CN112991550B - Obstacle position detection method and device based on pseudo point cloud and electronic equipment - Google Patents
Obstacle position detection method and device based on pseudo point cloud and electronic equipment Download PDFInfo
- Publication number
- CN112991550B CN112991550B CN202110353738.6A CN202110353738A CN112991550B CN 112991550 B CN112991550 B CN 112991550B CN 202110353738 A CN202110353738 A CN 202110353738A CN 112991550 B CN112991550 B CN 112991550B
- Authority
- CN
- China
- Prior art keywords
- dimensional structure
- point cloud
- information
- target obstacle
- obstacle
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 43
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 230000004888 barrier function Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Graphics (AREA)
- Software Systems (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides an obstacle position detection method, device and electronic equipment based on pseudo point cloud, and relates to the technical field of vehicle driving, wherein the method comprises the steps of obtaining pseudo point cloud information aiming at the same target obstacle and sparse Lei Dadian cloud from low-beam radar; clustering the sparse Lei Dadian cloud to obtain a first three-dimensional structure and a centroid position of the first three-dimensional structure, wherein the first three-dimensional structure is the smallest three-dimensional structure capable of comprising a clustering result; determining a second three-dimensional structure from the pseudo point cloud information by a three-dimensional bounding box detection method, wherein the second three-dimensional structure is the minimum three-dimensional structure capable of comprising a target obstacle; and matching the first three-dimensional structure with the second three-dimensional structure, if the matching result is larger than a preset volume threshold, determining the position of the target obstacle according to the mass center position of the first three-dimensional structure, combining the pseudo point cloud with the low-wire-harness radar point cloud to improve the accuracy of pseudo point cloud position detection, and obtaining more accurate obstacle position information under the condition of lower cost.
Description
Technical Field
The invention relates to the technical field of vehicle driving, in particular to a method and a device for detecting the position of an obstacle based on pseudo point cloud and electronic equipment.
Background
With the development of vehicle technology, accurate detection of the position of an obstacle is important, and the safety and reliability of automatic driving of a vehicle are affected.
At present, a laser radar is generally adopted to accurately detect the position of an obstacle, but the cost of the laser radar is relatively high, and the laser radar is not beneficial to wide application in the field of vehicle driving.
Disclosure of Invention
In view of the above, the present invention aims to provide a method, a device and an electronic device for detecting the position of an obstacle based on a pseudo point cloud, which combine the pseudo point cloud with a low-beam radar point cloud to improve the accuracy of the pseudo point cloud position detection, and obtain more accurate obstacle position information under the condition of lower cost.
In a first aspect, an embodiment provides a method for detecting a position of an obstacle based on a pseudo point cloud, including:
Acquiring pseudo point cloud information aiming at the same target obstacle and sparse Lei Dadian cloud from a low-beam radar;
Clustering the sparse radar point cloud to obtain a first three-dimensional structure and a mass center position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
Determining a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method, wherein the second three-dimensional structure is a minimum three-dimensional structure capable of comprising the target obstacle;
and matching the first three-dimensional structure with the second three-dimensional structure, and if the matching result is larger than a preset volume threshold value, determining the position of the target obstacle according to the centroid position of the first three-dimensional structure.
In an alternative embodiment, the method further comprises:
and if the matching result is smaller than or equal to a preset volume threshold value, determining the position of the target obstacle according to the centroid position of the second three-dimensional structure.
In an alternative embodiment, the method further comprises:
And if the matching result is smaller than or equal to the preset volume threshold value, re-executing the step of acquiring the pseudo point cloud information aiming at the same target obstacle and the sparse Lei Dadian cloud from the low-beam radar.
In an alternative embodiment, the method further comprises:
determining width information and height information of a second three-dimensional structure from the pseudo point cloud information by a three-dimensional bounding box detection method;
And determining contour information of the target obstacle according to the width information and the height information.
In an optional embodiment, the step of clustering the sparse radar point cloud to obtain a first stereo structure and a centroid position of the first stereo structure includes:
clustering the sparse radar point cloud to obtain a clustering result;
Adding a bounding box to the clustering result to obtain a first three-dimensional structure;
And calculating the mass center position of the first three-dimensional structure according to the first three-dimensional structure.
In an alternative embodiment, the step of acquiring the pseudo point cloud information for the same target obstacle includes:
acquiring an image to be detected of a target obstacle from a binocular camera;
calculating depth information of each pixel point in the image to be detected based on binocular parallax of a binocular distance measuring method;
and converting the depth information into pseudo point cloud information according to a camera conversion principle and a geometric relation.
In an alternative embodiment, the step of matching the first and second stereo structures comprises:
And carrying out three-dimensional cross-correlation calculation on the volumes of the first three-dimensional structure and the second three-dimensional structure.
In a second aspect, an embodiment provides an obstacle position detection device based on a pseudo point cloud, the device including:
the acquisition module is used for acquiring pseudo point cloud information aiming at the same target obstacle and sparse Lei Dadian clouds from the low-beam radar;
The clustering module is used for carrying out clustering operation on the sparse radar point cloud to obtain a first three-dimensional structure and a mass center position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
The detection module is used for determining a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method, wherein the second three-dimensional structure is a minimum three-dimensional structure capable of comprising the target obstacle;
And the determining module is used for matching the first three-dimensional structure with the second three-dimensional structure, and determining the position of the target obstacle according to the centroid position of the first three-dimensional structure if the matching result is larger than a preset volume threshold value.
In a third aspect, an embodiment provides an electronic device, including a memory, a processor, where the memory stores a computer program executable on the processor, and where the processor implements the steps of the method according to any of the foregoing embodiments when the computer program is executed.
In a fourth aspect, embodiments provide a machine-readable storage medium storing machine-executable instructions that, when invoked and executed by a processor, cause the processor to implement the steps of the method of any of the preceding embodiments.
According to the obstacle position detection method, the obstacle position detection device and the electronic equipment based on the pseudo point cloud, the low-wire harness radar point cloud is clustered to obtain the first three-dimensional structure, the pseudo point cloud detects through the bounding box to determine the second three-dimensional structure, the first three-dimensional structure is matched with the second three-dimensional structure, if the matching result of the first three-dimensional structure and the second three-dimensional structure exceeds the preset volume threshold, the first three-dimensional structure and the second three-dimensional structure are combined, the position of the pseudo point cloud on the target obstacle is updated through the centroid of the first three-dimensional structure of the low-wire harness radar point cloud, the detection precision of the pseudo point cloud on the target obstacle is improved, and the pseudo point cloud is combined with the low-wire harness radar point cloud to improve the accuracy of pseudo point cloud position detection, so that more accurate obstacle position information is obtained under the condition of low cost.
Additional features and advantages of the disclosure will be set forth in the description which follows, or in part will be obvious from the description, or may be learned by practice of the techniques of the disclosure.
The foregoing objects, features and advantages of the disclosure will be more readily apparent from the following detailed description of the preferred embodiments taken in conjunction with the accompanying drawings.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for detecting an obstacle position based on a pseudo point cloud according to an embodiment of the present invention;
FIG. 2 is a functional block diagram of an obstacle position detecting device based on pseudo point cloud according to an embodiment of the present invention;
Fig. 3 is a schematic diagram of a hardware architecture of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the current vehicle driving field, in order to ensure the safety of vehicle driving, a laser radar is generally adopted to accurately detect the position of an obstacle, but the laser radar has higher cost and is not beneficial to the wide application of vehicle driving.
In order to save the cost, the false point cloud information similar to the coordinates (x, y, z) of the point cloud of the laser radar, which is acquired for the position of the obstacle, can be used for replacing the acquired information of the laser radar, and the algorithm has the advantages of being very low in cost, and the false point cloud information similar to the laser radar can be generated only by splicing two cameras into a binocular system. However, the algorithm has the defect that the accuracy of the pseudo point cloud is not as high as that of the laser radar point cloud, the position of the obstacle cannot be accurately identified, and the safety of a user in the vehicle driving process cannot be ensured.
Based on the above, the obstacle position detection method, the obstacle position detection device and the electronic equipment based on the pseudo point cloud, provided by the embodiment of the invention, combine the pseudo point cloud with the low-harness radar point cloud to improve the accuracy of pseudo point cloud position detection, and obtain more accurate obstacle position information under the condition of lower cost.
For the convenience of understanding the present embodiment, first, the method for detecting the position of the obstacle based on the pseudo point cloud disclosed in the embodiment of the present invention is described in detail, and the method may be applied to control devices such as a PC personal computer, a controller, a server, an intelligent terminal, etc., and may be independently set or may be integrally set in a vehicle control terminal or a road facility.
It should be noted that the low-beam lidar is low in cost, but unlike the high-beam lidar, it has a dense point cloud, so it cannot reflect the shape information of the obstacle. According to the embodiment of the invention, the low-cost low-harness laser radar is combined with the pseudo point cloud, so that the detection precision of the pseudo point cloud is improved, and the accurate obstacle position is detected under the condition of low cost.
As an alternative embodiment, the vehicle control terminal realizes accurate detection of the position of the target obstacle by combining the pseudo-point cloud information and the sparse point cloud from the low-beam radar, namely, the control terminal communicates with the binocular camera and the low-beam radar, so as to obtain the pseudo-point cloud information and the washing point cloud information of the target obstacle, wherein the binocular camera and the low-beam radar can be installed on each vehicle. Such as the front part of the vehicle and the top part of the vehicle, is used for collecting obstacles appearing in the traveling direction of the vehicle or is arranged in road facilities and used for collecting obstacles possibly encountered by each vehicle passing through the current road; as another alternative embodiment, the road facility can communicate with vehicles running on the road, the accurate position of the target obstacle is determined by collecting the pseudo point cloud information and the sparse point cloud information of the target obstacle in the view of the vehicles running on the road, and the accurate position of the target obstacle is sent to the vehicles running on the road, so that the vehicles can operate correspondingly to avoid the obstacle, and the driving safety is guaranteed under the condition of lower cost. The target obstacle may be understood as an obstacle that may affect the safety of a traveling vehicle in a road, including an obstacle such as a larger stone, an obstacle vehicle, other vehicles that travel in the current road as such, or pedestrians that travel or stand on the current road.
Fig. 1 is a flowchart of a method for detecting an obstacle position based on a pseudo point cloud according to an embodiment of the present invention.
As shown in fig. 1, the method comprises the steps of:
step S102, pseudo point cloud information aiming at the same target obstacle is obtained, and sparse Lei Dadian cloud from a low-beam radar is obtained;
Wherein, the vehicle is running in the road and probably runs into a plurality of barriers, and binocular camera and low pencil radar at this moment gather to same barrier, as an optional embodiment, can gather one by one to all barriers according to predetermineeing the order to confirm the position of every barrier.
For example, the step of acquiring pseudo point cloud information for the same target obstacle may include acquiring an image to be detected of the target obstacle from a binocular camera; calculating depth information of each pixel point in the image to be detected based on binocular parallax of a binocular distance measuring method; and converting the depth information into pseudo point cloud information according to a camera conversion principle and a geometric relation.
As an optional embodiment, the embodiment of the invention collects the point cloud information corresponding to the target obstacle in the current running vehicle running direction through two devices, namely a binocular camera and a low-beam radar.
Step S104, clustering the sparse radar point cloud to obtain a first three-dimensional structure and a mass center position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
here, by performing clustering operation on the sparse radar point cloud, a three-dimensional structure wrapping the clustering results can be generated, and the centroid position of the three-dimensional structure can be determined.
Step S106, determining a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method, wherein the second three-dimensional structure is the smallest three-dimensional structure capable of comprising the target obstacle;
the pseudo point cloud information is detected by adopting a three-dimensional bounding box boundingbox, and a second three-dimensional structure is framed. The three-dimensional structure refers to a three-dimensional shape under a three-dimensional space coordinate system, such as a cube, a cuboid, a cylinder, and the like. The specific shape of the stereoscopic structure is not limited here, and depends on the pseudo point cloud information and the clustering result.
Step S108, matching the first three-dimensional structure with the second three-dimensional structure, and if the matching result is larger than a preset volume threshold, determining the position of the target obstacle according to the centroid position of the first three-dimensional structure.
As an alternative embodiment, the step of matching the first stereo structure with the second stereo structure may be performed by calculating a three-dimensional intersection ratio (Intersection-over-Union, IOU) of the volumes of the first stereo structure and the second stereo structure, that is, performing an intersection operation on the volumes of the first stereo structure and the second stereo structure, to obtain a matching result, where the matching result is a stereo volume portion where the first stereo structure and the second stereo structure intersect.
The preset volume threshold value may be one third, one half, etc. of the whole volume of the first three-dimensional structure or the second three-dimensional structure, for example, the first three-dimensional structure and the second three-dimensional structure are subjected to intersection operation, and the obtained volume intersection part occupies one half of the whole volume of the preset first three-dimensional structure, so that the pseudo point cloud information acquired by the binocular camera system can be matched with the sparse point cloud acquired by the low-beam laser radar, and the position of a more accurate target obstacle can be further determined by utilizing the centroid position of the first three-dimensional structure.
In a preferred embodiment of practical application, clustering the low-beam radar point clouds to obtain a first three-dimensional structure, determining a second three-dimensional structure through bounding box detection by the pseudo-point clouds, matching the first three-dimensional structure with the second three-dimensional structure, combining the first three-dimensional structure with the second three-dimensional structure if a matching result exceeds a preset volume threshold, updating the position of the pseudo-point clouds to a target obstacle through the mass center of the first three-dimensional structure of the low-beam radar point clouds, improving the detection precision of the pseudo-point clouds to the target obstacle, combining the pseudo-point clouds with the low-beam radar point clouds to improve the detection accuracy of the pseudo-point cloud position, and obtaining more accurate obstacle position information under the condition of lower cost.
Here, because the low-beam radar and the binocular camera acquire the same target obstacle, under general conditions, the matching degree of the first three-dimensional structure and the second three-dimensional structure of the low-beam radar and the binocular camera can both meet the preset volume threshold, but if an abnormal condition occurs, the matching result of the low-beam radar and the binocular camera is smaller, and then the point cloud of the low-beam radar and the binocular camera cannot be combined for application. In an alternative embodiment, the method further comprises:
step 1.1), if the matching result is less than or equal to the preset volume threshold, re-executing the step of acquiring the pseudo point cloud information aiming at the same target obstacle and the sparse Lei Dadian cloud from the low-beam radar.
The low-beam radar and the binocular camera in this case may fail for some reasons, so that they do not collect corresponding to the same target obstacle, and as an alternative embodiment, the low-beam radar and the binocular camera both collect the target obstacle again, obtain corresponding point cloud information, and perform matching according to the steps of the foregoing embodiments, so as to prevent failure of matching due to abnormality occurring in the collection process.
In an alternative implementation, the method in the foregoing embodiment further includes:
Step 2.1), if the matching result is smaller than or equal to a preset volume threshold, determining the position of the target obstacle according to the centroid position of the second three-dimensional structure.
On the basis of the foregoing embodiment, as another optional embodiment, if the matching result is still smaller than or equal to the preset volume threshold after the re-matching, at this time, possibly due to low collection accuracy of the low-beam radar and the binocular camera device itself or other anomalies, in order to ensure that the vehicle in the road normally runs, the centroid corresponding to the second three-dimensional structure of the pseudo point cloud determines the position of the target obstacle, so that the position of the obstacle can be determined quickly, and potential safety hazards of the vehicle during running can be avoided.
Because the road conditions of the vehicle running are complex and changeable, and the vehicle types and volumes of the vehicles on the road are different, in order to ensure the running safety, after knowing the position of the obstacle vehicle or the obstacle, the size of the target obstacle is obtained, so that the current vehicle can perform more detailed operation, and in an optional embodiment, the method further comprises:
step 3.1), determining width information and height information of a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method;
and 3.2) determining the outline information of the target obstacle according to the width information and the height information.
The more detailed operation performed by the current vehicle includes determining what speed the current vehicle adopts to perform, and which steering operation is kept a safe distance from the target obstacle according to the position of the current vehicle, the position and size of the target obstacle, the speed and size of the current vehicle and the distance between the target obstacle and the current vehicle.
As an alternative embodiment, the more accurate centroid position determined by the sparse point cloud of the low-beam radar is used to improve the accuracy of the pseudo point cloud position detection, and step S104 may further be implemented by the following steps, specifically including:
step 4.1), clustering the sparse radar point cloud to obtain a clustering result;
step 4.2), adding a bounding box to the clustering result to obtain a first three-dimensional structure; and enabling the first three-dimensional structure to wrap and cluster the result point cloud as much as possible.
And 4.3) calculating the mass center position of the first three-dimensional structure according to the first three-dimensional structure.
The method for calculating the centroid position coordinates of the three-dimensional structure is not particularly limited, and the centroid position determination method can be realized, and as an alternative embodiment, in the three-dimensional coordinate system, the average value of the coordinates of each point in the X, Y, Z three-axis direction of the first three-dimensional structure is calculated through the three-dimensional coordinates of each point in the three-dimensional structure, and the centroid position coordinates are determined according to the coordinates.
The embodiment of the invention can enable the pseudo point cloud with low cost to be denser, and solves the problem that the pseudo point cloud with low cost is not accurate enough for detecting the position of the obstacle.
As shown in fig. 2, an embodiment of the present invention further provides an obstacle position detection device 200 based on a pseudo point cloud, where the device includes:
the acquisition module 201 acquires pseudo point cloud information for the same target obstacle and sparse Lei Dadian cloud from a low-beam radar;
the clustering module 202 performs clustering operation on the sparse radar point cloud to obtain a first three-dimensional structure and a centroid position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
The detection module 203 determines a second three-dimensional structure from the pseudo point cloud information by a three-dimensional bounding box detection method, wherein the second three-dimensional structure is a minimum three-dimensional structure capable of including the target obstacle;
The determining module 204 matches the first stereo structure with the second stereo structure, and if the matching result is greater than a preset volume threshold, determines the position of the target obstacle according to the centroid position of the first stereo structure.
In a preferred embodiment of practical application, combining the pseudo point cloud with the low-beam laser radar, performing cross-correlation calculation on a first three-dimensional structure corresponding to the clustering result of the low-beam laser radar and a second three-dimensional structure of the pseudo point cloud after the object detection of the bounding box, and updating the position of the obstacle obtained by the detection of the pseudo point cloud into the centroid position of the first three-dimensional structure corresponding to the corresponding low-beam laser radar according to the matching result after the cross-correlation.
In an alternative embodiment, the determining module is further configured to determine the location of the target obstacle according to the centroid location of the second stereo structure if the matching result is less than or equal to a preset volume threshold.
In an alternative embodiment, the obtaining module is further configured to re-perform the step of obtaining the pseudo point cloud information for the same target obstacle and the sparse Lei Dadian cloud from the low-beam radar if the matching result is less than or equal to the preset volume threshold.
In an optional embodiment, the detection module is further configured to determine width information and height information of the second stereo structure from the pseudo point cloud information by using a three-dimensional bounding box detection method; and determining contour information of the target obstacle according to the width information and the height information.
In an optional embodiment, the clustering module is further configured to perform a clustering operation on the sparse radar point cloud to obtain a clustering result; adding a bounding box to the clustering result to obtain a first three-dimensional structure; and calculating the mass center position of the first three-dimensional structure according to the first three-dimensional structure.
In an alternative embodiment, the obtaining module is further configured to obtain an image to be detected of the target obstacle from the binocular camera; calculating depth information of each pixel point in the image to be detected based on binocular parallax of a binocular distance measuring method; and converting the depth information into pseudo point cloud information according to a camera conversion principle and a geometric relation.
In an alternative embodiment, the determining module is further configured to perform a three-dimensional cross-ratio calculation on the volumes of the first and second stereo structures.
Fig. 3 is a schematic hardware architecture of an electronic device 300 according to an embodiment of the present invention. Referring to fig. 3, the electronic device 300 includes: a machine-readable storage medium 301 and a processor 302, and may also include a non-volatile storage medium 303, a communication interface 304, and a bus 305; wherein the machine-readable storage medium 301, the processor 302, the non-volatile storage medium 303, and the communication interface 304 communicate with each other via a bus 305. The processor 302 may perform the above embodiments describe a pseudo point cloud based obstacle position detection method by reading and executing machine executable instructions of obstacle position detection in a machine readable storage medium 301.
The machine-readable storage medium referred to herein may be any electronic, magnetic, optical, or other physical storage device that can contain or store information, such as executable instructions, data, or the like. For example, a machine-readable storage medium may be: RAM (Radom Access Memory, random access memory), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), any type of storage disk (e.g., optical disk, dvd, etc.), or a similar storage medium, or a combination thereof.
The non-volatile medium may be a non-volatile memory, a flash memory, a storage drive (e.g., hard drive), any type of storage disk (e.g., optical disk, dvd, etc.), or a similar non-volatile storage medium, or a combination thereof.
It can be understood that the specific operation method of each functional module in this embodiment may refer to the detailed description of the corresponding steps in the above method embodiment, and the detailed description is not repeated here.
The embodiment of the invention provides a computer readable storage medium, in which a computer program is stored, and when the computer program code is executed, the method for detecting an obstacle position based on a pseudo point cloud according to any one of the embodiments is implemented, and the specific implementation can be referred to the method embodiment and will not be repeated herein.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described system and apparatus may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.
In addition, in the description of embodiments of the present invention, unless explicitly stated and limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above examples are only specific embodiments of the present invention, and are not intended to limit the scope of the present invention, but it should be understood by those skilled in the art that the present invention is not limited thereto, and that the present invention is described in detail with reference to the foregoing examples: any person skilled in the art may modify or easily conceive of the technical solution described in the foregoing embodiments, or perform equivalent substitution of some of the technical features, while remaining within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.
Claims (8)
1. The obstacle position detection method based on the pseudo point cloud is characterized by comprising the following steps of:
Acquiring pseudo point cloud information aiming at the same target obstacle and sparse Lei Dadian cloud from a low-beam radar; the step of acquiring the pseudo point cloud information aiming at the same target obstacle comprises the following steps: acquiring an image to be detected of a target obstacle from a binocular camera; calculating depth information of each pixel point in the image to be detected based on binocular parallax of a binocular distance measuring method; converting the depth information into pseudo point cloud information according to a camera conversion principle and a geometric relationship;
Clustering the sparse radar point cloud to obtain a first three-dimensional structure and a mass center position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
Determining a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method, wherein the second three-dimensional structure is a minimum three-dimensional structure capable of comprising the target obstacle;
And carrying out three-dimensional cross-over ratio calculation on the volume of the first three-dimensional structure and the volume of the second three-dimensional structure to obtain a matching result, and if the matching result is larger than a preset volume threshold value, determining the position of the target obstacle according to the mass center position of the first three-dimensional structure.
2. The method according to claim 1, wherein the method further comprises:
and if the matching result is smaller than or equal to a preset volume threshold value, determining the position of the target obstacle according to the centroid position of the second three-dimensional structure.
3. The method according to claim 2, wherein the method further comprises:
And if the matching result is smaller than or equal to the preset volume threshold value, re-executing the step of acquiring the pseudo point cloud information aiming at the same target obstacle and the sparse Lei Dadian cloud from the low-beam radar.
4. The method according to claim 1, wherein the method further comprises:
determining width information and height information of a second three-dimensional structure from the pseudo point cloud information by a three-dimensional bounding box detection method;
And determining contour information of the target obstacle according to the width information and the height information.
5. The method of claim 1, wherein the step of clustering the sparse radar point cloud to obtain a first three-dimensional structure and a centroid position of the first three-dimensional structure comprises:
clustering the sparse radar point cloud to obtain a clustering result;
Adding a bounding box to the clustering result to obtain a first three-dimensional structure;
And calculating the mass center position of the first three-dimensional structure according to the first three-dimensional structure.
6. An obstacle position detection device based on a pseudo point cloud, the device comprising:
the acquisition module is used for acquiring pseudo point cloud information aiming at the same target obstacle and sparse Lei Dadian clouds from the low-beam radar;
The acquisition module is also used for acquiring an image to be detected of the target obstacle from the binocular camera; calculating depth information of each pixel point in the image to be detected based on binocular parallax of a binocular distance measuring method; converting the depth information into pseudo point cloud information according to a camera conversion principle and a geometric relationship;
The clustering module is used for carrying out clustering operation on the sparse radar point cloud to obtain a first three-dimensional structure and a mass center position of the first three-dimensional structure, wherein the first three-dimensional structure is a minimum three-dimensional structure capable of comprising a clustering result;
The detection module is used for determining a second three-dimensional structure from the pseudo point cloud information through a three-dimensional bounding box detection method, wherein the second three-dimensional structure is a minimum three-dimensional structure capable of comprising the target obstacle;
And the determining module is used for carrying out three-dimensional cross-over ratio calculation on the volume of the first three-dimensional structure and the volume of the second three-dimensional structure to obtain a matching result, and determining the position of the target obstacle according to the centroid position of the first three-dimensional structure if the matching result is larger than a preset volume threshold value.
7. An electronic device comprising a memory, a processor, the memory having stored therein a computer program executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method of any of the preceding claims 1 to 5.
8. A machine-readable storage medium storing machine-executable instructions which, when invoked and executed by a processor, cause the processor to perform the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110353738.6A CN112991550B (en) | 2021-03-31 | 2021-03-31 | Obstacle position detection method and device based on pseudo point cloud and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110353738.6A CN112991550B (en) | 2021-03-31 | 2021-03-31 | Obstacle position detection method and device based on pseudo point cloud and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112991550A CN112991550A (en) | 2021-06-18 |
| CN112991550B true CN112991550B (en) | 2024-06-18 |
Family
ID=76338795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110353738.6A Active CN112991550B (en) | 2021-03-31 | 2021-03-31 | Obstacle position detection method and device based on pseudo point cloud and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112991550B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113569812A (en) * | 2021-08-31 | 2021-10-29 | 东软睿驰汽车技术（沈阳）有限公司 | Unknown obstacle identification method and device and electronic equipment |
| CN114627466A (en) * | 2022-03-25 | 2022-06-14 | 山东省国土测绘院 | Automatic positioning and detecting method and system for inclined real scene model floater |
| CN116224367A (en) * | 2022-10-12 | 2023-06-06 | 深圳市速腾聚创科技有限公司 | Obstacle detection method and device, medium and electronic equipment |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612728B (en) * | 2020-05-25 | 2023-07-25 | 北京交通大学 | 3D point cloud densification method and device based on binocular RGB image |
| CN111797745B (en) * | 2020-06-28 | 2024-08-13 | 中嘉锦诚(北京)科技有限公司 | Training and predicting method, device, equipment and medium for object detection model |
| CN112101222A (en) * | 2020-09-16 | 2020-12-18 | 中国海洋大学 | Sea surface three-dimensional target detection method based on unmanned ship multi-mode sensor |
-
2021
- 2021-03-31 CN CN202110353738.6A patent/CN112991550B/en active Active
Non-Patent Citations (2)
| Title |
|---|
| Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation;Jiaming Sun等;《arXiv》;1-10 * |
| 面向多源信息融合的智能汽车容错感知方法研究;傅骏伟;《中国博士学位论文全文数据库 (工程科技Ⅱ辑)》(第1期);C035-29 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112991550A (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112991550B (en) | Obstacle position detection method and device based on pseudo point cloud and electronic equipment | |
| CN112419494B (en) | Obstacle detection and marking method and device for automatic driving and storage medium | |
| JP6227318B2 (en) | Detection device for detecting obstacles and steep slopes, and vehicle equipped with the same | |
| CN110674705B (en) | Small-sized obstacle detection method and device based on multi-line laser radar | |
| KR101163042B1 (en) | Apparatus and Method for Extracting Vehicle | |
| KR20150074544A (en) | Method of tracking vehicle | |
| JP2015125760A (en) | Mine work machine | |
| CN113768419B (en) | Method and device for determining sweeping direction of sweeper and sweeper | |
| CN111386530B (en) | Vehicle detection method and apparatus | |
| JP5821276B2 (en) | Detection device for position and traveling direction of moving body | |
| CN112455502A (en) | Train positioning method and device based on laser radar | |
| CN111497741A (en) | Collision early warning method and device | |
| CN110426714B (en) | Obstacle identification method | |
| CN108399360A (en) | A kind of continuous type obstacle detection method, device and terminal | |
| CN118037790A (en) | Point cloud processing method and device, computer equipment and storage medium | |
| CN113569812A (en) | Unknown obstacle identification method and device and electronic equipment | |
| Leu et al. | High speed stereo vision based automotive collision warning system | |
| CN117470258A (en) | Map construction method, device, equipment and medium | |
| CN113436309A (en) | Scene reconstruction method, system and device and sweeping robot | |
| CN116358528A (en) | Map updating method, map updating device, self-mobile device and storage medium | |
| CN110015288B (en) | Method and device for detecting library position and electronic equipment | |
| CN114333390B (en) | Method, device and system for detecting shared vehicle parking event | |
| CN112380963B (en) | Depth information determining method and device based on panoramic looking-around system | |
| CN115236672A (en) | Obstacle information generation method, device, equipment and computer readable storage medium | |
| CN113030976A (en) | Method for eliminating interference of metal well lid on millimeter wave radar by using laser radar |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |