WO2021169772A1

WO2021169772A1 - Method and system for marking scene

Info

Publication number: WO2021169772A1
Application number: PCT/CN2021/075491
Authority: WO
Inventors: 于毅欣
Original assignee: 于毅欣
Priority date: 2020-02-27
Filing date: 2021-02-05
Publication date: 2021-09-02
Also published as: CN113779162A

Abstract

Disclosed is a method for marking a scene, the method comprising: collecting and creating basic scene information for describing the basic situation of a scene; collecting and creating passage information of the scene and/or a road for storing rules for passing a region; collecting and creating layer information for storing information of one layer or one layer plane which is vertically superimposed; and collecting and creating communication information for storing information of a connection relationship between the objects described above. The method can support the marking of a stereoscopic scene to guide navigation information or a navigation process so as to find a path or plan a path for navigation. The method arranges feature pictures and records coordinates of each feature picture so as to internally provide a basis for correcting the coordinates again.

Description

Method and system for scene marking

Technical field

The invention relates to the field of computer technology and artificial intelligence technology, and in particular to a method and system for marking scenes.

Background technique

GPS (Global Positioning System) often can only return 2D coordinate information. Many current scene description methods are relatively general. The present invention aims to refine a more detailed marking method of a three-dimensional scene to complete solid or The relatively fixed three-dimensional objective scene information is described and collected, and can support a larger range of surveying, mapping, description, and storage. Another feature of the present invention is that it can support marking the three-dimensional scene to guide the navigation information or the navigation process, so as to find the way or plan the route for navigation. Another feature of the present invention is that it can add a target object that can be positioned more accurately to the scene by adding a feature picture.

Summary of the invention

According to one aspect of the present disclosure, a method for scene marking is characterized in that it includes:

Collect and create the basic information of the scene to describe the basic situation of the scene.

Collect and create the passage information of the scene or \and the road, which is used to store the passage rules of an area.

Collect and create layer information, which is used to store the information of a layer or a layer that is vertically superimposed.

Collect and create Unicom information to store the information about the connection relationship between the above objects.

Collect and create characteristic picture information, where the characteristic picture is the picture obtained by shooting, or the picture or laser picture used for posting after printing, and is used to mark a coordinate in the target area and related information to identify the image information.

Preferably, edge vertex information is created for the scene, and the edge vertex information is coordinates relative to the origin of a certain coordinate system or coordinates obtained by global positioning technology.

Preferably, the containing box (bounding box) of this scene or sub-scene is generated according to the edge vertices.

Preferably, as an optional implementation manner, one way to determine the division of an area into sub-scene is based on the size of the area, when the size satisfies a value greater than a certain value (a value from 1 to 100 million square meters, for example, 2 Square meters), the continuous area can be regarded as a sub-scene.

Preferably, as an optional implementation, one way to determine the division of an area into sub-scene is based on the size of the area and the shape of the containing box (bounding box) (the aspect ratio should be less than a certain value and greater than a certain value). ), when the size is greater than a certain value and the aspect ratio is less than a certain value (for example, 10) and larger than a certain value (for example, 0.1), the continuous area can be regarded as a sub-scene.

Preferably, as an optional implementation manner, a point (for example, the center point of the containing box) can be selected in the containing box (bounding box) of this scene to connect the surrounding vertices to establish a triangle set or fan set.

Preferably, as an optional implementation manner, a vertex ( According to a fixed distance or a fixed line segment length ratio, refer to Fig. 1), it is used to connect with the edge vertices to form an edge buffer area and the obtained triangle list or fan list of the edge buffer area.

Preferably, as an optional implementation manner, the edge vertex of the scene can be connected to the diagonal line, and a vertex can be added in the direction approaching the interior of the graph (according to a fixed distance or a fixed line length ratio, refer to Figure 4) is used to connect with edge vertices to form an edge buffer area and the obtained triangle list or fan list of the edge buffer area.

Preferably, as an optional implementation manner, a vertex ( According to a fixed distance or a fixed line length ratio), it is used to connect with the edge vertices to form the edge buffer area and the obtained triangle list or fan list of the edge buffer area, which can be schematically shown with reference to FIG. 1.

Preferably, the basic information of the scene includes the unique ID information of the scene.

Preferably, the path information includes edge vertex information of the path.

Preferably, the path information contains passability information, which can be one-way pass information or two-way pass information, and can be 3D link information (a collection of 3D points, or one-way linked list information of 3D points or A doubly linked list of 3D points).

Preferably, as an optional implementation manner, the passability information may be composed of a 3D pass direction vector, a pass width vector perpendicular to the 3D pass direction vector, and pass height vector information (a total of 3 vectors).

Preferably, as an optional implementation manner, the passage information may be composed of a 3D coordinate, a passage width vector perpendicular to the passage direction and a passage height vector information (2 vectors in total) at this coordinate.

Preferably, the path information includes the maximum and minimum passing speeds of the road surface information.

Preferably, the path information includes unique ID information of the road surface information.

Preferably, the path information includes containment box information.

Preferably, the information contained in each layer of information may include a containment box (bounding box), a unique ID of this layer of information, and edge vertex information.

Preferably, the information contained in each layer information may include how many layers there are in this layer information, and the layer where this layer information is located (this value may be a negative number to indicate underground).

Preferably, as an optional implementation manner, the information that each of the layer information may contain may have the carrying strength value of the layer information or the strength type that can be carried.

Preferably, the Unicom information is information mainly used to store the connection relationship between logical objects in the scene (basic information of the scene, road access information, layer information, etc.).

Preferably, the unique ID information and edge vertex information of the Unicom information can be included in the Unicom information.

Preferably, the unicom information may include a list of objects connected to the unicom information, and each connected object adds a type to indicate whether it is scene, path information, layer information or unicom information.

Preferably, the computer equipment and storage medium are characterized by being able to store information.

Preferably, as an optional implementation manner, all the marking information of the scene can be stored in a file or in a folder.

Preferably, as an optional implementation manner, various information types of the scene can be divided into different files and stored separately.

Preferably, as an optional implementation manner, a virtual containment box can be created around a certain scene, and all scenes or sub-scenes or marking information in the scenes inside the containment box are stored in a folder, or the same In one file.

Preferably, the scene information can be a container of other information, and other information contained in the scene can also be retrieved from a storage list or storage container of various information according to the bounding box of the scene information.

Preferably, sub-scene information can be divided into the scene information.

Preferably, as an optional implementation manner, a data type or identification can be added to distinguish the edge vertex information, path information, layer information, and communication information.

Preferably, as an optional implementation manner, the edge vertex information, path information, layer information, and communication information may be stored in different files to distinguish them.

The edge vertices are coordinates relative to the origin of a certain coordinate system or coordinates obtained by global positioning technology, and are mainly used to describe the edge of the scene in question.

Optionally, the difference between the road surface information and the layer information is: the road surface information is the information of the horizontal plane, or the information recognized as the horizontal plane, and the layer information is the information that is superimposed or overlapped in the vertical direction.

Preferably, the unique ID is a unique ID of each object.

Preferably, as an optional implementation manner, when measuring or making coordinate information, various coordinates can be collected according to the origin of a local coordinate system first, and then the coordinates of the local coordinate system and the global coordinate system can be used. Differences to modify the coordinate data collected this time.

Preferably, the coordinates of the center point of each feature picture are collected and used as the coordinates of the feature picture information.

Preferably, a unique ID is created for each feature picture.

Preferably, each feature picture and its corresponding unique ID are stored in the database.

Preferably, as an optional implementation manner, the ID corresponding to each feature picture can be stored in the scene, path information, layer information, or China Unicom information where the feature picture is located.

Description of the drawings

In order to explain the technical solutions in the embodiments of the present invention more clearly, the following will briefly introduce the drawings needed in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

FIG. 1 schematically shows a method for determining the position of a selected point of an edge buffer area or a schematic diagram of a position determination law, a top view.

Fig. 2 schematically shows a layout diagram of channel information and Unicom information contained in a scene, top view.

Fig. 3 schematically shows a schematic diagram of a coordinate point and a passing width vector in a path information, a top view.

Fig. 4 schematically shows another method for determining the position of selected points or the law of determining the position of an edge buffer area, a top view.

Detailed ways

The following examples are used to illustrate the present invention, but not to limit the scope of the present invention.

Example 1

Step 1: Choose a place to create a scene, this place can be a company, first choose a point to be the origin of the local coordinate system in the scene (for example, the corner of the upper left corner is the coordinate origin), and then the local coordinate system The axis is the same as the axis of the global coordinate system, and then the edge vertex information (list information) of the scene is obtained by measurement, and a containment box information is obtained from the edge vertex information, and then a unique ID is generated for the scene object.

Step 2: Find each place that can pass (vehicles or people pass) in this scene, create path information, create a unique ID for each path information, collect edge vertex information (can be list information), and according to the edge vertex information Obtain the containment box information of the path information, and obtain the passability information (select a point in the path, measure the relative coordinates of this point, pass the height vector and pass the width vector to obtain the information or information list).

Step 3: In the place where a floor appears, create a floor information, for example, the floor information of the second floor includes the edge vertex information of this floor, the containment box, the unique ID, the number of floors of this floor (2), and the total floor of this building Number (for example, 33 floors), bearing strength value (for example, if it is able to withstand a magnitude 7 earthquake, mark 7, and if it can stop a helicopter, mark 100).

Step 4: For places that can be connected to other objects (such as elevators or doors, mainly elevators), establish a connection information, such as a 33-story elevator, and collect the edge vertex information (may be list information) of this connection information, Calculate the bounding box information, and connect the list of various types of information (the information of 33 floors, mainly unique ID and type value, here can compare a type value for various types, for example, the scene information is 1, the pass information is 2. The layer information is 3, the Unicom information is 4, the feature picture information is 5, etc.).

Step 5: Post or collect feature pictures in the scene, collect the coordinates of each feature picture, create a unique ID for each feature picture, and store the correspondence between this feature picture and its unique ID in the database.

Further, as an optional implementation manner, the unique ID in each feature picture information can be stored in the scene, path information, layer information, or China Unicom information where the feature picture is located.

Step 6: Since all the coordinates just now are calculated relative to the origin of the local coordinate system, and the axis of the local coordinate system is still the same as the axis of the global coordinate system, then calculate or measure the global coordinate system of the origin of the local coordinate system Coordinates. Add this coordinate to each local coordinate to get the conversion from local coordinate to global coordinate. You can also adjust it according to the actual situation.

Step 7: After collecting the relevant information of this scene, store it according to the previous storage method. If a certain area (for example, 100 kilometers long and wide or 10,000 square kilometers) is stored on a single machine, it will be stored according to the current Find the corresponding storage medium in the area where the scene information is stored. When storing, you can create a separate folder for this scene, and store each type of information in a file, or you can store several types of information in one. In a unified file.

The above are specific embodiments of the present invention, but the protection scope of the present invention should not be limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention shall be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be defined by the claims. The scope of protection shall prevail.

Claims

A method for marking scenes is characterized in that it includes:

Collect and create the basic information of the scene to describe the basic situation of the scene;

Collect and create scenes or road access information to store the passage rules of an area;

Collect and create layer information, which is used to store the information of a layer or layer that is vertically superimposed;

Collect and create Unicom information to store the information about the connection relationship between the above objects;

Collect and create characteristic picture information, where the characteristic picture is the picture obtained by shooting, or the picture or laser picture used for posting after printing, and is used to mark a coordinate in the target area and related information to identify the image information.
The method according to claim 1, wherein collecting and creating basic information of the scene further comprises:

Create edge vertex information for the scene, where the edge vertex information is coordinates relative to the origin of a certain coordinate system or coordinates obtained by global positioning technology;

Generate the containment box (bounding box) of this scene or sub-scene according to the edge vertices;

As an optional implementation manner, one way to determine the division of an area into sub-scene is based on the size of the area. When the size is greater than a certain value (a value from 1 to 100 million square meters, for example, 2 square meters), Then this continuous area can be regarded as a sub-scene;

As an optional implementation, one way to determine whether an area can be divided into sub-scene is based on the size of the area and the shape of the containing box (bounding box) (the aspect ratio should be less than a certain value and greater than a certain value) , When the size is greater than a certain value and the aspect ratio is less than a certain value (for example, 10) and larger than a certain value (for example, 0.1), the continuous area can be regarded as a sub-scene;

As an optional implementation manner, a point (for example, the center point of the containing box) can be selected in the containing box (bounding box) of this scene to connect the surrounding vertices to establish a triangle set or fan set;

As an optional implementation, you can add a vertex (according to a fixed Distance or fixed line length ratio), used to connect with the edge vertices to form the edge buffer area and the obtained triangle list or fan list of the edge buffer area;

As an optional implementation mode, the edge vertex of this scene can be connected to the diagonal line, and a vertex (according to a fixed distance or a fixed line length ratio) can be added in the direction approaching to the inside of the graph to be used with The edge vertices are connected to form an edge buffer area and the obtained triangle list or fan list of the edge buffer area;

The basic information of the scene includes the unique ID information of the scene.
The method according to claim 1, characterized in that collecting and creating the passage information of the scene or the road to store the passage rules of an area, further comprising:

The path information includes edge vertex information of the path;

The path information contains transit information, which can be one-way transit information or two-way transit information, and it can be 3D link information (a collection of 3D points, or one-way linked list information of 3D points or 3D point information). Doubly linked list);

As an optional implementation manner, the passability information may be composed of a 3D pass direction vector, a pass width vector perpendicular to the 3D pass direction vector, and pass height vector information (a total of 3 vectors);

As an optional implementation manner, the passage information may be composed of a 3D coordinate, a passage width vector perpendicular to the passage direction and a passage height vector information (total 2 vectors) at this coordinate;

The path information includes the maximum and minimum passing speeds of the road surface information;

The path information includes the unique ID information of the road surface information;

The path information includes containment box information.
The method according to claim 1, characterized in that collecting and creating layer information to store the information of one layer or one layer vertically superimposed, further comprising:

The information contained in each layer of information may include a containment box (bounding box), a unique ID of this layer of information, and edge vertex information;

The information contained in each layer information can include how many layers there are in this layer information, and the layer where this layer information is located (this value can be a negative number to indicate underground);

As an optional implementation manner, the information that each of the layer information may contain may include the carrying strength value of the layer information or the strength type that can be carried.
The method according to claim 1, characterized in that collecting and creating Unicom information to store the information of the connection relationship between the objects, further comprising:

The Unicom information is mainly used to store the information of the connection relationship between logical objects in the scene (basic information of the scene, road access information, layer information, etc.);

The unique ID information and edge vertex information of the Unicom information that can be included in the Unicom information;

The unicom information may include a list of objects connected to the unicom information, and each connected object adds a type to indicate whether it is scene, path information, layer information or unicom information.
The method according to claim 1, wherein the characteristic picture is a picture obtained by shooting, or a picture or a laser picture for posting after printing, and is used to mark a coordinate and related information in the target area. The recognized image information also includes:

Collect the coordinates of the center point of each feature picture as the coordinates of the feature picture information;

Create a unique ID for each feature picture;

Store each feature picture and its corresponding unique ID in the database;

As an optional implementation manner, the ID corresponding to each feature picture may be stored in the scene, path information, layer information, or China Unicom information described in this feature picture.
A system for marking roads, which is characterized in that it comprises:

Computer equipment and storage media are characterized by the ability to store information;

As an optional implementation manner, all the marking information of the scene can be stored in a file or in a folder;

As an optional implementation manner, it may be divided into different files according to various information types of the scene to be stored separately;

As an optional implementation manner, a virtual containment box can be created around a certain scene, and all scenes or sub-scenes or marking information in the scenes inside the containment box are stored in a folder, or in the same file .
A computer readable and writable medium on which a computer program and related data are stored, and is characterized in that when the program is executed by a processor, the related computing functions and content of the present invention are realized.
An electronic device, characterized in that it comprises:

One or more processors;

The storage device is used to store one or more programs.