WO2009092327A1

WO2009092327A1 - Map path seeking method and system

Info

Publication number: WO2009092327A1
Application number: PCT/CN2009/070182
Authority: WO
Inventors: Chao Peng
Original assignee: Tencent Technology (Shenzhen) Company Limited
Priority date: 2008-01-17
Filing date: 2009-01-16
Publication date: 2009-07-30
Also published as: CN101241507B; CN101241507A

Abstract

A map path seeking method and system are disclosed. The method includes: dividing a plane map into at least two regions which are equivalent in size in advance, and determining points of path on borders of the region; when the path seeking being required, performing the path seeking based on the region, and acquiring a region path; according to the region path, performing triangular plane-based path seeking, and generating a final path.

Description

Map searching method and system

The present invention relates to network information acquisition technology, and in particular to a map path finding method and system. Background of the invention

With the continuous development of online games, in games based on three-dimensional (3D) scenes, it is often necessary to realize the movement of virtual objects in the scene, thereby involving the problem of finding paths in the map corresponding to the scene. In map finding technology, A* path finding technology is an artificial intelligence technology widely used by people. A* path finding technology implements a heuristic search, which is to evaluate each time using a valuation function for each decision. The value of a program, which in turn determines the solution that needs to be tried first. Specifically, the so-called heuristic search is to select a heuristic function when the current search node selects the next node downward, and select the node with the least cost, such as the node closest to the current node, and jump to the selected node. On the node, the selected node continues to search as the current search node until it reaches the target node.

At present, the commonly used path finding methods in 3D scenes are: A* path finding method based on triangle face and A* two layer path finding method. Specifically, the triangle-based A* pathfinding method first sets an Open Table that stores all the probed but unsearched nodes, and a Close Table that stores the searched nodes. When starting the pathfinding, starting from the starting node, each node is added to the Open Table as the current node. If the target node is not reached and the Open Table is not empty, each possible node is searched for its possible child node one by one. And calculate the valuation of each child node. After the estimation is completed, the current node is placed in the Close Table, and the new current node is selected from the estimated child nodes, and then the new current node is searched for its possible The child nodes, and calculate the valuation of each child node, and so on, until the path found is output. As can be seen from the above process, when the map corresponding to a scene is composed of a large number of dense triangular faces when searching based on triangular faces, if the starting node is far away from the target node, since each of the detected or already Searched nodes, then, the search space needs to save a large number of extended nodes, the space overhead will be extremely large, and many of them are actually unrelated nodes, thus not only occupying a lot of unnecessary space overhead, resulting in waste; And will seriously affect the efficiency of pathfinding.

As for the A* two-layer pathfinding method, a low-density diamond mesh is created on the map, and the center point of the diamond mesh is fixedly selected as the waypoint. The macro pathfinding method is first used to determine the rough path, and then in each path. Micro-path finding is used on the path, but the search node used in the path-finding process is the center point of the grid. It has certain limitations in the scope of application. It can only process maps of two-dimensional shape rules, but not for maps. 2D or 3D irregularly shaped maps. Summary of the invention

An embodiment of the present invention provides a method for searching for a map, including:

Dividing the planar map into at least two equal-sized regions in advance, and determining the waypoints on the boundaries of the regions;

When path finding is required, path finding is performed based on the area to obtain a regional path;

Based on the area path, a triangle-based path finding is performed to generate a final path.

The embodiment of the invention further provides a map path finding system, comprising:

a region generating module, configured to divide the plane map into at least two equal-sized regions; a waypoint generating module, configured to determine a waypoint on a boundary of the region;

The path finding module is configured to perform path finding based on the area, obtain a regional path, and perform a path based on the triangular face according to the regional path to generate a final path. The map finding method and system provided by the embodiment of the present invention divides the planar map to be processed into a plurality of equal-sized regions in advance, and generates a waypoint on the boundary of the region; when searching for a path, first performs path-finding based on the region After obtaining the regional path, the triangle-based pathfinding is further used in the specific scenario according to the determined regional path to generate a final moving path. Therefore, the embodiments of the present invention have the following advantages and features:

1) In the embodiment of the present invention, the path is first searched based on the area, and then the waypoint is performed based on the road points involved in the determined area. Although the A* path finding algorithm is also used, the first path finding is based on the area. The search node is equivalent to one area. The second path-finding only involves the range covered by the determined area path. There are no more unrelated nodes. Obviously, the path-finding process greatly reduces the storage space required and reduces the storage space. Space overhead.

2) The second-level path-finding method adopted by the present invention, the first level first performs rough pathfinding by region, and narrows the search range; the second level performs fine path finding one by one based on the determined regions, according to the triangular face-based The way to find the way, to obtain the final moving path, is equivalent to the high-density triangular surface thinning of 4 bar, sharing the time and space overhead, effectively reducing the space-time complexity of the entire path-finding process, and greatly improving the path-finding efficiency. .

3) Since the map is divided into regions, and the road points used are on the regional boundary, the location of the road points is related to the shape of the map, so the present invention can be applied to processing maps of various shapes, for example, there is no A map of connected areas, etc., has a wider range of applications and is more practical.

4) In the present invention, the waypoint obtains the focus of the edge of the divided area and the edge of the polygon map, and can dynamically generate the waypoint instead of being fixed, so that the implementation is more flexible, convenient, and the actual path. Closer, the resulting moving path is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

1 is a schematic diagram of a general implementation flow of a method for searching for a map in an embodiment of the present invention; FIG. 2 is a schematic diagram of a 3D scene map projected onto a plane; 3 is a schematic diagram of dividing the map shown in FIG. 2 into a navigation grid;

Figure 4 is a schematic view of the map shown in Figure 2 after determining the waypoint;

FIG. 5 is a schematic diagram of an implementation process of generating a navigation grid according to an embodiment of the present invention; FIG. 6a to FIG. 6g are schematic diagrams of implementation processes of an embodiment of the present invention. Mode for carrying out the invention

In the embodiment of the present invention, the planar map to be processed is divided into at least two equal-sized regions in advance, and the waypoints of the region are determined on the boundary of each region; when path finding is required, the region is first searched The road obtains the regional path from the starting point to the ending point; after that, based on the obtained regional path, the triangular path based path finding is used to generate the final path.

Each area may be referred to as a navigation grid. Correspondingly, the road point of the area is the midpoint of the line segment where each side of the navigation grid coincides with the plane map, and all the way points on each side of each navigation grid constitute the navigation grid. Corresponding road point sequence, the area path is to treat each navigation grid as a point, first determine all the navigation grids from the navigation grid to which the starting point belongs to the navigation grid to which the destination belongs, and then according to the relationship between the navigation grids determined and the navigation grid The entry point, obtain a navigation grid path composed of road points, that is, the area path.

Here, the scene corresponding to the planar map to be processed may be a two-dimensional or three-dimensional scene, such as a 3D community map in the game. For a 3D scene, the map navigation piece of the 3D scene is projected onto the plane, and the obtained polygon plane map is used as the plane map to be processed. For the 2D plane scene, the plane map can be directly used as the plane map to be processed. The navigation grid can take different shapes, such as squares, rectangles, hexagons, etc., as long as the regions are equal in size; preferably, the entire planar map is divided into at least two regions of equal size by equally spaced vertical and horizontal lines. Square.

FIG. 1 is a schematic diagram showing the overall implementation process of a map path finding method in an embodiment of the present invention. As shown in FIG. 1, the map path finding method in the embodiment of the present invention includes the following steps. Step 101: Acquire a plane map to be processed in advance, and divide the acquired plane map into at least two navigation grids of equal size.

If it is necessary to process a map corresponding to a 3D scene of a plane map, such as a community map in a game, first project a map of the map of the 3D scene onto a plane to obtain a plane map. As shown in Fig. 2, Fig. 2 is a multi-plane map obtained by projecting a 3D community map navigation piece onto a plane.

In this step, the plane map may be divided in different manners, and the divided navigation grids may have different shapes, as long as the area corresponding to each navigation grid is equal. Preferably, the polygon is divided into a plurality of squares by equally spaced vertical and horizontal lines, as shown in Fig. 3, and Fig. 3 is a plane map of Fig. 2 divided into equal-sized squares.

Step 102: Predetermine a sequence of waypoints corresponding to each navigation grid, and establish and save a connection relationship between each waypoint in each route sequence.

In this embodiment, the midpoint of the line segment that coincides with a plane of the navigation map and the plane map is determined as the waypoint of the navigation grid. As shown in Fig. 4, Fig. 4 is a map after the map shown in Fig. 2 is determined. The black points in Figure 4 are waypoints, and all waypoints on each side of the navigation grid form a sequence of waypoints for the navigation grid. In this embodiment, a structure or a data structure is created for each navigation grid, and the structure includes related information of the navigation grid and its corresponding waypoint sequence.

The connectivity relationship between the various waypoints in the sequence of the waypoints, that is, the reachable relationship between the various waypoints in the sequence of the waypoints, can be established based on the pathfinding based on the triangular faces. The preservation of the connected relationship can be realized by using the collected data structure. The so-called parallel check refers to a number of disjoint sets, which can realize the operations of faster merging, judging the set of elements, and the like, mainly involving merging and searching of the set.

In the present embodiment, steps 101 and 102 are performed in advance before the path finding, and only need to be performed once, and need not be performed before each path finding.

Step 103: When a pathfinding is required, a navigation grid based A* path finding is used to obtain a navigation path. Specifically, when path finding is needed, the coordinate positions of the start point and the end point are first obtained, and the navigation grid to which the start point and the end point belong respectively can be located according to the coordinate position; after that, the navigation grid is used as the unit point, and the start navigation grid is the starting point. Terminate the navigation grid as the end point and perform A* path finding.

The A* pathfinding process based on the navigation grid includes the following steps:

A. Construct an Open table, according to the connection relationship between the pre-established waypoints, access the adjacent navigation grid reachable by the starting navigation grid, and put the two-group <navigation grid, entry point> as an element into the table.

Here, the adjacent navigation grids reachable by the starting navigation grid need to call the prior art triangle-based path finding to determine the direction of expansion.

B. If the Open table is not empty, take out the element with the lowest valuation value in the table and put the current navigation grid into the Close table. Set the extracted navigation grid to the current navigation grid.

C. If the current navigation grid is equal to the destination navigation grid, the pathfinding ends, and step E is performed; otherwise, step D is performed;

D. Access the adjacent navigation grid reachable by the current navigation grid. The elements in the unvisited state (navigation grid, entry point) are added to the Open table, and the steps are returned.

E. If the end point is found, a path from the start point to the end point formed by the waypoint is generated along each state point.

Step 104: Perform a triangle-based A* path finding according to the obtained navigation grid path to generate a final moving path.

The navigation grid path obtained in step 103 is a rough path composed of waypoints and cannot be directly used as a basis for movement. In this step, during the moving process, the A* path finding based on the triangular surface is further adopted to obtain a refined moving path.

In the embodiment of the present invention, the divided area is referred to as a navigation grid because each navigation grid corresponds to a navigation area, and has a navigation function, that is, adjacent navigation grids have an association relationship. Based on this, a process of generating a navigation grid in the embodiment of the present invention is as shown in the figure. As shown in 5, the following steps are included:

Step 501: Determine coordinates of the upper left corner of the current navigation grid;

Because the map has a certain coordinate range, and the size of each navigation grid is known at the time of division, according to the coordinates of the upper left corner of the navigation grid, the position of the navigation grid in the map can be determined, and the boundary of the navigation grid and the map can also be determined. relationship.

Step 502: Obtain a right navigation road boundary point and a lower boundary road point of the current navigation grid;

In this step, it is not necessary to obtain the waypoints of the upper and left boundaries of the current navigation grid. When two navigation grids have a common edge, that is, when two navigation grids are adjacent, the waypoints on the common edge can be shared by the two navigation grids. By extracting the waypoint on the right border of the left navigation grid and the waypoint on the lower boundary of the upper navigation grid in step 503, the upper boundary of the current navigation grid and the waypoint on the left boundary can be directly obtained. In this way, the efficiency of obtaining navigational waypoints is improved, and the complexity of the operation is reduced.

Step 503: Set the association between the current navigation grid and the navigation bar on the left side and the navigation grid on the left side, and obtain the waypoints on the upper boundary and the left boundary;

For the current navigation grid, the waypoints of the upper navigation grid and the left navigation grid have been acquired and acquired. Here, as long as the current navigation grid is actively associated with the upper navigation grid and the left navigation grid, the current navigation grid can be obtained. The waypoints on the boundary and the left border.

Step 504: Generate a sequence of waypoints of the current navigation grid, and establish a connection relationship between the road points.

Here, the sequence of waypoints is a set of all waypoints on each boundary of the current navigation grid; the establishing the connection relationship between the respective waypoints is implemented by a triangle-based A* pathfinding method.

The method for generating the navigation grid shown in Figure 5 is to generate a navigation grid from the upper left corner to the right and down. In the actual application, the navigation grid can also be generated from the upper right corner to the left and down, or from the lower left corner to the right. Generate a navigation grid up, or generate a navigation grid from the lower right corner to the left and up. The processing is similar. Just replace the angular coordinates and boundaries selected in the above steps, for example: Determine the coordinates of the upper right corner, get the left and lower road points, and so on.

The implementation of the map finder method in the embodiment of the present invention will be described in detail below with reference to specific embodiments and the accompanying drawings.

In this embodiment, the polygon map is divided into at least two squares by using equal-space vertical and horizontal lines in advance, and the sequence of road points corresponding to each navigation grid is determined, and each road point sequence in each road point sequence is established and saved. Connectivity.

Figures 6a through 6g show the process of implementing map path finding in this embodiment. Among them, the starting point for the path finding is S point, and the ending point is T point. As shown in Figure 6, the black points in the figure are waypoints. The S and T points are respectively represented by a five-pointed star, and the gray-filled square is the searched square node.

Figure 6a shows the initial state, and the path finding process from point S to point T is as follows.

Step 11: Positioning starting point S and ending point T The navigation grid to which T belongs is Gs and Gt, respectively, as shown in Figure 6b.

Step 12: Perform A* pathfinding, expand the navigation grid GO according to the waypoint that S point can reach, and put the GO into the priority queue, as shown in Figure 6c.

Here, the navigation grid GO is the navigation grid expanded from the first navigation grid Gs. How to determine the extension direction of the initial navigation grid Gs can be determined according to the existing triangle-based A* pathfinding technique. Specifically, in the embodiment, for the navigation grid Gs, according to the A* path finding based on the triangular surface, the starting point S can only move to the left, and the right side is not connected, so the navigation grid GO is extended from the navigation grid Gs to the left. If you can move from the starting point S to the left and right sides, according to the A* path finding technique, you can expand a navigation grid from the navigation grid Gs to the left and right, and then continue to perform A* path finding based on the navigation grid. .

Step 13: Based on the A* path finding technique, expand the navigation grids Gl, G2, and G3 from the navigation grid GO, as shown in Figure 6d.

Step 14: According to the heuristic search rule, that is, according to the distance to the navigation grid Gt, the navigation grid G1 is preferentially extended, and the navigation grid G4 is obtained, as shown in Fig. 6e. Step 15: According to the same principle as step 14, the navigation grid G4 is extended to obtain the navigation grid G5, and then reaches Gt, as shown in Fig. 6f. At this point, it is determined that Gs, G0, Gl, G4, G5, and Gt are all navigation grids of the navigation path path.

Step 16: According to the association relationship between each navigation grid and the entry point, a route sequence is generated, and then the navigation path from S point to T point is obtained. As shown in FIG. 6f, Gs is adjacent to GO, and the entry point of Gs to GO is point 1; GO is adjacent to G1, and the entry point of GO to G1 is point 2; G1 and G4 are adjacent to each other. The entry point of G1 to G4 is point 3; G4 is adjacent to G5, the entry point of G4 to G5 is point 4; G5 is adjacent to Gt, and the entry point of G5 to Gt is point 5, therefore, generated The sequence of waypoints is {S, 1, 2, 3, 4, 5, T}, and the obtained navigation grid path is S 1 2 3 4 5 T. As shown in Fig. 6g, the sections between the road points are long. path.

Step 17: During the movement, according to the obtained navigation grid path, a triangle-based A* pathfinding is performed to generate a final moving path.

To implement the above method, an embodiment of the present invention further provides a map path finding system, including: an area generating module, a waypoint management module, and a path finding module.

The area generation module is configured to divide the plane map into at least two equal-sized areas. The waypoint generating module is configured to obtain, as a waypoint, a midpoint of a line segment in which each side of the area coincides with the plane map for each of the divided regions, generate a waypoint sequence for the waypoints in each of the regions, and establish a sequence of the waypoints The connection between the waypoints.

The path finding module is used for path finding using A* path finding technology, and further includes an area finding module and a triangle face finding module. The regional path-finding module uses A* path-finding technology to find paths in the area, and determines an area path composed of road points according to the area of the route and the entry point between the areas; the triangular surface finding module, in the moving process Medium, based on the regional path, performs a triangle-based A* pathfinding to obtain the final moving path.

The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention. Range of protection.

Claims

Rights request

A map finding method, characterized in that

The map path finding method according to claim 1, wherein the dividing area comprises: dividing the plane map into at least two squares of equal size by using equally spaced vertical and horizontal lines.

The method of map finding according to claim 2, wherein the determining the waypoint on the boundary of the area comprises:

Determining the coordinates of the upper left corner of the current region;

Obtaining a waypoint on a right boundary of the current area and a waypoint on a lower boundary;

And setting a relationship between the current area and the upper area and the left area, obtaining a waypoint on a lower boundary of the upper area as a waypoint on an upper boundary of the current area, and acquiring a path on a right boundary of the left area The point serves as a waypoint on the left border of the current area.

The map path finding method according to claim 1, wherein the determining the waypoint on the boundary of the area comprises: determining a midpoint of a line segment in which each side of the area coincides with a plane map as a road point.

The method of claim 1 according to claim 1, wherein the path finding is performed based on the area, and obtaining the area path comprises:

Path finding based on the area, determining the area from the area to which the start point belongs to the area to which the end point belongs;

- point, get The path of the area is obtained.

The map path finding method according to any one of claims 1 to 5, wherein the path finding adopts an A* path finding technique.

The map path finding method according to any one of claims 1 to 5, wherein before the dividing the area, the method further comprises: projecting a map navigation piece corresponding to the three-dimensional scene onto a plane to form the plane map. .

8. A map path finding system, comprising:

The path finding module is configured to perform path finding based on the area, obtain a regional path, and perform a path based on the triangular surface according to the area path to generate a final path.

The map path finding system according to claim 8, wherein the path finding module comprises:

a regional path finding module, configured to perform path finding based on the area, determine an area from the area to which the starting point belongs to the area to which the end point belongs; obtain the location according to the determined relationship between the areas and the determined entry point between the areas Regional path

A triangular surface path finding module is configured to perform a triangle face based path finding according to the area path to generate a final path.