CN112037120A - Method and device for labeling road plane elements in 3D point cloud data and storage medium - Google Patents
Method and device for labeling road plane elements in 3D point cloud data and storage medium Download PDFInfo
- Publication number
- CN112037120A CN112037120A CN202010757801.8A CN202010757801A CN112037120A CN 112037120 A CN112037120 A CN 112037120A CN 202010757801 A CN202010757801 A CN 202010757801A CN 112037120 A CN112037120 A CN 112037120A
- Authority
- CN
- China
- Prior art keywords
- points
- plane
- point
- cloud data
- point cloud
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000002372 labelling Methods 0.000 title claims abstract description 63
- 238000006073 displacement reaction Methods 0.000 claims description 14
- 239000003550 marker Substances 0.000 description 15
- 238000010586 diagram Methods 0.000 description 14
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/06—Topological mapping of higher dimensional structures onto lower dimensional surfaces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/40—Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Processing Or Creating Images (AREA)
Abstract
The application discloses a method and a device for labeling road plane elements in 3D point cloud data and a storage medium, which are used for at least solving the problem that the road plane elements in the 3D point cloud data cannot be efficiently labeled in the related technology. The method comprises the following steps: the marking device displays 3D point cloud data of a frame of road scene; receiving an input road plane element type and an input element area parameter; generating an element area on an XY plane according to the element area parameter; projecting data points in the displayed 3D point cloud data onto an XY plane; the data point whose corresponding projected point is included in the element region is determined as a point belonging to the received road plane element type.
Description
Technical Field
The present application relates to the field of data annotation, and in particular, to a method and an apparatus for annotating road plane elements in 3D point cloud data, and a storage medium.
Background
In the related art, a annotator annotates 3D point cloud data displayed on a front end of a browser or the like. When labeling an object in 3D point cloud data, a plurality of points constituting the object are generally labeled one by one, that is, a single-point labeling is performed. When marking, the marking person marks corresponding colors or types to the single points.
Disclosure of Invention
Some embodiments of the present application provide a method and an apparatus for labeling road plane elements in 3D point cloud data, and a storage medium, so as to at least solve the problem in the related art that road plane elements in 3D point cloud data cannot be efficiently labeled.
In one aspect, some embodiments of the present application provide a method for labeling road plane elements in 3D point cloud data, including:
the marking device displays 3D point cloud data of a frame of road scene;
receiving an input road plane element type and an input element area parameter;
generating an element area on an XY plane according to the element area parameter;
projecting data points in the displayed 3D point cloud data onto an XY plane;
the data point whose corresponding projected point is included in the element region is determined as a point belonging to the received road plane element type.
In another aspect, some embodiments of the present application provide an apparatus for labeling road plane elements in 3D point cloud data, including: the system comprises a processor and at least one memory, wherein at least one machine executable instruction is stored in the at least one memory, and the processor executes the at least one machine executable instruction to execute the method for labeling the road plane elements in the 3D point cloud data.
In another aspect, some embodiments of the present application provide a non-volatile storage medium having at least one machine executable instruction stored therein, where the at least one machine executable instruction is executed by a processor to implement the method for labeling road plane elements in 3D point cloud data as described above.
According to the technical scheme provided by the embodiment of the application, the element area on the XY plane is generated according to the input element area parameters, the data points in the 3D point cloud data are projected onto the XY plane, and the data points of the corresponding projection points included in the element area are determined to be the points belonging to the input road plane element type, so that the plane elements on the road can be efficiently marked, and the speed and the efficiency of marking objects in the 3D point cloud data are improved.
Drawings
The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiment(s) of the application and together with the description serve to explain the application and not limit the application.
Fig. 1 is a block diagram of a structure of a device for labeling road plane elements in 3D point cloud data according to an embodiment of the present disclosure;
fig. 2 is an architecture diagram illustrating labeling processing of road plane elements in 3D point cloud data according to an embodiment of the present disclosure;
fig. 3 is a processing flow chart of a method for labeling road plane elements in 3D point cloud data according to an embodiment of the present disclosure;
FIG. 4 is a flowchart of a process of step 305 in FIG. 3 according to an embodiment of the present disclosure;
FIG. 5 is another flowchart of the processing of step 305 in FIG. 3 according to an embodiment of the present disclosure;
FIG. 6 is another flowchart of the processing of step 305 in FIG. 3 according to an embodiment of the present disclosure;
FIG. 7 is another flowchart of the processing of step 305 of FIG. 3 according to an embodiment of the present disclosure;
FIG. 8a is a schematic diagram of a generation rule provided in an embodiment of the present application;
FIG. 8b is a schematic diagram of another generation rule provided in the embodiments of the present application;
FIG. 8c is a schematic diagram of another generation rule provided in the embodiments of the present application;
FIG. 8d is a schematic diagram of another generation rule provided in the embodiments of the present application;
fig. 8e is a schematic diagram of another generation rule provided in the embodiment of the present application;
fig. 9 is a flowchart of a process of generating a curve edge of an element region according to an embodiment of the present application;
FIG. 10 is a schematic diagram of an area for generating a curve edge element according to an embodiment of the present disclosure;
fig. 11 is another processing flow chart of a method for labeling road plane elements in 3D point cloud data according to an embodiment of the present disclosure;
fig. 12 is another processing flow chart of a method for labeling road plane elements in 3D point cloud data according to an embodiment of the present disclosure;
fig. 13 is another processing flow chart of a method for labeling road plane elements in 3D point cloud data according to the embodiment of the present disclosure;
FIG. 14 is a flowchart illustrating a process of a method for labeling road surface elements in 3D point cloud data according to an embodiment of the present disclosure in an application scenario;
FIG. 15a is a schematic illustration of a data annotation applying the process shown in FIG. 14;
FIG. 15b is a schematic illustration of another data annotation applying the process shown in FIG. 14.
Detailed Description
In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Each data point in the 3D point cloud data is typically recorded with a three-dimensional coordinate value (x, y, z) for that point and a serial number for that point. In the related art, when a marker marks data points in 3D point cloud data, only the data points can be marked one by one. For example, when a annotator annotates an element on a road plane, the annotator first needs to identify the corresponding element and individually annotate the data points constituting the element. The labeling operations may include operations to stain, categorize, etc. the data points, where the colors and corresponding categories are predefined. The labeling operation has the problems of low labeling speed and low labeling efficiency.
The technical scheme provided by the embodiment of the application can be used for high-speed and high-efficiency marking of road plane elements in the 3D point cloud data.
The road plane elements may include areas on the road plane that distinguish road structures, road functions, etc., which correspond to travel behaviors that specify vehicles. For example, in some embodiments, road plane elements may include elements of road travelable areas, non-travelable areas, lane lines, road edges, ramps, carrousels, and the like. The road surface element may also include other elements or regions, which are not specifically limited in the embodiments of the present application.
Some embodiments of the present application provide a labeling scheme for road plane elements in 3D point cloud data. Fig. 1 shows a structure of a labeling apparatus provided in an embodiment of the present application, where the apparatus 1 includes a processor 11 and at least one memory 12.
In some embodiments, the at least one memory 12 may be a storage device of various modalities, such as a transitory or non-transitory storage medium, a volatile or non-volatile storage medium. At least one machine executable instruction may be stored in the memory 12, and the at least one machine executable instruction, after being executed by the processor 11, implements the labeling processing of the road plane elements in the 3D point cloud data provided in the embodiments of the present application.
In some embodiments, the annotation device 1 may be located on the server side. In other embodiments, the annotation device 1 may also be located in a cloud server. In other embodiments, the annotation device 1 may also be located in the client.
As shown in fig. 2, the labeling process for road plane elements in 3D point cloud data provided by the embodiment of the present application may include a front-end process 12 and a back-end process 14. The relevant 3D point cloud data or other data is displayed by the front end process 12 and the relevant data or information entered by the annotator is received, for example, the front end process 12 may be a process implemented by a web page or a process implemented by a separate application interface. The back-end processing 14 performs corresponding labeling processing according to the relevant data and information received by the front-end processing 12. After the annotation process is completed, the annotation device 1 may further provide the annotation result to other processes or applications on the client, the server, or the cloud server.
The following describes a labeling process of road plane elements in 3D point cloud data by the labeling device 1.
Fig. 3 shows a processing flow of a labeling device for performing labeling processing, that is, a processing flow of a method for labeling road plane elements in 3D point cloud data according to some embodiments of the present application, including:
301, displaying 3D point cloud data of a frame of road scene by a marking device;
step 305 of generating an element region on the XY plane based on the element region parameter;
307, projecting data points in the displayed 3D point cloud data onto an XY plane;
According to the method shown in fig. 3, the labeling device generates an element region on the XY plane according to the input element region parameters, projects data points in the 3D point cloud data onto the XY plane, and determines data points whose corresponding projected points are included in the element region as points belonging to the received road plane element type. Therefore, the labeling device can automatically label the data points in the 3D point cloud belonging to the road plane element type without receiving the one-by-one labeling of the data points in the 3D point cloud data by a labeling operator and only receiving the type and the element area parameters of the road plane element input by the labeling operator, and can remarkably improve the labeling speed and efficiency of the road plane element in the 3D point cloud data.
In some embodiments of the present application, when the annotation device is located in the server or the cloud server, the annotation device receives the road plane element type and the element region parameter from the client after displaying the 3D point cloud data. The road plane element type and the element area parameter received by the client can be input by a annotator.
In some embodiments of the present application, when the annotation device is located at the client, the annotation device displays the 3D point cloud data through a human-computer interface, and receives the road plane element type and the element region parameter input by the annotator through the human-computer interface.
Further, on the basis of the above embodiment, in step 301, when the annotation device displays the 3D point cloud data, the 3D point cloud data may be displayed according to a designated display direction. The designated display direction may be a preset display direction or a display direction input by a annotator.
For example, in some embodiments, after the labeling device reads a frame of 3D point cloud data, the frame of 3D point cloud data may be displayed according to a preset display direction. For another example, in some embodiments, when the annotator needs to carefully observe the scene or object represented by the 3D point cloud data, a desired display direction can be selected and input, and the annotating device displays the 3D point cloud data according to the received display direction, so as to facilitate observation and identification by the annotator.
After a marker observes a scene expressed by the displayed 3D point cloud data, objects in the scene are marked. When the annotator annotates a road plane element, the type of the road plane element to be annotated and the element region parameters can be input.
In an embodiment of the present application, in step 303, the type of the road plane element received by the annotation device includes at least one of the following: lane lines, road drivable zones, road undriven zones, ramps, carrousels, road edges, etc. The road plane element may further include other elements according to the requirements of the application scenario, and the application is not specifically limited herein. For different types of road plane elements, element area parameters and element area generation rules can be preset, so that the labeling device generates element areas according to the received element area parameters and the corresponding generation rules. The received element region parameters may include a region shape for identifying the shape of the road plane element and a predetermined number of points in 3D space for indicating a location at which the region shape is generated.
The annotator can input the type of the road plane element to be annotated and the element area parameters in various ways. For example, when the labeling device is located at the client, a pull-down menu or a button can be provided on the human-computer interface to allow a labeling operator to select the type of the road plane element to be labeled and the area shape in the element area parameter, the labeling operator can also click the coordinates of a predetermined number of points in the element area parameter in a three-dimensional coordinate space displaying the 3D point cloud data, and the labeling device receives the selection result of the labeling operator through the human-computer interface; or when the marking device is positioned at the server side, the type of the selectable road plane element and the area shape in the element area parameter can be sent to the client side, the marker selects the type of the road plane element to be marked and the area shape in the element area parameter from the selection on the client side according to the received option content, selects the coordinates of a predetermined number of points included in the element area parameter, and feeds back the selection result to the marking device.
In the embodiment of the present application, in the step 305, the labeling device may generate the element region on the XY plane according to at least the following various manners.
First, in some embodiments, as shown in fig. 4, the process of generating the element region on the XY plane includes:
and step 402, projecting the generated space plane onto an XY plane to obtain a corresponding element area. Illustratively, the labeling device orthogonally projects the generated spatial plane onto the XY plane to obtain the corresponding element region.
Where the projection direction is a direction parallel to the Z-axis in the three-dimensional space coordinate system, the data point may be orthogonally projected onto the XY plane. The orthogonal projection processing in the embodiment of the present application obtains coordinates (x, y, 0) on the XY plane by performing transformation processing on three-dimensional coordinates (x, y, z) of the data point.
In a second way, in some embodiments, as shown in fig. 5, the process of generating the element region on the XY plane includes:
For example, when performing perspective projection, the current display direction may be determined as the projection direction.
In general, the perspective projection process is to perform an arithmetic transformation on three-dimensional coordinates (x, y, z) of one point to obtain projection point coordinates (x ', y ', z ') that satisfy a projection condition (e.g., a projection direction). In the present application, a data point is perspective-projected onto an XY plane, and the coordinates of the obtained projection point are (x ', y', 0). In the embodiment of the present application, the process of performing perspective projection on the spatial region according to the display direction may be implemented by a perspective algorithm. The perspective algorithm may adopt a perspective algorithm before the present application or a perspective algorithm after the present application, and the embodiment of the present application is not particularly limited thereto.
In a third mode, in some embodiments, as shown in fig. 6, the process of generating the element region on the XY plane includes:
601, projecting the received points in the 3D space with the preset number onto an XY plane to obtain corresponding projection points; exemplarily, the labeling device orthogonally projects the received predetermined number of points in the 3D space onto the XY plane to obtain corresponding projection points;
step 602 is to generate an element region on the corresponding XY plane based on the projected points and a predetermined generation rule corresponding to the received region shape.
The orthogonal projection is as described above and will not be described in detail here.
Means four, in some embodiments, as shown in fig. 7, the process of generating the element region on the XY plane includes:
for example, when the perspective projection processing is performed, the display direction may be determined as the projection direction.
As described above, the process of perspective-projecting the points in space according to the display direction can be realized by a perspective-projection algorithm.
Step 702 generates an element region on the corresponding XY plane based on the projected points and a predetermined generation rule corresponding to the received region shape.
The above methods are only exemplary to illustrate several ways of generating the element region on the XY plane, and the present application is not limited to these methods, and may also be generated by other methods according to the requirements of the application scenario, and the present application is not particularly limited to this.
In the embodiment of the present application, the element region parameter may have a preset condition and have a corresponding generation rule.
For example, in some embodiments of the present application, when a annotator needs to label a rectangular region, the element region parameters received by the labeling device may include: the region is rectangular and has two points, and the generation rule corresponding to the rectangle includes: and determining two points as diagonal points, and generating a rectangular area according to the two diagonal points. As shown in fig. 8a, a (x1, y1) and b (x2, y2) are two points for generating a rectangle, and a and b are taken as points on a diagonal line, so that points a '(x 2, y1) and b' (x1, y2) on the other diagonal line of the rectangle are obtained, and the rectangle R can be generated through the points a, a ', b'.
For example, in some embodiments of the present application, when a annotator needs to label a rectangular region, the element region parameters received by the labeling device may include: the region is in a rectangular shape and comprises two points, and the generation rule corresponding to the rectangular shape comprises the steps of determining the two points as two end points and generating a straight line segment between the two end points; and vertically translating and copying the linear line segment on the preset side of the generated linear line segment according to the preset displacement width corresponding to the road plane element type, and generating a rectangular area by taking the generated linear line segment and the linear line segment obtained by translation and copying as the sides of a rectangle. The preset displacement width can be set according to a proportional relation between the size of the data points in the 3D point cloud data and the size of the actual object. As shown in fig. 8b, c and d are two points for generating a rectangle, a straight line cd between c and d is generated, and the copied straight line cd is vertically translated at a preset side according to a displacement width corresponding to the type of the road plane element to obtain a straight line c'd ', and a rectangular region R ' is obtained by the straight lines cd and c'd '.
For example, in some embodiments of the present application, when the annotator needs to annotate a polygonal area, the parameter of the element area received by the annotation device may include: the area shape is a polygon and a plurality of points, and the generation rule corresponding to the polygon includes: and according to the sequence relation of the received points, sequentially generating a straight line segment between two adjacent points, and generating a straight line segment between the starting point and the end point to obtain a polygonal area. As shown in fig. 8c, o, P, q, r, and s are points having a receiving sequence relationship, o is a starting point, and s is an end point, and a connecting line between adjacent points and a connecting line between the starting point o and the end point s are generated between the points, i.e. the polygon P is obtained.
For example, in some embodiments of the present application, when the annotator needs to annotate a circular region, the element region parameters received by the annotation device may include: the region shape is a circle and two points, and the generation rule corresponding to the circle comprises the following steps: and according to a preset point relation, one point of the two points is used as a central point, and the other point is used as a point on the circumference, so that a circular area is generated. For example, the order of receiving the points may be such that the first point is the center point and the second point is the point on the circumference, or the reverse of this. As shown in fig. 8d, x and y are two points having a receiving order relationship, and a circle C with a radius xy can be obtained by taking x as a central point and y as a point on the circumference.
For example, in some embodiments of the present application, when the annotator needs to annotate a straight line with a predetermined width, the element region parameters received by the annotation device may include: the region shape is a straight line having a predetermined width, two points, and the generation rule corresponding to the straight line having the predetermined width includes: respectively generating two circles by taking the two points as circle centers and the preset width as a radius; generating a straight line by taking two circle centers as end points, and vertically translating and copying the straight line to two sides of the straight line to obtain a straight line intersected with the circumferences of two circles; thereby obtaining a straight line having a predetermined width composed of two circles and two straight lines intersecting the circumference. As shown in fig. 8e, u and v are two points received by the labeling device, two circles C1 and C2 are generated with u and v as the center and a predetermined width w as the radius, respectively, u and v are connected to obtain a straight line uv, and the straight line is vertically translated and copied to both sides of the straight line uv to obtain straight lines u 'v' and u "v" intersecting the circumferences of the two circles, thereby obtaining a closed region u 'u "v" v', which is a straight line S having a predetermined width.
The generation rules of different shapes of element regions are exemplarily described above, and the embodiment of the present application is not limited to the above embodiments, and may include other shapes of element region parameters and generation rules. The element region parameters, the generation rules, and the process of generating region elements may also be updated and modified, and may be imported by a annotator or a user to adapt to different scenes and annotation needs, or may be imported by a developer of the annotation system 1 through upgrading.
In the above embodiment, the generated element regions such as rectangles and polygons are all element regions with straight sides. Further, in some embodiments of the present application, an element region with a curved edge may be obtained by processing, and the irregular-shaped road plane element is expressed by the element region with the curved edge.
As shown in fig. 9, the processing for obtaining the element region of the curved edge by the labeling device includes:
after the element region is generated by the marking device, the element region can be displayed so as to be convenient for a marker to observe; when a marker needs to modify the sides of an element area such as a rectangle or a polygon into a curve, the marker selects a curve side button or a key or inputs a corresponding curve side command, and a marking device receives a curve side request;
after the annotator selects the curve edge request, an edge of the element region can be selected in various ways, for example, the annotating device marks an identifier for each edge of the element region, and the annotator selects the corresponding edge by selecting the identifier; or the marking device displays the element area, the marker clicks the side of the element area to be selected, and the marking device compares the coordinate value of the position clicked by the marker with the coordinate value of a point on each side, or compares the position clicked by the marker with the distance between each side to determine one side selected by the marker; the embodiment of the application can also comprise a plurality of selection modes which are not listed;
the marking device generates at least one control point on the received selected edge, and a marker can modify the shape of the straight line edge by modifying the position of the control point; wherein the position of at least one control point generated on one side may be predetermined, for example, in the case of generating one control point, one control point may be generated at a position of one-half of the side, and in the case of generating two control points, one control point may be generated at a position of one-third and a position of two-thirds of the side, respectively;
the annotator can select a control point and select a new position of the control point, namely a displacement position; for example, the labeling device may mark each control point with an identifier, and the labeler selects the corresponding control point by selecting the identifier; or the control point is displayed on the side selected by the annotator by the annotating device, the annotator clicks the control point to be selected, and the control point selected by the annotator is determined by the annotator through position comparison by the annotating device.
Further, after the marking device determines the selected control point, receiving the displacement position of the control point input by a marker, wherein the marker can input the coordinate value of the displacement position and can also click a position in a coordinate system for displaying the 3D point cloud data as the displacement position; after receiving the displacement position of one control point, the marking device modifies the position of the control point into the displacement position;
and 904, generating a Bezier curve between the two end points according to the positions of the two end points of the edge and the position of the at least one control point.
The processing for generating the bezier curve may be performed according to an existing algorithm or formula, and the present application is not limited in detail herein. Fig. 9 is merely an exemplary process for modifying a straight line edge into a curved line edge, and in different application scenarios, another algorithm or process may be used to generate a curved line edge, which is not limited in this application. An example of modifying one side of the rectangular element region into a curved side in one exemplary embodiment is shown in fig. 10. In this example, the annotator chooses to modify the edge ij to a curve edge, the annotator generates one control point k1 on the edge ij, the annotator modifies the position of the control point k1 to the position of k2, and the annotator generates a bezier curve from the points i, k2, and j.
After generating the element area on the XY plane, the labeling device needs to project the data points in the 3D point cloud data onto the XY plane to determine which projection points corresponding to the data points fall into the element area.
In the embodiment of the present application, the process of projecting the data points in the displayed 3D point cloud data onto the XY plane in step 307 may include two implementations, as shown in fig. 11, including step 307': orthogonally projecting data points in the displayed 3D point cloud data onto an XY plane; alternatively, as shown in fig. 12, a step 307 ″ is included: the marking device perspectively projects data points in the 3D point cloud data to an XY plane according to the display direction of the currently displayed 3D point cloud data; for example, the perspective projection processing is performed with the current display direction as the projection direction.
The processes of fig. 4 to 7 include a process of projecting the generated spatial plane region onto the XY plane to obtain an element region on the XY plane (step 402 and step 502), and a process of projecting a data point in the 3D space in the received element region parameter onto the XY plane (step 601 and step 701). When the orthogonal projection is adopted in one of the modes of fig. 4 to 7, the orthogonal projection may be adopted correspondingly in step 307', and when the perspective projection is adopted in one of the modes of fig. 4 to 7, the perspective projection may be adopted correspondingly in step 307 ″.
In the embodiments of the present application, other projection modes may also be used, and the present application is not limited specifically herein.
In an embodiment of the present application, step 309 may include the process of: determining projection points included in the element region, determining data points in the 3D point cloud data corresponding to the projection points, and determining the determined data points as points belonging to the received road plane element type. Wherein, determining the projected point included in the element region may be achieved by comparing the coordinate value of the element region and the coordinate value of the projected point.
In some embodiments of the present application, as shown in fig. 13, the annotating device may further perform step 304 after receiving the road plane element type in step 303: establishing a corresponding group for the received road plane element type; also, after step 309, step 311 may be further performed: the data points in the 3D point cloud data determined to belong to the type are sorted into respective groupings.
Through grouping operation, it can be determined which data points in the 3D point cloud data are specifically included in different road plane elements, so that an effective data base can be provided for subsequent processing or application.
According to the technical scheme provided by the embodiment of the application, the element area on the XY plane is generated according to the input element area parameters, the data points in the 3D point cloud data are projected onto the XY plane, and the data points of the corresponding projection points included in the element area are determined to be the points belonging to the input road plane element type, so that the plane elements on the road can be efficiently marked, and the speed and the efficiency of marking objects in the 3D point cloud data are improved.
Some embodiments of the present application also provide a non-volatile storage medium having at least one machine executable instruction stored therein, the at least one machine executable instruction being executed by a processor to implement one or more of the processes of fig. 3-7, 9, and 11-13.
Fig. 14 is a processing flow diagram illustrating a labeling method provided in an application scenario according to an embodiment of the present application, including:
1401, displaying 3D point cloud data of a frame of road scene on a human-computer interface by a marking device; for example, the annotator selects to display the 3D point cloud data in a direction parallel to the Z-axis in the 3D coordinate system, as shown in the left diagram in fig. 15 a;
for example, the points selected by the annotator are the two diagonal points of the rectangular region, as shown by C1 and C2 in FIG. 15 a;
1406, the marking device orthogonally projects data points in the 3D point cloud data onto an XY plane;
In another application scenario, a similar processing procedure as that shown in fig. 14 is applied to label the travelable region of the road in the 3D point cloud data. An example of the data points included in the 3D point cloud data to mark the travelable region XS1 is shown in fig. 15 b. In fig. 15b, the left diagram shows a frame of 3D point cloud data, the middle diagram includes an element region superimposed on the 3D point cloud data, i.e., a travelable region XS1, and the right diagram includes data points labeled as belonging to a travelable region XS 1.
According to the technical scheme provided by the embodiment of the application, the element area on the XY plane is generated according to the input element area parameters, the data points in the 3D point cloud data are projected onto the XY plane, and the data points of the corresponding projection points included in the element area are determined to be the points belonging to the input road plane element type, so that the plane elements on the road can be efficiently marked, and the speed and the efficiency of marking objects in the 3D point cloud data are improved.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.
Claims (17)
1. A method for labeling road plane elements in 3D point cloud data is characterized by comprising the following steps:
the marking device displays 3D point cloud data of a frame of road scene;
receiving an input road plane element type and an input element area parameter;
generating an element area on an XY plane according to the element area parameter;
projecting data points in the displayed 3D point cloud data onto an XY plane;
the data point whose corresponding projected point is included in the element region is determined as a point belonging to the received road plane element type.
2. The method of claim 1, wherein the annotation device comprises a human machine interface; the marking device displays 3D point cloud data of a frame of road scene through a human-computer interface; receiving an input road plane element type through a human-computer interface, and receiving an input element region parameter; or,
the marking device receives the road plane element type and the element area parameter from the client.
3. The method of claim 1, wherein the annotating device displays 3D point cloud data of a frame of road scene comprising:
displaying 3D point cloud data of a frame of road scene according to a preset display direction; or,
and receiving an input display direction, and displaying the 3D point cloud data of a frame of road scene according to the received display direction.
4. The method of claim 1, wherein the received element region parameters comprise: a region shape and a predetermined number of points in 3D space for generating the region shape;
generating an element region on the XY plane from the element region parameters, including:
generating a corresponding space plane according to the received points in the 3D space with the preset number and a preset generation rule corresponding to the received region shape, and projecting the generated space plane onto an XY plane to obtain a corresponding element region; or,
projecting the received points in the 3D space with the preset number onto an XY plane to obtain corresponding projection points; and generating the element area on the corresponding XY plane according to the projection point and a preset generation rule corresponding to the received area shape.
5. The method of claim 4, wherein projecting the generated spatial region onto an XY plane results in a corresponding elemental region, comprising:
orthogonally projecting the generated space plane to an XY plane to obtain a corresponding element area; or,
and according to the display direction of the currently displayed 3D point cloud data, the generated space area is subjected to perspective projection on an XY plane to obtain a corresponding element area.
6. The method of claim 4, wherein projecting the received predetermined number of points in 3D space onto an XY plane results in corresponding projected points, comprising:
orthogonally projecting the received points in the 3D space with the preset number onto an XY plane to obtain corresponding projection points; or,
and according to the display direction of the currently displayed 3D point cloud data, carrying out perspective projection on the received points in the 3D space with the preset number onto an XY plane to obtain corresponding projection points.
7. The method of claim 4, wherein the received element region parameters comprise: the shape of the region is rectangular and two points;
the generation rule corresponding to the rectangle includes: and determining two points as diagonal points, and generating a rectangular area according to the two diagonal points.
8. The method of claim 4, wherein the received element region parameters comprise: the shape of the region is rectangular and two points;
the generation rule corresponding to the rectangle includes: determining two points as two end points and generating a straight line segment between the two end points; and vertically translating and copying the linear line segment on the preset side of the generated linear line segment according to the preset displacement width corresponding to the road plane element type, and generating a rectangular area by taking the generated linear line segment and the linear line segment obtained by translation and copying as the sides of a rectangle.
9. The method of claim 4, wherein the received element region parameters comprise: the area is in the shape of a polygon and a plurality of points;
the generation rule corresponding to the polygon includes: and according to the sequence relation of the received points, sequentially generating a straight line segment between two adjacent points, and generating a straight line segment between the starting point and the end point to obtain a polygonal area.
10. The method of claim 4, wherein the received element region parameters comprise: the shape of the area is circular and two points;
the generation rule corresponding to the circle includes: and according to a preset point relation, one point of the two points is used as a central point, and the other point is used as a point on the circumference, so that a circular area is generated.
11. The method of claim 4, wherein the received element region parameters comprise: the shape of the region is a straight line with a preset width and two points;
the generation rule corresponding to a straight line having a predetermined width includes: respectively generating two circles by taking the two points as circle centers and the preset width as a radius; and generating a straight line by taking the two circle centers as end points, vertically translating and copying the straight line to two sides of the straight line respectively to obtain a straight line intersected with the circumferences of the two circles, and obtaining a straight line with a preset width formed by the two circles and the two straight lines intersected with the circumferences.
12. The method according to claim 1, further comprising, after generating the element region on the XY plane:
receiving a curve edge request;
receiving an edge of the selected element region, and generating at least one control point on the selected edge;
for a control point, receiving an input displacement position of the control point, and modifying the position of the control point into the received displacement position;
and generating a Bezier curve between the two end points according to the positions of the two end points of the edge and the position of the at least one control point.
13. The method of claim 1, wherein projecting data points in the displayed 3D point cloud data onto an XY plane comprises:
orthogonally projecting data points in the displayed 3D point cloud data onto an XY plane; or
And projecting the data points in the 3D point cloud data onto an XY plane according to the display direction of the currently displayed 3D point cloud data.
14. The method of claim 1, wherein after receiving the input road plane element type, the method further comprises: establishing a group corresponding to the received road plane element type;
after determining that the corresponding projected point is included in the element region as a point belonging to the received road plane element type, the method further comprises: the points belonging to the received road plane element type are classified into corresponding groups.
15. The method of claim 1, wherein the road plane element type comprises at least one of: lane lines, road travelable areas, road non-travelable areas, ramps, carrousels, road edges.
16. An apparatus for labeling road surface elements in 3D point cloud data, comprising a processor and at least one memory, at least one machine executable instruction being stored in the at least one memory, the processor executing the at least one machine executable instruction to perform the method according to any one of claims 1 to 15.
17. A non-volatile storage medium having stored thereon at least one machine executable instruction, the at least one machine executable instruction when executed by a processor implementing a method as claimed in any one of claims 1 to 15.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010757801.8A CN112037120A (en) | 2020-07-31 | 2020-07-31 | Method and device for labeling road plane elements in 3D point cloud data and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010757801.8A CN112037120A (en) | 2020-07-31 | 2020-07-31 | Method and device for labeling road plane elements in 3D point cloud data and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112037120A true CN112037120A (en) | 2020-12-04 |
Family
ID=73581944
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010757801.8A Pending CN112037120A (en) | 2020-07-31 | 2020-07-31 | Method and device for labeling road plane elements in 3D point cloud data and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112037120A (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108694882A (en) * | 2017-04-11 | 2018-10-23 | 百度在线网络技术(北京)有限公司 | Method, apparatus and equipment for marking map |
CN109658504A (en) * | 2018-10-31 | 2019-04-19 | 百度在线网络技术(北京)有限公司 | Map datum mask method, device, equipment and storage medium |
CN109683170A (en) * | 2018-12-27 | 2019-04-26 | 驭势科技(北京)有限公司 | A kind of image traveling area marking method, apparatus, mobile unit and storage medium |
CN109978955A (en) * | 2019-03-11 | 2019-07-05 | 武汉环宇智行科技有限公司 | A kind of efficient mask method for combining laser point cloud and image |
CN110135396A (en) * | 2019-05-27 | 2019-08-16 | 百度在线网络技术(北京)有限公司 | Recognition methods, device, equipment and the medium of surface mark |
CN110136273A (en) * | 2019-03-29 | 2019-08-16 | 初速度(苏州)科技有限公司 | A kind of sample data mask method and device in machine learning |
CN110163904A (en) * | 2018-09-11 | 2019-08-23 | 腾讯大地通途(北京)科技有限公司 | Object marking method, control method for movement, device, equipment and storage medium |
CN110264468A (en) * | 2019-08-14 | 2019-09-20 | 长沙智能驾驶研究院有限公司 | Point cloud data mark, parted pattern determination, object detection method and relevant device |
CN110858415A (en) * | 2018-08-24 | 2020-03-03 | 北京图森未来科技有限公司 | Method and device for labeling object in 3D point cloud data |
CN111080805A (en) * | 2019-11-26 | 2020-04-28 | 北京云聚智慧科技有限公司 | Method and device for generating three-dimensional block diagram of marked object, electronic equipment and storage medium |
CN111210456A (en) * | 2019-12-31 | 2020-05-29 | 武汉中海庭数据技术有限公司 | High-precision direction arrow extraction method and system based on point cloud |
-
2020
- 2020-07-31 CN CN202010757801.8A patent/CN112037120A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108694882A (en) * | 2017-04-11 | 2018-10-23 | 百度在线网络技术(北京)有限公司 | Method, apparatus and equipment for marking map |
CN110858415A (en) * | 2018-08-24 | 2020-03-03 | 北京图森未来科技有限公司 | Method and device for labeling object in 3D point cloud data |
CN110163904A (en) * | 2018-09-11 | 2019-08-23 | 腾讯大地通途(北京)科技有限公司 | Object marking method, control method for movement, device, equipment and storage medium |
CN109658504A (en) * | 2018-10-31 | 2019-04-19 | 百度在线网络技术(北京)有限公司 | Map datum mask method, device, equipment and storage medium |
CN109683170A (en) * | 2018-12-27 | 2019-04-26 | 驭势科技(北京)有限公司 | A kind of image traveling area marking method, apparatus, mobile unit and storage medium |
CN109978955A (en) * | 2019-03-11 | 2019-07-05 | 武汉环宇智行科技有限公司 | A kind of efficient mask method for combining laser point cloud and image |
CN110136273A (en) * | 2019-03-29 | 2019-08-16 | 初速度(苏州)科技有限公司 | A kind of sample data mask method and device in machine learning |
CN110135396A (en) * | 2019-05-27 | 2019-08-16 | 百度在线网络技术(北京)有限公司 | Recognition methods, device, equipment and the medium of surface mark |
CN110264468A (en) * | 2019-08-14 | 2019-09-20 | 长沙智能驾驶研究院有限公司 | Point cloud data mark, parted pattern determination, object detection method and relevant device |
CN111080805A (en) * | 2019-11-26 | 2020-04-28 | 北京云聚智慧科技有限公司 | Method and device for generating three-dimensional block diagram of marked object, electronic equipment and storage medium |
CN111210456A (en) * | 2019-12-31 | 2020-05-29 | 武汉中海庭数据技术有限公司 | High-precision direction arrow extraction method and system based on point cloud |
Non-Patent Citations (1)
Title |
---|
MING CHENG 等: "Extraction and Classification of Road Markings Using Mobile Laser Scanning Point Clouds", IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101561830B (en) | Dimension marking method for CAD and device thereof | |
Mahoney | Image chunking: Defining spatial building blocks for scene analysis | |
US6526166B1 (en) | Using a reference cube for capture of 3D geometry | |
CN104732565B (en) | A kind of security protection layout drawing practice based on monitoring range the visual design | |
CN108334802A (en) | The localization method and device of roadway characteristic object | |
CN107870911B (en) | Data visualization display method and equipment | |
CN115762218A (en) | Parking space map generation method, device, equipment and storage medium | |
CN111695497A (en) | Pedestrian identification method, medium, terminal and device based on motion information | |
CN113610963B (en) | Three-dimensional network topology drawing method, device, computing equipment and storage medium | |
KR930009765B1 (en) | Method and apparatus for obtaining image data used for filling inner or outer region of graphic figure | |
CN112037120A (en) | Method and device for labeling road plane elements in 3D point cloud data and storage medium | |
CN113126120A (en) | Data annotation method, device, equipment, storage medium and computer program product | |
CN113345052A (en) | Classified data multi-view visualization coloring method and system based on similarity significance | |
CN115879186B (en) | Method, device, equipment and storage medium for determining placement position of part number | |
CN115847384B (en) | Mechanical arm safety plane information display method and related products | |
CN110852138B (en) | Method and device for labeling object in image data | |
CN113284052A (en) | Image processing method and apparatus | |
CN116468870A (en) | Three-dimensional visual modeling method and system for urban road | |
CN116248935A (en) | Focus switching method, device, interactive network television system and computer readable medium | |
CN115641436A (en) | Region segmentation method and device and robot device | |
JP2002074381A (en) | Device and method for editing graphic | |
CN114911433A (en) | Intelligent CAD drawing framing method and system and readable storage medium | |
CN114839976A (en) | Path planning method for farmland with complex boundary and farmland machine operating system | |
CN114092754A (en) | Method and device for labeling object in 3D point cloud data and storage medium | |
CN112036443A (en) | Method and device for labeling object contour in image data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |