the dotted line is before supplementary lighting. It can be seen that the data is easily in the state of 0 before supplementary lighting. To the right of the dotted line is after supplementary lighting, the data fluctuation decreases and the state of 0 no longer appears. It can be seen that after installing the supplementary lighting device, the collected data can be optimized and the fluctuations caused by environmental impacts can be reduced.

[0074] As Figure 6As shown in the figure, the monitoring target 101 includes a target plate 111, a target rod 112, and a fixed base 113. The fixed base 113 is made of concrete, and a weed control fabric 114 is also provided around the fixed base 113 and surrounds the fixed base 113.

[0075] Most of the slopes 200 to be monitored are covered with weeds. Weeds grow rapidly and are extremely likely to cover the monitoring target 101. The traditional method is to increase the height to avoid occlusion, such as increasing the height of the concrete base at the bottom of the monitoring target 101, or increasing the height of the target rod 112. However, these two methods not only have high costs, but also cannot truly solve the problem of weed growth. The high position of the target will also lead to unstable monitoring data. By using the method of combining the weed control fabric 114 with the fixed base 113, in this embodiment, specifically, a green weed control fabric 114 with a size of 2m × 1m is used to cover the periphery around the fixed base 113 of the target, which essentially blocks the growth of weeds and has low costs. By setting around the fixed base 113, the problem of weed growth on the slope 200 is effectively solved, and the monitoring is prevented from being ineffective due to occlusion of the target.

[0076] Embodiment 3

[0077] As Figure 2 As shown in the figure, this embodiment is a monitoring system for a slope 200, including a monocular image collector 103, multiple monitoring targets 101, a control host, and a storage device; the monitoring targets 101 are scattered on the slope 200 to be monitored, and monitoring points are provided on the target surface of the monitoring target 101; the monocular image collector 103 is fixed outside the slope 200 to be monitored and frames all the monitoring targets 101 within its field of view, and is used to obtain digital images; the control host establishes the pixel coordinates of each monitoring point in the digital image, and obtains the displacement change of the monitoring target 101 according to the displacement change of the monitoring point in the pixel coordinates, so as to realize the monitoring and early warning of the slope 200. A monitoring method for the slope 200 is stored in the storage device, and the controller reads this monitoring method to control the operation of the monitoring system.

[0078] The monitoring system further includes: a communication device, a power storage component, and an external solar energy system. The monitoring system is connected to the Internet of Things, and the data in the storage device is transmitted to a specific server through the communication device. The monitoring system adopts a method of using a single monocular camera with multiple monitoring targets 101. By increasing the targets, the monitoring points can be increased. Compared with the traditional method in which the camera device and the monitoring points correspond one by one, simply increasing the number of targets to achieve the same function greatly reduces the cost of the entire monitoring system, and each monitoring point formed by the target can achieve good and effective monitoring effects within the field of view. The distance between the monocular camera and the monitoring target 101 is 150m to 250m.

[0079] The target surface of the monitoring target 101 includes: a first unit 121 and a second unit 122; the first unit 121 and the second unit 122 adopt a pair of contrasting colors; the first unit 121 and the second unit 122 are irregularly distributed on the target surface, so that the monitoring points in the pixel coordinates change with any displacement. The first unit 121 and the second unit 122 can specifically be square blocks, the first unit 121 is white, and the second unit 122 is black. Through reasonable target design, the accuracy of target center positioning and edge detection can be improved, which plays a great role in improving the monitoring accuracy. The checkerboard-style target is prone to inaccurate feature point recognition during the monitoring process. That is, when the pixel point changes from one cross point to another cross point, the monocular image collector 103 will judge that there is no pixel change due to the continuity of the pixels, ignoring the actual displacement and generating errors. Therefore, the target surface of the monitoring target 101 is composed of the unevenly distributed first unit 121 and second unit 122. As Figure 7 shown, in order to further form irregularity, it can also be rotated. After rotation, the units at some edges no longer maintain a square shape, but it can increase the recognition.

[0080] The monitoring system further includes an anchoring target 102; the anchoring target 102 is arranged outside the slope 200 and is located between the monocular image collector 103 and the monitoring target 101; the anchoring target 102 is within the field of view of the monocular image collector 103; the monocular image collector 103 is directly facing the anchoring target 102; the distance between the monocular image collector 103 and the anchoring target 102 is less than the distance between the monocular image collector 103 and the monitoring target 101. In the calculation of the existing monitoring target 101, it is usually carried out under the ideal condition that the monocular image collector 103 is stable and motionless. However, in the actual operation process, due to natural factors such as the influence of vehicle driving on the road 100 and the instability of the roadbed, the monocular image collector 103 will shake. If not corrected, huge deviations will occur, seriously affecting the monitoring accuracy and accuracy. As for the camera or camera of the monocular image collector 103, the installation of the monitoring target 101 will inevitably be affected by the geographical environment, the width of the road 100, and the distance between the monocular image collector 103 and the target. Therefore, when setting the anchoring target 102, it is necessary to further limit its distance to improve the consistency of the environment between the anchoring target 102 and the monocular image collector 103.

[0081] The monitoring method includes the following steps: setting a number of targets on the slope 200; establishing a coordinate system on the target surface of the target to obtain the world coordinates of the target; obtaining the images containing each target surface and establishing a pixel coordinate system on the images; obtaining the displacement of the target in the world coordinates through the change of the target in the pixel coordinates; triggering the slope 200 warning through the displacement of the target.

[0082] The target surface is a vertical plane and is square. Taking the upper left corner of each target surface as the origin, a world coordinate system is established. The world coordinate system is a three-dimensional coordinate system, and the points in the world coordinates are . Any point on the target surface can be represented by . The target surface forms an original image when imaged in the monocular lens. A two-dimensional coordinate system is established in the original image to obtain the camera coordinates. Any point on the original image can be represented by the camera coordinates . As shown in Figure 3 , in this embodiment, a slope 200 beside a road 100 is monitored. The side close to the slope 200 to be monitored is the left side of the road 100, and the side far from the slope 200 to be monitored is the right side of the road 100. The monocular image collector 103 is arranged on the right side of the road 100 and fixed on the concrete road shoulder. A plurality of monitoring target markers 101 are provided on the slope 200 to be monitored. In this embodiment, 6 are adopted and evenly distributed on the slope 200 to be monitored. The target surfaces of each monitoring target marker 101 are parallel to each other and face the monocular image collector 103. The monocular image collector 103 includes all 6 monitoring target markers 101 within the field of view. A world coordinate system is established on the target surface of each monitoring target marker 101. The monocular image collector 103 may specifically be a monocular camera. An image including 6 monitoring target markers 101 is acquired through the monocular image collector 103, and a pixel coordinate system is established in the image, and the images of the target surfaces of each monitoring target marker 101 are all represented by pixel coordinates. The size of the target surface is 20 cm × 20 cm. The association between the world coordinates and the pixel coordinates in the image is specifically reflected in the internal parameters and external parameters of the monocular image collector 103.

[0083] The step of "acquiring an image containing each target surface and establishing a pixel coordinate system on the image" is specifically as follows: performing a rigid body transformation on the world coordinates of any point on the target surface to convert them into acquisition coordinates; performing a perspective projection on the acquisition coordinates to convert them into image coordinates; and performing a secondary transformation on the image coordinates to obtain the pixel coordinates.

[0084] The point with world coordinates on the target surface is . The image optical signal of the target surface enters the monocular image collector 103 to form an original image. By establishing a two-dimensional acquisition coordinate system for the original image in the monocular image collector 103, the points on the image of each target surface in the original image are . By adjusting the optical center of the acquisition coordinate system and combining a rigid body transformation, specifically a rotation transformation and a horizontal transformation, each target surface is included in the original image. The original image enters the monocular image collector 103 and generates a digital negative through perspective projection. Specifically, the original image is made to perform a pinhole imaging through the monocular image collector 103 to form a digital negative carrying the images of each target marker. A two-dimensional image coordinate is established on the image on the digital negative, and the points of each target surface in the image are . Through secondary coordinate transformation, two-dimensional pixel coordinates are established in the image, and the points in the image coordinates are converted into points on the pixel coordinates. . Since the pixel coordinates are sequentially converted through the world coordinates, the acquisition coordinates, and the image coordinates, the points on the world coordinates (i.e., the points on the target surface) correspond one-to-one with the points on the pixel coordinates, and the correlation between them can be obtained.

[0085] Let be the rotation matrix, be the translation matrix, be the 0 matrix, and the points in the world coordinates are converted into points on the acquisition coordinates through the transformation matrix ;

[0086] Let be the focal length, and the points on the acquisition coordinates are converted into points on the image coordinates through the transformation matrix ;

[0087] Let , be the physical size of the pixel point at, also known as the focal length, which is a fixed value. The points on the pixel coordinates are converted into points on the pixel coordinates through the transformation matrix ;

[0088] The relational formula between the world coordinates and the pixel coordinates is: , where the coordinates of the pixel are ; the coordinate values in the world coordinates are , is the scale factor, is the center of the pixel coordinate system.

[0089] In this embodiment, combined with the parameters of the monocular image collector, it can be known that: , , the pixel is 3840*2160, .

[0090] As an option, the formula can be further described as:

[0091] , where the matrix is the conversion matrix from the world coordinates to the pixel coordinates.

[0092] This embodiment obtains the matrix by the following steps. Taking a triangle as the calibration ruler, according to the actual lengths of the three line segments of the triangle in the world coordinate system and the actual lengths in the pixel coordinate system, the matrix is obtained. As shown in the appendix Figure 8As shown, the pixel-level center position of the light spot is obtained by using the deep learning object recognition algorithm, so as to obtain the length of the calibration ruler in the pixel coordinate system. In this embodiment, the calibration ruler has a size of right-angled side * right-angled side * hypotenuse = 600 * 600 * 850.

[0093] Specifically, The matrix is obtained in the following way:

[0094] class calibParam{

[0095] public double[] pwv; / / pixel -> world transformation matrix: 9 doubles

[0096] public double[] wpv; / / world -> pixel transformation matrix: 9 doubles

[0097] public double r; / / rotate

[0098] After the step of "obtaining the displacement of the target in the world coordinate", the following steps are further included: setting the anchor target 102 within the visual range of the image; obtaining the correction parameters through the anchor target 102; and correcting the displacement data through the correction parameters.

[0099] In the actual process, the image acquisition device (the monocular image collector 103) is fixed on the concrete road shoulder and is easily affected by the surrounding environment and shaken due to force majeure. There are many reasons for the shaking, such as wind force, natural vibration, or vibration caused by the driving of large vehicles, etc. The shaking of the image acquisition device itself has a great impact on high-precision image monitoring and is one of the important sources of monitoring data deviation. Moreover, the deviation reflected in the amplitude change in the world coordinate system will be larger than the change value of the target itself, seriously affecting the result and causing misjudgment. The target includes the monitoring target 101 and the anchor target 102. The target surface of the anchor target 102 is the same as that of the monitoring target 101. The distance between the anchor target 102 and the image acquisition device is less than the distance between the monitoring target 101 and the image acquisition device. In this embodiment, the anchor target 102 is set on the concrete road shoulder on the left side of the road 100. The monocular image collector 103 is facing the anchor target 102 directly. When the anchor target 102 shakes due to force majeure factors, a deviation will occur and will be reflected in the pixel coordinates as. Taking [correction parameter] as the correction parameter, by correcting the displacement of the monitoring points of the monitoring target 101, assuming the pixel coordinates of the monitoring points are , the correction method is , and the accurate corrected data can be obtained. Figure 5It shows the difference in the fluctuation amplitude of the monitoring data before and after data correction by adding the anchoring target 102. The upper curve represents the displacement change of the X-axis in the world coordinate, and the lower curve represents the displacement change of the Y-axis in the world coordinate. The ordinate of both is displacement, and the abscissa is the corresponding time. The state before correction is on the left side of the dashed line, and the state after correction is on the right side of the dashed line. After correction, the amplitude of the data is significantly reduced. It can be seen that correcting through the correction parameters of the anchoring target 102 can improve the problems caused by the instability of the monocular camera.

[0100] The step of "triggering the warning of the slope 200 through the displacement of the target" is specifically as follows:

[0101] At time continuously obtain images; obtain the displacement change of the target in the world coordinate from images; the displacement change includes: axis change and axis change; the axis change of the target is , the axis change of the target is ; obtain the amplitude of the = , axis amplitude , ; obtain the range , ; set the warning threshold , when and / or , and ; obtain the local slope , , where is or , is the time. If , issue a warning.

[0102] In this embodiment, the center point of the target surface is used as the monitoring point, and the displacement of the monitoring point represents the displacement of the target. The displacement change of the monitoring point in the world coordinate is obtained by monitoring the change of the monitoring point in the pixel coordinate. Taking the change amount of displacement as the ordinate and time as the abscissa, the amplitude dot plots of the axis and axis are respectively obtained, and local curves are fitted in the amplitude dot plots to obtain the local slope . There are 6 monitoring targets 101 in this embodiment, and the 6 monitoring targets 101 form 6 monitoring points. Any monitoring point meeting the conditions can trigger a warning. The warning can specifically be a hierarchical warning, and the warning value Including: , wherein , when the amplitude , trigger -level early warning. After triggering -level early warning, when , trigger -level early warning again.

[0103] The monitoring system further includes a supplementary lighting device; the supplementary lighting device uses an array infrared light source with a wavelength of 800nm to 900nm, and the orientation of the array infrared light source is the same as that of the monocular image collector 103. An indispensable monitoring means is to have a monitoring function when the light is insufficient. Usually, an infrared lamp is installed at the camera end for supplementary lighting. However, in the actual monitoring of the slope 200, the distance between the camera and the target is basically in the range of 150m to 250m. There are many monitoring points and a wide monitoring range. There are some problems that make the monitoring effect unable to meet the all-day monitoring requirements. This is also the original intention of adding the supplementary lighting device.

[0104] The addition of the supplementary lighting device will also increase the power consumption of the overall system accordingly. Therefore, it is necessary to control the supplementary lighting according to the ambient brightness. The power supply stability of the solar system can reflect the state of the light brightness. Combining with the voltage prompt that appears before the storage battery is disconnected, the control of the supplementary lighting system can be simply realized. The monitoring target 101 includes a target board 111, a target rod 112 and a fixed base 113. The fixed base 113 is made of concrete, and a weeding cloth 114 is also provided around the fixed base 113. The fixed base 113 of the target is improved, and the problem that the target is easily covered by vegetation is solved in a low-cost way. Compared with the existing technology that calculates the monitoring in an ideal state, a more practical environmental impact correction is proposed, making the data more reliable, thus solving the interference and data deviation under natural factors. By introducing a supplementary lighting design with a specific infrared wavelength, the deficiencies in night monitoring are solved.

[0105] Obviously, the above-mentioned embodiments of the present invention are merely examples for clearly illustrating the technical solutions of the present invention, rather than limitations on the specific implementation manners of the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the claims of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A method for monitoring a slope, characterized in that: The following steps are involved: Setting up several targets on the slope; Establish a coordinate system on the target surface to obtain the world coordinates of the target; Acquire an image containing each target surface and establish a pixel coordinate system on the image; Obtain the displacement of the target in the world coordinates through the change of the target in the pixel coordinates; The triggering of the slope warning is controlled by the displacement of the target, specifically: Continuously acquire n images within time T; Obtain the displacement change of the target in the world coordinate from n images; Displacement changes include: changes in the X-axis and changes in the Y-axis; The X-axis of the target changes from X0, X1...X n , the Y axis of the target changes to Y0, Y1...Y n ; Get the magnitude of the X axis Y-axis amplitude Get the range of amplitude Assume the warning threshold D, when and / or And T≥12H; Get the local slope k, where x i for or y i For time; If k∈-(0.1~8.5), issue a warning.

2. A slope monitoring method according to claim 1, characterized in that: The step of "obtaining an image containing each target surface and establishing a pixel coordinate system on the image" is specifically as follows: Perform rigid body transformation on the world coordinates of any point on the target surface and convert them into acquisition coordinates; Perform perspective projection on the collected coordinates and convert them into image coordinates; The image coordinates are converted twice to obtain pixel coordinates.

3. A slope monitoring method according to claim 2, characterized in that: Let R be the rotation matrix, T be the translation matrix, 0 be the zero matrix, and the point in world coordinates is transformed by the matrix Convert to points on acquisition coordinates; Let f be the focal length, and the points on the acquisition coordinates are transformed by the matrix Convert to a point on the image coordinates; let dx, dy be the physical size of the pixel point, and the point on the pixel coordinates is transformed by the transformation matrix Convert to a point in pixel coordinates; The relationship between world coordinates and pixel coordinates is: Among them, the coordinates of the pixel are (x, y); the coordinate value in the world coordinate is (X, Y, Z), s is the scale factor, and (x0, y0) is the center of the pixel coordinate system.

4. A slope monitoring method according to claim 1, characterized in that: After the step "obtaining the displacement of the target in the world coordinates", the following steps are also included: Setting anchor targets within the visual range of the image; Obtaining correction parameters by anchoring the target; Correct the displacement data by using the correction parameters.

5. A monitoring system for slopes according to the monitoring method according to any one of claims 1 to 4, characterized in that: It includes a monocular image collector, multiple monitoring targets, a control host and storage; The monitoring targets are dispersedly arranged on the slopes to be monitored, and the target surfaces of the monitoring targets are provided with monitoring points; The monocular image collector is fixed outside the slope to be monitored and frames all monitoring targets within its field of view to obtain digital images; The control host establishes the pixel coordinates of each monitoring point in the digital image, obtains the displacement change of the monitoring target according to the displacement change of the monitoring point in the pixel coordinates, and realizes the monitoring and early warning of the slope.

6. A monitoring system for slopes according to claim 5, characterized in that: The target surface of the monitoring target includes: a first unit and a second unit; the first unit and the second unit use a pair of contrasting colors; the first unit and the second unit are irregularly distributed on the target surface, so that the monitoring point in the pixel coordinate changes with the change of any displacement.

7. A monitoring system for slopes according to claim 5, characterized in that: It also includes an anchoring target; the anchoring target is arranged outside the slope and between the monocular image collector and the monitoring target; The anchoring target is located within the field of view of the monocular image collector; The monocular image collector faces the anchoring target.

8. A monitoring system for slopes according to claim 5, characterized in that: It also includes a fill light device; the fill light device uses an array infrared light source with a wavelength of 800nm ​​to 900nm, and the array infrared light source is in the same direction as the monocular image collector.

9. A monitoring system for slopes according to claim 5, characterized in that: The monitoring target comprises a target plate, a target pole and a fixed foundation, wherein the fixed foundation is made of concrete and a weed control cloth is arranged around the fixed foundation.

Images