CN106709439A - Monocline structure landform automatic identification method - Google Patents

Abstract

The invention discloses an automatic identification method of monoclinic rock formation structure and landform, which comprises the following steps: (1) extracting bedrock stratum objects, establishing a relationship with the extracted adjacent objects according to the rules of adjacency, inclination and inclination consistency, and drawing Adjacency ARG graph; (2) Based on the adjacency ARG graph, the scene model is constructed according to the adjacency relationship, and the scene model is simplified; (3) The simplified scene model is screened according to the fold screening rules, and the model with an incomplete fold structure is retained. It is a monoclinic rock structure. The invention realizes the automatic recognition of the structure and topography of the monoclinic rock layer.

Description

A method for automatic identification of monoclinic stratum structure and geomorphology

technical field

The invention relates to the application field of geographic information technology, in particular to a method for automatically identifying monoclinic rock layer structure and geomorphology through spatial structure pattern matching based on digital geological maps.

Background technique

The horizontal rock layer is inclined due to the influence of the crustal changes, forming a certain angle with the horizontal plane, and the rock layer is inclined in the same direction, which is called monocline structure. The common monoclinic rock structure landforms include Zhubei Mountain and Shanmian Mountain. The monocline may appear in the damaged anticline wing, or around the destroyed dome structure, the periphery of the basin, tilted horizontal rock formations or tilted faults, etc. The monoclinic strata mainly manifests in three forms: one is a wing of a fold structure;

Traditional structural identification is mainly manual operation, which requires a large workload and takes a long time, and is limited by the technical and understanding level of the cartographers. The identification of monocline rock structure and geomorphology mainly depends on geological data or field survey results. Finding monoclines on geological maps is mainly to find a series of inclined strata on the side of folds or faults. Chen Ying et al. (Chen Ying, Li Anbo, Yao Mengmeng, et al. Automatic recognition of folded landform types based on spatial structure pattern matching[J]. Geo-Information Science Journal, 2016, 18(11)) through the method of spatial structure pattern matching, A method for automatic identification of anticline and syncline structural landforms based on digital geological surface data has been realized. However, this structural model only considers symmetrical and repetitive structures, and does not recognize the monoclinic rock structure and geomorphology well.

The formation of monoclinic rock structure and landform is closely related to the surrounding landform changes and structural evolution. The study of monocline rock structure and landform can provide effective support for geological map recognition and structural evolution reasoning. premise.

Contents of the invention

Purpose of the invention: The present invention aims at the problems existing in the prior art, and provides an automatic identification method of monoclinic strata structure and landform. This method adopts digital geological maps, and aims at the law of consistent dip of monocline strata, and basically realizes the automatic recognition of monocline strata structure and geomorphology in bedrock and mountainous areas by means of structural pattern matching. Compared with the traditional manual identification method, automatic identification based on geological maps is realized, and it also provides help for non-professionals to read and recognize geological maps.

Technical solution: The method for automatic identification of monoclinic rock structure and landform according to the present invention comprises the following steps:

(1) Extract bedrock stratum objects (excluding intrusive rock mass), establish a relationship with the extracted adjacent objects according to the rules of adjacency, inclination and inclination consistency, and draw an adjacency ARG graph (attributed relational graph);

(2) Based on the adjacency ARG graph, the scene model is constructed according to the adjacency relationship, and the scene model is simplified;

(3) Filter the simplified scene model according to the fold screening rules, and keep the model with incomplete fold structure, that is, the monocline rock structure.

Wherein, step (1) specifically includes:

(1-1) Set the angle threshold α as the basis for judging whether the difference between the two azimuth angles or strike angles is within the acceptable range, and generally take the value [22.5, 90); set the dip angle threshold β as the basis for judging whether the rock formation meets the dip standard The basis, the general value is [15, 90), and the specific value can be determined according to the user's needs;

(1-2) Load the stratigraphic element layer in vector format, and get all the surface element sets Stra={s _i |i=1,2,3,...,n}; where, s _i represents the i-th surface element, The surface elements include the serial number attribute Id, the stratigraphic age attribute Age, the occurrence dip attribute OccT and the occurrence dip attribute OccA, and n is the number of surface elements;

(1-3) Calculate the centroid points of surface elements respectively, and obtain the set of centroid points of stratigraphic elements OriGrav={og _i (xog _i ,yog _i )|i=1,2,3,...,n}; where, og _i represents the centroid point element of the i-th surface element, (xog _i , yog _i ) is the coordinate of og _i , and the centroid point element inherits the number attribute Id of the surface element, the stratigraphic age attribute Age, the occurrence tendency attribute OccT and the occurrence Inclination attribute OccA;

(1-4) Screening of centroid point elements: For the set OriGrav, according to the stratum age attribute Age, select the centroid point elements of non-Quaternary and non-intrusive rock mass formations, and create a centroid point element set Grav={g _i (xg _i ,yg _i )|i＝1,2,3,…,c}; among them, g _i is the retained i-th centroid element image, (xg _i ,yg _i ) is the coordinate of g _i , and c is the reserved The number of centroid point features;

(1-5) Create the relationship between stratum objects: for the centroid point element set Grav, take the centroid point elements g _i and g _j , j=1,2,3,...,c, j>i, and judge whether the following conditions are met:

a) The occurrence dip angle attribute OccA of the centroid point elements g _i and g _j is greater than or equal to the dip angle threshold β;

b) The stratigraphic age attributes Age of centroid point elements g _i and g _j are different;

c) The stratigraphic layer elements s _i and s _j corresponding to centroid point elements g _i and g _j are adjacent;

d) The difference between the occurrence tendency attribute OccT of centroid point elements g _i and g _j does not exceed the angle threshold α;

e) The difference between the direction of the line where the centroid elements g _i and g _j are located and the occurrence tendency attribute OccT of the centroid element g _i or g _j does not exceed the angle threshold α;

Create relation r _k <g _i , g _j > for centroid point elements g _i and g _j satisfying all the above conditions, and store relation r _k in the left

The number attribute Id of the object g _i and the right object g _j ;

(1-6) Complete the creation of the relationship between all centroid elements, and get the relationship set Rel={r _k <g _i , g _j >|k=1,2,3,...,m}, m is the created relationship quantity;

(1-7) Draw the vertex set according to the centroid point element, and draw the edge according to the relationship between the elements, then complete the adjacency ARG graph based on the centroid point element set Grav and the relationship set Rel.

Wherein, step (2) specifically includes:

(2-1) Based on the adjacency ARG graph, create a scene model for all connected edges based on the adjacency relationship, and obtain the scene model set OriM={OM _i |i=1,2,3,...,p}, OM _i represents the first There are i scene models, p is the number of scene models, and the scene model OM _i includes a set of vertices and a set of edges. For example, construct a scene model OM for edges r _u <g _i , g _j > and r _v <g _i , g _k >(u≠v) with adjacent relationships, and the scene model OM includes a set of vertices OMGra _v = {g _i , g ,g _k }, and the edge set OMRel={r _u ,r _v }, and calculate and record the number of edges g _i , g, g _k are 2, 1, 1 respectively;

(2-2) Find endpoints: In the vertex set of the scene model, there are two types of endpoints, simplified endpoints and ordinary endpoints. Follow the steps below to search for simplified endpoints and ordinary endpoints in turn:

1) Find vertices that satisfy all of the following conditions as simplified endpoints:

a) The number of edges passing through the vertex is greater than or equal to 2;

b) There is a situation with the same stratum age attribute in the set of endpoints connected to the vertex;

c) There is at least one pair of edges with an included angle less than 90° in the set of edges connected to the vertex;

2) Find the vertex that satisfies the condition that the number of connected edges is equal to 1 as a common endpoint;

(2-3) Recursively rebuild the model: use all simplified endpoints and unmarked common endpoints as starting endpoints in turn to recurse according to the following steps, and for scene models without simplified endpoints, directly use ordinary endpoints as starting endpoints to recurse according to the following steps ;

1) Add the starting endpoint g _i to the vertex chain set Link, and regard the other vertices g _ij connected to it as the children of g _i , j=1,2,3,...,t _i , t _i is the same as the end point g _i Number of connected vertices:

2) If the Link length<2, then add the vertex g _ij to the Link, and perform step 4) judgment on it; if the Link length>=2, then perform step 3) judgment on it;

3) Judging the vertices g _ij by rules, j=1,2,3,...,t _i : if the vertices g _ij satisfy the conditions at the same time ① is a simplified endpoint or is not marked; ② does not exist in the Link; ③ the last two points in the Link The difference between the azimuth angle of the edge composed of two vertices and the azimuth angle of the last vertex in the Link pointing to g _ij is less than the preset angle threshold α, then add the vertex g _ij to the Link, mark the vertex g _ij , and continue to step 4) to judge; otherwise, Then continue to judge the next child vertex g _i(j+1) in this step until all children of g _i are judged and skip to step 5);

4) Judging the endpoint of vertex g _ij : if the vertex g _ij is a simplified endpoint or an ordinary endpoint, the Link ends, and the model M is reconstructed according to the current Link set, and the current vertex g _ij is removed from the Link; otherwise, the vertex g ij is removed from the Link. g _ij is regarded as g _i , repeat step 2) judgment for its child g _ij ;

5) After traversing all the children of the current parent vertex g _i , remove the vertex g _i from the Link, and continue to repeat step 2) for the next vertex g _(i+1 ) or the next starting point until the last After the endpoint is judged, the recursion ends;

(2-4) Screen the recursive scene model according to the following rules:

1) Eliminate scene models that only contain two vertices and one edge in the model;

2) For multiple scene models reconstructed with the same simplified endpoint as the starting point or ending point, the following principles are used to eliminate them:

a) If the number of elements in the vertex set of the model is the same, and the stratum age attributes corresponding to the vertices in the set are the same respectively, then the corresponding scene model with a large stratum layer area is retained;

b) If the vertex sets of the models are not exactly the same, then keep the scene model with more set elements;

(2-5) Renumber the screened scene models to obtain a model set AllM={M ₁ ,M ₂ ,M ₃ ,...,M _q }, where q is the number of scene models.

Wherein, step (3) specifically includes:

(3-1) Voronoi unit division of the model: Carry out Voronoi unit division based on the scene model obtained in step (2) (Wang Xinsheng, Liu Jiyuan, Zhuang Dafang, etc. Approximation method of Voronoi diagram for arbitrary occurrence elements based on GIS[J]. Geography Science Advances, 2004, 23(4)), the divided Voronoi polygon unit inherits the model number attribute;

(3-2) Perform fold screening on the model: compare the scene models M _i and M _j in adjacent Voronoi polygonal units, and if the following conditions are met at the same time, it is determined that the two scene models form complete folds, and these two models are eliminated :

a) The vertex sets of the scene model M _i and M _j are consistent;

b) According to the consistent vertex order, the direction of the two scene models is opposite;

(3-3) The scene model retained after the fold screening is regarded as a monoclinic model, and the monoclinic model set FinM={FM _h |h=1,2,3,...,w} is obtained after the full image screening, where w is a monoclinic model The number of oblique models, FM _h represents the hth monoclinic model, FM _h includes the vertex set FMGrav _h and the edge set FMRel _h ;

(3-4) For each scene model in the screened model set FinM, merge the stratum level elements to obtain the corresponding single-slope elements to form a single-slope element set Poly={p _r | _r =1,2,3, …,w}, p _r is the single slope element of the scene model FM _r .

Beneficial effects: Compared with the prior art, the present invention has the remarkable advantages that: the present invention adopts digital geological maps, aims at the law of consistent dip of monocline formations, and realizes monocline single-cline strata in the bedrock mountainous area by means of structural pattern matching. Automatic identification of oblique strata structural landforms. Compared with the traditional manual identification method, automatic identification based on geological maps is realized, and it also provides help for non-professionals to read and recognize geological maps. The formation of monoclinic rock structure and landform is closely related to the surrounding landform changes and structural evolution. The study of monocline rock structure and landform can provide effective support for geological map recognition and structural evolution reasoning. premise.

Description of drawings

Fig. 1 is the flowchart of the inventive method;

Fig. 2 is the schematic diagram of embodiment geological surface layer data;

Fig. 3 is a schematic diagram of embodiment centroid point feature layer data;

Fig. 4 is a schematic diagram of the construction relationship of centroid point elements in the embodiment;

Fig. 5 is a schematic diagram of example (a) scene model layer (b) vertex set attribute (c) edge set attribute;

Figure 6 is an example diagram of a scene model in a complex situation

Fig. 7 is a schematic diagram of model data after simplified screening;

Fig. 8 is a schematic diagram of Voronoi subdivision of model data;

Figure 9 is a schematic diagram of a complete fold model;

Figure 10 is a schematic diagram of the identified single-slope layer data.

detailed description

The effect of the present invention will be further illustrated below in conjunction with the accompanying drawings and by describing an example of automatically identifying monoclinic rock formation structure and topography. Method flowchart of the present invention as shown in Figure 1, comprises the following steps:

(1) Adjacency ARG graph generation

Step 11: Set the angle threshold α to 45°, and the inclination threshold β to 15°;

Step 12: In this embodiment, the shp format data of the geological surface layer of Zijin Mountain in Nanjing is selected as an example, as shown in Figure 2, the layer element layer of this map in vector format is loaded, and all surface element sets Stra={s _i are obtained |i=1,2,3,...,78}. The serial number attribute Id, the stratigraphic age attribute Age, the occurrence dip attribute OccT and the occurrence dip attribute OccA of the storage area elements are stored. The data adopts the sixteen azimuth model to represent the tendency of occurrence, and converts it into a numerical value. Here, it is replaced by the central azimuth angle of the azimuth interval;

Step 13: Calculate the centroid points of the surface elements respectively, and obtain the set of centroid points of the stratum level elements OriGrav={og _i (xog _i , yog _i )|i=1,2,3,...,78}. The centroid point element inherits the serial number attribute Id of the surface element, the stratum age attribute Age, the occurrence tendency attribute OccT, and the occurrence dip angle attribute OccA;

Step 14: Screen the centroid points, and select the centroid point elements of non-Quaternary and non-intrusive rock mass strata according to the stratum age attribute Age, and obtain the centroid point object set Grav={g _i (xg _i ,yg _i )|i= 1,2,3,...,62}, the filtered centroid point object layer is shown in Figure 3;

Step 15: Create relationships between stratigraphic objects. For the set of centroid point objects Grav, take the centroid point objects g _i and g _j (j=1,2,3,...,62, j>i) to judge whether the following conditions are met:

a) The occurrence dip angle attribute OccA of the centroid point elements g _i and g _j is greater than or equal to the dip angle threshold β;

b) The stratigraphic age attributes Age of centroid point elements g _i and g _j are different;

c) The stratigraphic layer elements s _i and s _j corresponding to centroid point elements g _i and g _j are adjacent;

d) The difference between the occurrence tendency attribute OccT of centroid point elements g _i and g _j does not exceed the angle threshold α;

e) The difference between the direction of the line where the centroid elements g _i and g _j are located and the occurrence tendency attribute OccT of the centroid element g _i or g _j does not exceed the angle threshold α;

Create relation r _k <g _i , g _j > for centroid point elements g _i and g _j satisfying all the above conditions, and store relation r _k in the left

The numbering attribute Id of the object g _i and the right object g _j , get the set Rel={r _k <g _i , g _j >|k=1,2,3,...,21}, according to the quality

The heart point object and relationship collection draw vertices and edges to complete the ARG graph, as shown in Figure 4.

(2) Scene model construction and simplification

Step 21: Based on the ARG graph, create a scene model based on the adjacent relationship. Taking the model OM ₅ as an example (as shown in Figure 5(a)), its vertex set OMGrav ₅ includes 8 objects OMGrav ₅ = {g ₂₃ , g ₂₄ , g ₂₇ , g ₂₈ , g ₃₀ , g ₃₁ , g ₃₂ , g ₃₃ }, as shown in Figure 5(b), its edge set OMRel ₅ includes 7 edges OMRel ₅ = {r ₉ <g ₂₃ , g ₂₈ >, r ₁₀ <g ₂₃ , g ₃₀ >, r ₁₁ <g ₂₃ , g ₃₁ >, r ₁₂ <g ₂₃ , g ₃₃ >, r ₁₃ <g ₂₇ , g ₂₈ >, r ₁₄ <g ₃₁ , g ₃₂ >, r ₁₅ <g ₂₄ , g ₂₇ >}, attribute As shown in Figure 5(c). The number of edges connected by vertices in the model is 4, 1, 2, 2, 1, 2, 1, 1;

Step 22: Construct the scene model of the whole image, and obtain the scene model set OriM={OM _i |i=1,2,3,...,8};

Step 23: Simplify the scene model. There may be cases where the structure of the scene model is complicated due to the influence of later faults (as shown in Figure 6), and the scene model needs to be simplified. Taking the scene model OM ₅ as an example, the process of simplifying the scene model is described in steps 24-26. .

Step 24: Find Endpoints.

1) In the vertex set OMGrav ₅ of the scene model OM ₅ , the vertex g ₂₃ satisfies the following conditions, then the vertex g ₂₃ is a simplified endpoint:

a) The number of edges passing through the vertex g ₂₃ is 4, greater than 2;

b) In the set of vertices connected to vertex g ₂₃ , the stratigraphic age attributes of g ₃₁ and _g33 are consistent

c) There are edges with angles less than 90° in all edge sets including vertex g ₂₃ ;

2) Secondly, according to the condition that the number of edges passed through is equal to 1, common endpoints are searched for, then vertices g ₂₄ , g ₃₀ , g ₃₂ , and g ₃₃ are common endpoints;

Step 25: Recursively rebuild the model. Create a vertex chain set Link, and perform recursion on the simplified endpoint g ₂₃ of the scene model. First, the simplified endpoint g ₂₃ is added to Link, and the other vertices g _23j (j=1,2,3,…,4) connected by edges are g ₂₈ , g ₃₀ , g ₃₁ , g ₃₃ respectively, and the recursion is as follows:

1) At this time, only g ₂₃ is included in the Link, and the length is less than 2, then the vertex g ₂₈ is added to the Link, and g ₂₈ is marked;

2) judge the endpoint of vertex g ₂₈ , g ₂₈ is neither a simplified endpoint nor a common endpoint, then find its other vertices g _28j (j=1,2) connected by edges are g ₂₃ , g ₂₇ respectively;

3) At this time, the Link contains g ₂₃ and g ₂₈ , and the length is equal to 2, then the rule judgment is performed on the vertex g ₂₃ . If the vertex g _ij satisfies the conditions at the same time ① is a simplified endpoint or is not marked; ② does not exist in the Link; ③ the difference between the azimuth angle of the edge formed by the last two vertices in the Link and the azimuth angle of the last vertex in the Link pointing to g _ij is less than If the angle threshold α is preset, the vertex g _ij is added to Link, and the vertex g _ij is marked. For the vertex g ₂₃ , it is a simplified endpoint and already exists in the Link, so the condition ② is not satisfied, skip this point, and then make a rule judgment on g ₂₇ ;

4) For the vertex g ₂₇ , which is not marked and does not exist in the Link, the calculated angle from g ₂₃ to g ₂₈ is 341.3°, and the angle from g ₂₈ to g ₂₇ is 90.3°. The difference between the two angles is greater than 45°. If the condition ③ is met, then skip this point, and the child of g ₂₈ has finished judging at this time;

5) Vertex g ₂₈ is removed from the Link. At this time, the Link only contains g ₂₃ and the length is less than 2, then the vertex g ₃₀ is added to the Link, marking g ₃₀ ;

6) Judge the endpoint of vertex g ₃₀ , if g ₃₀ is a common endpoint, then the Link ends, and a new model M is created, MGrav={g ₂₃ , g ₃₀ }, MRel={r ₁₀ <g ₂₃ , g ₃₀ >}. Remove g ₃₀ from the Link. At this time, the Link only contains g ₂₃ and the length is less than 2, then add the vertex g ₃₁ to the Link and mark g ₃₁ ;

7) To judge the endpoint of vertex g ₃₁ , g ₃₁ is neither a simplified endpoint nor a common endpoint, then find its other vertices g _31j (j=1, 2) connected by edges as g ₂₃ and g ₃₂ respectively;

8) At this time, the Link contains g ₂₃ and g ₃₁ , and the length is equal to 2, then the rule judgment is performed on the vertex g ₂₃ . For the vertex g ₂₃ , which is a simplified endpoint, if there is an unsatisfied condition ② in the Link, skip this point, and then make a rule judgment on g ₃₂ ;

9) For the vertex g ₃₂ , which is unmarked and does not exist in the Link, the calculated angle from g ₂₃ to g ₃₁ is 251.4°, and the angle from g ₃₁ to g ₃₂ is 256.0°, and the difference between the two angles is less than the threshold value of 45°. Then add vertex g ₃₂ to Link, and mark g ₃₂ ;

10) Judging the endpoint of vertex g ₃₂ , g ₃₂ is a common endpoint, then the Link ends, and a new model M is created, MGrav={g ₂₃ , g ₃₁ , g ₃₂ }, MRel={r ₁₁ <g ₂₃ , g ₃₁ > , r ₁₄ <g ₃₁ , g ₃₂ >}. Remove g ₃₂ from the Link, at this point, the judgment of the child of g ₃₁ ends;

11) Vertex g ₃₁ is removed from the Link. At this time, only g ₂₃ is included in the Link, and the length is less than 2, then the vertex g ₃₃ is added to the Link, and g ₃₃ is marked;

12) Judge the endpoint of vertex g ₃₃ , if g ₃₃ is a common endpoint, then the Link ends, and a new model M is created, MGrav={g ₂₃ , g ₃₃ }, MRel={r ₁₂ <g ₂₃ , g ₃₃ >}. Remove g ₃₃ from Link, at this time, the judgment of the child of g ₂₃ ends;

13) Traverse unmarked common endpoints. According to the judging rules and procedures, judge the unmarked child whose common end point is g ₂₄ , and obtain the model M, which includes the vertex set MGrav={g ₂₄ , g ₂₇ , g ₂₈ }, and the edge set MRel={r ₁₃ < g ₂₇ , g ₂₈ >, r ₁₅ <g ₂₄ , g ₂₇ >};

14) The common endpoints g ₃₀ , g ₃₂ , and g ₃₃ are all marked and no longer traversed, and the recursion of the current scene model ends. The sample scene model OriM ₅ is simplified to get 4 models as follows:

M ₁ ={MGrav ₁ , MRel ₁ }, MGrav ₁ ={g ₂₃ , g ₃₀ }, MRel ₁ ={r ₁₀ <g ₂₃ , g ₃₀ >}

M ₂ ={MGrav ₂ , MRel ₂ }, MGrav ₂ ={g ₂₃ , g ₃₁ , g ₃₂ }, MRel ₂ ={r ₁₁ <g ₂₃ , g ₃₁ >, r ₁₄ <g ₃₁ , g ₃₂ >}

M ₃ ={MGrav ₃ , MRel ₃ }, MGrav ₃ ={g ₂₃ , g ₃₃ }, MRel ₃ ={r ₁₂ <g ₂₃ , g ₃₃ >}

M ₄ ={MGrav ₄ , MRel ₄ }, MGrav ₄ ={g ₂₄ , g ₂₇ , g ₂₈ }, MRel ₄ ={r ₁₃ <g ₂₇ , g ₂₈ >, r ₁₅ <g ₂₄ , g ₂₇ >}

Step 26: Screen the 4 simplified models of the scene model OM ₅ according to the rules. Among them, the models M ₁ and M ₃ only contain two vertices and one edge, then remove the models M ₁ and M ₃ , and finally keep M ₂ and M ₄ two models;

Step 27: After simplifying and filtering the whole image, 5 models are obtained, and the renumbered models are marked as shown in Figure 7.

(3) Monoclinic pattern matching and screening

Step 31: Carry out Voronoi unit subdivision according to the combed model, and the subdivided polygon inherits the model number attribute, and the generated Voronoi polygon element is shown in Figure 8;

Step 32: Screen the adjacent models in the Voronoi diagram. Compare the scene models M _i and M _j in the adjacent Voronoi polygonal units, and if the following conditions are met at the same time, it is determined that the two scene models form a complete fold, and these two models are eliminated:

a) The vertex sets of the scene model M _i and M _j are consistent;

b) According to the consistent vertex order, the direction of the two scene models is opposite (as shown in Figure 9);

There are no complete folds in the example, and there are no monoclinic objects that need to be eliminated. Therefore, the final model set FinM={FM _h |h=1,2,3,...,5} is obtained, which is consistent with the number of model M;

Step 33: For the model FM _h that satisfies the monocline structure, merge the stratum layer element objects corresponding to the vertex set to create a new monocline element p _h . Merge the elements of the stratum layer on the filtered model set FinM to obtain the single-slope element set Poly={p _h |h=1,2,3,...,5}, as shown in Figure 10.

What is disclosed above is only a preferred embodiment of the present invention, which cannot limit the scope of rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope of the present invention.

Claims (4)

1. an automatic identification method of monoclinic strata structure landform, it is characterized in that the method comprises the following steps: (1) Extract bedrock stratum objects, establish relationships with the extracted adjacent objects according to the rules of adjacency, inclination and dip consistency, and draw an adjacency ARG diagram; (2) Based on the adjacency ARG graph, the scene model is constructed according to the adjacency relationship, and the scene model is simplified; (3) Filter the simplified scene model according to the fold screening rules, and keep the model with incomplete fold structure, that is, the monocline rock structure. 2. the automatic recognition method of monoclinic strata structure topography according to claim 1, is characterized in that: step (1) specifically comprises: (1-1) The angle threshold α is set as the basis for judging whether the difference between two azimuth angles or strike angles is within an acceptable range, and the dip angle threshold β is set as the basis for judging whether the rock formation reaches the inclination standard; (1-2) Load the stratigraphic element layer in vector format, and get all the surface element sets Stra={s _i |i=1,2,3,...,n}; where, s _i represents the i-th surface element, The surface elements include the serial number attribute Id, the stratigraphic age attribute Age, the occurrence dip attribute OccT and the occurrence dip attribute OccA, and n is the number of surface elements; (1-3) Calculate the centroid points of surface elements respectively, and obtain the set of centroid points of stratigraphic elements OriGrav={og _i (xog _i ,yog _i )|i=1,2,3,...,n}; where, og _i represents the centroid point element of the i-th surface element, (xog _i , yog _i ) is the coordinate of og _i , and the centroid point element inherits the number attribute Id of the surface element, the stratigraphic age attribute Age, the occurrence tendency attribute OccT and the occurrence Inclination attribute OccA; (1-4) Screening of centroid point elements: For the set OriGrav, according to the stratum age attribute Age, select the centroid point elements of non-Quaternary and non-intrusive rock mass formations, and create a centroid point element set Grav={g _i (xg _i ,yg _i )|i＝1,2,3,…,c}; among them, g _i is the retained i-th centroid element image, (xg _i ,yg _i ) is the coordinate of g _i , and c is the reserved The number of centroid point features; (1-5) Create the relationship between stratum objects: for the centroid point element set Grav, take the centroid point elements g _i and g _j , j=1,2,3,...,c, j>i, and judge whether the following conditions are met: a) The occurrence dip angle attribute OccA of the centroid point elements g _i and g _j is greater than or equal to the dip angle threshold β; b) The stratigraphic age attributes Age of centroid point elements g _i and g _j are different; c) The stratigraphic layer elements s _i and s _j corresponding to centroid point elements g _i and g _j are adjacent; d) The difference between the occurrence tendency attribute OccT of centroid point elements g _i and g _j does not exceed the angle threshold α; e) The difference between the direction of the line where the centroid elements g _i and g _j are located and the occurrence tendency attribute OccT of the centroid element g _i or g _j does not exceed the angle threshold α; Create a relationship r _k <g _i , g _j > for the centroid point elements g _i and g _j that meet all the above conditions, and store the number attribute Id of the left object g _i and the right object g _j in the relationship r _k ; (1-6) Complete the creation of the relationship between all centroid elements, and get the relationship set Rel={r _k <g _i , g _j >|k=1,2,3,...,m}, m is the created relationship quantity; (1-7) Draw the vertex set according to the centroid point element, and draw the edge according to the relationship between the elements, then complete the adjacency ARG graph based on the centroid point element set Grav and the relationship set Rel. 3. the automatic identification method of monoclinic strata structure topography according to claim 1, is characterized in that: step (2) specifically comprises: (2-1) Based on the adjacency ARG graph, create a scene model for all connected edges based on the adjacency relationship, and obtain the scene model set OriM={OM _i |i=1,2,3,...,p}, OM _i represents the first i scene models, p is the number of scene models, and the scene model OM _i includes a set of vertices and a set of edges; (2-2) Find endpoints: In the vertex set of the scene model, there are two types of endpoints, simplified endpoints and ordinary endpoints. Follow the steps below to search for simplified endpoints and ordinary endpoints in turn: 1) Find vertices that satisfy all of the following conditions as simplified endpoints: a) The number of edges passing through the vertex is greater than or equal to 2; b) There is a situation with the same stratum age attribute in the set of endpoints connected to the vertex; c) There is at least one pair of edges with an included angle less than 90° in the set of edges connected to the vertex; 2) Find the vertex that satisfies the condition that the number of connected edges is equal to 1 as a common endpoint; (2-3) Recursively rebuild the model: use all simplified endpoints and unmarked common endpoints as starting endpoints in turn to recurse according to the following steps, and for scene models without simplified endpoints, directly use ordinary endpoints as starting endpoints to recurse according to the following steps ; 1) Add the starting endpoint g _i to the vertex chain set Link, and regard the other vertices g _ij connected to it as the children of g _i , j=1,2,3,...,t _i , t _i is the same as the end point g _i Number of connected vertices: 2) If the Link length<2, then add the vertex g _ij to the Link, and perform step 4) judgment on it; if the Link length>=2, then perform step 3) judgment on it; 3) Judging the vertices g _ij by rules, j=1,2,3,...,t _i : if the vertices g _ij satisfy the conditions at the same time ① is a simplified endpoint or is not marked; ② does not exist in the Link; ③ the last two points in the Link The difference between the azimuth angle of the edge composed of two vertices and the azimuth angle of the last vertex in the Link pointing to g _ij is less than the preset angle threshold α, then add the vertex g _ij to the Link, mark the vertex g _ij , and continue to step 4) to judge; otherwise, Then continue to judge the next child vertex g _i(j+1) in this step until all children of g _i are judged and skip to step 5); 4) Judging the endpoint of vertex g _ij : if the vertex g _ij is a simplified endpoint or an ordinary endpoint, the Link ends, and the model M is reconstructed according to the current Link set, and the current vertex g _ij is removed from the Link; otherwise, the vertex g ij is removed from the Link. g _ij is regarded as g _i , repeat step 2) judgment for its child g _ij ; 5) After traversing all the children of the current parent vertex g _i , remove the vertex g _i from the Link, and continue to repeat step 2) for the next vertex g _(i+1 ) or the next starting point until the last After the endpoint is judged, the recursion ends; (2-4) Screen the recursive scene model according to the following rules: 1) Eliminate scene models that only contain two vertices and one edge in the model; 2) For multiple scene models reconstructed with the same simplified endpoint as the starting point or ending point, the following principles are used to eliminate them: a) If the number of elements in the vertex set of the model is the same, and the stratum age attributes corresponding to the vertices in the set are the same respectively, then the corresponding scene model with a large stratum layer area is retained; b) If the vertex sets of the models are not exactly the same, then keep the scene model with more set elements; (2-5) Renumber the screened scene models to obtain a model set AllM={M ₁ ,M ₂ ,M ₃ ,...,M _q }, where q is the number of scene models. 4. the automatic recognition method of monoclinic strata structure topography according to claim 1, is characterized in that: step (3) specifically comprises: (3-1) Model Vonoroi unit division: carry out Voronoi unit division according to the scene model that step (2) obtains, the Voronoi polygonal unit after the division inherits the model serial number attribute; (3-2) Perform fold screening on the model: compare the scene models M _i and M _j in adjacent Voronoi polygonal units, and if the following conditions are met at the same time, it is determined that the two scene models form complete folds, and these two models are eliminated : a) The vertex sets of the scene model M _i and M _j are consistent; b) According to the consistent vertex order, the direction of the two scene models is opposite; (3-3) The scene model retained after the fold screening is regarded as a monoclinic model, and the monoclinic model set FinM={FM _h |h=1,2,3,...,w} is obtained after the full image screening, where w is a monoclinic model The number of oblique models, FM _h represents the hth monoclinic model, FM _h includes the vertex set FMGrav _h and the edge set FMRel _h ; (3-4) For each scene model in the screened model set FinM, merge the stratum level elements to obtain the corresponding single-slope elements to form a single-slope element set Poly={p _h|h =1,2,3, …,w}, _{ph is the single-slope element of the scene model FM h .}