Open slope deformation monitoring method and system based on radar cooperation

Abstract

The invention relates to the technical field of radar monitoring, and particularly discloses an open-air side slope deformation monitoring method and system based on radar collaboration.The open-air side slope deformation monitoring method comprises the steps that space-based radar data and ground-based radar data of a target open-air side slope are collected through a space-based radar and a ground-based radar, and after noise removal and coordinate transformation are conducted on the ground-based radar data, the ground-based radar data are obtained; taking the space-based radar data as an example, carrying out geometric correction and resolution guide reconstruction on the space-based radar data to enable the space-based radar data to be matched with the ground-based radar data precision, and then carrying out data registration and fusion processing on the ground-based radar data and the space-based radar data after preprocessing; and time sequence analysis is carried out on the sky-ground radar fusion data to determine the deformation trend and rate of the target open-air slope, and the potential slope instability risk is prompted through intelligent early warning. In this way, the respective limitations of the ground-based radar and the space-based radar can be overcome, and efficient and accurate monitoring of the deformation of the open-air slope and intelligent early warning prompting of the slope instability risk are achieved.

Description

Technical Field

[0001] The present application relates to the field of radar monitoring technology, and more specifically, to an open-pit slope deformation monitoring method and system based on radar collaboration. Background Art

[0002] The stability of open-pit slopes is crucial to many fields such as mining and civil engineering construction. As an important part of mine safety management, slope deformation monitoring is mainly used to timely discover and deal with potential slope instability risks. Traditional slope deformation monitoring methods such as total station measurement and level measurement have problems such as limited measurement range, great influence of terrain and environment, and difficulty in achieving real-time continuous monitoring.

[0003] With the advancement of monitoring technology and the development of slope stability theory, radar-based monitoring technology has gradually become an important means of slope deformation monitoring. Radar technology has the advantages of wide monitoring range, all-weather, all-day, long-distance, and non-contact monitoring, and is especially suitable for slope deformation monitoring under complex terrain and environmental conditions. However, although ground-based radar can perform high-precision deformation monitoring of slopes within a certain range, its monitoring ability is limited for slopes in large areas, long distances, and areas blocked by complex terrain. Space-based radar has the advantage of a wide coverage range, but its spatial resolution is relatively low and is greatly affected by atmospheric factors such as clouds. In other words, single space-based or ground-based radar monitoring is difficult to meet the needs of comprehensive, accurate, and real-time deformation monitoring of open-air slopes.

[0004] Therefore, we expect a radar-coordinated open-pit slope deformation monitoring method and system that can integrate the advantages of ground-based radar and space-based radar to achieve efficient monitoring of open-pit slope deformation. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an open-air slope deformation monitoring method and system based on radar collaboration, which uses space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope, and after removing noise and converting the coordinates of the ground-based radar data, it is used as an example to perform geometric correction and resolution-guided reconstruction on the space-based radar data, so that the space-based radar data matches the accuracy of the ground-based radar data, and then the pre-processed ground-based radar data and space-based radar data are subjected to data registration and fusion processing, and the deformation trend and rate of the target open-air slope are determined by performing time series analysis on the fused data of the ground and space-based radars, and intelligent warning prompts are given for potential slope instability risks. In this way, the respective limitations of ground-based radar and space-based radar can be overcome, and efficient and accurate monitoring of open-air slope deformation and intelligent warning prompts for slope instability risks can be achieved. Provide more accurate data support with analysis.

[0006] According to one aspect of the present application, a method for monitoring deformation of an open-pit slope based on radar collaboration is provided, which comprises: Use space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope; Performing noise removal and coordinate conversion on the ground-based radar data to obtain pre-processed ground-based radar data; Performing geometric correction and resolution enhancement on the space-based radar data to obtain pre-processed space-based radar data; Performing data registration on the preprocessed ground-based radar data and the preprocessed space-based radar data to obtain registered ground-based radar data and registered space-based radar data; Based on wavelet transform, the registered ground-based radar data and the registered space-based radar data are fused to obtain space-based radar fusion data; Performing time series analysis on the ground-based radar fusion data to determine the deformation trend and rate of the target open-pit slope; Based on the comparison between the deformation trend and rate of the target open-pit slope and a preset threshold, it is determined whether to generate an early warning prompt.

[0007] According to another aspect of the present application, there is provided an open-pit slope deformation monitoring system based on radar collaboration, which comprises: A radar data acquisition module, used to collect space-based radar data and ground-based radar data of a target open-air slope by using space-based radar and ground-based radar; A ground-based radar data preprocessing module, used for performing noise removal and coordinate conversion on the ground-based radar data to obtain preprocessed ground-based radar data; A space-based radar data preprocessing module, used for performing geometric correction and resolution enhancement on the space-based radar data to obtain preprocessed space-based radar data; A radar data registration module, used for performing data registration on the preprocessed ground-based radar data and the preprocessed space-based radar data to obtain registered ground-based radar data and registered space-based radar data; A radar data fusion module, used for fusing the registered ground-based radar data and the registered space-based radar data based on wavelet transform to obtain space-based radar fusion data; A deformation analysis module, used for performing time series analysis on the ground-ground radar fusion data to determine the deformation trend and rate of the target open-pit slope; The early warning decision module is used to determine whether to generate an early warning prompt based on the comparison between the deformation trend and rate of the target open-pit slope and a preset threshold.

[0008] Compared with the prior art, the radar-coordinated open-air slope deformation monitoring method and system provided in the present application utilizes space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope. After noise removal and coordinate conversion of the ground-based radar data, the space-based radar data is used as an example to perform geometric correction and resolution-guided reconstruction on the space-based radar data so that the space-based radar data matches the accuracy of the ground-based radar data. Then, the pre-processed ground-based radar data and space-based radar data are subjected to data registration and fusion processing, and the deformation trend and rate of the target open-air slope are determined by time series analysis of the fused data of the ground-based radar and the space-based radar, and intelligent warning prompts are given for potential slope instability risks. In this way, the respective limitations of ground-based radar and space-based radar can be overcome, and efficient and accurate monitoring of open-air slope deformation and intelligent warning prompts for slope instability risks can be achieved. More accurate data support is provided for analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 It is a flow chart of an open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0011] Figure 2 Schematic diagram of data flow of an open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0012] Figure 3 It is a flowchart of sub-step S3 of the open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0013] Figure 4 It is a flowchart of sub-step S32 of the open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0014] Figure 5 It is a flowchart of sub-step S323 of the open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0015] Figure 6 It is a flowchart of sub-step S3232 of the open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application.

[0016] Figure 7It is a block diagram of an open-pit slope deformation monitoring system based on radar collaboration according to an embodiment of the present application. DETAILED DESCRIPTION

[0017] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0018] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0019] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0020] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0021] It is worth noting that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where the data is located, and with the authorization given by the owner of the corresponding device.

[0022] In response to the technical problems described in the above background technology, this application proposes a method for monitoring the deformation of an open-air slope based on radar collaboration, which uses space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope. After removing noise and converting coordinates of the ground-based radar data, it is used as an example to perform geometric correction and resolution-guided reconstruction on the space-based radar data so that the space-based radar data matches the accuracy of the ground-based radar data. Then, the pre-processed ground-based radar data and space-based radar data are subjected to data registration and fusion processing, and the deformation trend and rate of the target open-air slope are determined by performing time series analysis on the fused data of the ground and space-based radars, and intelligent warning prompts are given for potential slope instability risks. In this way, the respective limitations of ground-based radar and space-based radar can be overcome, and efficient and accurate monitoring of open-air slope deformation and intelligent warning prompts for slope instability risks can be achieved. More accurate data support is provided for analysis.

[0023] Figure 1 It is a flow chart of an open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application. Figure 2 FIG. 1 is a data flow diagram of an open-pit slope deformation monitoring method based on radar collaboration according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the open-pit slope deformation monitoring method based on radar collaboration includes the following steps: S1, using space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-pit slope; S2, removing noise and transforming coordinates of the ground-based radar data to obtain pre-processed ground-based radar data; S3, geometrically correcting and enhancing the resolution of the space-based radar data to obtain pre-processed space-based radar data; S4, performing data registration on the pre-processed ground-based radar data and the pre-processed space-based radar data to obtain registered ground-based radar data and registered space-based radar data; S5, fusing the registered ground-based radar data and the registered space-based radar data based on wavelet transform to obtain space-to-ground radar fusion data; S6, performing time series analysis on the space-to-ground radar fusion data to determine the deformation trend and rate of the target open-pit slope; S7, determining whether to generate an early warning prompt based on the comparison between the deformation trend and rate of the target open-pit slope and a preset threshold.

[0024] In the above-mentioned open-air slope deformation monitoring method based on radar collaboration, the step S1 uses the space-based radar and the ground-based radar to collect the space-based radar data and the ground-based radar data of the target open-air slope. It should be understood that the space-based radar is carried by the satellite, which has the advantage of wide coverage, and can perform macroscopic observation of a large area of ​​open-air slopes to obtain overall deformation trend information. The ground-based radar is installed on the stable ground around the slope, and can monitor the key areas of the slope at close range and high resolution to capture local subtle deformation changes. By combining the two, the deformation data of the open-air slope can be collected in an all-round and multi-level manner, making up for the limitations of single radar monitoring. In actual operation, the ground-based radar uses a high-resolution and high-precision millimeter-wave radar, which is installed on the stable ground around the slope to ensure that it can cover the key areas of the slope, and by calibrating the parameters of the ground-based radar, including the optimization of the parameters such as the transmission frequency, antenna gain, and scanning angle, it can achieve the best monitoring performance; at the same time, the satellite is used to carry the space-based radar, and the appropriate satellite orbit is selected to ensure that the space-based radar can regularly scan and observe the target slope area. Specifically, the ground-based radar scans the slope at a preset time interval, for example, every 10 minutes, collects information such as the distance and angle of the slope surface, and transmits the collected data to the ground data processing center in real time through wired or wireless communication. At the same time, the scanning angle of the radar is adjusted so that it can cover the main slope surface and potential dangerous areas of the slope. After calibration, the transmission frequency of the ground-based radar is set to 35GHz, the antenna gain is 40dB, the scanning angle range is 0°-90°, the horizontal resolution reaches 0.5 meters, and the vertical resolution is 0.3 meters. During its orbital operation, when the space-based radar passes over the target slope, it conducts a large-area scanning observation of the slope to collect the macroscopic deformation information of the slope. The data collected by the space-based radar is transmitted to the ground receiving station via a satellite communication link, and then forwarded to the ground data processing center.

[0025] In the above-mentioned open-air slope deformation monitoring method based on radar collaboration, the step S2 is to remove noise and coordinate transform the ground-based radar data to obtain the pre-processed ground-based radar data. Specifically, since the collected ground-based radar data inevitably contains noise, such as electromagnetic interference, meteorological factors, etc., it may seriously affect the accuracy of the data and the reliability of the subsequent analysis results. Therefore, it is necessary to further remove noise from the ground-based radar data. In an embodiment of the present application, a filtering algorithm such as Kalman filtering is used to remove noise signals generated by factors such as environmental interference to improve the accuracy of the data. Kalman filtering is an optimal estimation filter based on a state space model. Through two steps of prediction and updating, the optimal estimation of the current moment is obtained according to the state estimation of the previous moment and the measured value of the current moment. For example, for the distance data sequence collected by the ground-based radar, Kalman filtering can remove noise by establishing a state equation (describing the dynamic change of the distance data) and a measurement equation (linking the actually measured distance data with the state variable). After multiple iterative calculations, the noise in the data can be significantly suppressed, thereby providing a more reliable data basis for subsequent analysis. It should be understood that the data collected by the ground-based radar is based on its own radar coordinate system, and in practical applications, it needs to be integrated and analyzed with other geographic information data. Therefore, the present application also performs coordinate conversion on the ground-based radar data after noise removal to convert the data in the radar coordinate system into data in the geographic coordinate system (such as the WGS84 coordinate system). Specifically, based on the known radar installation position and attitude angle, through simple trigonometric functions and matrix operations, the data collected by the ground-based radar can be converted from the radar coordinate system to the geographic coordinate system, realizing the unification and standardization of the data, so that the ground-based radar data can be integrated with other geographic information data in a unified geographic space framework, which is convenient for subsequent comprehensive analysis of radar data.

[0026] In the above-mentioned open-pit slope deformation monitoring method based on radar collaboration, the step S3 is to perform geometric correction and resolution enhancement on the space-based radar data to obtain pre-processed space-based radar data. Figure 3 FIG. 4 is a flowchart of sub-step S3 of the method for monitoring deformation of open-pit slopes based on radar collaboration according to an embodiment of the present application. Figure 3 As shown, the step S3 includes the steps of: S31, geometrically correcting the space-based radar data to obtain geometrically corrected space-based radar data; S32, taking the pre-processed ground-based radar data as an example image, performing resolution enhancement on the geometrically corrected space-based radar data to obtain the pre-processed space-based radar data.

[0027] Specifically, the step S31 is to perform geometric correction on the space-based radar data to obtain geometrically corrected space-based radar data. It should be understood that due to factors such as satellite attitude changes and terrain undulations, the space-based radar data will have geometric deformation in the image, which makes it unable to accurately reflect the actual geographical location and shape of the slope, affecting the accuracy of the monitoring results. Therefore, the present application first performs geometric correction on the space-based radar data to compensate for the geometric deformation of the image caused by factors such as satellite attitudes and terrain undulations, so that the space-based radar image can accurately reflect the actual geographical location and shape of the slope. Specifically, geometric correction is to compensate for the geometric deformation in the image by using ground control points (points with known accurate geographical coordinates) through the establishment of a mathematical model. That is, for the pixels of the space-based radar image with geometric deformation, a polynomial transformation relationship (such as a quadratic polynomial model) is established between the pixels and the actual geographical coordinates, and the polynomial coefficients are solved according to the coordinate information of the ground control points, and then the entire image is corrected using the polynomial coefficients obtained by the solution to obtain the geometrically corrected space-based radar data, so that it can accurately reflect the actual geographical location and shape of the slope.

[0028] Specifically, in step S32, the preprocessed ground-based radar data is used as an example image, and the resolution of the geometrically corrected space-based radar data is enhanced to obtain the preprocessed space-based radar data. It should be understood that the present application takes into account that the spatial resolution of space-based radar data is relatively low and it is difficult to meet the needs of monitoring local subtle deformations of slopes. Therefore, it is necessary to further perform resolution enhancement processing on the geometrically corrected space-based radar data. Here, in order to enable space-based radar data to be better coordinated with ground-based radar data for analysis, the present application uses the preprocessed ground-based radar data as a high-resolution example image to guide the resolution enhancement of the geometrically corrected space-based radar data, so that the space-based radar data can improve the resolution of local areas while maintaining the overall deformation trend information. Among them, Figure 4 FIG. 4 is a flowchart of sub-step S32 of the method for monitoring deformation of open-pit slopes based on radar collaboration according to an embodiment of the present application. Figure 4 As shown, the step S32 includes the steps of: S321, extracting radar data reference image format features from the pre-processed ground-based radar data to obtain a radar data image reference format coding feature map; S322, extracting image features from the geometrically corrected space-based radar data to obtain a space-based radar data image feature coding feature map; S323, performing causal inference-driven feature joint perception on the space-based radar data image feature coding feature map and the radar data image reference format coding feature map to obtain an example-guided space-based radar data image joint perception coding feature map; S324, generating the pre-processed space-based radar data based on the example-guided space-based radar data image joint perception coding feature map.

[0029] More specifically, the step S321 extracts radar data reference image format features from the pre-processed ground-based radar data to obtain a radar data image reference format coding feature map. It should be understood that due to the high resolution of ground-based radars and the accurate detection of local details of slopes, in order to introduce the advantages of ground-based radar data in terms of resolution, texture, edges, etc. into space-based radar data, the present application uses a neural network model based on deep learning technology to extract image format features from the pre-processed ground-based radar data. In an embodiment of the present application, a convolutional neural network (CNN) is used as a feature extractor to process the pre-processed ground-based radar data, so as to utilize the advantages of convolutional neural networks in image feature extraction, extract image features of ground-based radar data, such as high-resolution detail information such as edges and textures, and generate a radar data image reference format coding feature map, thereby providing a high-quality feature reference standard for subsequent resolution enhancement of space-based radar data.

[0030] More specifically, step S322 extracts image features from the geometrically corrected space-based radar data to obtain a space-based radar data image feature encoding feature map. It should be understood that in order to deeply explore the slope feature information carried by the space-based radar data itself, so as to synergistically enhance it with the high-resolution features of the ground-based radar data, the present application also uses a convolutional neural network model to extract features from the geometrically corrected space-based radar data to capture macroscopic morphological information such as the overall shape of the slope and large-area terrain changes, and obtain a space-based radar data image feature encoding feature map. It is worth noting that here, since the characteristics of space-based radar data are different from those of ground-based radar data, when using a convolutional neural network model to extract features from image data, the network structure and parameters of the convolutional neural network model can be adjusted in a targeted manner to optimize the feature extraction effect.

[0031] More specifically, in step S323, the feature joint perception driven by causal inference is performed on the feature coding feature map of the space-based radar data image and the feature coding feature map of the reference format of the radar data image to obtain an example-guided space-based radar data image joint perception coding feature map. Specifically, in order to effectively utilize the high-resolution feature information provided by the feature coding feature map of the reference format of the radar data image, and to perform feature-level resolution collaborative enhancement on the macro-morphological information in the feature coding feature map of the space-based radar data image, the present application proposes a feature joint perception method driven by causal inference, which captures the local feature correlation and difference between the feature coding feature map of the space-based radar data image and the feature coding feature map of the reference format of the radar data image, and performs joint perception on local feature pairs with strong correlation, so as to integrate the high-resolution detail information of the ground-based radar data into the macro-features of the space-based radar data, while maintaining the overall characteristics of the space-based radar data itself, thereby achieving collaborative enhancement of resolution at the feature level, and finally generating an example-guided space-based radar data image joint perception coding feature map. Among them, Figure 5 FIG. 4 is a flowchart of sub-step S323 of the method for monitoring deformation of open-pit slopes based on radar collaboration according to an embodiment of the present application. Figure 5 As shown, the step S323 includes the steps of: S3231, performing feature decoupling along the channel dimension on the feature coding feature map of the space-based radar data image and the feature coding feature map of the reference format of the radar data image to obtain a set of local feature vectors of the space-based radar data image and a set of local feature vectors of the reference format of the radar data image; S3232, performing strong causal association perceptual pairing on the set of local feature vectors of the space-based radar data image and the set of local feature vectors of the reference format of the radar data image to obtain a set of {space-based radar data image local feature vector, radar data image reference format local feature vector} feature pairs constituting a causal chain unit; S3233, performing joint perceptual aggregation processing on the set of {space-based radar data image local feature vector, radar data image reference format local feature vector} feature pairs constituting a causal chain unit to obtain the example guided space-based radar data image joint perceptual coding feature map.

[0032] In a specific example of the present application, the step S3231 is expressed by the formula:

[0033] in, represents the feature decoupling operation, and respectively represent the space-based radar data image feature coding feature map and the radar data image reference format coding feature map, , , and They represent the first, second, and third local feature vectors in the set of space-based radar data images. and Local feature vectors of space-based radar data images, , , and They represent the first, second, and third local feature vectors in the radar data image reference format. and A radar data image reference format local feature vector.

[0034] That is, the feature coupling of the space-based radar data image feature encoding feature map and the radar data image reference format encoding feature map is performed along the channel dimension, and they are decomposed into a series of independent space-based radar data image local feature vectors and radar data image reference format local feature vectors. Here, through the decoupling operation, highly coupled feature elements can be separated, which helps to clarify the boundaries of the semantic expressions of each channel and reduce redundant information, thereby improving the ability to understand different feature information in the image and the accuracy and efficiency of feature migration.

[0035] Figure 6 FIG. 4 is a flowchart of sub-step S3232 of the method for monitoring deformation of open-pit slopes based on radar collaboration according to an embodiment of the present application. Figure 6 As shown, the step S3232 includes the steps of: S32321, extracting candidate space-based radar data image local feature vectors and candidate radar data image reference format local feature vectors from the set of space-based radar data image local feature vectors and the set of radar data image reference format local feature vectors respectively; S32322, performing semantic alignment processing on the candidate space-based radar data image local feature vectors and the candidate radar data image reference format local feature vectors to obtain aligned candidate space-based radar data image local feature vectors and aligned candidate radar data image reference format local feature vectors; S32323, determining whether the candidate space-based radar data image local feature vectors and the candidate radar data image reference format local feature vectors constitute a causal chain unit based on the causal correlation strength between the aligned candidate space-based radar data image local feature vectors and the aligned candidate radar data image reference format local feature vectors.

[0036] In a specific example of the present application, the step S32322 includes: constructing a semantic alignment modulation flow field between the local feature vector of the candidate space-based radar data image and the local feature vector of the candidate radar data image reference format; based on the semantic alignment modulation flow field, performing feature alignment mapping on the local feature vector of the candidate space-based radar data image and the local feature vector of the candidate radar data image reference format to obtain the aligned local feature vector of the candidate space-based radar data image and the aligned local feature vector of the candidate radar data image reference format. The above step S32322 is expressed by the formula:

[0037] in, represents the transpose of a vector, represents matrix multiplication, represents a 3×3 convolution operation, represents a 1×1 convolution operation, represents the semantic alignment modulation flow field, and They respectively represent the local feature vector of the aligned candidate space-based radar data image and the local feature vector of the aligned candidate radar data image reference format.

[0038] That is, considering that the space-based radar data image features and the ground-based radar data image features have different data sources and semantic backgrounds, direct matching will lead to semantic mismatch or offset due to inconsistent information distribution. Therefore, before screening the causal chain unit, the present application first constructs a semantic alignment modulation flow field between the local feature vector of the space-based radar data image and the local feature vector of the radar data image reference format, performs feature mapping modulation on the two, and realizes feature alignment processing, thereby obtaining the aligned candidate space-based radar data image local feature vector and the aligned candidate radar data image reference format local feature vector.

[0039] In particular, in a preferred example of the present application, the step S32323 includes: inputting the aligned candidate space-based radar data image local feature vector and the aligned candidate radar data image reference format local feature vector into a causal chain unit screening network to obtain a ground-based-space-based radar data causal chain unit closed-loop strength factor, which is expressed as:

[0040] in, represents the arccosine function, It means calculating the square of the vector norm. express and The closed-loop strength factor of the causal chain unit between ground-based and space-based radar data.

[0041] Here, the ground-based-space-based radar data causal chain unit closed-loop strength factor includes a knowledge storage part based on the feature difference norm between the aligned candidate space-based radar data image local feature vector and the aligned candidate radar data image reference format local feature vector. And their respective norms are expressed as , The environmental interaction part under the distribution evolution knowledge is to cope with the forgetting of causal evolution knowledge and the loss of autocorrelation plasticity. Furthermore, the positive contribution of distribution evolution knowledge to the strength of causal chain closure (as a molecule) and the negative environmental interaction , Based on the reverse contribution of (as the denominator), the causal strength of geometric correlation is quantified ( Function) is used to intuitively reflect the directional similarity of the causal closed loop strength, so as to obtain the closed loop strength factor of the causal chain unit of ground-based and space-based radar data for screening highly correlated and causally related feature pairs.

[0042] In a specific example of the present application, the step S32323 further includes: based on the comparison between the closed-loop strength factor of the ground-based-space-based radar data causal chain unit and a preset threshold, determining whether the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector constitute a causal chain unit, which is expressed as follows:

[0043] in, Indicates the preset threshold value, Represents the feature pair of {local feature vector of space-based radar data image, local feature vector of radar data image reference format} that constitutes the causal chain unit.

[0044] That is, the causal chain unit screening network is used to screen whether the feature pairs meet the causal coupling conditions. In this way, by comparing the closed-loop strength factor of the ground-based-space-based radar data causal chain unit between the feature pairs with the preset threshold, it is possible to capture those highly correlated and causally related feature pairs, filter out potential noise feature pairs, and retain those important features with high causality.

[0045] In a specific example of the present application, the step S3233 includes: inputting each of the feature pairs of {local feature vector of space-based radar data image, local feature vector of radar data image reference format} constituting the causal chain unit into the semantic dynamic response network of the causal chain unit to obtain a set of local joint perceptual coding feature vectors of the example-guided space-based radar data image; and performing feature aggregation on the set of local joint perceptual coding feature vectors of the example-guided space-based radar data image to obtain the joint perceptual coding feature map of the example-guided space-based radar data image. The above step S3233 is expressed by the formula:

[0046] in, and represents different weight parameters, and represents different weight matrices, Indicates point reduction, Indicates point addition, Represents the unit that constitutes the causal chain and The local joint perceptual coding feature vector of the example-guided space-based radar data image obtained after joint perceptual coding.

[0047] That is, after completing the screening of the causal chain unit, the present application further performs joint perceptual coding on each feature pair of {local feature vector of space-based radar data image, local feature vector of radar data image reference format} constituting the causal chain unit, and organically integrates the individual semantics and causal relationship between the feature pairs through high-order dynamic interaction modeling. Specifically, the semantic dynamic response network of the causal chain unit captures the complex semantic associations and potential causal relationships between feature pairs by performing multi-dimensional feature interactions on the input feature pairs and introducing the powerful fitting ability of neural networks, so as to realize the migration mapping of the local features in the feature coding feature map of the space-based radar data image to the local features in the feature coding feature map of the reference format of the radar data image, and generate a set of example-guided local joint perceptual coding feature vectors of space-based radar data images. This dynamic enhancement method based on the causal chain unit effectively improves the relationship modeling capability between the local features of the space-based radar data image and the local features of the reference format of the radar data image, so that the example-guided local joint perceptual coding feature vector of the space-based radar data image has stronger expression ability across semantic levels and causal dimensions. Finally, feature aggregation is performed on the set of strong causal joint perceptual coding feature vectors between the example-query image features to obtain a joint perceptual coding feature representation that integrates the strong causal relationship between the space-based radar data image and the radar data image, that is, the example-guided space-based radar data image joint perceptual coding feature map.

[0048] More specifically, step S324 includes: inputting the example-guided space-based radar data image joint perception coding feature map into a resolution enhancement module based on a diffusion model to obtain the pre-processed space-based radar data. It should be understood that the diffusion model is a generation model based on probability theory, which has achieved remarkable results in the field of image generation and restoration in recent years. In the present application, the example-guided space-based radar data image joint perception coding feature map combines the high-resolution detail information of ground-based radar data and the macroscopic morphological information of space-based radar data, and uses it as the input of the diffusion model. The diffusion model learns the potential feature distribution of the example-guided space-based radar data image joint perception coding feature map to gradually map it to a higher resolution feature space, restore clear image pixel information, thereby generating a high-resolution space-based radar image and realizing resolution enhancement processing of space-based radar data.

[0049] Here, the space-based radar data image feature coding feature map and the radar data image reference format coding feature map respectively represent the image semantic coding features of the preprocessed ground-based radar data and the image semantic coding features of the geometrically corrected space-based radar data. When jointly perceiving their features, considering that the preprocessed ground-based radar data and the geometrically corrected space-based radar data have significant differences in the data source domain and the preprocessing mechanism such as geometric correction is superimposed, the obtained example-guided space-based radar data image joint perception coding feature map will have a non-smooth feature distribution spatial structure manifold, thereby affecting the data quality of the preprocessed space-based radar data obtained by inputting the resolution enhancement module based on the diffusion model.

[0050] Therefore, in a preferred example, before the example guided space-based radar data image joint perceptual coding feature map is input into a resolution enhancement module based on a diffusion model to obtain the pre-processed space-based radar data, the example guided space-based radar data image joint perceptual coding feature map is optimized, and the optimization process includes: Based on the sum of the absolute values ​​of the feature values ​​at each position in the example-guided space-based radar data image joint perception coding feature map, an example-guided space-based radar data image joint perception coding association density benchmark value is constructed, which is expressed as:

[0051] in, represents the feature values ​​of each position in the joint perceptual coding feature map of the space-based radar data image guided by the example, represents the total number of features, Indicates the example guided space-based radar data image joint perception coding correlation density benchmark value; Based on the sum of the squares of the eigenvalues ​​at each position in the example guided space-based radar data image joint perception coding feature map, an example guided space-based radar data image joint perception coding energy distribution reference value is constructed, which is expressed as:

[0052] in, Indicates the benchmark value of energy distribution of joint perception coding of example-guided space-based radar data image; Based on the example guided space-based radar data image joint perception coding association density benchmark value, the example guided space-based radar data image joint perception coding feature map is visualized based on the association mapping benchmark to obtain the first example guided space-based radar data image joint perception coding global coupling coefficient, which is expressed as:

[0053]

[0054] in, represents the characteristic scale of the example guided space-based radar data image joint perceptual coding feature map, , , Respectively represent the width, height and number of channels of the example guided space-based radar data image joint perception coding feature map, represents the natural exponential function, represents the first example guided space-based radar data image joint perception coding global coupling coefficient; Based on the energy distribution reference value of the example guided space-based radar data image joint perceptual coding, feature coupling explicit modeling is performed on the example guided space-based radar data image joint perceptual coding feature map to obtain a second example guided space-based radar data image joint perceptual coding global coupling coefficient, which is expressed as:

[0055] in, represents the global coupling coefficient of the second example guided space-based radar data image joint perception coding; Based on the global coupling coefficient of the first example guided space-based radar data image joint perceptual coding and the global coupling coefficient of the second example guided space-based radar data image joint perceptual coding, generalized topological adaptive enhancement is performed on the example guided space-based radar data image joint perceptual coding feature map to obtain an optimized eigenvalue corresponding to each eigenvalue in the example guided space-based radar data image joint perceptual coding feature map, expressed as:

[0056] in, represents the first weight hyperparameter, represents the second weight hyperparameter, represents the third weight hyperparameter, which can be set according to experience. , , , can be adjusted according to actual conditions, and is not specifically limited in this preferred embodiment. Indicates the optimized eigenvalue corresponding to each eigenvalue in the example-guided space-based radar data image joint perception coding feature map; The optimized feature values ​​are combined into an optimized example-guided space-based radar data image joint perception coding feature map.

[0057] Accordingly, in this preferred embodiment, a global feature coupling mechanism is constructed by mapping the total feature response magnitude of the example-guided space-based radar data image joint perceptual coding feature map relative to the spatial correlation of the example-guided space-based radar data image joint perceptual coding feature map, so as to strengthen the neighborhood constraint strength of the example-guided space-based radar data image joint perceptual coding feature map and the geometric polymorphism of the feature semantics deconstructed by the heterogeneous feature anchor prediction of the example-guided space-based radar data image joint perceptual coding feature map, thereby enhancing the generalized topological adaptability of the feature set of the example-guided space-based radar data image joint perceptual coding feature map. In this way, the data quality of the pre-processed space-based radar data obtained by the resolution enhancement module based on the diffusion model of the example-guided space-based radar data image joint perceptual coding feature map is improved.

[0058] In the above-mentioned open-pit slope deformation monitoring method based on radar collaboration, the step S4 performs data registration on the pre-processed ground-based radar data and the pre-processed space-based radar data to obtain registered ground-based radar data and registered space-based radar data. Specifically, since the pre-processed ground-based radar and space-based radar data are also affected by factors such as the angle of the acquisition equipment, there may be deviations in the spatial positions between the two. Therefore, in order to realize the fusion processing of ground-based radar and space-based radar data, the present application first registers the pre-processed ground-based radar data and space-based radar data based on the geographic information system (GIS) data of the slope, and uses feature point matching, edge extraction and other technologies to accurately align the two types of data in space to ensure data consistency. Specifically, in an embodiment of the present application, the Canny operator is used to extract edge features of the slope in ground-based radar and space-based radar data, respectively. Then, through optimization algorithms such as the least squares method, the best matching relationship between the edge features in the two types of data is found, and the registration error is controlled within 1 meter, so that the two types of data are accurately aligned in space, and the registered ground-based radar data and the registered space-based radar data are obtained.

[0059] Specifically, for ground-based radar data, due to its high resolution, it can provide more detailed terrain information. Therefore, when applying the Canny operator, a relatively high threshold can be set to suppress the impact of noise while maintaining sufficient edge details. For space-based radar data, considering its lower spatial resolution and possible atmospheric interference factors, more attention should be paid to smoothing in the edge detection process to reduce misjudgment. Appropriately lowering the threshold will help capture more effective edge information. In addition, it is necessary to flexibly adjust the various parameters of the Canny operator, such as the standard deviation of the Gaussian filter, the size of the gradient calculation window, etc., in combination with factors such as lighting conditions and terrain complexity in specific application scenarios to ensure that the extracted edge features are both clear and accurate.

[0060] After edge feature extraction is completed, the stage of finding the best matching relationship begins. This process not only relies on the least squares method, but can also be combined with other advanced optimization algorithms, such as genetic algorithms or particle swarm optimization algorithms, to improve matching efficiency and accuracy. These algorithms can quickly find the global optimal solution in a huge solution space by simulating natural selection or group intelligence behavior. For example, when using a genetic algorithm, a set of initial solutions (i.e., possible combinations of spatial transformation parameters) is first randomly generated, and then this set of solutions is continuously evolved through selection, crossover, and mutation operations until the predetermined stop condition is met. The transformation parameters corresponding to the individual with the highest fitness in each generation of the population are regarded as the current optimal solution, and as the number of iterations increases, the true optimal matching relationship is gradually approached. This method is particularly suitable for solving complex optimization problems with multiple variables and nonlinearity, and can effectively overcome the problem that the traditional least squares method is prone to falling into the local optimal solution in some cases.

[0061] In addition to directly comparing the positional differences of corresponding points between the two data sets, evaluation indicators based on geometric similarity, such as Hausdorff distance or Frechet distance, can also be introduced. These indicators can measure the similarity between the two data sets as a whole and help identify those subtle mismatches that are difficult to detect by position deviation alone. At the same time, combined with specific geographical background knowledge, such as known distribution of land objects, geological structure trends and other information, it can assist in judging the rationality of the registration results and avoid unreasonable registration caused by relying solely on mathematical models.

[0062] In order to ensure that the registration error is controlled within 1 meter, quality control measures must be strictly implemented throughout the data registration process. On the one hand, strengthen the work in the data preprocessing stage, especially the noise removal and coordinate conversion steps, to ensure the quality of the input data; on the other hand, monitor the error change trend in real time during the registration process. Once the error is found to be beyond the set range, immediately suspend the current operation, review and adjust the relevant parameter settings. In addition, regularly review the data that has been registered, and further confirm the accuracy of the registration results by comparing and verifying with field measurement results or other independent data sources. If significant deviations are found, it is necessary to trace back to the original data collection stage, investigate the causes that may lead to increased errors, and take corresponding remedial measures.

[0063] In the above-mentioned open-pit slope deformation monitoring method based on radar collaboration, the step S5 is based on wavelet transform to fuse the registered ground-based radar data and the registered space-based radar data to obtain ground-based and space-based radar fusion data. That is, in order to combine the high-precision local deformation information of the ground-based radar and the large-area macro deformation information of the space-based radar, the present application adopts a fusion algorithm based on wavelet transform to fuse the registered ground-based radar data and the registered space-based radar data to obtain ground-based and space-based radar fusion data. Specifically, for the high-precision local deformation information of the ground-based radar and the large-area macro deformation information of the space-based radar, the wavelet transform can decompose the data into sub-bands of different scales and directions, so that the two are fused in the wavelet domain according to certain fusion rules (such as the regional energy maximum rule). For example, in the high-frequency sub-band after wavelet decomposition, the coefficients with larger energy in the ground-based radar data are selected, and in the low-frequency sub-band, the two data are fused according to their weights (which can be determined based on factors such as data reliability and resolution) to obtain the fused wavelet coefficients. The fused slope deformation information, i.e., the ground-based radar fusion data, is then obtained through inverse wavelet transform. This helps to give full play to the advantages of the two radars and improve the accuracy and reliability of deformation monitoring.

[0064] Specifically, since different types of radar data have their own spectral characteristics, it is necessary to select the most suitable wavelet basis function based on these characteristics. The selection criteria include factors such as orthogonality, compact support and symmetry of the wavelet basis function, ensuring that the detailed information in the data can be effectively captured while reducing unnecessary redundancy.

[0065] After selecting the wavelet basis function, the next step is to perform multi-scale wavelet decomposition on the two types of data. In this process, the original data is decomposed into approximate coefficients and detail coefficients at multiple scales. The approximate coefficient represents the general outline or low-frequency part of the data, while the detail coefficient contains more local change information, that is, the high-frequency part. For ground-based radar data, due to its high spatial resolution, it usually contains rich terrain detail information; while space-based radar data, although with low spatial resolution, has advantages in large-scale coverage. Through multi-scale decomposition, the characteristic performance of the two types of data at different scales can be clearly distinguished, laying the foundation for further data fusion.

[0066] In the fusion stage, the weighted average method or the method based on regional energy is used to determine the best combination of approximation coefficients and detail coefficients at each scale. The weighted average method is relatively simple and direct. It assigns appropriate weights to the approximation coefficients and detail coefficients of ground-based radar data and space-based radar data at the same scale, and then calculates the weighted average as the fusion result. This method is suitable for situations where the two types of data are relatively balanced in quality. However, in practical applications, more factors often need to be considered, such as data in certain specific areas are more reliable or contain more important information. At this time, the method based on regional energy is more applicable. This method automatically adjusts the weight distribution by calculating the energy distribution of data in different areas at each scale, so that the final fusion result can better reflect the actual situation. For example, in areas with edges or drastic changes, the detail coefficients of ground-based radar data may be more preferred because such data can provide higher resolution; while in relatively flat or large areas of consistency, the approximation coefficients of space-based radar data can be relied on more to obtain a larger coverage.

[0067] After determining the fusion strategy, the next step is to reconstruct the fused approximate coefficients and detail coefficients to restore the complete ground-based radar fusion data. This step requires the precise execution of the inverse wavelet transform process, that is, to recombine the decomposed coefficients into the original data form. In order to ensure the quality of the reconstruction, the mathematical definition of the selected wavelet basis function must be strictly followed, and the reconstruction operations at each level must be performed in the correct order. During the entire reconstruction process, any slight error may lead to deviations in the final result, so special attention should be paid to the stability and accuracy of the numerical calculation.

[0068] In the above-mentioned open-air slope deformation monitoring method based on radar collaboration, the step S6 performs time series analysis on the ground-based radar fusion data to determine the deformation trend and rate of the target open-air slope. That is, in order to reveal the changing pattern of slope deformation over time, the present application performs time series analysis on the ground-based radar fusion data and uses a trend analysis method to determine the deformation trend and rate of the target open-air slope. Specifically, commonly used trend analysis methods include moving average method and least squares fitting. In an embodiment of the present application, first, the area of ​​interest or pixel point in the ground-based radar fusion data is determined, and the deformation variables of the corresponding positions in the ground-based radar fusion data at adjacent time points are calculated, and the deformation variables are divided by the time interval to obtain a time series of deformation rate. Then, the time series of deformation rate is smoothed using the moving average method to remove noise and outliers to reveal the long-term trend of slope deformation over time. Next, the least squares method is used to fit the deformation rate trend line to describe the changing trend of slope deformation over time in the form of a straight line or curve. The deformation prediction model based on the ARIMA model is used to predict the deformation trend of the slope in the next week, thereby providing an important basis for slope stability and safety assessment.

[0069] In the above-mentioned open-air slope deformation monitoring method based on radar collaboration, the step S7 determines whether to generate an early warning prompt based on the comparison between the deformation trend and rate of the target open-air slope and the preset threshold. Specifically, in order to timely discover the risk of slope instability and ensure the safety of personnel and facilities, when the deformation trend and rate exceed the preset safety threshold, it indicates that the slope may be unstable, and the system will automatically trigger an early warning prompt. The early warning prompt can be conveyed to relevant personnel in a variety of ways, such as e-mail, SMS notification, system interface pop-up window, etc., so as to take timely protective measures such as evacuation, suspension of operations, reinforcement, etc. Here, the early warning prompt not only contains the specific data of the slope deformation, but also attaches a visual chart of the deformation trend to help personnel quickly understand the current status of the slope deformation and possible future development trends, so as to make timely and effective countermeasures.

[0070] In summary, the radar-coordinated open-air slope deformation monitoring method based on the embodiment of the present application is explained, which uses space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope, and after removing noise and converting the coordinates of the ground-based radar data, it is used as an example to perform geometric correction and resolution-guided reconstruction on the space-based radar data, so that the space-based radar data matches the accuracy of the ground-based radar data, and then the pre-processed ground-based radar data and space-based radar data are subjected to data registration and fusion processing, and the deformation trend and rate of the target open-air slope are determined by time series analysis of the fused data of the ground and the sky, and intelligent early warning prompts are given for potential slope instability risks. In this way, the respective limitations of ground-based radar and space-based radar can be overcome, and efficient and accurate monitoring of open-air slope deformation and intelligent early warning prompts for slope instability risks can be achieved. Provide more accurate data support with analysis.

[0071] Furthermore, a radar-coordinated open-pit slope deformation monitoring system is also provided.

[0072] Figure 7 FIG. 1 is a block diagram of an open-pit slope deformation monitoring system based on radar collaboration according to an embodiment of the present application. Figure 7 As shown, according to the embodiment of the present application, the open-air slope deformation monitoring system 100 based on radar collaboration includes: a radar data acquisition module 110, which is used to collect space-based radar data and ground-based radar data of the target open-air slope by using space-based radar and ground-based radar; a ground-based radar data preprocessing module 120, which is used to remove noise and transform coordinates of the ground-based radar data to obtain preprocessed ground-based radar data; a space-based radar data preprocessing module 130, which is used to perform geometric correction and resolution enhancement on the space-based radar data to obtain preprocessed space-based radar data; a radar data registration module 140, which is used to perform geometric correction and resolution enhancement on the space-based radar data to obtain preprocessed space-based radar data. The ground-based radar data and the pre-processed sky-based radar data are registered to obtain the registered ground-based radar data and the registered sky-based radar data; a radar data fusion module 150 is used to fuse the registered ground-based radar data and the registered sky-based radar data based on wavelet transform to obtain the ground-based and sky-based radar fusion data; a deformation analysis module 160 is used to perform time series analysis on the sky-based and sky-based radar fusion data to determine the deformation trend and rate of the target open-air slope; an early warning decision module 170 is used to determine whether to generate an early warning prompt based on the comparison between the deformation trend and rate of the target open-air slope and a preset threshold.

[0073] Here, those skilled in the art can understand that the specific operations of each module in the above-mentioned open-pit slope deformation monitoring system based on radar collaboration have been referred to above. Figures 1 to 6 The description of the radar-coordinated open-pit slope deformation monitoring method has been introduced in detail, and therefore, its repeated description will be omitted.

[0074] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0075] In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0076] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0077] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0078] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for monitoring deformation of open-pit slopes based on radar collaboration, characterized in that: include: Use space-based radar and ground-based radar to collect space-based radar data and ground-based radar data of the target open-air slope; Performing noise removal and coordinate conversion on the ground-based radar data to obtain pre-processed ground-based radar data; Performing geometric correction and resolution enhancement on the space-based radar data to obtain pre-processed space-based radar data; Performing data registration on the preprocessed ground-based radar data and the preprocessed space-based radar data to obtain registered ground-based radar data and registered space-based radar data; Based on wavelet transform, the registered ground-based radar data and the registered space-based radar data are fused to obtain space-based radar fusion data; Performing time series analysis on the ground-ground radar fusion data to determine the deformation trend and rate of the target open-pit slope; Based on the comparison between the deformation trend and rate of the target open-pit slope and a preset threshold, it is determined whether to generate an early warning prompt.

2. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 1 is characterized in that: Performing geometric correction and resolution enhancement on the space-based radar data to obtain pre-processed space-based radar data, including: Performing geometric correction on the space-based radar data to obtain geometrically corrected space-based radar data; The preprocessed ground-based radar data is used as an example image, and resolution enhancement is performed on the geometrically corrected space-based radar data to obtain the preprocessed space-based radar data.

3. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 2 is characterized in that: Taking the pre-processed ground-based radar data as an example image, performing resolution enhancement on the geometrically corrected space-based radar data to obtain the pre-processed space-based radar data, including: Extracting radar data reference image format features from the preprocessed ground-based radar data to obtain a radar data image reference format coding feature map; Extracting image features from the geometrically corrected space-based radar data to obtain a space-based radar data image feature encoding feature map; Performing causal inference-driven feature joint perception on the space-based radar data image feature encoding feature map and the radar data image reference format encoding feature map to obtain an example-guided space-based radar data image joint perception encoding feature map; Based on the example-guided space-based radar data image joint perception coding feature map, the pre-processed space-based radar data is generated.

4. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 3 is characterized in that: Performing causal inference-driven feature joint perception on the space-based radar data image feature encoding feature map and the radar data image reference format encoding feature map to obtain an example-guided space-based radar data image joint perception encoding feature map, including: Performing feature decoupling along a channel dimension on the feature encoding feature map of the space-based radar data image and the feature encoding feature map of the radar data image reference format to obtain a set of local feature vectors of the space-based radar data image and a set of local feature vectors of the radar data image reference format; Performing strong causal association perception pairing on the set of local feature vectors of the space-based radar data image and the set of local feature vectors of the radar data image reference format to obtain a set of feature pairs of {local feature vectors of space-based radar data image, local feature vectors of radar data image reference format} constituting a causal chain unit; The set of feature pairs of {space-based radar data image local feature vector, radar data image reference format local feature vector} constituting the causal chain unit is subjected to joint perception aggregation processing to obtain the example guided space-based radar data image joint perception coding feature map.

5. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 4 is characterized in that: The set of local feature vectors of the space-based radar data image and the set of local feature vectors of the radar data image reference format are paired with each other in a strong causal association perception manner to obtain a set of feature pairs of {local feature vectors of space-based radar data image, local feature vectors of radar data image reference format} constituting a causal chain unit, including: Extracting candidate space-based radar data image local feature vectors and candidate radar data image reference format local feature vectors from the set of space-based radar data image local feature vectors and the set of radar data image reference format local feature vectors respectively; Performing semantic alignment processing on the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector to obtain an aligned candidate space-based radar data image local feature vector and an aligned candidate radar data image reference format local feature vector; Based on the causal correlation strength between the aligned alternative space-based radar data image local feature vector and the aligned alternative radar data image reference format local feature vector, it is determined whether the alternative space-based radar data image local feature vector and the alternative radar data image reference format local feature vector constitute a causal chain unit.

6. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 5 is characterized in that: The method of performing semantic alignment processing on the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector to obtain an aligned candidate space-based radar data image local feature vector and an aligned candidate radar data image reference format local feature vector comprises: Constructing a semantically aligned modulation flow field between the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector; Based on the semantically aligned modulated flow field, feature alignment mapping is performed on the local feature vector of the alternative space-based radar data image and the local feature vector of the alternative radar data image reference format to obtain the aligned local feature vector of the alternative space-based radar data image and the aligned local feature vector of the alternative radar data image reference format.

7. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 6 is characterized in that: Determining whether the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector constitute a causal chain unit based on the causal correlation strength between the aligned candidate space-based radar data image local feature vector and the aligned candidate radar data image reference format local feature vector, including: Inputting the aligned candidate space-based radar data image local feature vector and the aligned candidate radar data image reference format local feature vector into a causal chain unit screening network to obtain a ground-based-space-based radar data causal chain unit closed-loop strength factor; Based on the comparison between the closed-loop strength factor of the ground-based-space-based radar data causal chain unit and a preset threshold, it is determined whether the candidate space-based radar data image local feature vector and the candidate radar data image reference format local feature vector constitute a causal chain unit.

8. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 7 is characterized in that: The set of feature pairs of {space-based radar data image local feature vector, radar data image reference format local feature vector} constituting the causal chain unit is subjected to joint perception aggregation processing to obtain the example guided space-based radar data image joint perception coding feature map, including: Inputting each of the feature pairs of {local feature vector of space-based radar data image, local feature vector of radar data image reference format} constituting the causal chain unit into the semantic dynamic response network of the causal chain unit respectively to obtain a set of example-guided local joint perception coding feature vectors of space-based radar data image; Feature aggregation is performed on a set of local joint perceptual coding feature vectors of the example guided space-based radar data image to obtain the example guided space-based radar data image joint perceptual coding feature map.

9. The method for monitoring deformation of open-pit slopes based on radar collaboration according to claim 8 is characterized in that: Based on the example guided space-based radar data image joint perceptual coding feature map, generating the pre-processed space-based radar data, including: The example-guided space-based radar data image joint perceptual coding feature map is input into a resolution enhancement module based on a diffusion model to obtain the pre-processed space-based radar data.

10. An open-pit slope deformation monitoring system based on radar collaboration, characterized in that: include: A radar data acquisition module, used to collect space-based radar data and ground-based radar data of a target open-air slope by using space-based radar and ground-based radar; A ground-based radar data preprocessing module, used for performing noise removal and coordinate conversion on the ground-based radar data to obtain preprocessed ground-based radar data; A space-based radar data preprocessing module, used for performing geometric correction and resolution enhancement on the space-based radar data to obtain preprocessed space-based radar data; A radar data registration module, used for performing data registration on the preprocessed ground-based radar data and the preprocessed space-based radar data to obtain registered ground-based radar data and registered space-based radar data; A radar data fusion module, used for fusing the registered ground-based radar data and the registered space-based radar data based on wavelet transform to obtain space-based radar fusion data; A deformation analysis module, used for performing time series analysis on the ground-ground radar fusion data to determine the deformation trend and rate of the target open-pit slope; The early warning decision module is used to determine whether to generate an early warning prompt based on the comparison between the deformation trend and rate of the target open-pit slope and a preset threshold.

Images