CN111178214B - High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology - Google Patents
High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology Download PDFInfo
- Publication number
- CN111178214B CN111178214B CN201911337549.9A CN201911337549A CN111178214B CN 111178214 B CN111178214 B CN 111178214B CN 201911337549 A CN201911337549 A CN 201911337549A CN 111178214 B CN111178214 B CN 111178214B
- Authority
- CN
- China
- Prior art keywords
- dangerous rock
- rock mass
- point cloud
- point
- aerial vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a rapid identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology, which comprises the following steps: generating a point cloud model by using a high-definition photo of an unmanned aerial vehicle, and carrying out point cloud denoising and sparse processing; calculating general characteristic points of the point cloud model, and clustering the general characteristic points into a plurality of general characteristic point sets by using a noise application space clustering algorithm based on density; classifying the general characteristic point set by using a support vector machine to determine a boundary point set of the dangerous rock mass; in the neighborhood of each boundary point set, fitting points on a slope where a dangerous rock mass is located into a plane to cut a point cloud model, and dividing the remaining point cloud model into a plurality of independent point cloud areas by using a density-based noise application spatial clustering algorithm; determining the region of the dangerous rock mass according to the geometric characteristics of each independent point cloud region; and carrying out primary stability evaluation on the dangerous rock mass. The method can realize quick identification and stable preliminary judgment of the dangerous rock mass, and provides technical support for engineering construction.
Description
Technical Field
The invention belongs to the technical field of dangerous rock mass identification, and particularly relates to a rapid identification method for a high and steep slope dangerous rock mass based on an unmanned aerial vehicle photography technology.
Background
Dangerous rock mass on a high and steep slope is one of the hazards frequently encountered in the construction of hydropower engineering, and the existence of the dangerous rock mass not only influences the safety of the engineering, but also influences the construction progress. Therefore, the early-stage investigation and the preliminary evaluation of the dangerous rock mass have important significance for the overall construction of the project.
The use of drones and the development of new aerial and ground sensors have enabled a significant increase in the measurement capacity for complex and hazardous environments. The main advantage of unmanned aerial vehicle photography in high and steep slope surveying is its superior spatial resolution and the ability to carry multiple sensors. Simultaneously, unmanned aerial vehicle removes in a flexible way, and it is low to control the space requirement, compares with satellite-borne image acquisition, can avoid the hindrance of unfavorable weather condition and complicated topography, can reach the region that someone piloted aircraft can't arrive. Furthermore, the drone photography system can be stored and deployed at minimal cost. By the characteristics, the unmanned aerial vehicle can rapidly and efficiently survey dangerous rock masses on high and steep slopes.
However, because the dangerous rock mass has large shape and size difference and the surface of the slope where the dangerous rock mass is located has complex shape, the boundary line of the dangerous rock mass is not clear and is easily confused with the boundaries such as gullies and trees. For dangerous rock masses with high and steep slopes, an economically feasible method capable of being rapidly identified and evaluated is still lacking due to the complex form and the inconvenience in traffic.
Therefore, how to provide a rapid identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology is a problem to be solved urgently by technical personnel in the field.
Disclosure of Invention
In view of the above, the invention provides a rapid identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology, which is based on a high-precision image obtained by unmanned aerial vehicle photography, and can realize rapid identification and stable preliminary judgment on the dangerous rock mass and provide technical support for engineering construction by applying density-based noise application spatial clustering, a class of support vector machines, stress tensor reduction calculation and other theoretical methods to a generated point cloud model.
In order to achieve the purpose, the invention adopts the following technical scheme:
a rapid identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology comprises the following steps:
1) Generating a point cloud model by using a high-definition photo of an unmanned aerial vehicle, and performing point cloud denoising and sparse processing;
2) Calculating general characteristic points of the point cloud model, and clustering the general characteristic points into a plurality of general characteristic point sets by using a noise application space clustering algorithm based on density;
3) Classifying the general characteristic point set by using a support vector machine to determine a boundary point set of the dangerous rock mass;
4) In the neighborhood of each boundary point set, fitting points on the slope where the dangerous rock mass is located into a plane to cut the point cloud model; deleting the intersection points between the planes and the point cloud model from the point cloud, and dividing the rest point cloud model into a plurality of independent point cloud areas by using a noise application space clustering algorithm based on density; determining a region representing the dangerous rock mass according to the geometric characteristics of each independent point cloud region;
5) And carrying out primary stability evaluation on the dangerous rock mass.
Preferably, in the step 1), the method for generating the point cloud model by using the high-definition photo of the unmanned aerial vehicle and performing point cloud denoising and sparse processing comprises the following steps:
and generating a dense point cloud model through unmanned aerial vehicle photography and SFM technology, and denoising and sparse operation are carried out on the dense point cloud.
Preferably, in step 2), the general feature points are calculated as follows:
s1, estimating the discrete main curvature and the main direction of each point according to the Weingerten mapping matrix property;
s2, judging whether the minimum principal curvature of the point on the normal section line can be a minimum value or not in a range of r meters away from the left side and the right side of each point along the direction of the minimum principal curvature of each point; if so, the point belongs to a general feature point.
Preferably, in the step 3), the method for classifying the general feature point set by using the support vector machine to determine the boundary point set of the dangerous rock mass includes:
firstly, a plurality of typical dangerous rock masses are defined from a slope point cloud model through manual visual interpretation; meanwhile, selecting a point set belonging to the boundary of the dangerous rock mass in the general characteristic point set as sample data of SVM training; then, calculating characteristic parameters of all the general characteristic point sets; secondly, training the SVM by adopting a typical dangerous rock boundary point set, deducing the distance value from all general characteristic point sets to an SVM decision surface, and taking the distance as the fraction of each point set; finally, the minimum score of the sample is used as a threshold value, and the general characteristic point set with the score larger than the threshold value is used as the boundary point set of the dangerous rock mass.
Preferably, in the step 4), the method for cutting the slope point cloud model and extracting the possibly dangerous rock mass comprises the following steps:
firstly, calculating neighborhood points of each dangerous rock mass boundary point set according to a neighborhood algorithm; then, fitting points on the slope surface in the neighborhood points into a plane, and cutting the point cloud model; then, removing the intersection points of the planes and the model from the point cloud model, and performing density-based noise application spatial clustering on the remaining point clouds to obtain a clustering result which is a plurality of isolated point cloud areas; finally, calculating the volume-surface area ratio of a plurality of typical dangerous rock masses, and taking the minimum volume-surface area ratio as a threshold value; and comparing the threshold with the volume-surface area ratio of each isolated point cloud area obtained by clustering, and if the volume-surface area ratio of a certain area is greater than the threshold, determining that the isolated point cloud area is a dangerous rock mass area.
Preferably, the area points of the dangerous rock boundary point set include points on the surface of the dangerous rock body and points on the slope where the dangerous rock body is located.
Preferably, in the step 5), the preliminary stability evaluation method for the dangerous rock mass includes:
firstly, determining the inclination angle of the bottom sliding surface of each dangerous rock mass according to the analysis of the declination projection; secondly, projecting the surface of the dangerous rock body on a bottom sliding surface and dividing the surface of the dangerous rock body into grids with proper sizes, wherein each grid corresponds to a rock column from the surface of the dangerous rock body to the bottom sliding surface; then, estimating the ratio of shear stress and normal stress caused by the self weight of each rock column on the sliding surface of the bottom surface by using a stress tensor reduction theory so as to replace the average stress ratio of a grid region corresponding to the rock column; and finally, determining the position of the damaged grid on the sliding surface of the bottom surface according to a molar-coulomb damage criterion, calculating the ratio of the area of all the damaged grids to the total area of the sliding surface of the bottom surface, and grading the stability of the dangerous rock mass according to the ratio.
The invention has the beneficial effects that:
according to the method, a high-precision image obtained by unmanned aerial vehicle photography is taken as a basis, the generated point cloud model is subjected to density-based noise application spatial clustering, a class of support vector machines, stress tensor calculation reduction and other theoretical methods, so that the dangerous rock mass with a high and steep slope, which is difficult to reach by manpower, can be quickly identified and stably evaluated, meanwhile, point cloud clustering and machine learning algorithms are introduced to assist and replace manual identification of the dangerous rock mass, so that subjective errors caused by complete manual judgment of the dangerous rock mass are avoided, and meanwhile, the stability of the dangerous rock mass can be preliminarily evaluated and graded only by the point cloud model of the dangerous rock mass. Compared with the traditional method, the method has the following advantages: 1. the investigation of dangerous rock masses in a large area is possible; 2. the dangerous rock body boundary is automatically identified, and meanwhile, the geometric information of the dangerous rock body can be obtained; 3. only dangerous rock mass DEM model and bottom slip surface information are needed, and dangerous rock mass stability can be preliminarily graded.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of the process of the present invention.
FIG. 2 is a schematic structural diagram of the sparse point cloud model of the present invention.
FIG. 3 is a schematic diagram of the density and actual distribution of the general characteristic points of the present invention.
FIG. 4 is a schematic diagram of a critical rock boundary point set separated by the SVM of the present invention.
Fig. 5 is a schematic diagram of a point cloud area corresponding to a dangerous rock mass.
FIG. 6 is a schematic diagram showing distribution and stability grading of dangerous rock masses on two rivers and water and high and steep slopes on two banks.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to the attached drawing 1, the invention provides a rapid identification method of high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology, which comprises the following steps:
the first step is as follows: and generating a dense point cloud model by unmanned aerial vehicle photography and SFM technology, and denoising and sparsely operating the dense point cloud.
The second step: and calculating general characteristic points of the point cloud model, and clustering the general characteristic points into a plurality of general characteristic point sets by using a noise application space clustering algorithm based on density. The general feature point calculation process is as follows:
s1, estimating the discrete main curvature and the main direction of each point according to the Weingerten mapping matrix property;
s2, judging whether the minimum principal curvature of the point on the normal sectional line can be a minimum value or not along the direction of the minimum principal curvature of each point and within r meters away from the left side and the right side of the point; if so, the point belongs to a general feature point.
The third step: classifying the general characteristic point set by using a support vector machine to determine a boundary point set of the dangerous rock mass;
the method for classifying the general characteristic point set by using the support vector machine to determine the boundary point set of the dangerous rock mass comprises the following steps: firstly, a plurality of typical dangerous rock masses are defined from a slope point cloud model through manual visual interpretation; meanwhile, selecting a point set belonging to the boundary of the dangerous rock mass in the general characteristic point set as sample data of SVM training; then, calculating characteristic parameters of all general characteristic point sets (including point sets serving as samples); secondly, training the SVM by adopting a typical dangerous rock boundary point set, deducing the distance value from all general characteristic point sets to an SVM decision surface, and taking the distance as the fraction of each point set; finally, the minimum score of the sample is used as a threshold value, and the general characteristic point set with the score larger than the threshold value is used as the boundary point set of the dangerous rock mass.
The fourth step: in the neighborhood of each boundary point set, fitting points on the slope where the dangerous rock mass is located into a plane to cut the point cloud model; deleting the intersection points between the planes and the point cloud model from the point cloud, and dividing the rest point cloud model into a plurality of independent point cloud areas by using a density-based noise application space clustering (DBSCAN) algorithm; determining a region representing the dangerous rock mass according to the geometric characteristics of each independent point cloud region;
the method for cutting the slope point cloud model and extracting the possibly dangerous rock mass comprises the following steps: firstly, calculating a neighborhood point of each dangerous rock body boundary point set according to a neighborhood algorithm, wherein the neighborhood points of the dangerous rock body boundary point set comprise points on the surface of a dangerous rock body and points on a slope surface where the dangerous rock body is located; then, fitting points on the slope surface in the neighborhood points into a plane, and cutting the point cloud model; then, removing the intersection points of the planes and the model from the point cloud model, and carrying out density-based noise application spatial clustering on the residual point cloud, wherein the clustering result is a plurality of isolated point cloud areas; finally, calculating the volume-surface area ratio of a plurality of typical dangerous rock masses, and taking the minimum volume-surface area ratio as a threshold value; and comparing the threshold with the volume-surface area ratio of each isolated point cloud region obtained by clustering, and if the volume-surface area ratio of a certain region is greater than the threshold, determining that the isolated point cloud region is a dangerous rock mass region.
The fifth step: and carrying out primary stability evaluation on the dangerous rock mass. The method comprises the following steps: firstly, determining the inclination angle of the bottom sliding surface of each dangerous rock mass according to the analysis of the equatorial projection; secondly, projecting the surface of the dangerous rock body on a bottom sliding surface and dividing the surface of the dangerous rock body into grids with proper sizes, wherein each grid corresponds to a rock column from the surface of the dangerous rock body to the bottom sliding surface; then, estimating the ratio of shear stress and normal stress caused by the self weight of each rock column on the bottom sliding surface by using a stress tensor reduction theory so as to replace the average stress ratio of a grid area corresponding to the rock column; and finally, determining the grid position where the bottom sliding surface is damaged according to a molar-coulomb damage criterion, calculating the ratio of the area of all damaged grids to the total area of the bottom sliding surface, and classifying the stability of the dangerous rock mass according to the ratio because the ratio reflects the damage degree of the rock on the bottom sliding surface of the dangerous rock mass so as to evaluate the safety of the dangerous rock mass.
Application example:
taking the dangerous rock masses at two banks of the hydropower station at two river mouths as an example, the unmanned aerial vehicle shooting is carried out on the dangerous rock masses. The high-definition aerial photo is processed in the first step to generate a sparse point cloud model, and the density of the sparse point cloud model is calculated according to the number of points in one point neighborhood: there are 3-5 neighbor points in a 1m neighborhood of a point, see fig. 2. Determining m (m = 10) nearest neighbor points of each point p, and calculating the minimum discrete principal curvature of the point, wherein m is a threshold value of the number of neighbor points of each point when the discrete principal curvature of each point is estimated; then estimating the minimum discrete principal curvature of two points which are along the direction of the minimum principal curvature of the point and are r (r =6 m) meters away from the left side and the right side of the point, comparing the minimum discrete principal curvature of the point with the minimum discrete principal curvature values of the left point and the right point corresponding to the point, and if the minimum discrete principal curvature values of the point are smaller than the corresponding values of the left point and the right point, the point is a common characteristic point; wherein, r is the distance threshold value between the point for comparison and the general characteristic point when the general characteristic point is judged; finally, according to the density and the actual distribution of the general characteristic points, referring to the attached figure 3, setting clustering parameters Eps =2.3m and MinPts =5, and carrying out DBSCAN clustering on the clustering parameters; when Eps is DBSCAN cluster, a neighborhood distance threshold of a certain point is set; when MinPts is a DBSCAN cluster, the distance of a certain point is a threshold value of the number of samples in the neighborhood of Eps. Thirdly, calculating n1 (n 1= 45) neighborhood points of each feature point in each general feature point set (clustering result in the second step) by using a nearest neighbor algorithm, and solving a union set of the neighborhood points as a neighborhood point set of the general feature point set; wherein n1 is a neighborhood point number threshold of a certain point when determining the neighborhood of the general feature point set. Each general feature point set can be fitted into a curve, and the neighborhood point set is divided into a left region and a right region. Calculating the surface appearance, theta, of the left and right regions, respectively j (0≤θ j The included angle between the left area and the right area is less than or equal to 180 degrees; theta 1 And theta 2 (0≤θ 1 ,θ 2 The inclination angles of the left and right areas are respectively less than or equal to 90 degrees; theta d (0≤θ d An included angle between the inclination angles of the left and right regions (theta) is less than or equal to 180 DEG j ,θ 1 ,θ 2 ,θ d ) Is a feature vector for each general set of feature points. And then, manually defining point cloud areas corresponding to 7 typical dangerous rock bodies, selecting two general feature point sets on each dangerous rock body boundary, and taking feature vectors of the 14 general feature point sets as SVM training samples. Selecting SVM parameters gamma =0.25 and v =0.4, screening a dangerous rock mass boundary point set, and referring to the result in figure 4, wherein gamma and v are variable parameters in a type of support vector machine. And fourthly, removing the intersection point of the point cloud model and the cutting plane from the point cloud model, setting clustering parameters Eps =1.5m and MinPts =3 to perform DBSCAN clustering on the residual point cloud, and performing region screening on the point cloud corresponding to the possible dangerous rock mass, referring to the attached figure 5. The fifth step, assuming the maximum principal stress σ of the bottom slip surface caused by the self-weight of the rock 1 Equal to the dead weight, then the minimum principal stress σ 3 Is σ 1 One third of (1), i.e. = -1/3 σ 1 . Then, 3R values (R is respectively equal to 0.2,0.3 and 0.4) are selected to calculate the shear stress and normal stress ratio distribution of the bottom slip surface of the dangerous rock mass by using a stress tensor reduction method (wherein the physical meaning of R is as follows: setting l = -sigma) 3 The negative of the minimum principal stress, the three principal stresses are added with l and divided by (σ) 1 + l) to obtain new principal stress sigma 'after reduction' 1 ＝1，
σ′ 3 =0, let R = σ' 2 R is called the primary stress ratio). Since the change in the R value does not greatly affect the distribution of the high stress ratio region data on the circle of molar stress, the middle value (R = 0.3) of the assumed R value was used for the evaluation of the stability of the dangerous rock mass. FIG. 6 is a diagram of distribution and stable classification of bank slope dangerous rock masses.
According to the method, a high-precision image obtained by unmanned aerial vehicle photography is taken as a basis, the generated point cloud model is subjected to density-based noise application spatial clustering, a class of support vector machines, stress tensor calculation reduction and other theoretical methods, so that the dangerous rock mass with a high and steep slope, which is difficult to reach by manpower, can be quickly identified and stably evaluated, meanwhile, point cloud clustering and machine learning algorithms are introduced to assist and replace manual identification of the dangerous rock mass, so that subjective errors caused by complete manual judgment of the dangerous rock mass are avoided, and meanwhile, the stability of the dangerous rock mass can be preliminarily evaluated and graded only by the point cloud model of the dangerous rock mass. Compared with the traditional method, the method has the following advantages: 1. the investigation of dangerous rock masses in a large area is possible; 2. the boundary of the dangerous rock mass is automatically identified, and meanwhile, the geometric information of the dangerous rock mass can be obtained; 3. only dangerous rock mass DEM model and bottom slip surface information are needed, and dangerous rock mass stability can be preliminarily graded.
The rapid identification and the preliminary evaluation of the dangerous rock mass of the high and steep slope have important significance for the overall construction of a hydraulic engineering project. Due to the severe terrain and inconvenient traffic, the manual measurement method is time-consuming and labor-consuming, and has great potential safety hazard. The invention provides a rapid identification method for dangerous rock masses with high steep slopes based on unmanned aerial vehicle photography technology, which is based on high-precision images obtained by unmanned aerial vehicle photography, can realize rapid identification and stable primary judgment of dangerous rock masses by applying density-based noise application space clustering (DBSCAN), a first-class Support Vector Machine (SVM), stress tensor reduction calculation and other theoretical methods to a generated point cloud model, and provides technical support for engineering construction.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed in the embodiment corresponds to the method disclosed in the embodiment, so that the description is simple, and the relevant points can be referred to the description of the method part.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (7)
1. A quick identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology is characterized by comprising the following steps:
1) Generating a point cloud model by using a high-definition photo of an unmanned aerial vehicle, and performing point cloud denoising and sparse processing;
2) Calculating general characteristic points of the point cloud model, and clustering the general characteristic points into a plurality of general characteristic point sets by using a noise application space clustering algorithm based on density;
3) Classifying the general characteristic point set by using a support vector machine to determine a boundary point set of the dangerous rock mass;
4) In the neighborhood of each boundary point set, fitting points on a slope where the dangerous rock mass is located into a plane to cut a point cloud model; deleting the intersection points between the planes and the point cloud model from the point cloud, and dividing the rest point cloud model into a plurality of independent point cloud areas by using a noise application space clustering algorithm based on density; determining a region representing the dangerous rock mass according to the geometric characteristics of each independent point cloud region;
5) And carrying out primary stability evaluation on the dangerous rock mass.
2. The rapid identification method for dangerous rock mass with high steep slope based on unmanned aerial vehicle photography technology as claimed in claim 1, wherein in step 1), the method for generating point cloud model by using unmanned aerial vehicle high definition photo and performing point cloud denoising and sparse processing comprises:
and generating a dense point cloud model through unmanned aerial vehicle photography and SFM technology, and denoising and sparse operation are carried out on the dense point cloud.
3. The rapid identification method for the high and steep slope dangerous rock mass based on the unmanned aerial vehicle photography technology as claimed in claim 1, wherein in the step 2), the calculation process of the general characteristic points is as follows:
s1, estimating the discrete main curvature and the main direction of each point according to the Weingarten mapping matrix property;
s2, judging whether the minimum principal curvature of the point on the normal section line can be a minimum value or not in a range of r meters away from the left side and the right side of each point along the direction of the minimum principal curvature of each point; if so, the point belongs to a general feature point.
4. The rapid identification method for high and steep slope dangerous rock mass based on unmanned aerial vehicle photography technology according to claim 1 or 3, wherein in the step 3), a support vector machine is used for classifying a general feature point set, and the method for determining the boundary point set of the dangerous rock mass comprises the following steps:
firstly, a plurality of typical dangerous rock masses are defined from a slope point cloud model through manual visual interpretation; meanwhile, selecting a point set belonging to the boundary of the dangerous rock mass in the general characteristic point set as sample data of SVM training; then, calculating characteristic parameters of all the general characteristic point sets; secondly, training the SVM by adopting a typical dangerous rock boundary point set, deducing the distance value from all general characteristic point sets to an SVM decision surface, and taking the distance as the fraction of each point set; finally, the minimum score of the sample is used as a threshold value, and the general characteristic point set with the score larger than the threshold value is used as the boundary point set of the dangerous rock mass.
5. The rapid identification method for dangerous rock mass with high steep slope based on unmanned aerial vehicle photography technology according to claim 4, characterized in that in the step 4), the point cloud model of the slope is cut, and the method for extracting the dangerous rock mass with high possibility comprises the following steps:
firstly, calculating a neighborhood point of each dangerous rock boundary point set according to a neighborhood algorithm; then, fitting points on the slope surface in the neighborhood points into a plane, and cutting the point cloud model; then, removing the intersection points of the planes and the model from the point cloud model, and carrying out density-based noise application spatial clustering on the residual point cloud, wherein the clustering result is a plurality of isolated point cloud areas; finally, calculating the volume-surface area ratio of a plurality of typical dangerous rock masses, and taking the minimum volume-surface area ratio as a threshold value; and comparing the threshold with the volume-surface area ratio of each isolated point cloud region obtained by clustering, and if the volume-surface area ratio of a certain region is greater than the threshold, determining that the isolated point cloud region is a dangerous rock mass region.
6. The rapid identification method for dangerous rock mass with high steep side slope based on unmanned aerial vehicle photography technology according to claim 5, characterized in that the field points of the dangerous rock mass boundary point set include both points on the surface of the dangerous rock mass and points on the slope where the dangerous rock mass is located.
7. The method for rapidly identifying the dangerous rock mass with the high steep slope based on the unmanned aerial vehicle photography technology as claimed in claim 1 or 5, wherein in the step 5), the method for primarily evaluating the stability of the dangerous rock mass comprises the following steps:
firstly, determining the inclination angle of the bottom sliding surface of each dangerous rock mass according to the analysis of the declination projection; secondly, projecting the surface of the dangerous rock body on a bottom sliding surface and dividing the surface of the dangerous rock body into grids with proper sizes, wherein each grid corresponds to a rock column from the surface of the dangerous rock body to the bottom sliding surface; then, estimating the ratio of shear stress and normal stress caused by the self weight of each rock column on the bottom sliding surface by using a stress tensor reduction theory so as to replace the average stress ratio of a grid area corresponding to the rock column; and finally, determining the position of the damaged grid on the bottom slip surface according to a molar-coulomb damage criterion, calculating the ratio of the area of all the damaged grids to the total area of the bottom slip surface, and grading the stability of the dangerous rock mass according to the ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911337549.9A CN111178214B (en) | 2019-12-23 | 2019-12-23 | High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911337549.9A CN111178214B (en) | 2019-12-23 | 2019-12-23 | High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111178214A CN111178214A (en) | 2020-05-19 |
| CN111178214B true CN111178214B (en) | 2023-04-18 |
Family
ID=70646317
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911337549.9A Active CN111178214B (en) | 2019-12-23 | 2019-12-23 | High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111178214B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112528746A (en) * | 2020-11-11 | 2021-03-19 | 中南大学 | Dangerous rock falling rock protective net setting method |
| CN114266987B (en) * | 2022-03-03 | 2022-05-20 | 水利部长江勘测技术研究所 | Intelligent identification method for high slope dangerous rock mass of unmanned aerial vehicle |
| CN115861296B (en) * | 2023-02-14 | 2023-05-16 | 湖北工业大学 | Automatic high-steep slope dangerous rock mass identification method and system based on unmanned aerial vehicle point cloud |
| CN117556703B (en) * | 2023-11-23 | 2024-06-14 | 中铁大桥局集团有限公司 | Method, device and equipment for identifying rock mass structural surface of side slope and readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529455A (en) * | 2013-10-21 | 2014-01-22 | 中铁第四勘察设计院集团有限公司 | Three-dimensional investigation method for dangerous falling rock based on air-borne laser radar |
| CN105180890A (en) * | 2015-07-28 | 2015-12-23 | 南京工业大学 | Rock structural surface occurrence measuring method integrated with laser-point cloud and digital imaging |
| CN106896213A (en) * | 2017-02-22 | 2017-06-27 | 中国地质大学（武汉） | A kind of rock mass discontinuity Intelligent Recognition and information extracting method based on cloud data |
| CN107123134A (en) * | 2017-05-03 | 2017-09-01 | 长安大学 | A kind of Dangerous Rock Body landslide monitoring method of feature based |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6817721B2 (en) * | 2016-05-20 | 2021-01-20 | アジア航測株式会社 | Topographic change analysis method |
| JP6771320B2 (en) * | 2016-06-23 | 2020-10-21 | 大成建設株式会社 | Surveying photography method for dam construction, finished form control chart generation method for dam construction using this, concrete placement amount calculation method, finished form management map generation system, concrete placement amount calculation system, and finished form control chart Generation support program |
| US20190026400A1 (en) * | 2017-07-18 | 2019-01-24 | Fuscoe Engineering, Inc. | Three-dimensional modeling from point cloud data migration |
| CA2986526C (en) * | 2017-07-31 | 2023-10-17 | Queen's University At Kingston | Autorotating unmanned aerial vehicle surveying platform |
| CN108020212A (en) * | 2017-12-06 | 2018-05-11 | 长江三峡勘测研究院有限公司（武汉） | A kind of small scale mapping method based on unmanned plane Yu CORS technologies |
| CN109685821A (en) * | 2018-12-26 | 2019-04-26 | 中国科学院大学 | Region growing 3D rock mass point cloud plane extracting method based on high quality voxel |
| CN110298103A (en) * | 2019-06-25 | 2019-10-01 | 中国电建集团成都勘测设计研究院有限公司 | The steep Dangerous Rock Body investigation method of height based on unmanned aerial vehicle onboard three-dimensional laser scanner |
-
2019
- 2019-12-23 CN CN201911337549.9A patent/CN111178214B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103529455A (en) * | 2013-10-21 | 2014-01-22 | 中铁第四勘察设计院集团有限公司 | Three-dimensional investigation method for dangerous falling rock based on air-borne laser radar |
| CN105180890A (en) * | 2015-07-28 | 2015-12-23 | 南京工业大学 | Rock structural surface occurrence measuring method integrated with laser-point cloud and digital imaging |
| CN106896213A (en) * | 2017-02-22 | 2017-06-27 | 中国地质大学（武汉） | A kind of rock mass discontinuity Intelligent Recognition and information extracting method based on cloud data |
| CN107123134A (en) * | 2017-05-03 | 2017-09-01 | 长安大学 | A kind of Dangerous Rock Body landslide monitoring method of feature based |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111178214A (en) | 2020-05-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111178214B (en) | High and steep slope dangerous rock mass rapid identification method based on unmanned aerial vehicle photography technology | |
| CN111091105B (en) | Remote sensing image target detection method based on new frame regression loss function | |
| US10235566B2 (en) | Determining stockpile volume based on digital aerial images and three-dimensional representations of a site | |
| CN111738165B (en) | Method for extracting individual plant canopy from high-resolution unmanned aerial vehicle visible light remote sensing image | |
| CN100538399C (en) | Cutting apart and classification and variety detection integration method of high-resolution remote sensing image | |
| US7995055B1 (en) | Classifying objects in a scene | |
| CN108038846A (en) | Transmission line equipment image defect detection method and system based on multilayer convolutional neural networks | |
| CN106780524A (en) | A kind of three-dimensional point cloud road boundary extraction method | |
| CN101901343A (en) | Remote sensing image road extracting method based on stereo constraint | |
| Jung et al. | Automated and efficient powerline extraction from laser scanning data using a voxel-based subsampling with hierarchical approach | |
| CN113065455B (en) | Landslide risk inspection method and system based on deep learning | |
| CN102903102A (en) | Non-local-based triple Markov random field synthetic aperture radar (SAR) image segmentation method | |
| CN106157283A (en) | The detection method of lane segmentation thing and device | |
| CN110363299B (en) | Spatial case reasoning method for outcrop rock stratum layering | |
| Seo et al. | Exploiting publicly available cartographic resources for aerial image analysis | |
| Kumar et al. | An automated road roughness detection from mobile laser scanning data | |
| Sun et al. | Building outline extraction from aerial imagery and digital surface model with a frame field learning framework | |
| Ghani et al. | Accuracy Assessment of Urban Growth Pattern Classification Methods Using Confusion Matrix and ROC Analysis | |
| Popescu et al. | Aerial image segmentation by use of textural features | |
| Guili et al. | A man-made object detection algorithm based on contour complexity evaluation | |
| Demir et al. | A nonparametric fuzzy shoreline extraction approach from Sentinel-1A by integration of RASAT pan-sharpened imagery | |
| Cui et al. | Identification of unstable bedrock promontory on steep slope based on UAV photogrammetry | |
| Harshit et al. | Geometric features interpretation of photogrammetric point cloud from Unmanned Aerial Vehicle | |
| CN109697754B (en) | 3D rock mass point cloud characteristic surface extraction method based on principal direction estimation | |
| Kim et al. | Building footprints extraction of dense residential areas from Lidar data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |