CN113703455A - Semantic information labeling method of laser point cloud and related equipment - Google Patents
Semantic information labeling method of laser point cloud and related equipment Download PDFInfo
- Publication number
- CN113703455A CN113703455A CN202110994744.XA CN202110994744A CN113703455A CN 113703455 A CN113703455 A CN 113703455A CN 202110994744 A CN202110994744 A CN 202110994744A CN 113703455 A CN113703455 A CN 113703455A
- Authority
- CN
- China
- Prior art keywords
- polygon
- target
- point
- intersection point
- list
- 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
- 238000002372 labelling Methods 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 47
- 238000012545 processing Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 abstract description 9
- 238000004891 communication Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 4
- 230000006399 behavior Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008447 perception Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0246—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0257—Control of position or course in two dimensions specially adapted to land vehicles using a radar
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
Abstract
The application discloses a semantic information labeling method of laser point cloud and related equipment, comprising the following steps: acquiring a first polygon list marked as an interesting area and a second polygon list marked as a non-driving area in a laser point cloud image; judging whether each first polygon in the first polygon list is overlapped with any second polygon in the second polygon list; if yes, aiming at the overlapped target first polygon and the target second polygon, generating a new first polygon to replace the target first polygon in the first polygon list according to the difference set of the target first polygon and the target second polygon; and marking each first polygon in the final first polygon list as a travelable area until all first polygons in the first polygon list do not overlap with any second polygon in the second polygons. The calculation required in the calculation process is not complex and the calculation amount is not large, and the semantic information marking of the laser point cloud can be quickly and efficiently realized.
Description
Technical Field
The application relates to the technical field of automatic driving, in particular to a semantic information labeling method of laser point cloud and related equipment.
Background
In recent years, with the development of artificial intelligence technology, various landing scenes thereof are gradually put into use, such as AI medical treatment, face recognition payment technology, intelligent robots, and automatic driving technology. The automatic driving technology is an important technology, is still in a testing stage of continuous iteration of the technology, and is still continuously invested in resources at home and abroad to carry out related research and development work.
The automatic driving technology is a complex engineering system, all-around driving safety can be guaranteed only by mutual cooperation of all modules, and the modules are roughly divided into four modules of environment perception, behavior decision, path planning and behavior control.
The environment sensing module comprises a static object, a dynamic object and a road surface. For a moving object, not only object detection but also a trajectory thereof is tracked, and a next position of the object is predicted based on a tracking result. This is essential in urban roads, and especially in cities with disordered traffic conditions, environmental perception for moving objects is particularly important. When the unmanned vehicle runs on the road, as the pedestrians can continuously pass through the front of the vehicle, if the pedestrians or the vehicle stops once the pedestrians or the vehicle in front is detected, the pedestrians or the vehicle can never reasonably pass through all intersections. The driver will roughly estimate the next position from the pedestrian's trajectory of movement and then calculate the safety space from the pedestrian's speed, i.e. the path planning process. The process is not only directed to single object detection and tracking, but also to multi-target detection and tracking in most cases. The unmanned vehicle must also be able to do this.
The behavior decision module needs to collect all important peripheral information of the vehicle, including not only the current speed, direction, position and driving lane of the unmanned vehicle itself, but also obstacle information within a certain distance of the unmanned vehicle, and needs to predict the trajectories of other obstacles. The behavior decision is used as a module for collecting information in an automatic driving system, and the problem to be solved is to take the output of an environment sensing module as input and decide the driving strategy of the unmanned vehicle according to the knowledge of the information.
The path planning module finds a better passable path according to a certain path searching algorithm based on the surrounding environment information provided by the environment sensing module and the determined position of the vehicle in the environment, and then realizes that the unmanned vehicle automatically navigates to the determined position. According to the integrity of the ambient working environment information sensed by the environment sensing module, the path planning method comprises a local path planning method which depends on the environment information acquired by the environment sensing module in real time as fusion environment information and a global path planning method which depends on global ambient environment information.
The motion control module includes longitudinal and lateral vehicle controls. Lateral control is through a given speed, and the vehicle can move along a predetermined trajectory by controlling travel and steering. The longitudinal control is to set a proper speed in the driving process of the vehicle, and the cooperation of horizontal direction control is needed mostly, so that the stability, the comfort and especially the safety of the automatic driving vehicle are achieved together, and excellent riding experience is provided for customers.
From the above analysis, the environmental perception model plays a very critical role under the whole system of unmanned driving. As road scenes become more complex, it is difficult to understand the scene well if only 2D or 3D boxes can be recognized. And acquiring label data, which is the most critical part for constructing the environment perception model. Thus, semantic information for the point granularity of each point cloud is urgently needed.
However, currently, semantic information of the point granularity of each point cloud is mostly marked manually, so that the marking cost is high, the marking speed is slow, and finally the label data size of the laser point cloud is small. How to increase the semantic information labeling speed of the laser point cloud point granularity is very important.
Disclosure of Invention
In view of this, the present application provides a semantic information labeling method for laser point cloud and related devices, which can solve the problems of high labeling cost and low labeling speed of laser point cloud data to a certain extent.
In order to achieve the above object, a first aspect of the present application provides a semantic information labeling method for laser point cloud, including:
acquiring at least one first polygon marked as an interesting area in a laser point cloud image, wherein the at least one first polygon forms a first polygon list;
acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and marked as a non-driving region, wherein the at least one second polygon forms a second polygon list;
for each first polygon in the first polygon list, judging whether the first polygon is overlapped with any second polygon in the second polygon list;
if yes, aiming at the overlapped target first polygon and the overlapped target second polygon, generating a new first polygon to replace the target first polygon in a first polygon list according to a difference set of the target first polygon and the target second polygon, wherein the new first polygon and the target second polygon are not overlapped;
and returning to execute the step of judging whether the first polygon overlaps with any one second polygon in the second polygon list or not for each first polygon in the first polygon list, and marking each first polygon in the final first polygon list as a travelable area when all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list.
Preferably, the first polygon and the second polygon are both simple polygons;
a process of generating a new first polygon to replace the target first polygon in the first polygon list according to a difference set between the target first polygon and the target second polygon, comprising:
judging whether the difference set forms a non-simple polygon or not;
if so, splitting the non-simple polygon corresponding to the difference set into at least two new first polygons; adding the at least two new first polygons to the first polygon list and removing the target first polygon from the first polygon list;
if not, generating at least one new first polygon according to the difference set; adding the new first polygon to the first polygon list and removing the target first polygon from the first polygon list.
A process of splitting the non-simple polygon corresponding to the difference set into at least two new first polygons, comprising:
determining a reference point from within said target second polygon;
setting a reference straight line, wherein the reference straight line passes through the reference point and intersects the target first polygon and the target second polygon at a plurality of points;
determining a first intersection point, a second intersection point, a third intersection point and a fourth intersection point from the plurality of points; the first intersection point and the second intersection point are located on the first side of the reference point, and the third intersection point and the fourth intersection point are located on the second side of the reference point; the first intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the first side, and the fourth intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the second side; the second intersection point is the intersection point which is farthest from the reference point in the intersection points of the first side of the target second polygon, and the third intersection point is the intersection point which is farthest from the reference point in the intersection points of the second side of the target second polygon;
and taking a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries, and splitting the non-simple polygon corresponding to the difference set into two new first polygons.
Preferably, the process of splitting the non-simple polygon corresponding to the difference set into two new first polygons by using a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries includes:
taking the first intersection point as a starting point, and picking up a first vertex of the target first polygon along the clockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a second vertex of the target second polygon along the counterclockwise direction of the target second polygon until the second intersection point; the first intersection point, the first vertex, the fourth intersection point, the third intersection point, the second vertex and the second intersection point are connected in sequence to obtain a new first polygon;
taking the first intersection point as a starting point, and picking up a third vertex of the target first polygon along the counterclockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a fourth vertex of the target second polygon along the clockwise direction of the target second polygon until the third intersection point is reached; and sequentially connecting the first intersection point, the third vertex, the fourth intersection point, the third intersection point, the fourth vertex and the second intersection point to obtain another new first polygon.
Preferably, the process of determining a reference point from within said target second polygon comprises:
when the target second polygon is a triangle: determining a centroid of the target second polygon as the reference point;
when the target second polygon is not a triangle:
when the target second polygon is a triangle: determining a centroid of the target second polygon as the reference point;
when the target second polygon is not a triangle:
determining three target vertexes which are adjacent in sequence from the vertexes of the target second polygon, and generating a triangle by taking the target vertexes as the vertexes, wherein other vertexes of the target second polygon are all in the triangle;
determining a point from the edge of the triangle inside the target second polygon as a candidate reference point;
judging whether the candidate reference point is inside the target second polygon;
if so, determining the candidate reference point as the reference point;
if not, returning to the step of determining three target vertexes which are sequentially adjacent from the vertexes of the target second polygon until the reference point is determined.
Preferably, the process of determining a point as a candidate reference point from the edge of the triangle inside the target second polygon comprises:
and determining the middle point of the edge of the triangle positioned in the target second polygon as the candidate reference point.
Preferably, the process of determining whether the difference set constitutes a non-simple polygon comprises:
generating at least one polygon from the difference set;
for each polygon, determining whether the polygon has the following conditions:
one vertex of the polygon is on one edge of the polygon;
one side of the polygon intersects the other side;
and if so, determining that the difference set forms a non-simple polygon.
Preferably, the determining whether there is an overlap between the first polygon and any second polygon in the second polygon list includes:
determining, for each second polygon in a second polygon list, whether there is overlap of the first polygon with the second polygon by calculating an intersection ratio of the first polygon to the second polygon.
Preferably, the reference straight line is a straight line which is parallel to the horizontal axis or the vertical axis and passes through the reference point under the reference coordinate system.
The second aspect of the present application provides a semantic information labeling apparatus for laser point cloud point granularity, including:
the system comprises an interesting region acquisition unit, a laser point cloud image processing unit and a processing unit, wherein the interesting region acquisition unit is used for acquiring at least one first polygon marked as an interesting region in a laser point cloud image, and the at least one first polygon forms a first polygon list;
the non-driving area acquisition unit is used for acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and is marked as a non-driving area, and the at least one second polygon forms a second polygon list;
the area overlapping judging unit is used for judging whether the first polygon overlaps with any second polygon in the second polygon list or not for each first polygon in the first polygon list;
a region processing unit, configured to generate, for a target first polygon and a target second polygon that overlap, a new first polygon to replace the target first polygon in a first polygon list according to a difference set between the first polygon and the second polygon, where the new first polygon does not overlap with the target second polygon, and return to an operation directed to the region overlap determining unit;
and the travelable area labeling unit is used for labeling each first polygon in the final first polygon list as a travelable area when all first polygons in the first polygon list do not overlap with any second polygon in the second polygons.
The third aspect of the present application provides a semantic information labeling apparatus for laser point cloud point granularity, including: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program and realizing the steps of the semantic information labeling method of the laser point cloud.
A fourth aspect of the present application provides a storage medium, on which a computer program is stored, wherein the computer program, when being executed by a processor, implements the steps of the above method for labeling semantic information of laser point cloud.
According to the technical scheme, only the data marked as the region of interest and the non-driving region in the region of interest in the laser point cloud image are acquired, and the region of interest and the non-driving region are represented in a polygonal form. The region of interest is usually a road on which a target is to travel, the non-travelable region is usually an obstacle region such as a pedestrian or other vehicle on the road or a region of a pothole on the road, and the required labeling workload is relatively small, so that a large amount of polygon data of the region of interest and the non-travelable region can be easily acquired.
After the region of interest and the non-drivable region are characterized in a polygonal form, a first polygon of the region of interest is saved in a first polygon list, and a second polygon of the non-drivable region is saved in a second polygon list. Then, whether the polygon in the first polygon list overlaps with any polygon in the second polygon list is judged. And for the overlapped target first polygon and the target second polygon, generating a new first polygon to replace the target first polygon in the first polygon list according to the difference set of the first polygon and the second polygon. And the new first polygon is not overlapped with the target second polygon, so that the generated new first polygon is ensured to have no area incapable of driving.
And then returning to the step of judging whether the polygons in the first polygon list overlap with any polygon in the second polygon list, and judging the polygons (including the original first polygon and the generated new first polygon) in the first polygon list one by one until all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list, which means that the regions where the non-travelable regions exist in the first polygons in the first polygon list are all removed, and at this time, each first polygon in the final first polygon list represents the travelable region in the region of interest, namely, the travelable region can be marked.
The above-described process of acquiring the travelable region requires no complicated calculation and is not large in calculation amount. According to the technical scheme, the semantic information marking of the laser point cloud can be rapidly and efficiently realized, and the dilemma that the label data volume of the laser point cloud is small can be solved to a certain extent.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a schematic diagram of a semantic information labeling method for laser point cloud disclosed in an embodiment of the present application;
FIG. 2 illustrates a schematic diagram of splitting a non-simple polygon disclosed in an embodiment of the present application;
FIG. 3 illustrates a schematic diagram of determining reference points from within a polygon as disclosed in an embodiment of the present application;
fig. 4 is a schematic diagram of a semantic information labeling apparatus for laser point cloud disclosed in the embodiment of the present application;
fig. 5 is a schematic diagram of semantic information labeling equipment for laser point cloud disclosed in the embodiment of the present application.
Detailed Description
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.
The semantic information labeling method for laser point cloud provided by the embodiment of the application is described below. Referring to fig. 1, a semantic information labeling method for laser point cloud provided in the embodiment of the present application may include the following steps:
step S100, a first polygon list characterizing the region of interest is obtained.
Specifically, at least one first polygon marked as a region of interest in the laser point cloud image is acquired, and a first polygon list is composed of the first polygons.
The laser point cloud image can be obtained by analyzing the Rosbag file, and the laser point cloud image can be marked by a marking tool.
Specifically, the collected rossbag file may be preprocessed by the tool, and divided into an independent Lidar frame (Lidar frame) and an Image frame (Image frame) corresponding to the Lidar frame, and the Lidar frame and the Image frame are synchronized (sync) so that the specific Image frame corresponds to the Lidar frame. The Image frame is an Image of one frame, and the Lidar frame is a PCD file and is used for storing point cloud data of the laser radar, including information such as coordinates of points and intensity (intensity) of the points.
The annotator marks out a Region of Interest (ROI) in a Polygon (Polygon) form through an annotation tool according to various information such as point cloud forms, point cloud motion states, picture assistance and the like in each frame, and subsequent annotation contents can only concern the range within the ROI.
The Polygon may be represented by a list of points, such as Polygon1, which is surrounded by Polygon1 ═ point1, point2, point3, point4, …, and point n.
In step S200, a second polygon list representing the non-travelable area is obtained.
Specifically, at least one second polygon which is located in the region of interest in the laser point cloud image and is marked as a non-driving area is obtained, and a second polygon list is formed by the second polygons.
The non-drivable zones may include, among other things, areas occupied by various types of obstacles, such as vehicles, pedestrians, warning barricades, and the like. The specific labeling method and process are similar to those in step S100, and are not described herein again. When the ROI region and the non-drivable region are acquired, the non-drivable region is excluded from the ROI region, and the remaining region is the drivable region.
In step S300, it is determined whether there is an overlap between a first polygon in the first polygon list and any second polygon in the second polygon list.
Specifically, for each first polygon in the first polygon list, it is determined whether there is an overlap of the first polygon with any one of the second polygons in the second polygon list. The overlap between two polygons may be a partial overlap of the two polygons, or one of the polygons may completely contain the other polygon.
On the one hand, a certain first polygon in the first polygon list overlaps with a certain second polygon in the second polygon list, which means that a region (i.e., ROI region) in which the first polygon is located has a non-drivable region, and further, the included non-drivable region needs to be removed from the first polygon.
On the other hand, all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list, which means that there is no area in which all the first polygons in the first polygon list are located (i.e., ROI area), i.e., all the areas of the first polygons in the first polygon list constitute a travelable area.
Therefore, if there is a first polygon overlapping with a second polygon in the first polygon list, step S400 is performed; otherwise, if there is no overlap between all the first polygons in the first polygon list and any second polygon in the second polygons, step S500 is executed.
In step S400, a new first polygon is generated to replace the overlapped target first polygon in the first polygon list.
Specifically, for a target first polygon and a target second polygon which overlap, a new first polygon is generated according to a difference set of the target first polygon and the target second polygon, and the target first polygon in the first polygon list is replaced by the new first polygon. Wherein there is no overlap of the new first polygon with the target second polygon. Then, the process returns to step S300.
Wherein the difference refers to a new region obtained by removing a region belonging to the second polygon from the region of the first polygon.
It is understood that the new first polygon generated according to the difference set between the target first polygon and the target second polygon may be only one new first polygon, or two or more new first polygons.
In step S500, each first polygon in the final first polygon list is labeled as a travelable region.
Specifically, through the processing of steps S300 and S400, all of the first polygons in the first polygon list that overlap with the second polygons in the second polygon list have been replaced with new first polygons that do not overlap with any of the second polygons in the second polygon list. That is, the area in which the first polygon in the first polygon list (final first polygon list) at this time is located is not the no-travel area. Therefore, each first polygon in the final first polygon list may be labeled as a travelable region.
The calculation required in the process of acquiring the travelable region in the embodiment of the application is not complex and the calculation amount is not large. Through the technical scheme of the embodiment of the application, the semantic information marking of the laser point cloud can be quickly and efficiently realized, and the dilemma that the label data volume of the laser point cloud is small can be solved to a certain extent.
In some embodiments of the present application, the aforementioned first polygon and the second polygon are both simple polygons. Based on this, the step S400 of generating a new first polygon to replace the target first polygon in the first polygon list according to the difference set between the target first polygon and the target second polygon may include:
a1, judging whether the difference set forms a non-simple polygon; if yes, A2 is executed; if not, A3 is executed.
Where the difference set can be seen as a set of points. The judgment of the difference set can be performed according to the conditions that the edges and the vertexes in the simple polygon need to satisfy. If the difference set itself clearly constitutes a simple polygon, the corresponding simple polygon can be generated directly by a 3. Otherwise, it needs to be further processed by a2 to split the difference set into simple polygons.
A2, splitting the non-simple polygon corresponding to the difference set into at least two new first polygons; these new first polygons are added to the first polygon list and the target first polygon is removed from the first polygon list.
A3, generating at least one new first polygon according to the difference set; adding the new first polygon to the first polygon list and removing the target first polygon from the first polygon list.
It is understood that the first polygons generated in a2 and A3 also belong to simple polygons.
In some embodiments of the present application, the process of splitting the non-simple polygon corresponding to the difference set into at least two new first polygons by a2 may include:
b1, determining a reference point from inside the target second polygon.
B2, setting a reference straight line passing through the reference point and intersecting the target first polygon and the target second polygon at a plurality of points.
B3, determining a first intersection point, a second intersection point, a third intersection point and a fourth intersection point from the points in B2; wherein the first and second intersection points are located on a first side of the reference point and the third and fourth intersection points are located on a second side of the reference point. It will be appreciated that the first and second sides belong to different sides of the reference point, relative to the two sides of the reference point.
The first intersection point is the intersection point which is closest to the reference point in the intersection points of the first side of the target first polygon, and the fourth intersection point is the intersection point which is closest to the reference point in the intersection points of the second side of the target first polygon; the second intersection point is the intersection point which is farthest from the reference point in the intersection points of the first side of the target second polygon, and the third intersection point is the intersection point which is farthest from the reference point in the intersection points of the second side of the target second polygon.
And B4, dividing the non-simple polygon corresponding to the difference set into two new first polygons by using a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries.
In some embodiments of the present application, the reference straight line of B2 may be a straight line parallel to a horizontal axis or a vertical axis in a reference coordinate system.
For example, referring to fig. 2, for the target first polygon ABCDEFGH, and the target second polygon a1b1c1d1e1f1g1, the reference point O is determined from the inside of the target second polygon.
Taking a straight line L passing through the point O as a reference straight line, the straight line L intersects the target first polygon ABCDEFGH and the target second polygon a1b1c1d1e1f1g1 at a plurality of points, respectively.
The point M and the point N on the left side of the point O are respectively used as a first intersection point and a second intersection point, and the point P and the point Q on the right side of the point O are respectively used as a third intersection point and a fourth intersection point. Wherein, the point M is the intersection point which is closest to the point O among the intersection points of the straight line L and the target first polygon ABCDEFGH and positioned on the left side of the point O; the point Q is the intersection point which is closest to the point O in the intersection points of the straight line L and the target first polygon ABCDEFGH and positioned on the right side of the point O; the point N is the intersection point farthest from the point O among the intersection points of the straight line L and the target second polygon a1b1c1d1e1f1g1, which are located on the left side of the point O; the point P is the farthest point from the point O among the intersection points of the straight line L and the target second polygon a1b1c1d1e1f1g1 located on the right side of the point O.
The difference set between the target first polygon ABCDEFGH and the target second polygon a1b1c1d1e1f1g1 is split into two new first polygons with the line segment MN and the line segment PQ as a boundary.
In some embodiments of the present application, the splitting, by the B4, the non-simple polygon corresponding to the difference set into two new first polygons by using a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries may include:
c1, picking up the first vertex of the target first polygon along the clockwise direction of the target first polygon with the first intersection point as the starting point until reaching the fourth intersection point; taking the third intersection point as a starting point, and picking up a second vertex of the target second polygon along the counterclockwise direction of the target second polygon until the second intersection point; sequentially connecting the first intersection point, the first vertex, the fourth intersection point, the third intersection point, the second vertex and the second intersection point to obtain a new first polygon;
c2, picking up a third vertex of the target first polygon along the counterclockwise direction of the target first polygon with the first intersection point as a starting point until a fourth intersection point; taking the third intersection point as a starting point, and picking up a fourth vertex of the target second polygon along the clockwise direction of the target second polygon until the third intersection point is reached; and sequentially connecting the first intersection point, the third vertex, the fourth intersection point, the third intersection point, the fourth vertex and the second intersection point to obtain another new first polygon.
For example, referring to fig. 2, in one aspect, starting from the first intersection point M, the first vertex C, B, A, H, G, F of the target first polygon ABCDEFGH is picked up along the clockwise direction of the target first polygon ABCDEFGH until reaching the fourth intersection point Q; the second vertices g1, a1 of the target second polygon a1b1c1d1e1f1g1 are picked up along the counterclockwise direction of the target second polygon a1b1c1d1e1f1g1 with the third intersection point P as a starting point until the second intersection point N. MCBAHGFQP g1a1N is the new first polygon.
On the other hand, the third vertex D, E of the target first polygon ABCDEFGH is picked up along the counterclockwise direction of the target first polygon ABCDEFGH with the first intersection point M as a starting point until the fourth intersection point Q; taking the third intersection point P as a starting point, the fourth vertices f1, e1, d1, c1, b1 of the target second polygon a1b1c1d1e1f1g1 are picked up along the clockwise direction of the target second polygon a1b1c1d1e1f1g1 until the second intersection point N. MDEQP f1e1d1c1b1N is another new first polygon.
The number of sides of the target polygon may be only three (in this case, the target polygon is a triangle), or may exceed three (in this case, the target polygon is not a triangle). Based on this, in some embodiments of the present application, the above process of B1 determining a reference point from inside the target second polygon may include:
1) when the target second polygon is a triangle:
the centroid of the target second polygon is determined as the reference point.
2) When the target second polygon is not a triangle:
d1, determining three target vertexes which are adjacent in sequence from the vertexes of the target second polygon, and generating a triangle by using the target vertexes as vertexes. Wherein the other vertices of the target second polygon are all inside the triangle.
D2, determining a point as a candidate reference point from the edge of the triangle inside the target second polygon.
Because the triangle is formed by connecting three sequentially adjacent vertexes in the target second polygon, the edge formed by connecting the head point and the tail point is taken from the three sequentially adjacent vertexes as the edge inside the target second polygon.
It will be appreciated that the reference point may be the edge of the triangle that lies inside the target second polygon, excluding any points beyond the end points.
D3, determining whether the candidate reference point is inside the target second polygon. If so, determining the candidate reference point as a reference point; if not, return to execute D1 until a reference point is determined.
For example, referring to FIG. 3, for a polygon ABCD, three adjacent vertices A, B, C are first determined from the vertices of the polygon, and triangle ABC is generated by using the three adjacent vertices A, B, C as vertices.
At this time, since the vertex D of the polygon ABCD is located inside the triangle ABC. Three sequentially adjacent vertices, such as vertex B, C, D, are thus re-determined from the polygon ABCD, and the triangle BCD is generated with the re-determined three sequentially adjacent vertices B, C, D as vertices.
At this time, two points B and D are taken from the head and the tail, and one point O is selected from the line segment BD as a candidate reference point. Since the point O is inside the polygon ABCD, the point O can be determined as a reference point.
In some embodiments of the present application, the step of D5 selecting a point from the edge of the triangle inside the target second polygon, and determining the point as a candidate reference point may include:
and determining the middle point of the edge of the triangle positioned in the target second polygon as a candidate reference point.
In some embodiments of the present application, the above-mentioned process of determining whether the difference set constitutes a non-simple polygon by a1 may include:
e1, determining at least one polygon from the difference set.
E2, for each polygon in E1, determining whether the polygon has the following conditions:
one vertex of the polygon is on one edge of the polygon;
one edge of the polygon intersects the other edge.
If yes, determining that the difference set forms a non-simple polygon; if not, determining that the difference set does not form a non-simple polygon.
In some embodiments of the present application, the step S300 of determining whether there is an overlap between the first polygon and any one of the second polygons in the second polygon list may include:
for each second polygon in the second polygon list, determining whether there is an overlap of the first polygon with the second polygon by calculating an intersection ratio of the first polygon to the second polygon.
The semantic information labeling device for laser point clouds provided in the embodiment of the present application is described below, and the semantic information labeling device for laser point clouds described below and the semantic information labeling method for laser point clouds described above may be referred to in a corresponding manner.
Referring to fig. 4, the semantic information labeling apparatus for laser point cloud provided in the embodiment of the present application may include:
the region-of-interest acquiring unit 21 is configured to acquire at least one first polygon labeled as a region-of-interest in the laser point cloud image, where the at least one first polygon forms a first polygon list;
the non-travelable region acquiring unit 22 is configured to acquire at least one second polygon which is located in the region of interest in the laser point cloud image and is marked as a non-travelable region, and the at least one second polygon forms a second polygon list;
a region overlap judging unit 23, configured to judge, for each first polygon in the first polygon list, whether there is an overlap between the first polygon and any one second polygon in the second polygon list;
a region processing unit 24, configured to generate, for a target first polygon and a target second polygon with overlap, a new first polygon to replace the target first polygon in the first polygon list according to a difference set between the first polygon and the second polygon, where the new first polygon does not overlap with the target second polygon;
and a travelable region labeling unit 25 for labeling each first polygon in the final first polygon list as a travelable region when all first polygons in the first polygon list do not overlap with any second polygon in the second polygons.
In some embodiments of the present application, the first polygon and the second polygon are both simple polygons; the process of generating a new first polygon to replace the target first polygon in the first polygon list by the region processing unit 24 according to the difference set between the target first polygon and the target second polygon may include:
judging whether the difference set forms a non-simple polygon or not;
if so, splitting the non-simple polygon corresponding to the difference set into at least two new first polygons; adding the at least two new first polygons to the first polygon list and removing the target first polygon from the first polygon list;
if not, generating at least one new first polygon according to the difference set; adding the new first polygon to the first polygon list and removing the target first polygon from the first polygon list.
In some embodiments of the present application, the process of splitting the non-simple polygon corresponding to the difference set into at least two new first polygons by the region processing unit 24 may include:
determining a reference point from within said target second polygon;
setting a reference straight line, wherein the reference straight line passes through the reference point and intersects the target first polygon and the target second polygon at a plurality of points;
determining a first intersection point, a second intersection point, a third intersection point and a fourth intersection point from the plurality of points; the first intersection point and the second intersection point are located on the first side of the reference point, and the third intersection point and the fourth intersection point are located on the second side of the reference point; the first intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the first side, and the fourth intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the second side; the second intersection point is the intersection point which is farthest from the reference point in the intersection points of the first side of the target second polygon, and the third intersection point is the intersection point which is farthest from the reference point in the intersection points of the second side of the target second polygon;
and taking a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries, and splitting the non-simple polygon corresponding to the difference set into two new first polygons.
In some embodiments of the present application, the process of splitting the non-simple polygon corresponding to the difference set into two new first polygons by using the line segment formed by connecting the first intersection point and the second intersection point and the line segment formed by connecting the third intersection point and the fourth intersection point as boundaries by the region processing unit 24 may include:
taking the first intersection point as a starting point, and picking up a first vertex of the target first polygon along the clockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a second vertex of the target second polygon along the counterclockwise direction of the target second polygon until the second intersection point; the first intersection point, the first vertex, the fourth intersection point, the third intersection point, the second vertex and the second intersection point are connected in sequence to obtain a new first polygon;
taking the first intersection point as a starting point, and picking up a third vertex of the target first polygon along the counterclockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a fourth vertex of the target second polygon along the clockwise direction of the target second polygon until the third intersection point is reached; and sequentially connecting the first intersection point, the third vertex, the fourth intersection point, the third intersection point, the fourth vertex and the second intersection point to obtain another new first polygon.
In some embodiments of the present application, the process of the area processing unit 24 determining a reference point from inside the target second polygon may include:
when the target second polygon is a triangle: determining a centroid of the target second polygon as the reference point;
when the target second polygon is not a triangle:
when the target second polygon is a triangle: determining a centroid of the target second polygon as the reference point;
when the target second polygon is not a triangle:
determining three target vertexes which are adjacent in sequence from the vertexes of the target second polygon, and generating a triangle by taking the target vertexes as the vertexes, wherein other vertexes of the target second polygon are all in the triangle;
determining a point from the edge of the triangle inside the target second polygon as a candidate reference point;
judging whether the candidate reference point is inside the target second polygon;
if so, determining the candidate reference point as the reference point;
if not, returning to the step of determining three target vertexes which are sequentially adjacent from the vertexes of the target second polygon until the reference point is determined.
In some embodiments of the present application, the process of determining the reference point by the area processing unit 24 according to the edge of the triangle inside the target second polygon may include:
and determining the middle point of the edge as the reference point.
In some embodiments of the present application, the process of determining whether the difference set constitutes a non-simple polygon by the region processing unit 24 may include:
determining at least one polygon from the difference set;
for each polygon, determining whether the polygon has the following conditions:
one vertex of the polygon is on one edge of the polygon;
one side of the polygon intersects the other side;
and if so, determining that the difference set forms a non-simple polygon.
In some embodiments of the present application, the process of determining whether the first polygon overlaps with any one of the second polygons in the second polygon list by the region overlap determining unit 23 includes:
determining, for each second polygon in a second polygon list, whether there is overlap of the first polygon with the second polygon by calculating an intersection ratio of the first polygon to the second polygon.
The semantic information labeling device for the laser point cloud provided by the embodiment of the application can be applied to semantic information labeling equipment for the laser point cloud, such as a computer. Optionally, fig. 5 shows a hardware structure block diagram of a semantic information labeling device for laser point cloud, and referring to fig. 5, the hardware structure of the semantic information labeling device for laser point cloud may include: at least one processor 31, at least one communication interface 32, at least one memory 33 and at least one communication bus 34.
In the embodiment of the present application, the number of the processor 31, the communication interface 32, the memory 33 and the communication bus 34 is at least one, and the processor 31, the communication interface 32 and the memory 33 complete the communication with each other through the communication bus 34;
the processor 31 may be a central processing unit CPU, or an application Specific Integrated circuit asic, or one or more Integrated circuits configured to implement embodiments of the present application, etc.;
the memory 32 may comprise a high-speed RAM memory, and may further comprise a non-volatile memory (non-volatile memory) or the like, such as at least one disk memory;
wherein the memory 33 stores a program and the processor 31 may invoke the program stored in the memory 33, the program being for:
acquiring at least one first polygon marked as an interesting area in a laser point cloud image, wherein the at least one first polygon forms a first polygon list;
acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and marked as a non-driving region, wherein the at least one second polygon forms a second polygon list;
for each first polygon in the first polygon list, judging whether the first polygon is overlapped with any second polygon in the second polygon list;
if yes, aiming at the overlapped target first polygon and the overlapped target second polygon, generating a new first polygon to replace the target first polygon in a first polygon list according to a difference set of the target first polygon and the target second polygon, wherein the new first polygon and the target second polygon are not overlapped;
and returning to execute the step of judging whether the first polygon overlaps with any one second polygon in the second polygon list or not for each first polygon in the first polygon list, and marking each first polygon in the final first polygon list as a travelable area when all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list.
Alternatively, the detailed function and the extended function of the program may be as described above.
Embodiments of the present application further provide a storage medium, where a program suitable for execution by a processor may be stored, where the program is configured to:
acquiring at least one first polygon marked as an interesting area in a laser point cloud image, wherein the at least one first polygon forms a first polygon list;
acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and marked as a non-driving region, wherein the at least one second polygon forms a second polygon list;
for each first polygon in the first polygon list, judging whether the first polygon is overlapped with any second polygon in the second polygon list;
if yes, aiming at the overlapped target first polygon and the overlapped target second polygon, generating a new first polygon to replace the target first polygon in a first polygon list according to a difference set of the target first polygon and the target second polygon, wherein the new first polygon and the target second polygon are not overlapped;
and returning to execute the step of judging whether the first polygon overlaps with any one second polygon in the second polygon list or not for each first polygon in the first polygon list, and marking each first polygon in the final first polygon list as a travelable area when all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list.
Alternatively, the detailed function and the extended function of the program may be as described above.
In summary, the following steps:
according to the method and the device, only data marked as the region of interest and the non-driving region in the region of interest in the laser point cloud image are acquired, and the region of interest and the non-driving region are represented in a polygonal form. The region of interest is usually a road on which a target is to travel, the non-travelable region is usually an obstacle region such as a pedestrian or other vehicle on the road or a region of a pothole on the road, and the required labeling workload is relatively small, so that a large amount of polygon data of the region of interest and the non-travelable region can be easily acquired.
After the region of interest and the non-drivable region are characterized in a polygonal form, a first polygon of the region of interest is saved in a first polygon list, and a second polygon of the non-drivable region is saved in a second polygon list. Then, whether the polygon in the first polygon list overlaps with any polygon in the second polygon list is judged. And for the overlapped target first polygon and the target second polygon, generating a new first polygon to replace the target first polygon in the first polygon list according to the difference set of the first polygon and the second polygon. And the new first polygon is not overlapped with the target second polygon, so that the generated new first polygon is ensured to have no area incapable of driving.
And then returning to the step of judging whether the polygons in the first polygon list overlap with any polygon in the second polygon list, and judging the polygons (including the original first polygon and the generated new first polygon) in the first polygon list one by one until all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list, which means that the regions where the non-travelable regions exist in the first polygons in the first polygon list are all removed, and at this time, each first polygon in the final first polygon list represents the travelable region in the region of interest, namely, the travelable region can be marked.
The above-described process of acquiring the travelable region requires no complicated calculation and is not large in calculation amount. According to the technical scheme, the semantic information marking of the laser point cloud can be rapidly and efficiently realized, and the dilemma that the label data volume of the laser point cloud is small can be solved to a certain extent.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, the embodiments may be combined as needed, and the same and similar parts may be referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (12)
1. A semantic information labeling method of laser point cloud is characterized by comprising the following steps:
acquiring at least one first polygon marked as an interesting area in a laser point cloud image, wherein the at least one first polygon forms a first polygon list;
acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and marked as a non-driving region, wherein the at least one second polygon forms a second polygon list;
for each first polygon in the first polygon list, judging whether the first polygon is overlapped with any second polygon in the second polygon list;
if yes, aiming at the overlapped target first polygon and the overlapped target second polygon, generating a new first polygon to replace the target first polygon in a first polygon list according to a difference set of the target first polygon and the target second polygon, wherein the new first polygon and the target second polygon are not overlapped;
and returning to execute the step of judging whether the first polygon overlaps with any one second polygon in the second polygon list or not for each first polygon in the first polygon list, and marking each first polygon in the final first polygon list as a travelable area when all the first polygons in the first polygon list do not overlap with any second polygon in the second polygon list.
2. The method of claim 1, wherein the first polygon and the second polygon are both simple polygons;
a process of generating a new first polygon to replace the target first polygon in the first polygon list according to a difference set between the target first polygon and the target second polygon, comprising:
judging whether the difference set forms a non-simple polygon or not;
if so, splitting the non-simple polygon corresponding to the difference set into at least two new first polygons; adding the at least two new first polygons to the first polygon list and removing the target first polygon from the first polygon list;
if not, generating at least one new first polygon according to the difference set; adding the new first polygon to the first polygon list and removing the target first polygon from the first polygon list.
3. The method according to claim 2, wherein the process of splitting the non-simple polygon corresponding to the difference set into at least two new first polygons comprises:
determining a reference point from within said target second polygon;
setting a reference straight line, wherein the reference straight line passes through the reference point and intersects the target first polygon and the target second polygon at a plurality of points;
determining a first intersection point, a second intersection point, a third intersection point and a fourth intersection point from the plurality of points; the first intersection point and the second intersection point are located on the first side of the reference point, and the third intersection point and the fourth intersection point are located on the second side of the reference point; the first intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the first side, and the fourth intersection point is the intersection point which is closest to the reference point in the intersection points of the first target polygon on the second side; the second intersection point is the intersection point which is farthest from the reference point in the intersection points of the first side of the target second polygon, and the third intersection point is the intersection point which is farthest from the reference point in the intersection points of the second side of the target second polygon;
and taking a line segment formed by connecting the first intersection point and the second intersection point and a line segment formed by connecting the third intersection point and the fourth intersection point as boundaries, and splitting the non-simple polygon corresponding to the difference set into two new first polygons.
4. The method according to claim 3, wherein the splitting of the non-simple polygon corresponding to the difference set into two new first polygons with the line segment formed by connecting the first intersection point and the second intersection point and the line segment formed by connecting the third intersection point and the fourth intersection point as boundaries comprises:
taking the first intersection point as a starting point, and picking up a first vertex of the target first polygon along the clockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a second vertex of the target second polygon along the counterclockwise direction of the target second polygon until the second intersection point; the first intersection point, the first vertex, the fourth intersection point, the third intersection point, the second vertex and the second intersection point are connected in sequence to obtain a new first polygon;
taking the first intersection point as a starting point, and picking up a third vertex of the target first polygon along the counterclockwise direction of the target first polygon until a fourth intersection point; taking the third intersection point as a starting point, and picking up a fourth vertex of the target second polygon along the clockwise direction of the target second polygon until the third intersection point is reached; and sequentially connecting the first intersection point, the third vertex, the fourth intersection point, the third intersection point, the fourth vertex and the second intersection point to obtain another new first polygon.
5. The method of claim 3, wherein determining a reference point from within the target second polygon comprises:
when the target second polygon is a triangle: determining a centroid of the target second polygon as the reference point;
when the target second polygon is not a triangle:
determining three target vertexes which are adjacent in sequence from the vertexes of the target second polygon, and generating a triangle by taking the target vertexes as the vertexes, wherein other vertexes of the target second polygon are all in the triangle;
determining a point from the edge of the triangle inside the target second polygon as a candidate reference point;
judging whether the candidate reference point is inside the target second polygon;
if so, determining the candidate reference point as the reference point;
if not, returning to the step of determining three target vertexes which are sequentially adjacent from the vertexes of the target second polygon until the reference point is determined.
6. The method of claim 5, wherein determining a point from the edge of the triangle inside the target second polygon as a candidate reference point comprises:
and determining the middle point of the edge of the triangle positioned in the target second polygon as the candidate reference point.
7. The method of claim 2, wherein determining whether the difference set constitutes a non-simple polygon comprises:
generating at least one polygon from the difference set;
for each polygon, determining whether the polygon has the following conditions:
one vertex of the polygon is on one edge of the polygon;
one side of the polygon intersects the other side;
and if so, determining that the difference set forms a non-simple polygon.
8. The method according to claim 1, wherein said determining whether the first polygon overlaps with any of the second polygons in the second polygon list comprises:
determining, for each second polygon in a second polygon list, whether there is overlap of the first polygon with the second polygon by calculating an intersection ratio of the first polygon to the second polygon.
9. The method of claim 3, wherein the reference straight line is a straight line passing through the reference point and parallel to a horizontal axis or a vertical axis in the reference coordinate system.
10. The semantic information labeling device for the laser point cloud is characterized by comprising the following components:
the system comprises an interesting region acquisition unit, a laser point cloud image processing unit and a processing unit, wherein the interesting region acquisition unit is used for acquiring at least one first polygon marked as an interesting region in a laser point cloud image, and the at least one first polygon forms a first polygon list;
the non-driving area acquisition unit is used for acquiring at least one second polygon which is positioned in the region of interest in the laser point cloud image and is marked as a non-driving area, and the at least one second polygon forms a second polygon list;
the area overlapping judging unit is used for judging whether the first polygon overlaps with any second polygon in the second polygon list or not for each first polygon in the first polygon list;
a region processing unit, configured to generate, for a target first polygon and a target second polygon that overlap, a new first polygon to replace the target first polygon in a first polygon list according to a difference set between the first polygon and the second polygon, where the new first polygon does not overlap with the target second polygon, and return to an operation directed to the region overlap determining unit;
and the travelable area labeling unit is used for labeling each first polygon in the final first polygon list as a travelable area when all first polygons in the first polygon list do not overlap with any second polygon in the second polygons.
11. The semantic information labeling equipment for the laser point cloud is characterized by comprising the following components: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program and realizing the steps of the semantic information labeling method of the laser point cloud according to any one of claims 1 to 9.
12. A storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for semantic information labeling of a laser point cloud as claimed in any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110994744.XA CN113703455A (en) | 2021-08-27 | 2021-08-27 | Semantic information labeling method of laser point cloud and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110994744.XA CN113703455A (en) | 2021-08-27 | 2021-08-27 | Semantic information labeling method of laser point cloud and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113703455A true CN113703455A (en) | 2021-11-26 |
Family
ID=78655892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110994744.XA Pending CN113703455A (en) | 2021-08-27 | 2021-08-27 | Semantic information labeling method of laser point cloud and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113703455A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111562787A (en) * | 2020-05-28 | 2020-08-21 | 长沙中联重科环境产业有限公司 | Robot full-coverage path planning region dividing method, device, medium and equipment |
| CN112597846A (en) * | 2020-12-14 | 2021-04-02 | 合肥英睿系统技术有限公司 | Lane line detection method, lane line detection device, computer device, and storage medium |
| CN113110445A (en) * | 2021-04-13 | 2021-07-13 | 上海高仙自动化科技发展有限公司 | Robot path planning method and device, robot and storage medium |
| CN113189610A (en) * | 2021-04-28 | 2021-07-30 | 中国科学技术大学 | Map-enhanced autonomous driving multi-target tracking method and related equipment |
| CN113219472A (en) * | 2021-04-28 | 2021-08-06 | 合肥工业大学 | Distance measuring system and method |
-
2021
- 2021-08-27 CN CN202110994744.XA patent/CN113703455A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111562787A (en) * | 2020-05-28 | 2020-08-21 | 长沙中联重科环境产业有限公司 | Robot full-coverage path planning region dividing method, device, medium and equipment |
| CN112597846A (en) * | 2020-12-14 | 2021-04-02 | 合肥英睿系统技术有限公司 | Lane line detection method, lane line detection device, computer device, and storage medium |
| CN113110445A (en) * | 2021-04-13 | 2021-07-13 | 上海高仙自动化科技发展有限公司 | Robot path planning method and device, robot and storage medium |
| CN113189610A (en) * | 2021-04-28 | 2021-07-30 | 中国科学技术大学 | Map-enhanced autonomous driving multi-target tracking method and related equipment |
| CN113219472A (en) * | 2021-04-28 | 2021-08-06 | 合肥工业大学 | Distance measuring system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Van Brummelen et al. | Autonomous vehicle perception: The technology of today and tomorrow | |
| CN111874006B (en) | Route planning processing method and device | |
| US11493635B2 (en) | Ground intensity LIDAR localizer | |
| KR20220054278A (en) | Travelling track prediction method and device for vehicle | |
| JP6664470B2 (en) | High-accuracy map data processing method, apparatus, storage medium, and device | |
| CN111367282B (en) | Robot navigation method and system based on multimode perception and reinforcement learning | |
| US11670087B2 (en) | Training data generating method for image processing, image processing method, and devices thereof | |
| Meyer et al. | Deep semantic lane segmentation for mapless driving | |
| Gehrig et al. | Collision avoidance for vehicle-following systems | |
| Pfeiffer et al. | Modeling dynamic 3D environments by means of the stixel world | |
| CN106599108A (en) | Method for constructing multi-mode environmental map in three-dimensional environment | |
| CN104590259B (en) | A kind of transitional search method and system | |
| CN110705385B (en) | Method, device, equipment and medium for detecting angle of obstacle | |
| Ortega et al. | Overtaking maneuver scenario building for autonomous vehicles with PreScan software | |
| WO2022205618A1 (en) | Positioning method, driving control method, apparatus, computer device, and storage medium | |
| CN114488185A (en) | Robot navigation system method and system based on multi-line laser radar | |
| CN114463713A (en) | Information detection method and device of vehicle in 3D space and electronic equipment | |
| CN113282088A (en) | Unmanned driving method, device and equipment of engineering vehicle, storage medium and engineering vehicle | |
| Börcs et al. | A model-based approach for fast vehicle detection in continuously streamed urban LIDAR point clouds | |
| CN113158779A (en) | Walking method and device and computer storage medium | |
| CN113703455A (en) | Semantic information labeling method of laser point cloud and related equipment | |
| JP7140933B1 (en) | machine learning system | |
| Kwon et al. | Incremental end-to-end learning for lateral control in autonomous driving | |
| CN116311114A (en) | Method and device for generating drivable region, electronic equipment and storage medium | |
| Nagy et al. | Development of a lane keeping steering control by using camera vanishing point strategy |
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 |