A method for determining directional relationships between geographic entities
By switching between ADR and DRM models under different distance conditions and selecting reference objects based on the area of the minimum boundary rectangle, the problem of accuracy in determining the directional relationship between geographic entities was solved, and accurate judgment was achieved under different conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Chinese People's Liberation Army Cyberspace Force Information Engineering University
- Filing Date
- 2023-02-08
- Publication Date
- 2026-05-26
Smart Images
Figure CN116448051B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of geographic information technology, and specifically relates to a method for determining the directional relationship between geographic entities. Background Technology
[0002] The directional relationships between geographic entities are an important aspect of people's understanding of geospatial space. They are a fundamental and important type of spatial relationship, expressing the sequential relationships between geographic entities, and mainly include East (E), Southeast (SE), South (S), Southwest (SW), West (W), Northwest (NW), North (N), and Northeast (NE). Currently commonly used directional relationship models include the cone model, the centroid model, the MBR model, and the directional relationship matrix model.
[0003] like Figure 1 The image shows the Direction Relation Matrix (DRM) model proposed by Goyal RK. This DRM model can formally express the directional relationships between geographic entities. Figure 1 In this model, B is the reference object and A is the target object. The orientation of entity A is defined as the union of the regions of entity A within the extended rectangle bounded by B, such as... Figure 1 As shown, regions A, SE, and E all intersect. Therefore, the region with the largest intersection area is selected as the final location of A, meaning A is southeast of B. However, when the reference object and the target object are close, the DRM model may result in discrepancies between the predicted and actual results due to uneven shape distribution of the research objects. Figure 2 As shown, with province A as the reference and province B as the target, province B overlaps with the north, northeast, and east. The overlapping areas, from largest to smallest, are ③, ①, and ②. However, based on the direction relationship matrix, province B is judged to be located east of province A, which is obviously inconsistent with the actual situation.
[0004] Furthermore, in addition to the inaccuracy of the judgment results when using the DRM model to determine the directional relationship between geographic entities when the reference object and the target object are close, there are also cases where the judgment results are inaccurate when the distance is far and the shape distribution of the research object is uneven or the area of the two research objects differs greatly. Summary of the Invention
[0005] The purpose of this invention is to provide a method for determining the directional relationship between geographic entities, in order to solve the problem that when determining the directional relationship between geographic entities, if the reference object and the target object are close, the judgment method using the DRM model may result in inaccurate judgment.
[0006] To address the aforementioned technical problems, this invention provides a method for determining the directional relationship between geographic entities, comprising the following steps:
[0007] 1) Determine the distance between two geographic entities according to the set rules; the distance includes two cases: close distance and far distance;
[0008] 2) When the distance is close, the directional relationship model is used to determine the directional relationship between geographic entities and obtain the directional relationship between the two geographic entities.
[0009] Its beneficial effects are as follows: The method of the present invention uses an angle-direction relationship model to determine the directional relationship between two geographical entities when the two research objects are close to each other. Because the angle-direction relationship model can take into account the influence of the concave and convex shapes of the entities on the directional relationship determination, even when there is uneven distribution of the shape of the research objects themselves, the directional relationship between the two research objects can be accurately determined when they are close to each other. This effectively avoids the situation where the judgment result is wrong due to the uneven distribution of the shape of the research objects themselves when using the DRM model in this case.
[0010] Furthermore, in step 2), when the distance is far, the directional relationship matrix model is used to determine the directional relationship between geographic entities to obtain the directional relationship between the two geographic entities.
[0011] The angle-direction relationship model determines the directional relationship between entities by angle, which is easily affected by the distance between entities. When the distance between two entities is large, the judgment result may be inaccurate. Therefore, this invention uses the DRM model to make judgments when the two research objects are far apart. The judgment result of this DRM model is not affected by the distance between the two entities. Therefore, the process of using the DRM model to make judgments when the distance is large can effectively avoid the inaccuracy of the angle-direction relationship model due to the distance between the two research objects.
[0012] Furthermore, the reference object and the target object in the directional relationship matrix model are determined from the two geographic entities according to the selection rules.
[0013] Furthermore, the selection rule is as follows: compare the area of the minimum bounding rectangle formed by two geographic entities, take the geographic entity with the smaller area of the minimum bounding rectangle as the reference object, and take the other geographic entity as the target object.
[0014] This invention also takes into account the problem that the DRM model is susceptible to the shape distribution of the research object itself. When two research objects are far apart, the size of the minimum boundary rectangle area of the two research objects is first determined, and the research object with the smaller minimum boundary rectangle area is used as the reference object to ensure the accuracy of the judgment result.
[0015] Further, in step 1), the setting rule is: whether the minimum boundary rectangles formed by two geographic entities overlap; when there is overlap, the distance between the two geographic entities is close, and when there is no overlap, the distance between the two geographic entities is far.
[0016] That is, using the minimum boundary rectangle to delineate two geographic entities has already placed the two geographic entities within their respective relatively small ranges. If there is still an overlapping part in the regions of the two geographic entities constructed in this way, it proves that the distance between the two geographic entities is relatively close. Therefore, the method of the present invention can accurately determine the degree of distance between two geographic entities.
[0017] Further, in step 2), the angle and direction relationship model calculates the average value of the tilt angle of the line connecting the grid points in one geographic entity to the grid points in the other geographic entity by creating grid points covering the entire geographic entity region to be judged, obtains the angle value between the two geographic entities based on the average value of this tilt angle, and uses the angle and direction relationship model to determine the directional relationship between the two geographic entities.
[0018] By constructing grid points on two geographic entities whose directional relationship needs to be determined, and then connecting all grid points within one geographic entity's region with all grid points within the other geographic entity's region with straight lines, the average tilt angle of each straight line is obtained. The average value of this tilt angle is then used to reflect the directional relationship between the two geographic entities. This angular directional relationship model can take into account the influence of the entity's concave and convex shape on the directional relationship judgment, thus avoiding the problem of incorrect judgment results caused by the uneven shape of the geographic entity itself.
[0019] Furthermore, the average value θ of the tilt angle is obtained by the following formula:
[0020]
[0021] in, The angle of inclination is the line connecting grid points in one geographic entity to grid points in another geographic entity, and n is the product of the number of grid points in one geographic entity and the number of grid points in another geographic entity.
[0022] Furthermore, the angle value is obtained using the following formula:
[0023]
[0024] Where θ is the average value of the tilt angle. The angle of inclination is the line connecting grid points in one geographic entity to grid points in another geographic entity, and n is the product of the number of grid points in one geographic entity and the number of grid points in another geographic entity.
[0025] Furthermore, the angle-direction relationship model is an eight-direction cone model.
[0026] This invention achieves the determination of directional relationships between geographic entities based on an angle-direction relationship model through the above process. Furthermore, the angle-direction relationship model of this invention standardizes the calculated average tilt angle to satisfy the angle addition and subtraction rules. Therefore, based on the angle value obtained after this processing, the directional relationship between two geographic entities can be accurately determined by using an eight-direction cone model.
[0027] Furthermore, the directional relationship matrix model establishes a directional domain centered on the reference object by taking two geographic entities as the reference object and the target object, respectively. The directional domain with the largest overlap area between the target object and the directional domain is taken as the final directional relationship between the target object and the reference object. Attached Figure Description
[0028] Figure 1 This is a schematic diagram of the directional relationship matrix model;
[0029] Figure 2 This is a schematic diagram showing the location of Province B determined using a directional relationship matrix model with Province A as the reference.
[0030] Figure 3 This is a schematic diagram of the angular direction relationship model in this embodiment;
[0031] Figure 4 This is a schematic diagram of the eight-directional cone model of this embodiment;
[0032] Figure 5-1 This is a schematic diagram of the results obtained using the ADR model in this embodiment;
[0033] Figure 5-2 This is in this embodiment Figure 5-1 The example diagrams used;
[0034] Figure 6-1 This is a schematic diagram showing the orientation of street 03 determined using a directional relationship matrix model with street 16 as the reference object.
[0035] Figure 6-2 This is a schematic diagram showing the orientation of street 16 determined using a direction relationship matrix model with street 03 as the reference object.
[0036] Figure 7-1 This is a schematic diagram showing the orientation of Street 21 determined using a directional relationship matrix model with Street 16 as the reference.
[0037] Figure 7-2 This is a schematic diagram showing the orientation of Street 16 determined using a directional relationship matrix model with Street 21 as the reference.
[0038] Figure 8 This is a schematic diagram of the method flow in this embodiment;
[0039] Figure 9 This is a map of the experimental area used in the method described in this embodiment. Detailed Implementation
[0040] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.
[0041] Example of a method for determining directional relationships between geographic entities:
[0042] In order to overcome the problem of inaccurate results when using the DRM model to determine the directional relationship between the target object and the reference object when the reference object and the target object are close, this embodiment first determines the distance between the target object and the reference object when determining the directional relationship between geographic entities. If the target object and the reference object are close, the Angle Direction Relation (ADR) model is used to determine the directional relationship between the target object and the reference object.
[0043] The angular orientation relationship model refers to creating a uniform grid of points within the object of study to represent its positional distribution. The orientation angles of each grid point within the target object relative to each grid point within the reference object are... like Figure 3 As shown, the directional relationship between the target object and the reference object can be represented by the average value θ of the orientation angles of each grid point, as shown in equation (1):
[0044]
[0045] In equation (1), In this embodiment, the angle of inclination is the line connecting the grid points of the reference object and the grid points of the target object. pass The calculations show that (x1, y1) are the coordinates of a point in the reference object, and (x2, y2) are the coordinates of a point in the target object.
[0046] In this embodiment, the obtained average angle θ is standardized as shown in equation (2) to satisfy the angle addition and subtraction operation rules.
[0047]
[0048] Finally, the final direction is determined based on the model of the relationship between the corresponding angle and direction. In this embodiment, the following method is used: Figure 4 The final direction is obtained by determining the angle range of the eight-directional cone model shown.
[0049] The ADR model can consider the influence of the concave and convex shapes of entities on the determination of directional relationships, and can satisfy the relative relationship between the reference object and the target object, enabling reverse reasoning of relationships. Therefore, even when the distance between the reference object and the target object is relatively close, the directional relationship between geographic entities can still be accurately determined. Thus, the method in this embodiment first determines the distance relationship between two geographic entities when determining the directional relationship between geographic entities. When the distance is relatively close, the ADR model is directly applied to determine the directional relationship between geographic entities, thereby avoiding inaccurate judgment results caused by uneven distribution of research objects.
[0050] Because the ADR model determines the directional relationship of entities through angles, it is easily affected by the position of the entities, such as... Figure 5-1 The diagram shows the results obtained from the ADR model. The angles between object O and the directions of objects A and B are θ1 and θ2, respectively. Objects A and B are roughly on the same horizontal line, both located northeast of O. Object B is farther from object O than object A, therefore object B is located further east of object O. Figure 5-1 For example Figure 5-2 As shown, by Figure 5-2 It can be seen that both Street 15 and Street 03 are located northeast of Street 11. This result contradicts the result obtained by the ADR model, which states that Street 15 is located northeast of Street 11 (35°) and Street 03 is located east of Street 11 (17°). In other words, the result obtained by the ADR model is inconsistent with the actual situation.
[0051] Therefore, in this embodiment, to avoid inaccuracies in the results of using the ADR model when the distance between two research objects is large, the ADR model is not used to determine the directional relationship between geographic entities in this case. Instead, the DRM model, which is unaffected by distance, is used. The directional relationship matrix is divided into a coarse directional relationship matrix and a precise directional relationship matrix. The coarse directional relationship matrix is represented by equation (3):
[0052]
[0053] If the target object intersects with a certain direction, the corresponding matrix element value is T; otherwise, it is F. Figure 1 The result of equation (3) shown is: Then, the orientation relationship is described according to the orientation relationship matrix; the precise orientation relationship matrix is shown in equation (4):
[0054]
[0055] When a target object intersects with multiple directional regions, the overlapping area between the target object and each directional region is calculated, and the corresponding matrix element value is the ratio of the overlapping area to the total area of the target object. This model is based on the shape, size, and other characteristics of the entity itself, is easy to calculate and derive, and is currently a widely used directional relationship model.
[0056] However, when using this model, regardless of whether the two geographical entities are close or far apart, inaccurate results can occur if the shapes of the objects are unevenly distributed or if their areas differ significantly. Specifically, this is because directional relationships are relative spatial relationships. Generally, if the object is located in a certain direction of a reference object, the reference object should be located in the opposite direction of the target object; that is, directional relationships are deducible. However, this directional relationship matrix model also presents two situations that lead to inaccurate results when the two objects are far apart:
[0057] Scenario 1: The shape of the research object is not uniformly distributed.
[0058] In this case, an example of using the DRM model for judgment is... Figure 6-1 , Figure 6-2 As shown, due to the uneven distribution of the shape of street 16, the ratio of the actual area of street 16 to the area of its MBR (minimum bounding rectangle) is relatively small. When the reference object is street 16 and the target object is street 03, the result is that street 03 is located to the east of street 16. Figure 6-1 (as shown), and Figure 6-2 As shown, when the reference object is street 03 and the target object is street 16, the result is that street 16 is in the northwest direction of street 03. That is, when different research objects are used as reference objects, the two results are not opposite. Therefore, the two results of the DRM model in this case are not derivable. However, there is a result that can correctly reflect the directional relationship between street 16 and street 03. That is, when street 03 is used as the reference street and street 16 is used as the target street, the result that street 16 is in the northwest direction of street 03 is the correct result.
[0059] Scenario 2: The research objects have uneven shape distribution and their areas differ significantly.
[0060] In this case, an example of using the DRM model for judgment is... Figure 7-1 , Figure 7-2As shown, due to the uneven distribution of the shape of street 16 and the significant difference in area between street 16 and street 21, when the reference object is street 16 and the target object is street 21, the result is that street 21 is northeast of street 16 (e.g., Figure 7-1 (as shown), and Figure 7-2 As shown, when the reference object is street 21 and the target object is street 16, the result is that street 16 is to the west of street 21. That is, when different research objects are used as reference objects, the two results are not opposite. Therefore, the two results of the DRM model in this case are not derivable. However, there is a result that can correctly reflect the directional relationship between street 16 and street 21. That is, when street 16 is used as the reference object and street 21 is used as the target object, the result that street 21 is to the northeast of street 16 is correct.
[0061] Based on the analysis of Case 1 and Case 2, although the DRM model may lead to situations where the directional relationship between the two research objects is not opposite when the shape distribution of the research objects is uneven or the area of the reference object and the target object differs significantly, there are correct results among contradictory results. Based on the results of Case 1 and Case 2, the result obtained when the research object with the larger MBR (Minimum Boundary Rectangle) area is taken as the target object is the accurate result. Therefore, in this embodiment, when the two research objects are far apart, it is necessary to determine which of the two research objects should be used as the reference object when using the DRM model to determine the directional relationship between geographic entities. In this embodiment, the research object with the larger MBR (Minimum Boundary Rectangle) area is taken as the target object to realize the determination of the directional relationship between geographic entities.
[0062] like Figure 8 The diagram shown illustrates the DRM-ADR (Direction Relationship Matrix and Angular Direction Relationship) model used in this embodiment, which is also a flowchart of the method in this embodiment:
[0063] 1) Determine whether the MBRs of the two research subjects overlap;
[0064] 2) When the MBRs of two research objects overlap, the ADR model is used to determine the directional relationship between the two research objects; the process of determining the directional relationship between two research objects using the ADR model includes:
[0065] a) Create a grid covering the entire study area and generate grid points;
[0066] b) Based on grid points grid_A within region A of the study object and grid points grid_B within region B of the study object, calculate the inclination angle of the line connecting the points in grid_A and the points in grid_B. Calculate all The average value θ is standardized so that θ ∈ [0°, 360°];
[0067] c) Based on the results obtained after standardization, the final direction is determined using the angle range of the eight-direction cone model.
[0068] When the MBRs of two research objects do not overlap, the DRM model is used to determine the directional relationship between the two research objects; the process of determining the directional relationship between two research objects using the DRM model includes:
[0069] I) Determine the size of the area between the minimum bounding rectangle MBR_A of research object A and the area between the minimum bounding rectangle MBR_B of research object B;
[0070] II) Select the research object with the smaller area of the minimum boundary rectangle between research object A and research object B as the reference object, and the other research object as the target object. Then, establish a nine-square grid directional region with the MBR of the reference object as the center.
[0071] III) Calculate the overlap area between the target object and the direction region, and take the direction region with the largest overlap area as the final direction.
[0072] In other words, the method in this embodiment determines whether the distance between two reference objects is too far based on whether the MBR of the reference object and the MBR of the target object overlap. As another implementation, it can also determine whether the distance between two research objects is too far based on whether the minimum distance between the boundaries of the actual regions of the two reference objects is greater than a set distance value.
[0073] In this embodiment, when the MBR of the reference object and the MBR of the target object do not overlap, the two research objects are considered to be far apart, and the DRM model is used for judgment. At the same time, the relative size and shape of the research objects are considered. The research object with the smaller MBR area is used as the reference object, and the reference object with the larger MBR area is used as the target object. Then, the DRM model is used to determine the directional relationship. When the MBR of the reference object and the MBR of the target object overlap, this embodiment considers the two research objects to be close, and the ADR model is used to determine the directional relationship.
[0074] To verify the feasibility of the method in this embodiment, the directional relationships between streets in a certain area are determined. A street map of that area is shown below. Figure 9 As shown in Tables 1 and 2 below, the judgment results are as follows: "<Street 11, North direction, Street 16>" means that "Street 16" is in the "North direction" of "Street 11".
[0075] Table 1 Results of Determining the Directional Relationship between Street 11 and Other Streets
[0076]
[0077] Table 2 Results of Determining the Directional Relationship between Street 16 and Other Streets
[0078]
[0079] Based on these results, it can be confirmed that the method of this embodiment accurately determines the directional relationship between geographic entities. Specifically, this embodiment uses the ADR model to determine the directional relationship between two geographic entities when they are close together. Because the ADR model can consider the influence of the entity's shape on the directional relationship determination, even when the shape distribution of the entities is uneven, it can accurately determine the directional relationship between the two entities at close range. This effectively avoids the error in the judgment result caused by the uneven shape distribution of the entities when using the DRM model in this situation. This embodiment also uses the DRM model to determine the directional relationship when the two entities are far apart. The judgment result of the DRM model is not affected by the distance. Therefore, the process of using the DRM model to determine the directional relationship when the distance is far can effectively avoid the inaccuracy of the ADR model due to the distance between the two entities. This embodiment also considers the problem that the DRM model is easily affected by the shape distribution of the entities themselves. When the two entities are far apart, the size of the minimum boundary rectangle area of the two entities is first determined, and the entity with the smaller minimum boundary rectangle area is used as the reference object to ensure the accuracy of the judgment result. Furthermore, the method based on this embodiment effectively solves the problem of inaccurate judgment results caused by uneven shape distribution of the research object and the large distance between the reference object and the target object.
[0080] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. The scope of patent protection of the present invention shall be determined by the claims. Similarly, any equivalent structural changes made based on the description and drawings of the present invention shall also be included within the scope of protection of the present invention.
Claims
1. A method of determining a direction relationship between geographical entities, characterized by, Includes the following steps: 1) Determine the distance between two geographic entities according to the set rules; the distance includes two cases: close distance and far distance. 2) When the distance is considered close, create a uniform grid within the entire area of the two geographic entities to obtain each grid point, and calculate the inclination angle of the line connecting grid points in one geographic entity to grid points in the other geographic entity. The average value θ, based on this tilt angle The average value θ is used to obtain the angle between two geographic entities, and the directional relationship between the two geographic entities is determined using an angle-direction relationship model; tilt angle The formula for calculating the average value θ is: ； Where n is the product of the number of grid points in one geographic entity and the number of grid points in another geographic entity; When the distance is large, compare the areas of the minimum bounding rectangles formed by the two geographic entities. Take the geographic entity with the smaller area of the minimum bounding rectangle as the reference object and the other geographic entity as the target object. Establish directional regions with the reference object as the center, calculate the overlapping area between the target object and each directional region, and take the directional region with the largest overlapping area as the directional relationship between the target object and the reference object.
2. The method of judging the direction relation between geographical entities according to claim 1, characterized in that, The setting rule is as follows: based on the actual area of two geographical entities, the distance between the two geographical entities is determined by whether the minimum distance between the boundaries of the actual areas is greater than a set distance value.
3. The method for determining the directional relationship between geographical entities according to claim 1, characterized in that, In step 1), the setting rule is: whether the minimum boundary rectangles formed by two geographic entities overlap; when there is overlap, the distance between the two geographic entities is close, and when there is no overlap, the distance between the two geographic entities is far.
4. The method for determining the directional relationship between geographical entities according to claim 1, characterized in that, The directional area is a nine-grid directional area.
5. The method for determining the directional relationship between geographical entities according to claim 1, characterized in that, Incline angle The calculation formula is: ; Where (x1, y1) are the coordinates of a point in the reference object, and (x2, y2) are the coordinates of a point in the target object.
6. The method for determining the directional relationship between geographic entities according to claim 1, characterized in that, The angle value is obtained by the following formula: ； Where θ is the average value of the tilt angle. The angle of inclination is the line connecting grid points in one geographic entity to grid points in another geographic entity, and n is the product of the number of grid points in one geographic entity and the number of grid points in another geographic entity.
7. The method for determining the directional relationship between geographical entities according to claim 1, characterized in that, The angle-direction relationship model is an eight-direction cone model.