Geological disaster hidden danger recognition system and method integrating LiDAR and oblique photography

Abstract

The invention relates to the technical field of geological disaster recognition, and discloses a geological disaster hidden danger recognition system and method fusing LiDAR and oblique photography, and the system comprises a data collection module, a data processing module, an interpretation analysis module and a hidden danger recognition module which are in communication connection in sequence. The method corresponds to the system. According to the application, comprehensive acquisition of multi-source data is realized, LiDAR data and inclined image data are processed to generate various models, a basis is provided for subsequent analysis, the interpretation analysis module is combined with data and optical images obtained by penetrating vegetation through LiDAR, a landslide form is delineated, and disaster-inducing factors are extracted for interpretation, so that geological disaster features are deeply excavated, and the geological disaster detection efficiency is improved. According to the method, the interpretation mark is constructed, multi-source information is fused to identify potential hidden dangers, accurate identification of geological disaster hidden dangers is realized, the limitation of a traditional single monitoring means is overcome, the accuracy and reliability of identification are improved, powerful support is provided for geological disaster prevention and control, and the method has more advantages in pertinence and accuracy of identification of multiple geological disaster hidden dangers.

Description

Technical Field

[0001] The present application relates to the technical field of geological disaster identification, and specifically to a geological disaster hazard identification system and method that integrates LiDAR and oblique photography. Background Art

[0002] With the intensification of global climate change and human engineering activities, geological disasters occur frequently, and the monitoring and prevention of geological disasters have become increasingly important. Existing geological disaster monitoring technologies, such as the combination of manual on-site surveys and simple measuring instruments, have problems such as low efficiency, high risk, and inability to monitor in real time, and can no longer meet current needs. Furthermore, emerging single monitoring technologies, such as optical remote sensing and airborne LiDAR, have their own advantages, but they have limitations in characterizing the evolution of geological disasters, and the data reliability is also low. At the same time, the existing geological disaster monitoring system is also imperfect, lacking accurate predictions, digital rehearsals, and scientific plan generation mechanisms.

[0003] Chinese patent number CN118212747A discloses a multi-source information fusion intelligent monitoring and early warning system and method for slope disasters. Although it adopts multi-source information fusion technology to build a relatively complete monitoring and early warning system, it still has deficiencies in the interpretation of geological disaster hazards. The deficiencies are as follows: it mainly focuses on slope disasters, and does not construct detailed interpretation methods and signs for a variety of geological disasters, making it difficult to accurately identify different types of geological disaster hazards; for example, in the interpretation of disasters such as landslides and collapses, there is a lack of necessary technical solutions, and therefore, it cannot meet the needs of disaster hazard identification in complex geological environments.

[0004] In summary, a new technical solution for geological disaster risk identification that integrates LiDAR and oblique photography is urgently needed to solve the above technical problems. Summary of the invention

[0005] The purpose of this application is to provide a geological disaster hazard identification system and method that integrates LiDAR and oblique photography to solve the technical problems raised in the above background technology.

[0006] To achieve the above objectives, this application discloses the following technical solutions:

[0007] In the first aspect, the present application discloses a geological disaster hazard identification system integrating LiDAR and oblique photography, the system comprising a data acquisition module, a data processing module, an interpretation and analysis module and a hazard identification module which are sequentially connected in communication;

[0008] The data acquisition module is configured as follows: using a drone equipped with LiDAR and an oblique camera to collect laser point cloud data, optical image data and oblique image data of the geological disaster point area;

[0009] The data processing module is configured to: perform strip stitching, denoising, and filtering on the original LiDAR laser point cloud data, classify the point cloud in a human-computer interaction manner, generate a digital elevation model based on the processed laser point cloud data, and create a three-dimensional geological sand table by overlaying high-definition images and mountain shadows; process the oblique image data, and the processing at least includes equalizing illumination and color, image control point piercing, and aerial triangulation to produce a digital orthophoto map and a real scene three-dimensional model.

[0010] The interpretation and analysis module is configured to: based on the data obtained by LiDAR penetrating vegetation, remove the vegetation to generate a digital surface model, combine with optical remote sensing images to delineate the landslide morphology, and the landslide morphology at least includes the boundary, area, and volume of the landslide morphology; extract the disaster-causing factors in the LiDAR data, and the disaster-causing factors at least include slope, aspect, and ground roughness, and overlay them into a preset three-dimensional interpretation platform for shallow surface geological disaster interpretation; based on the terrain data generated by LiDAR, quantify the geological disasters, and the quantification at least includes extracting information on cracks and landslide bodies.

[0011] The hidden danger identification module is configured to: construct interpretation marks for geological disasters, and the geological disasters at least include landslides and collapses, obtain auxiliary identification information based on the digital elevation model and the digital surface model, and the auxiliary identification information at least includes mountain shadows, slope, aspect, and terrain openness factors; combine with preset disaster-forming background conditions, fuse the optical image data, the auxiliary identification information, and historical deformation characteristics, and analogously analyze the topographic and geomorphic characteristics of the areas where disasters have occurred under the same geological conditions around the geological disaster point area to identify potential geological disaster hidden dangers.

[0012] Preferably, when the data acquisition module acquires data, it realizes the preliminary fusion of LiDAR and oblique photography based on time synchronization and spatial registration, and the preliminary fusion includes: during the flight of the UAV, LiDAR and the oblique camera use the same time reference for data acquisition in terms of time, and in terms of space, a preset positioning and orientation system is used to determine the positions and postures of LiDAR and the oblique camera, and the acquired data is unified into the same geographic coordinate system.

[0013] Preferably, when the data processing module makes a digital orthophoto map, it performs ortho-rectification on the oblique image data by using the digital elevation model based on the oblique image rectification formula; wherein, the oblique image rectification formula is:

[0014]

[0015] Among them, (x, y) are the coordinates of the image point, (x0, y0) are the coordinates of the principal point of the image, f is the focal length of the camera, (X, Y, Z) are the ground coordinates of the object point, and (X0, Y0, Z0) are the ground coordinates of the projection center.

[0016] Preferably, when delineating the landslide morphology, the interpretation and analysis module extracts the topographic features of the landslide based on the digital surface model, and the topographic features at least include slope features and aspect features. The texture features and color features of the landslide are extracted based on the optical influence data, and the topographic features, the texture features, and the color features are fused using the comprehensive feature calculation formula to obtain the comprehensive feature; among them, the comprehensive feature calculation formula is:

[0017] F = ω s *F s + ω a *F a + ω t *F t + ω c *F c

[0018] Among them, F s is the slope feature, F a is the aspect feature, F t is the texture feature, F c is the color feature, ω s 、ω a 、ω t and ω c are the weights of the corresponding features, and F is the calculated comprehensive feature; based on the preset comprehensive feature threshold table, the landslide boundary is delineated in combination with the calculated comprehensive feature, and the comprehensive feature threshold table stores the comprehensive feature corresponding to the landslide boundary.

[0019] Preferably, the interpretation and analysis module extracts the disaster-causing factors based on the LiDAR data and conducts shallow surface geological disaster interpretation based on the disaster-causing factors. The shallow surface geological disaster interpretation includes:

[0020] The orthophoto obtained by oblique photography is used as texture information and superimposed on the 3D interpretation platform constructed based on the LiDAR data. The texture and tone changes of different ground objects in the orthophoto are analyzed using the disaster-causing factors, and the possibility of shallow surface geological disasters is calculated in combination with the slope value calculated using the LiDAR data and the shallow layer disaster possibility calculation formula. The shallow layer disaster possibility calculation formula is:

[0021] P = k1 * S + k2 * T + k3 * C

[0022] Among them, S is the slope value, T is the texture change value of the ground object in the orthophoto, C is the tone change value of the ground object in the orthophoto, k1, k2, and k3 are the weights of the corresponding numerical values, P is the calculated possibility of shallow surface geological disasters occurring. When P≥P τ it is determined that there are potential hazards of shallow surface geological disasters in this area, and P τ is the preset possibility threshold of shallow surface geological disasters occurring.

[0023] Preferably, when the interpretation and analysis module extracts crack information, it performs edge detection based on LiDAR data to obtain a preliminary candidate area for cracks, and verifies the preliminary candidate area in combination with multi-angle images obtained by oblique photography and the comprehensive gradient amplitude calculation formula. The comprehensive gradient amplitude calculation formula is:

[0024]

[0025] Among them, G1 is the gradient amplitude obtained by performing edge detection based on LiDAR data, G2 is the gradient amplitude obtained based on multi-angle images, is the gradient amplitude compensation value obtained by regression analysis. This gradient amplitude compensation value is used to represent the compensation for extracting crack information based on historical crack information. G is the calculated comprehensive gradient amplitude. When G≥G τ this area is determined as a crack area, and G τ is the preset comprehensive gradient amplitude threshold.

[0026] Preferably, when the potential hazard identification module constructs an interpretation mark for landslides, it combines LiDAR data and oblique photography data, and uses the landslide possibility index calculation formula to determine the interpretation mark for landslides. The landslide possibility index calculation formula is:

[0027]

[0028] Among them, I1 is the topographic feature index of the landslide obtained based on LiDAR data and oblique photography data. This topographic feature index includes at least slope and elevation difference. I2 is the image feature index of the landslide obtained based on LiDAR data and oblique photography data. This image feature index includes at least tone and texture. is the landslide possibility index compensation value obtained by regression analysis. This landslide possibility index compensation value is used to represent the compensation for the landslide interpretation mark based on the historical landslide interpretation mark. α and β are the corresponding weights. L is the calculated landslide interpretation mark. Based on the preset landslide threshold range, the calculated landslide interpretation mark is mapped to different levels.

[0029] Preferably, when obtaining the terrain openness factor based on the digital elevation model and the digital surface model, the potential hazard identification module calculates the terrain openness factor by using the panoramic image information obtained from oblique photography and the terrain openness factor calculation formula. The terrain openness factor calculation formula is as follows:

[0030] O = γ * R h + δ * θ

[0031] Wherein, R h is the elevation change rate within the surrounding range corresponding to a point in the digital elevation model, θ is the visible range of this point in the panoramic image information, γ and δ are the corresponding weights, and O is the calculated terrain openness factor.

[0032] Preferably, the disaster-forming background conditions at least include geological structures, rock and soil types, topography and geomorphology, and the geological background corresponding to meteorology and hydrology.

[0033] In a second aspect, the present application discloses a geological disaster potential hazard identification method integrating LiDAR and oblique photography. This method is applicable to the geological disaster potential hazard identification system integrating LiDAR and oblique photography as described above. The method includes:

[0034] Using an unmanned aerial vehicle equipped with LiDAR and an oblique camera to collect laser point cloud data, optical image data, and oblique image data of the geological disaster point area;

[0035] Performing strip splicing, denoising, and filtering on the original laser point cloud data of LiDAR, classifying the point cloud by means of human-computer interaction, generating a digital elevation model based on the processed laser point cloud data, and creating a three-dimensional geological sand table by overlaying high-definition images and mountain shadows; processing the oblique image data, and this processing at least includes equalizing light and color, image control point piercing, and aerial triangulation, and making a digital orthophoto image and a real scene three-dimensional model;

[0036] Based on the data obtained by LiDAR penetrating vegetation, removing the vegetation to generate a digital surface model, combining with optical remote sensing images to delineate the landslide morphology, and the landslide morphology at least includes the boundary, area, and volume of the landslide morphology; extracting the disaster-causing factors in the data of LiDAR, and the disaster-causing factors at least include slope, aspect, and ground roughness, and overlaying them into a preset three-dimensional interpretation platform for shallow surface geological disaster interpretation; quantifying the geological disasters based on the terrain data generated by LiDAR, and this quantification at least includes extracting information on cracks and landslide bodies;

[0037] Construct an interpretation mark for geological disasters, where the geological disasters include at least landslides and collapses. Based on the digital elevation model and the digital surface model, obtain auxiliary identification information, and the auxiliary identification information includes at least mountain shadow, slope, aspect, and terrain openness factor; combine the preset disaster-forming background conditions, fuse the optical image data, the auxiliary identification information, and historical deformation characteristics, and analogously analyze the topographic and geomorphic characteristics of the areas where disasters have occurred under the same geological conditions around the geological disaster point area to identify potential geological disaster hazards.

[0038] Beneficial effects: The geological disaster hazard identification system and method that fuse LiDAR and oblique photography of the present application use the data acquisition module to carry a drone to obtain lidar point cloud, optical image, and oblique image data, realizing the comprehensive acquisition of multi-source data. The data processing module processes the LiDAR data and oblique image data to generate various models, providing a basis for subsequent analysis. The interpretation and analysis module combines the data obtained by LiDAR penetrating vegetation and optical images, delineates the landslide morphology, extracts disaster-causing factors for interpretation, and thus deeply explores the characteristics of geological disasters. The hazard identification module constructs an interpretation mark, fuses multi-source information to identify potential hazards, realizes the accurate identification of geological disaster hazards, overcomes the limitations of traditional single monitoring means, improves the accuracy and reliability of identification, provides strong support for geological disaster prevention and control, and has more advantages in the pertinence and accuracy of identifying various geological disaster hazards. Brief Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for description in the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0040] Figure 1 It is a structural block diagram of a geological disaster hazard identification system that fuses LiDAR and oblique photography provided by an embodiment of the present application;

[0041] Figure 2 It is a flow block diagram of a geological disaster hazard identification method that fuses LiDAR and oblique photography provided by an embodiment of the present application. Detailed Embodiments

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of them. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0043] In this text, the term "comprising" is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a series of elements includes not only those elements but also other elements not expressly listed, or elements that are inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the said element.

[0044] The first aspect of this embodiment discloses a geological disaster hidden danger identification system that integrates LiDAR and oblique photography as shown in Figure 1 which includes a data acquisition module, a data processing module, an interpretation and analysis module, and a hidden danger identification module that are sequentially communicatively connected;

[0045] The data acquisition module is configured to: use an unmanned aerial vehicle to carry LiDAR and an oblique camera to collect laser point cloud data, optical image data, and oblique image data of the geological disaster point area;

[0046] The data processing module is configured to: perform strip splicing, denoising, and filtering on the original laser point cloud data of LiDAR, classify the point cloud in a human-computer interaction manner, generate a digital elevation model based on the processed laser point cloud data, and create a three-dimensional geological sand table by overlaying high-definition images and mountain shadows; process the oblique image data, and this processing at least includes equalizing light and color, image control point piercing, and aerial triangulation to produce a digital orthophoto image and a real scene three-dimensional model;

[0047] The interpretation and analysis module is configured to: based on the data obtained by LiDAR penetrating vegetation, remove the vegetation to generate a digital surface model, combine with optical remote sensing images to delineate the landslide morphology, and the landslide morphology at least includes the boundary, area, and volume of the landslide morphology; extract the disaster-causing factors in the LiDAR data, and the disaster-causing factors at least include slope, aspect, and ground roughness, and overlay them into a preset three-dimensional interpretation platform for shallow surface geological disaster interpretation; based on the terrain data generated by LiDAR, quantify the geological disasters, and this quantification at least includes extracting the information of cracks and landslide bodies;

[0048] The hidden danger identification module is configured to: construct an interpretation mark for geological disasters, and the geological disasters at least include landslides and collapses, obtain auxiliary identification information based on the digital elevation model and the digital surface model, and the auxiliary identification information at least includes mountain shadows, slope, aspect, and terrain openness factors; combine with the preset disaster-forming background conditions, integrate optical image data, auxiliary identification information, and historical deformation characteristics, and analogously analyze the topographic and geomorphic characteristics of the areas where disasters have occurred under the same geological conditions around the geological disaster point area to identify potential geological disaster hidden dangers.

[0049] Through the above, this embodiment uses the data acquisition module carried by the drone to obtain LiDAR point cloud, optical image and oblique image data, realizing the comprehensive acquisition of multi-source data. The data processing module processes the LiDAR data and oblique image data to generate various models, providing a basis for subsequent analysis. The interpretation and analysis module combines the data obtained by LiDAR penetrating vegetation and optical images, delineates the landslide morphology and extracts disaster-causing factors for interpretation, so as to deeply explore the characteristics of geological disasters. The hidden danger identification module constructs interpretation marks, fuses multi-source information to identify potential hidden dangers, realizes the accurate identification of geological disaster hidden dangers, overcomes the limitations of traditional single monitoring means, improves the accuracy and reliability of identification, provides strong support for geological disaster prevention and control, and has more advantages in the pertinence and accuracy of identifying various geological disaster hidden dangers.

[0050] Specifically, when the data acquisition module collects data, it realizes the preliminary fusion of LiDAR and oblique photography based on time synchronization and spatial registration. This preliminary fusion includes: during the flight of the drone, LiDAR and the oblique camera use the same time reference for data acquisition in terms of time, and in terms of space, a preset positioning and orientation system is used to determine the positions and postures of LiDAR and the oblique camera, and the collected data is unified into the same geographic coordinate system.

[0051] Through the above, this embodiment realizes the preliminary fusion of LiDAR and oblique photography data acquisition based on the existing time synchronization and spatial registration technology. By using the same reference for acquisition in terms of time, the timeliness correspondence of the data is ensured; in terms of space, the coordinate system is unified by using the positioning and orientation system to ensure the accurate matching of the collected data in terms of spatial position, so that subsequent data processing and analysis can be carried out based on accurately aligned data, improving the availability of the data and the accuracy of the analysis results. In practical applications, when making digital orthophoto images and constructing 3D models, the advantages of the two types of data can be better integrated, improving the model accuracy and image quality, so as to further optimize the fusion method of data acquisition, laying a more solid data foundation for the accurate interpretation and identification of subsequent geological disaster hidden dangers, and enhancing the system's ability to process data in complex geological environments.

[0052] Specifically, when the data processing module makes digital orthophoto images, it performs orthorectification on the oblique image data based on the oblique image rectification formula by using the digital elevation model; among them, the oblique image rectification formula is:

[0053]

[0054] where (x, y) is the image point coordinate, (x0, y0) is the principal point coordinate of the image, f is the camera focal length, (X, Y, Z) is the ground coordinate of the object point, and (X0, Y0, Z0) is the ground coordinate of the projection center.

[0055] In a simple example, for instance, in a certain geological disaster monitoring area, the main point coordinates (x0, y0) are (100, 200), the camera focal length f is 50 mm, the ground coordinates (X, Y, Z) of the object point are (1000, 2000, 500), and the ground coordinates (X0, Y0, Z0) of the projection center are (900, 1900, 400). According to the oblique image correction formula, the image point coordinates (x, y) can be calculated. Through such calculations, the orthorectification of the oblique image data is carried out, effectively eliminating the geometric distortion of the oblique image and making the positions and shapes of the ground objects in the image more accurate.

[0056] Through the above, in this embodiment, the orthorectification of the oblique image data is carried out based on the digital elevation model and the oblique image correction formula, realizing the production of high-quality digital orthophotos. The correction process of this embodiment is based on calculations, using the terrain information of the digital elevation model to effectively eliminate the geometric distortion of the oblique image, making the positions and shapes of the ground objects in the image more accurate, providing a more intuitive and accurate image basis for subsequent geological disaster interpretation, facilitating interpreters to more clearly observe and analyze geological features, thus having more technical advantages in the digital orthophoto production link, being able to provide higher-precision image data, contributing to improving the accuracy of geological disaster hidden danger interpretation, especially playing a better role in the identification of geological disasters in complex terrain areas.

[0057] Specifically, when the interpretation and analysis module delineates the landslide morphology, it extracts the terrain features of the landslide based on the digital surface model. The terrain features at least include slope features and aspect features, extracts the texture features and color features of the landslide based on the optical image data, and uses the comprehensive feature calculation formula to fuse the terrain features, texture features, and color features to obtain the comprehensive features; among them, the comprehensive feature calculation formula is:

[0058] F = ω s *F s + ω a *F a + ω t *F t + ω c *F c

[0059] Among them, F s is the slope feature, F a is the aspect feature, F t is the texture feature, F c is the color feature, ω s 、ω a 、ω t and ω cω is the weight of each corresponding feature, and F is the calculated comprehensive feature; based on a preset comprehensive feature threshold table, the landslide boundary is delineated in combination with the calculated comprehensive feature, and the comprehensive feature threshold table stores the comprehensive feature corresponding to the landslide boundary.

[0060] In a specific example, in a certain landslide area, the slope feature F extracted based on the digital surface model s is 0.6 (the value range is 0 - 1, and the larger the value, the steeper the slope), the aspect feature F a is 0.4 (the value range is 0 - 1, which can be quantified according to the angle between the aspect and the due north direction, etc.), the texture feature F extracted based on optical image data t is 0.5, and the color feature F c is 0.3. The weights of each feature are respectively set as ω s = 0.4, ω a = 0.3, ω t = 0.2, and ω c = 0.1. According to the comprehensive feature calculation formula, the calculated comprehensive feature F = 0.47. In the preset comprehensive feature threshold table, the comprehensive feature threshold corresponding to the landslide boundary is 0.45. Since the calculated comprehensive feature, this area can be delineated as part of the landslide boundary accordingly.

[0061] Through the above, this embodiment is based on the digital surface model and optical image data, and fuses multiple features through the comprehensive feature calculation formula, realizing the accurate delineation of the landslide boundary. In this embodiment, based on the extraction of the terrain, texture, and color features of the landslide, the landslide features can be described from multiple dimensions, overcoming the limitations of single-feature interpretation. Delineating the boundary according to the comprehensive feature threshold table makes the delineation result more scientific and accurate. Exemplarily, in a complex geological environment, different features complement each other, which can avoid misjudgment caused by the inconspicuousness of some features. It provides a more detailed and accurate method in landslide interpretation, enhances the ability to identify landslide hazards, helps to more comprehensively evaluate landslide risks, and provides a more reliable basis for geological disaster prevention and control.

[0062] Specifically, the interpretation and analysis module extracts disaster-causing factors based on LiDAR data, and conducts shallow subsurface geological disaster interpretation based on the disaster-causing factors. The shallow subsurface geological disaster interpretation includes:

[0063] Overlay the orthophoto obtained by oblique photography as texture information on the 3D interpretation platform constructed based on LiDAR data, use the disaster-causing factors to analyze the texture and tone changes of different ground objects in the orthophoto, and combine the slope value calculated using LiDAR data and the shallow layer disaster possibility calculation formula to calculate the possibility of shallow subsurface geological disasters. The shallow layer disaster possibility calculation formula is:

[0064] P = k1 * S + k2 * T + k3 * C

[0065] Among them, S is the slope value, T is the texture change value of the ground objects in the orthophoto, C is the tone change value of the ground objects in the orthophoto, k1, k2, and k3 are the weights of the corresponding numerical values, P is the calculated possibility of shallow surface geological disasters occurring. When P≥P τ it is determined that there are potential hazards of shallow surface geological disasters in this area, and P τ is the preset possibility threshold of shallow surface geological disasters occurring.

[0066] Through the above, this embodiment realizes the effective interpretation of shallow surface geological disasters. By analyzing the texture and tone changes of the orthophoto, calculating the disaster possibility in combination with the slope value, and comprehensively considering various factors to judge potential hazards, it can more comprehensively consider the influencing factors of shallow surface geological disasters. It is more targeted and accurate in the interpretation of shallow surface geological disasters, makes up for the deficiencies in various geological disaster interpretation methods, improves the system's ability to identify potential hazards of shallow surface geological disasters, and provides more effective technical support for the early warning and prevention of geological disasters.

[0067] Specifically, when the interpretation and analysis module extracts crack information, it performs edge detection based on LiDAR data to obtain the preliminary candidate area of the crack, and verifies the preliminary candidate area in combination with the multi-angle images obtained by oblique photography and the comprehensive gradient amplitude calculation formula. The comprehensive gradient amplitude calculation formula is:

[0068]

[0069] Among them, G1 is the gradient amplitude obtained by performing edge detection based on LiDAR data, G2 is the gradient amplitude obtained based on the multi-angle images, is the gradient amplitude compensation value obtained based on regression analysis. This gradient amplitude compensation value is used to represent the compensation for extracting crack information based on historical crack information. G is the calculated comprehensive gradient amplitude. When G≥G τ it is determined that this area is the crack area. G τ is the preset comprehensive gradient amplitude threshold.

[0070] In a specific example, in a certain geological area, G1 = 100, G2 = 80, and the gradient amplitude compensation value obtained by performing regression analysis on historical crack information According to the comprehensive gradient amplitude calculation formula, the calculated comprehensive gradient amplitude G≈128.06. The preset comprehensive gradient amplitude threshold G τ = 120. Based on this, it is determined that this area is the crack area.

[0071] With the above, this embodiment uses edge detection of LiDAR-based data in combination with oblique photography multi-angle images, extracts and verifies crack information through a comprehensive gradient magnitude calculation formula, realizes accurate extraction of crack information, uses multi-angle images to supplement information, and combines the gradient magnitude compensation value to more accurately judge the crack area and avoid misjudgment. In a complex geological environment, multi-angle images can observe cracks from different perspectives and enhance the recognition of crack features. It improves the ability to identify crack features in geological disaster hazards, helps to timely discover potential geological disaster risks, and provides more accurate data support for geological disaster prevention and control.

[0072] Specifically, when constructing the interpretation signs for landslides, the hidden danger identification module combines LiDAR data and oblique photography data, and uses the landslide possibility index calculation formula to determine the interpretation signs for landslides. The landslide possibility index calculation formula is:

[0073]

[0074] where I1 is the topographic feature index of the landslide obtained from LiDAR data and oblique photography data, and this topographic feature index at least includes slope and elevation difference; I2 is the image feature index obtained from LiDAR data and oblique photography data, and this image feature index at least includes hue and texture. is the landslide possibility index compensation value obtained through regression analysis, and this landslide possibility index compensation value is used to represent the compensation for the landslide interpretation signs obtained from the historical landslide interpretation signs. α and β are corresponding weights, and L is the calculated landslide interpretation sign. Based on the preset landslide threshold range, the calculated landslide interpretation sign is mapped to different levels.

[0075] In a simple example, for a certain potential landslide area, the topographic feature index I1 of the landslide obtained from LiDAR data and oblique photography data = 0.55 (where, based on a slope of 35°, the converted value is 0.6; the elevation difference is 50 meters, and the converted value is 0.5, and the combined value is I1 = 0.55), the image feature index I2 = 0.45 (where, based on a hue feature value of 0.4 and a texture feature value of 0.5, the combined value is I2 = 0.45), and the landslide possibility index compensation value obtained through regression analysis Set α = 0.6, β = 0.4. According to the landslide possibility index calculation formula, the calculated landslide interpretation sign L = 0.52. The preset landslide threshold range is: low risk (0 - 0.4), medium risk (0.4 - 0.6), high risk (0.6 - 1). Then the landslide possibility of this area belongs to the medium risk level.

[0076] Through the above, in this embodiment, the landslide interpretation marks are constructed through the landslide possibility index calculation formula, realizing the scientific assessment and grading of landslide hazards. Considering the terrain and image feature indicators and combining with the landslide possibility index compensation value can more comprehensively and accurately reflect the degree of landslide hazards. Mapping the levels according to the preset threshold range facilitates the intuitive judgment of the landslide risk level. It is more scientific and comprehensive in constructing the landslide interpretation marks, improving the quantification degree of landslide hazard identification, providing a more valuable reference basis for geological disaster prevention and control decisions, helping to reasonably allocate prevention and control resources, and improving the prevention and control efficiency.

[0077] Specifically, when the hazard identification module obtains the terrain openness factor based on the digital elevation model and the digital surface model, it calculates the terrain openness factor by using the panoramic image information obtained by oblique photography and the terrain openness factor calculation formula. The terrain openness factor calculation formula is:

[0078] O = γ * R h + δ * θ

[0079] Wherein, R h is the elevation change rate within the surrounding range corresponding to a point in the digital elevation model, θ is the visible range of this point in the panoramic image information, γ and δ are the corresponding weights, and O is the calculated terrain openness factor.

[0080] Through the above, in this embodiment, by combining the oblique photography panoramic image information and the terrain openness factor calculation formula, the terrain openness factor is obtained based on the digital elevation model and the digital surface model, realizing the accurate acquisition of the auxiliary identification information of geological disaster hazards. The panoramic image provides richer visible range information. Combined with the elevation change rate calculation, it can more accurately reflect the impact of the terrain on geological disasters. For example, when judging the collapse hazard, the terrain openness factor can be used as an important reference. It is more comprehensive and accurate in obtaining the auxiliary identification information, enhancing the system's ability to identify geological disaster hazards, providing more powerful support for subsequent comprehensive analysis and hazard identification, and improving the scientific nature of geological disaster prevention and control.

[0081] Specifically, the disaster-forming background conditions at least include the geological background corresponding to geological structure, rock and soil type, topography and geomorphology, and meteorology and hydrology.

[0082] Through the above, this embodiment considers the geological background corresponding to geological structures, rock and soil types, topography, and meteorology and hydrology as the background conditions for disaster formation, achieving a more comprehensive and in-depth analysis of geological disaster hazards. Different disaster formation background conditions are interrelated and jointly affect the occurrence of geological disasters. By comprehensively considering these factors, the possibility and risk of geological disasters can be evaluated more accurately. For example, when analyzing landslide hazards, the stability of rock masses is judged in combination with geological structures, and the inducing effect of rainfall on landslides is considered in combination with meteorology and hydrology, thereby achieving a more systematic and perfect analysis of disaster formation background conditions, making up for the deficiency of the comparative document in the analysis of disaster formation background in the identification of various geological disaster hazards, and improving the accuracy and reliability of geological disaster hazard identification.

[0083] The second aspect of this embodiment discloses a geological disaster hazard identification method integrating LiDAR and oblique photography as shown in Figure 2 The method is applicable to the geological disaster hazard identification system integrating LiDAR and oblique photography as described above. The method includes:

[0084] Using an unmanned aerial vehicle to carry LiDAR and an oblique camera to collect laser point cloud data, optical image data, and oblique image data of the geological disaster point area;

[0085] Perform strip splicing, denoising, and filtering on the original laser point cloud data of LiDAR, classify the point cloud by means of human-computer interaction, generate a digital elevation model based on the processed laser point cloud data, and create a three-dimensional geological sand table by overlaying high-definition images and mountain shadows; process the oblique image data, and the processing at least includes equalizing light and color, image control point piercing, and aerial triangulation, and produce a digital orthophoto image and a real three-dimensional model;

[0086] Based on the data obtained by LiDAR penetrating vegetation, remove the vegetation to generate a digital surface model, combine it with optical remote sensing images to delineate the landslide morphology, and the landslide morphology at least includes the boundary, area, and volume of the landslide morphology; extract the disaster-causing factors in the LiDAR data, and the disaster-causing factors at least include slope, aspect, and ground roughness, overlay them into a preset three-dimensional interpretation platform for shallow surface geological disaster interpretation; based on the terrain data generated by LiDAR, quantify the geological disasters, and the quantification at least includes extracting the information of cracks and landslide bodies;

[0087] Construct an interpretation mark for geological disasters, and the geological disasters at least include landslides and collapses. Based on the digital elevation model and the digital surface model, obtain auxiliary identification information, and the auxiliary identification information at least includes mountain shadows, slope, aspect, and terrain openness factors; combine the preset disaster formation background conditions, integrate optical image data, auxiliary identification information, and historical deformation characteristics, and analogously analyze the topographic and geomorphic characteristics of the areas where disasters have occurred under the same geological conditions around the geological disaster point area to identify potential geological disaster hazards.

[0088] It should be noted that the geological disaster hidden danger identification method integrating LiDAR and oblique photography in this embodiment corresponds to the aforementioned geological disaster hidden danger identification system integrating LiDAR and oblique photography. Therefore, for the content not specifically described in the geological disaster hidden danger identification method integrating LiDAR and oblique photography in this embodiment, which may but is not limited to function definitions, working principles, technical effects, etc., reference can be made to the records of the aforementioned geological disaster hidden danger identification system integrating LiDAR and oblique photography, and will not be elaborated herein.

[0089] In summary, for the geological disaster hidden danger identification system and method integrating LiDAR and oblique photography in this embodiment, the data acquisition module is carried on an unmanned aerial vehicle to obtain laser point cloud, optical image, and oblique image data, realizing the comprehensive acquisition of multi-source data. The data processing module processes the LiDAR data and oblique image data to generate various models, providing a basis for subsequent analysis. The interpretation and analysis module combines the data obtained by LiDAR penetrating vegetation and optical images, delineates the landslide morphology, extracts disaster-causing factors for interpretation, thereby deeply exploring the characteristics of geological disasters. The hidden danger identification module constructs interpretation marks, integrates multi-source information to identify potential hidden dangers, realizes the accurate identification of geological disaster hidden dangers, overcomes the limitations of traditional single monitoring means, improves the accuracy and reliability of identification, provides strong support for geological disaster prevention and control, and has more advantages in the pertinence and accuracy of identifying various geological disaster hidden dangers.

[0090] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. The computer-readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM, or other optical disc storage, magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0091] Finally, it should be noted that the above are only the preferred embodiments of the present application and are not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A geological disaster hazard identification system integrating LiDAR and oblique photography, characterized in that: The system includes a data acquisition module, a data processing module, an interpretation and analysis module and a hidden danger identification module which are sequentially connected in communication; The data acquisition module is configured as follows: using a drone equipped with LiDAR and an oblique camera to collect laser point cloud data, optical image data and oblique image data of the geological disaster point area; The data processing module is configured to: perform flight strip stitching, denoising, and filtering on the original laser point cloud data of LiDAR, classify the point cloud using human-computer interaction, generate a digital elevation model based on the processed laser point cloud data, and superimpose high-definition images and mountain shadows to create a three-dimensional geological sand table; process the oblique image data, which at least includes uniform light and color, image control point puncture and aerial triangulation, to produce digital orthophotos and real-scene three-dimensional models; The interpretation and analysis module is configured as follows: based on the data obtained by LiDAR penetrating vegetation, the vegetation is removed to generate a digital surface model, and the landslide morphology is delineated in combination with optical remote sensing images, and the landslide morphology at least includes the boundary, area and volume of the landslide morphology; the hazard factors in the LiDAR data are extracted, and the hazard factors at least include the slope, slope aspect and ground roughness, and are superimposed into a preset three-dimensional interpretation platform to interpret shallow geological disasters; based on the terrain data generated by LiDAR, the geological disasters are quantified, and the quantification at least includes extracting information on cracks and landslide bodies; The hidden danger identification module is configured to: construct an interpretation mark for geological disasters, the geological disasters at least include landslides and collapses, and obtain auxiliary identification information based on the digital elevation model and the digital surface model, the auxiliary identification information at least includes mountain shadows, slopes, slope aspects and terrain openness factors; Combined with the preset disaster-prone background conditions, the optical image data, the auxiliary identification information and the historical deformation characteristics are integrated to analyze the topographic and geomorphic characteristics of the disaster-affected areas under the same geological conditions around the geological disaster point area by analogy, and identify potential geological disaster risks.

2. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1 is characterized in that: When collecting data, the data acquisition module realizes the preliminary fusion of LiDAR and oblique photography based on time synchronization and spatial registration. The preliminary fusion includes: during the flight of the UAV, the LiDAR and the oblique camera use the same time reference to collect data in time, and in space, use a preset positioning and orientation system to determine the position and posture of the LiDAR and the oblique camera, and unify the collected data into the same geographic coordinate system.

3. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1 is characterized in that: When making a digital orthophoto, the data processing module uses the digital elevation model to perform orthorectification on the oblique image data based on the oblique image correction formula; wherein the oblique image correction formula is: Among them, (x, y) are the image point coordinates, (x0, y0) are the image principal point coordinates, f is the camera focal length, (X, Y, Z) are the ground coordinates of the object point, and (X0, Y0, Z0) are the ground coordinates of the projection center.

4. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: When delineating the landslide morphology, the interpretation and analysis module extracts the terrain features of the landslide based on the digital surface model, the terrain features at least including slope features and slope aspect features, extracts the texture features and color features of the landslide based on the optical impact data, and obtains the comprehensive features by fusing the terrain features, the texture features and the color features using a comprehensive feature calculation formula; wherein the comprehensive feature calculation formula is: F=ω s *F s +oh a *F a +oh t *F t +oh c *F c Among them, F s is the slope characteristic, F a is the slope characteristic, F t is the texture feature, F c is the color feature, ω s ,ω a ,ω t and ω c is the weight of each corresponding feature, and F is the calculated comprehensive feature; based on the preset comprehensive feature threshold table, the landslide boundary is delineated in combination with the calculated comprehensive feature, and the comprehensive feature threshold table stores the comprehensive features corresponding to the landslide boundary.

5. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: The interpretation and analysis module extracts the disaster-causing factors based on the LiDAR data, and interprets the shallow geological disasters based on the disaster-causing factors. The shallow geological disaster interpretation includes: The orthophotos obtained by oblique photography are superimposed as texture information on the three-dimensional interpretation platform built based on LiDAR data. The disaster-causing factors are used to analyze the texture and color changes of different objects in the orthophotos. The possibility of shallow geological disasters is calculated by combining the slope value calculated using LiDAR data and the shallow disaster possibility calculation formula. The shallow disaster possibility calculation formula is: P=k1*S+k2*T+k3*C Among them, S is the slope value, T is the texture change value of the object in the orthophoto, C is the color change value of the object in the orthophoto, k1, k2 and k3 are the weights of the corresponding values, and P is the calculated possibility of shallow surface geological disasters. When P≥P τ When determining that there are shallow geological disaster risks in the area, P τ It is the preset threshold value of the possibility of shallow geological disasters.

6. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: When extracting crack information, the interpretation and analysis module performs edge detection based on LiDAR data to obtain a preliminary candidate area for cracks, and verifies the preliminary candidate area by combining multi-angle images obtained by oblique photography and a comprehensive gradient amplitude calculation formula, which is: Among them, G1 is the gradient amplitude obtained by edge detection based on LiDAR data, and G2 is the gradient amplitude obtained based on multi-angle images. is the gradient amplitude compensation value obtained based on regression analysis. The gradient amplitude compensation value is used to characterize the compensation for extracting fracture information based on historical fracture information. G is the calculated comprehensive gradient amplitude. When G ≥ G τ The area is determined as the crack area G τ is the preset comprehensive gradient amplitude threshold.

7. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: When constructing the interpretation mark for landslide, the hidden danger identification module integrates the LiDAR data and the oblique photography data, and uses the landslide possibility index calculation formula to determine the interpretation mark for landslide. The landslide possibility index calculation formula is: Wherein, I1 is the terrain characteristic index of the landslide obtained based on the LiDAR data and the oblique photography data, and the terrain characteristic index at least includes slope and height difference; I2 is the image characteristic index obtained based on the LiDAR data and the oblique photography data, and the image characteristic index at least includes hue and texture. It is the landslide possibility index compensation value obtained based on regression analysis. The landslide possibility index compensation value is used to characterize the compensation for the interpretation mark of the landslide based on the interpretation mark of the historical landslide. α and β are the corresponding weights. L is the calculated interpretation mark of the landslide. Based on the preset landslide threshold range, the calculated interpretation mark of the landslide is mapped to different levels.

8. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: When the hidden danger identification module obtains the terrain openness factor based on the digital elevation model and the digital surface model, it calculates the terrain openness factor using the panoramic image information obtained by oblique photography and the terrain openness factor calculation formula. The terrain openness factor calculation formula is: O=γ*R h +δ*θ Among them, R h is the elevation change rate in the surrounding area corresponding to a point in the digital elevation model, θ is the visible range of the point in the panoramic image information, γ and δ are the corresponding weights, and O is the calculated terrain openness factor.

9. The geological disaster hazard identification system integrating LiDAR and oblique photography according to claim 1, characterized in that: The disaster-prone background conditions at least include geological structure, rock and soil type, topography and geomorphology, and the geological background corresponding to meteorology and hydrology.

10. A method for identifying geological hazards by integrating LiDAR and oblique photography, the method being applicable to the geological hazards identification system integrating LiDAR and oblique photography as claimed in any one of claims 1 to 9, characterized in that: The method includes: Use drones equipped with LiDAR and oblique cameras to collect laser point cloud data, optical image data, and oblique image data in the geological disaster area; The original laser point cloud data of LiDAR is processed by flight strip stitching, denoising and filtering, and point cloud classification is performed using human-computer interaction. A digital elevation model is generated based on the processed laser point cloud data, and high-definition images and mountain shadows are superimposed to create a three-dimensional geological sand table; the oblique image data is processed, and the processing includes at least uniform light and color, image control point puncture and aerial triangulation, to produce digital orthophotos and real-scene three-dimensional models; Based on the data obtained by LiDAR penetrating vegetation, the vegetation is removed to generate a digital surface model, and the landslide morphology is delineated in combination with optical remote sensing images, and the landslide morphology at least includes the boundary, area and volume of the landslide morphology; the hazard factors in the LiDAR data are extracted, and the hazard factors at least include slope, slope aspect and ground roughness, and are superimposed on a preset three-dimensional interpretation platform to interpret shallow geological disasters; based on the terrain data generated by LiDAR, the geological disasters are quantified, and the quantification at least includes the extraction of information on cracks and landslide bodies; Construct interpretation signs for geological disasters, which at least include landslides and collapses, and obtain auxiliary identification information based on the digital elevation model and the digital surface model, wherein the auxiliary identification information at least includes mountain shadows, slopes, slope directions and terrain openness factors; combine the preset disaster-pregnant background conditions, fuse the optical image data, the auxiliary identification information and historical deformation characteristics, and analyze the topographic and geomorphic characteristics of the disaster-affected areas under the same geological conditions around the geological disaster point area by analogy, so as to identify potential geological disaster hazards.

Images