CN118135569A - Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar - Google Patents
Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar Download PDFInfo
- Publication number
- CN118135569A CN118135569A CN202410308395.5A CN202410308395A CN118135569A CN 118135569 A CN118135569 A CN 118135569A CN 202410308395 A CN202410308395 A CN 202410308395A CN 118135569 A CN118135569 A CN 118135569A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- segmentation
- frame
- laser radar
- map
- 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
- 230000011218 segmentation Effects 0.000 title claims abstract description 150
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000004458 analytical method Methods 0.000 claims abstract description 68
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 238000005065 mining Methods 0.000 claims abstract description 6
- 238000001514 detection method Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 7
- 230000004927 fusion Effects 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 5
- 238000000638 solvent extraction Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 7
- 238000003708 edge detection Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention relates to the technical field of automatic driving, in particular to a single-multi-frame collaborative passable point cloud segmentation method and device based on a laser radar, wherein the method comprises the following steps: inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result; inputting original laser radar data into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the mining is reduced, and carrying out trafficability segmentation on the point cloud map after the mining is reduced to obtain a grid-based trafficability analysis result; and fusing the point cloud segmentation result of the single frame and the grid-based trafficability analysis result to obtain the point cloud map with semantic information. Therefore, the problems of lack of point cloud data, interpretability, difficulty in real vehicle deployment, poor segmentation effect and the like of the existing automatic driving point cloud segmentation and road edge detection method are solved.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to a single-multi-frame collaborative passable point cloud segmentation method and device based on a laser radar.
Background
Semantic segmentation is one of the key tasks to achieve autopilot, which helps the autopilot system to better understand the autopilot Zhou Zhao environment by giving the awareness data a human scene understanding. The essence of laser point cloud segmentation is that data of input laser point cloud (a set of points) enters an algorithm, different definitions (such as a tree, an obstacle, a road and the like) are given to each point through the algorithm, and finally a point cloud scene which accords with human understanding is formed. Among them, semantic segmentation of laser point clouds has been a major difficulty due to uneven and discrete point cloud distribution of laser point clouds and lack of semantic information (such as textures and colors). And in addition to the traditional task of dividing the obstacle point cloud, in order to better assist in decision control of autopilot in urban scenarios, it is considered by those skilled in the art that extracting the road edges as passable boundaries is essential. The existing point cloud semantic segmentation technology is roughly divided into two directions according to the use method:
The first is based on deep learning (data-based approach): the semantic segmentation deep learning model is learned through a large number of marked point clouds, and the difference of different semantic segmentation learning places mainly comes from different point cloud data processing to extract more point cloud features such as PointNet ++, cylindar3D and the like. The method is popular nowadays because the segmentation effect is good, and the method occupies the head of the point cloud segmentation accuracy ranking list. However, the point cloud segmentation based on deep learning requires a large number of accurate labeling point clouds (datasets), but labeling point clouds is a very time-consuming and difficult task. In a task, the route edge needs to be extracted, and unfortunately, no route edge data set is disclosed in the market at present. Besides, the unexplained nature of the deep learning model makes it difficult to locate and solve problems at the operation stage. And the problem that the real vehicle is deployed even though the GPU is used on the real vehicle due to the huge model parameter quantity of the semantic segmentation model is still a great difficulty.
The second method is to give the point cloud corresponding semantic point cloud results by the geometric features of the artificial total point cloud, and the segmentation method based on the geometric features is usually low in accuracy and uneven segmentation caused by missed detection and false detection easily occurs, so that great problems are caused for the follow-up automatic driving system to understand the scene by using the uncertain segmentation results. In order to solve these problems, many scholars utilize a machine learning method to promote a segmentation algorithm such as a clustering algorithm and a fitting algorithm, etc., however, the clustering algorithm is easy to bring about the situation that the same object is clustered into different objects; fitting algorithms often require multiple attempts to obtain the best fit, both of which require computing a large number of point clouds, making operation difficult in real-time. And the method can also have incompatible conditions similar to a deep learning model on different laser radar platforms. Conventional algorithms based on the geometrical features of the point cloud have the disadvantage of not being small in terms of countering disturbances such as motion distortion of the vehicle, sensor noise and scene changes.
As a summary, point cloud segmentation and edge detection remain to be addressed as an important loop in automatic driving: the cloud data of the mark points are lacking, the interpretability (debuggeability) is poor, the deployment of a real vehicle is difficult, and the segmentation effect is poor.
Disclosure of Invention
The invention provides a single-multi-frame collaborative passable point cloud segmentation method and device based on a laser radar, which are characterized in that urban scenes are mainly structured buildings, so that an algorithm based on a point cloud geometric characteristic rule has universality, but the method and the device are very challenging in implementation, such as problems of summarizing geometric characteristic requirements, distance decoupling from data to a vehicle, interference of scenes to data, such as up-down slopes, and shielding, and the like caused by uneven density of the single-frame laser point cloud, so that the existing automatic driving point cloud segmentation and road edge detection method has the problems of lack of marked point cloud data, poor interpretability, difficult real vehicle deployment, poor segmentation effect and the like.
An embodiment of a first aspect of the present invention provides a single-multi-frame collaborative passable point cloud segmentation method based on a laser radar, including the following steps:
inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result;
Inputting the original laser radar data into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the mining is reduced, and carrying out trafficability segmentation on the point cloud map after the mining is reduced to obtain a grid-based trafficability analysis result;
and fusing the single-frame point cloud segmentation result and the grid-based trafficability analysis result to obtain a point cloud map with semantic information.
Optionally, inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result, including:
converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to regularize the point cloud to obtain near-ground point cloud;
performing annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
and carrying out adaptive road edge point detection on the ring-shaped structure loading point cloud to obtain a point cloud segmentation result of the single frame.
Optionally, the performing the trafficability segmentation on the point cloud map after the downsampling to obtain a trafficability analysis result includes:
rasterizing the downscaled point cloud map to obtain a rasterized point cloud map;
And carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a trafficability analysis result based on the grids.
Optionally, the fusing the single-frame point cloud segmentation result and the grid-based trafficability analysis result to obtain a point cloud map with semantic information includes:
Analyzing inconsistent situations of the point cloud segmentation result of the single frame and the grid-based trafficability analysis result;
and adding point cloud samples to the local grids according to the inconsistency, and carrying out refined analysis on the added point cloud samples to obtain the point cloud map with the semantic information.
An embodiment of a second aspect of the present invention provides a single-multi-frame cooperative passable point cloud segmentation apparatus based on a lidar, including:
The first segmentation module is used for inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result;
the second segmentation module is used for inputting the original laser radar data into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the downsampling, and carrying out trafficability segmentation on the point cloud map after the downsampling to obtain a grid-based trafficability analysis result;
And the fusion module is used for fusing the point cloud segmentation result of the single frame and the grid-based trafficability analysis result to obtain a point cloud map with semantic information.
Optionally, the first segmentation module includes:
the filtering and extracting unit is used for converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to perform point cloud regularization to obtain near-ground point cloud;
The segmentation unit is used for carrying out annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
and the road edge point detection unit is used for carrying out adaptive road edge point detection on the ring-shaped structure loading point cloud to obtain a point cloud segmentation result of the single frame.
Optionally, the second segmentation module includes:
The rasterizing unit is used for rasterizing the point cloud map after the downsampling to obtain a rasterized point cloud map;
And the trafficability analysis unit is used for carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a trafficability analysis result based on the grids.
Optionally, the fusion module includes:
the analysis unit is used for analyzing the inconsistency situation of the point cloud segmentation result of the single frame and the grid-based trafficability analysis result;
And the fine analysis unit is used for adding point cloud samples to the local grids according to the inconsistency condition, and carrying out fine analysis on the added point cloud samples to obtain the point cloud map with the semantic information.
An embodiment of a third aspect of the present invention provides an electronic device, including: the system comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the program to realize the single multi-frame collaborative passable point cloud segmentation method based on the laser radar.
An embodiment of a fourth aspect of the present invention provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements the above laser radar-based single-multi-frame collaborative passable point cloud segmentation method.
According to the single-multi-frame collaborative passable point cloud segmentation method and device based on the laser radar, through double-flow analysis of the single-frame point cloud and the local map, point cloud segmentation of automatic driving and passable boundary (road edge) segmentation under urban scenes can be completed in real time, and a semantic map of a high-accuracy and uniform passable area is obtained, so that vehicles can better understand scenes and plan safe routes, accuracy of a traditional geometric feature segmentation algorithm is improved, good segmentation results are provided for fine objects such as road edges, and the problem of non-uniformity of the segmented point cloud and difficulty in processing subsequent modules due to uncertainty of the results are greatly reduced.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The foregoing and/or additional aspects and advantages of the invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
Fig. 1 is a flowchart of a single-multi-frame collaborative passable point cloud segmentation method based on a laser radar according to an embodiment of the present invention;
Fig. 2 is a schematic diagram of input and output visualization in a single-multi-frame collaborative passable point cloud segmentation method based on a laser radar according to an embodiment of the present invention;
FIG. 3 is a detailed flowchart of another single-multi-frame collaborative passable point cloud segmentation method based on a lidar according to an embodiment of the present invention;
FIG. 4 is a block diagram of another method for partitioning a single-multi-frame collaborative passable point cloud based on a lidar according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of the different modules of FIG. 4;
fig. 6 is a schematic diagram of a point cloud segmentation effect of a single frame according to an embodiment of the present invention;
FIG. 7 is a single-multiframe semantic point cloud map provided by an embodiment of the present invention;
fig. 8 is a schematic block diagram of a single-multi-frame cooperative passable point cloud segmentation device based on a laser radar according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present invention and should not be construed as limiting the invention.
The following describes a single multi-frame collaborative passable point cloud segmentation method and device based on a laser radar according to an embodiment of the invention with reference to the accompanying drawings.
Fig. 1 is a flow chart of a single-multi-frame collaborative passable point cloud segmentation method based on a laser radar according to an embodiment of the present invention.
As shown in fig. 1, the single-multi-frame collaborative passable point cloud segmentation method based on the laser radar comprises the following steps:
in step S101, the original lidar data is input into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result.
Further, in one embodiment of the present invention, inputting original lidar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result, including:
Converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to perform point cloud regularization to obtain near-ground point cloud;
Performing annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
And carrying out adaptive road edge point detection on the loading point cloud of the annular structure to obtain a single-frame point cloud segmentation result.
Specifically, as shown in fig. 2 and 3, original laser radar data are converted to a local coordinate system to obtain a target laser point, point cloud data of each frame in the target laser point are input to a single-frame point cloud segmentation algorithm, point cloud regularization is performed on the point cloud data to facilitate point cloud searching, a near-ground point cloud is obtained, neighborhood geometric feature data of the near-ground point cloud are extracted in a self-adaptive mode, rough segmentation point clouds such as obstacles, road edges and grasslands are analyzed and extracted through special neighborhood general geometric features, namely annular sector space segmentation and road edge point detection are performed, and a single-frame point cloud segmentation result is obtained.
As shown in fig. 4 and 5, the point cloud regularization includes ring regularization and fan regularization, the two regularization needs to be performed simultaneously, the ring regularization comes from a point cloud coil of a scanning laser radar self mechanism, the point cloud coil is mainly responsible for detecting lawns, large obstacles and road edges, the fan regularization is to pack the point clouds with similar horizontal scanning angles into a wire harness, the wire harness is mainly responsible for protruding of a road surface, and the two regularization can be performed simultaneously to better search and analyze the relationship between points and surrounding points.
The adaptivity is to solve the problem that geometric features caused by the fact that the density of near point clouds is larger than that of far point clouds are influenced by the distance from a target point to a vehicle. Specifically, the geometric features and the point cloud density need to be decoupled, and the analysis area of the point cloud is related to the point cloud density (the distance between the point and the vehicle), so that the geometric features with similar sizes can be conveniently extracted by a single-frame point cloud segmentation algorithm.
The general geometric characteristics are different from the method for determining the target point by using an accurate threshold, the embodiment of the invention extracts the point cloud by using the change percentage of a special angle, the special angle can reduce the unstable influence of a decoupling method on the point cloud analysis area and the geometric characteristics, and the change percentage can ensure the feature universality under different laser radar platforms, namely similar change percentages can be obtained by changing certain parameters.
As shown in fig. 6, it can be seen that the point cloud, both near and far, is segmented, which illustrates that it greatly alleviates the side effects of near-dense and far-sparse point clouds, and can obtain a relatively accurate distinction between obstacles and travelable boundaries (curbs). However, the geometric-based algorithm has certain defects that the segmented point cloud is uneven and some false detection point clouds appear, and in addition, the complexity of the point cloud segmentation method is increased for higher accuracy due to the excessive number of single-frame point clouds, so that the running time is greatly increased. It is therefore also desirable to use the down-mined point cloud map to aid in further analysis of street scenes.
In step S102, the original laser radar data is input into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the downsampling, and the downsampling point cloud map is subjected to trafficability segmentation to obtain a grid-based trafficability analysis result.
Further, in an embodiment of the present invention, performing trafficability segmentation on the point cloud map after downsampling to obtain a trafficability analysis result includes:
rasterizing the point cloud map after the downsampling to obtain a rasterized point cloud map;
and carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a trafficability analysis result based on the grids.
Specifically, as shown in fig. 3 and 5, the original laser radar data may be input into an algorithm capable of making a map of a point cloud, such as a SLAM (instant localization and mapping algorithm), the original laser radar data is firstly converted into a local coordinate system to obtain a target laser point, the point cloud data of each frame in the target laser point is extracted according to the curvature, then the point cloud matching is performed according to Ping Miandian clouds and angular points, finally the downsampling point clouds are added into the point clouds of a world coordinate system or a global coordinate system, a local map is integrated, the downsampling global point cloud map is obtained through the map, and finally the downsampling point cloud map is subjected to gridding trafficability segmentation to obtain a trafficability analysis result based on grids, so that the street scene can be further analyzed.
In step S103, the point cloud segmentation result of the single frame and the grid-based trafficability analysis result are fused to obtain a point cloud map with semantic information.
Further, in an embodiment of the present invention, fusing a single frame point cloud segmentation result and a grid-based trafficability analysis result to obtain a point cloud map with semantic information, including:
Analyzing inconsistent conditions of a point cloud segmentation result of a single frame and a grid-based trafficability analysis result;
And adding point cloud samples to the local grids according to the inconsistent condition, and carrying out refined analysis on the added point cloud samples to obtain a point cloud map with semantic information, wherein the added point cloud samples come from dense single-frame division point cloud results.
Specifically, as shown in fig. 7, the inconsistent condition of the rough segmentation point cloud result obtained by single-frame point cloud segmentation and the grid analysis result is analyzed, samples are added to the local grid according to the inconsistent condition, and fine analysis is performed, so that an accurate and uniform result is obtained, and finally the segmented point cloud is output to obtain a semantic map.
For example, through the trafficable analysis after the point cloud map is rasterized and downsampled, a large obstacle different from a flat land and a flat land are segmented, so that a grid analysis result is obtained, and due to the serious downsampling of the point cloud map, the flat land actually comprises grasslands, flat lands and road edges, and then the point clouds of the grasslands and the road edges are extracted from the point clouds of the flat land.
By fusing the point cloud segmentation results and the grid analysis results of a single frame, the analysis results which are different due to downsampling can be known, so that geometrical characteristics are used for helping to distinguish grasslands and road edge point clouds, local point clouds are subjected to refined analysis, verification effect is achieved, uniform and accurate passable area segmentation is completed, and finally, the marked point clouds are output to form a point cloud map with semantic information.
It should be noted that, the embodiment of the invention does not need a CPU only by a GPU, and can obtain a reliable passable area in real time under the urban scene without using the GPU to help the intelligent vehicle to realize automatic driving, thereby greatly reducing the computational pressure of the GPU in the automatic driving task and being beneficial to the deployment of an automatic driving system on a real vehicle.
The single-multi-frame collaborative passable point cloud segmentation method based on the laser radar provided by the embodiment of the invention has the following steps of
The beneficial effects are that:
(1) High efficiency: the area needing fine analysis can be accurately found, and the result is obtained in real time and provided for a subsequent obstacle avoidance planning module;
(2) Feature commonality: the general characteristics can be obtained through geometric analysis of the structured scene, and the method is suitable for different urban scenes;
(3) Debuggeability: after debugging, the method can be used for different laser radar platforms;
(4) Plug and play: can cooperate with different laser point cloud SLAM (instant positioning and mapping algorithm).
Next, a single-multi-frame cooperative passable point cloud segmentation device based on the laser radar according to the embodiment of the invention is described with reference to the accompanying drawings.
Fig. 8 is a schematic block diagram of a laser radar-based single-multi-frame cooperative passable point cloud segmentation apparatus according to an embodiment of the present invention.
As shown in fig. 8, the single-multi-frame cooperative passable point cloud segmentation apparatus 80 based on the laser radar includes: a first segmentation module 801, a second segmentation module 802, and a fusion module 803.
The first segmentation module 801 is configured to input original lidar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result. The second segmentation module 802 is configured to input the original lidar data into a preset real-time positioning and mapping algorithm to obtain a reduced point cloud map, and perform trafficability segmentation on the reduced point cloud map to obtain a grid-based trafficability analysis result. And the fusion module 803 is used for fusing the point cloud segmentation result of the single frame and the grid-based trafficability analysis result to obtain the point cloud map with the semantic information.
Further, in one embodiment of the present invention, the first segmentation module 801 includes:
the filtering and extracting unit is used for converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to conduct point cloud regularization so as to obtain near-ground point cloud;
the segmentation unit is used for carrying out annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
and the road edge point detection unit is used for carrying out self-adaptive road edge point detection on the ring-shaped structure loading point cloud to obtain a single-frame point cloud segmentation result.
Further, in one embodiment of the present invention, the second segmentation module 802 includes:
The rasterizing unit is used for rasterizing the point cloud map after the downsampling to obtain a rasterized point cloud map;
and the trafficability analysis unit is used for carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a grid-based trafficability analysis result.
Further, in one embodiment of the present invention, the fusion module 803 includes:
the analysis unit is used for analyzing the inconsistency situation of the point cloud segmentation result of the single frame and the grid-based trafficability analysis result;
and the fine analysis unit is used for adding point cloud samples to the local grids according to the inconsistent condition, and carrying out fine analysis on the added point cloud samples to obtain a point cloud map with semantic information.
It should be noted that the explanation of the embodiment of the single-multi-frame co-operable point cloud segmentation method based on the laser radar is also applicable to the single-multi-frame co-operable point cloud segmentation device based on the laser radar of the embodiment, and is not repeated herein.
The single-multi-frame cooperative passable point cloud segmentation device based on the laser radar provided by the embodiment of the invention has the following advantages that
The beneficial effects are that:
(1) High efficiency: the area needing fine analysis can be accurately found, and the result is obtained in real time and provided for a subsequent obstacle avoidance planning module;
(2) Feature commonality: the general characteristics can be obtained through geometric analysis of the structured scene, and the method is suitable for different urban scenes;
(3) Debuggeability: after debugging, the method can be used for different laser radar platforms;
(4) Plug and play: can cooperate with different laser point cloud SLAM (instant positioning and mapping algorithm).
Fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. The electronic device may include:
memory 901, processor 902, and a computer program stored on memory 901 and executable on processor 902.
The processor 902 implements the single-multi-frame cooperative passable point cloud segmentation method based on the laser radar provided in the above embodiment when executing the program.
Further, the electronic device further includes:
a communication interface 903 for communication between the memory 901 and the processor 902.
Memory 901 for storing a computer program executable on processor 902.
Memory 901 may comprise high-speed RAM memory or may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
If the memory 901, the processor 902, and the communication interface 903 are implemented independently, the communication interface 903, the memory 901, and the processor 902 may be connected to each other through a bus and perform communication with each other. The bus may be an industry standard architecture (Industry Standard Architecture, abbreviated ISA) bus, an external device interconnect (PERIPHERAL COMPONENT, abbreviated PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, abbreviated EISA) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, only one thick line is shown in fig. 9, but not only one bus or one type of bus.
Alternatively, in a specific implementation, if the memory 901, the processor 902, and the communication interface 903 are integrated on a chip, the memory 901, the processor 902, and the communication interface 903 may communicate with each other through internal interfaces.
The processor 902 may be a central processing unit (Central Processing Unit, abbreviated as CPU), or an Application SPECIFIC INTEGRATED Circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the invention.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, realizes the single-multi-frame collaborative passable point cloud segmentation method based on the laser radar.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "N" means at least two, for example, two, three, etc., unless specifically defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and additional implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order from that shown or discussed, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present invention.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or N wires, a portable computer cartridge (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product.
The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like. While embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the invention, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the invention.
Claims (10)
1. A single-multi-frame collaborative passable point cloud segmentation method based on a laser radar is characterized by comprising the following steps:
inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result;
Inputting the original laser radar data into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the mining is reduced, and carrying out trafficability segmentation on the point cloud map after the mining is reduced to obtain a grid-based trafficability analysis result;
and fusing the single-frame point cloud segmentation result and the grid-based trafficability analysis result to obtain a point cloud map with semantic information.
2. The method for partitioning single multi-frame collaborative passable point clouds based on the laser radar according to claim 1, wherein the step of inputting original laser radar data into a single-frame point cloud partitioning algorithm to obtain a single-frame point cloud partitioning result comprises the steps of:
converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to regularize the point cloud to obtain near-ground point cloud;
performing annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
and carrying out adaptive road edge point detection on the ring-shaped structure loading point cloud to obtain a point cloud segmentation result of the single frame.
3. The laser radar-based single-multi-frame collaborative passable point cloud segmentation method according to claim 1, wherein the step of carrying out passable segmentation on the downscaled point cloud map to obtain a passable analysis result comprises the following steps:
rasterizing the downscaled point cloud map to obtain a rasterized point cloud map;
And carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a trafficability analysis result based on the grids.
4. The laser radar-based single-multi-frame collaborative passable point cloud segmentation method according to claim 1, wherein the fusing the single-frame point cloud segmentation result and the grid-based passable analysis result to obtain a point cloud map with semantic information comprises:
Analyzing inconsistent situations of the point cloud segmentation result of the single frame and the grid-based trafficability analysis result;
and adding point cloud samples to the local grids according to the inconsistency, and carrying out refined analysis on the added point cloud samples to obtain the point cloud map with the semantic information.
5. Single multiframe is trafficable point cloud segmentation device in coordination based on laser radar, its characterized in that includes:
The first segmentation module is used for inputting the original laser radar data into a single-frame point cloud segmentation algorithm to obtain a single-frame point cloud segmentation result;
the second segmentation module is used for inputting the original laser radar data into a preset real-time positioning and mapping algorithm to obtain a point cloud map after the downsampling, and carrying out trafficability segmentation on the point cloud map after the downsampling to obtain a grid-based trafficability analysis result;
And the fusion module is used for fusing the point cloud segmentation result of the single frame and the grid-based trafficability analysis result to obtain a point cloud map with semantic information.
6. The lidar-based single-multi-frame co-operable point cloud segmentation apparatus of claim 5, wherein the first segmentation module comprises:
the filtering and extracting unit is used for converting and inputting the original laser radar data into a single-frame point cloud segmentation algorithm to perform point cloud regularization to obtain near-ground point cloud;
The segmentation unit is used for carrying out annular sector space segmentation on the near-ground point cloud to obtain an annular structure loading point cloud;
and the road edge point detection unit is used for carrying out adaptive road edge point detection on the ring-shaped structure loading point cloud to obtain a point cloud segmentation result of the single frame.
7. The lidar-based single-multi-frame co-operable point cloud segmentation apparatus as set forth in claim 5, wherein the second segmentation module comprises:
The rasterizing unit is used for rasterizing the point cloud map after the downsampling to obtain a rasterized point cloud map;
And the trafficability analysis unit is used for carrying out trafficability analysis on each grid in the rasterized point cloud map to obtain a trafficability analysis result based on the grids.
8. The laser radar-based single-multi-frame cooperative passable point cloud segmentation apparatus of claim 5, wherein the fusion module comprises:
the analysis unit is used for analyzing the inconsistency situation of the point cloud segmentation result of the single frame and the grid-based trafficability analysis result;
And the fine analysis unit is used for adding point cloud samples to the local grids according to the inconsistency condition, and carrying out fine analysis on the added point cloud samples to obtain the point cloud map with the semantic information.
9. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the lidar-based single-multi-frame co-operable point cloud segmentation method of any of claims 1-4.
10. A computer readable storage medium having stored thereon a computer program, the program being executable by a processor for implementing the lidar-based single-multi-frame co-operable point cloud segmentation method according to any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410308395.5A CN118135569A (en) | 2024-03-18 | 2024-03-18 | Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410308395.5A CN118135569A (en) | 2024-03-18 | 2024-03-18 | Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118135569A true CN118135569A (en) | 2024-06-04 |
Family
ID=91246149
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410308395.5A Pending CN118135569A (en) | 2024-03-18 | 2024-03-18 | Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118135569A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118298183A (en) * | 2024-06-05 | 2024-07-05 | 江西师范大学 | High-precision semantic segmentation method and system for vehicle-mounted laser point cloud without labeling |
-
2024
- 2024-03-18 CN CN202410308395.5A patent/CN118135569A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118298183A (en) * | 2024-06-05 | 2024-07-05 | 江西师范大学 | High-precision semantic segmentation method and system for vehicle-mounted laser point cloud without labeling |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11709058B2 (en) | Path planning method and device and mobile device | |
| JP6561199B2 (en) | Urban road recognition method, apparatus, storage medium and equipment based on laser point cloud | |
| CN110687539B (en) | Parking space detection method, device, medium and equipment | |
| Moras et al. | Credibilist occupancy grids for vehicle perception in dynamic environments | |
| CN112101092A (en) | Automatic driving environment sensing method and system | |
| CN116129376A (en) | Road edge detection method and device | |
| US10423840B1 (en) | Post-processing method and device for detecting lanes to plan the drive path of autonomous vehicle by using segmentation score map and clustering map | |
| CN118135569A (en) | Single-multi-frame collaborative passable point cloud segmentation method and device based on laser radar | |
| CN111338361A (en) | Obstacle avoidance method, device, equipment and medium for low-speed unmanned vehicle | |
| CN110674705A (en) | Small-sized obstacle detection method and device based on multi-line laser radar | |
| CN110900602B (en) | Positioning recovery method and device, robot and storage medium | |
| CN115638787B (en) | Digital map generation method, computer readable storage medium and electronic device | |
| CN112912894B (en) | Road boundary identification method and device | |
| CN111308500A (en) | Obstacle sensing method and device based on single-line laser radar and computer terminal | |
| CN112348000A (en) | Obstacle recognition method, device, system and storage medium | |
| CN112183381A (en) | Method and device for detecting driving area of vehicle | |
| CN110909656A (en) | Pedestrian detection method and system with integration of radar and camera | |
| CN112435336B (en) | Curve type identification method and device, electronic equipment and storage medium | |
| CN117671644A (en) | Signboard detection method and device and vehicle | |
| CN117789160A (en) | Multi-mode fusion target detection method and system based on cluster optimization | |
| US20220229186A1 (en) | Object shape detection apparatus and method | |
| CN116311114A (en) | Method and device for generating drivable region, electronic equipment and storage medium | |
| CN115578703A (en) | Laser perception fusion optimization method, device and equipment and readable storage medium | |
| CN114882458A (en) | Target tracking method, system, medium and device | |
| CN111338336B (en) | Automatic driving method and device |
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 |