CN108957432B - Road edge detection method and device, computer equipment and storage medium - Google Patents
Road edge detection method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN108957432B CN108957432B CN201810311877.0A CN201810311877A CN108957432B CN 108957432 B CN108957432 B CN 108957432B CN 201810311877 A CN201810311877 A CN 201810311877A CN 108957432 B CN108957432 B CN 108957432B
- Authority
- CN
- China
- Prior art keywords
- scanning
- road
- scanning points
- road surface
- points
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Abstract
The application relates to a road edge detection method, a road edge detection device, computer equipment and a storage medium. The method comprises the following steps: scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point; carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points; extracting pavement scanning points from the multiple scanning points by using the normal features, and calculating the pavement height corresponding to the pavement scanning points; acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points; and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points. The road edge detection method provided by the application is small in noise interference, and the road edge information detection is more accurate.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for detecting a road edge, a computer device, and a storage medium.
Background
The road edge is an important component of an urban road, and along with the development of technologies such as unmanned driving and the like and the drawing requirement aiming at a high-precision map, a road edge detection technology appears.
Among the road edge detection techniques, a road edge detection method based on laser scanning is widely adopted due to its advantage of being not affected by weather and light. However, the conventional road edge detection method based on laser scanning obtains the road edge information by performing quadratic curve fitting on the road edge candidate points, and in the fitting process, problems such as that the road edge candidate points cannot be identified and overfitting are easily caused, and in the identification process, the detected road edge information is easily influenced by noise, so that the detected road edge information is inaccurate. Therefore, the traditional road edge detection method has the problem of inaccurate road edge information detection.
Disclosure of Invention
In view of the above, it is necessary to provide a method, an apparatus, a computer device and a storage medium for detecting a road edge with more accurate road edge information detection.
A method of road edge detection, the method comprising:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
extracting pavement scanning points from the multiple scanning points by using the normal features, and calculating the pavement height corresponding to the pavement scanning points;
acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
In one embodiment, the tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain the line features and normal features of the plurality of scanning points includes: encoding the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points; tensor voting is carried out on a plurality of tensors; carrying out tensor decomposition according to the result of tensor voting to obtain the rod tensor and the plate tensor of a plurality of scanning points; and extracting features according to the rod tensor and the plate tensor of the plurality of scanning points to obtain line features and normal features of the plurality of scanning points.
In one embodiment, the method further comprises establishing a three-dimensional coordinate system by taking a direction vertical to the road surface and upward as a Z axis; the step of extracting the road surface scanning points from the plurality of scanning points by using the normal features and calculating the road surface height corresponding to the road surface scanning points comprises the following steps: acquiring an included angle between the normal characteristic and a Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; and acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis.
In one embodiment, the step of filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line feature includes: and when the road surface height is greater than the height threshold value or the line characteristic value of the road surface scanning point is greater than the line characteristic threshold value, filtering out the corresponding road surface scanning point.
In one embodiment, the echo signal also carries a relative displacement of the scanning point, and the method further includes: measuring relative displacement of the vehicle between the multiple scans of the laser beam; determining the relative position of a scanning point obtained by multiple times of scanning according to the relative displacement; and correcting the road edge information according to the relative position of the scanning point.
In one embodiment, the method further comprises: measuring linear acceleration and angular acceleration of the vehicle; when the linear acceleration is larger than a linear acceleration threshold value, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than an angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
In one embodiment, the method further comprises: counting the road height corresponding to the road scanning points, and dividing a plurality of road scanning points into a plurality of intervals; calculating the average road height corresponding to the road scanning points in the multiple sections; and correcting the road edge information according to the average road surface height.
A road edge detection device, the device comprising:
the acquisition module is used for scanning a target environment for multiple times by utilizing a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
the tensor voting module is used for carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
the data processing module is used for extracting road surface scanning points from the plurality of scanning points by using the normal features and calculating the road surface height corresponding to the road surface scanning points;
the filtering module is used for acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points; and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
A computer device comprising a memory, the memory storing a computer program, a processor implementing the following steps when the processor executes the computer program:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
extracting pavement scanning points from the multiple scanning points by using the normal features, and calculating the pavement height corresponding to the pavement scanning points;
acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
extracting pavement scanning points from the multiple scanning points by using the normal features, and calculating the pavement height corresponding to the pavement scanning points;
acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
According to the road edge detection method, the device, the computer equipment and the storage medium, the laser beam is used for scanning the target environment for multiple times to obtain the multiple scanning points reflected back by the target environment, tensor voting is carried out on the scanning points to extract the line characteristics and the normal characteristics of the scanning points, the line characteristics and the normal characteristics of the scanning points are further analyzed to obtain the road edge information, the problems that identification cannot be achieved and overfitting cannot be achieved in the secondary curve fitting process in the traditional method are solved, the influence of noise is greatly reduced by using tensor voting, and the obtained road edge information is more accurate.
Drawings
FIG. 1 is a diagram of an exemplary implementation of a method for road edge detection;
FIG. 2 is a flow diagram of a method for road edge detection in one embodiment;
FIG. 3 is a schematic flow chart of a method for road edge detection in another embodiment;
FIG. 4 is a block diagram of a road edge detection device according to an embodiment;
FIG. 5 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The road edge detection method provided by the application can be applied to the application environment shown in fig. 1. Lidar 102 is communicatively coupled to a terminal 104. The communication connection between laser radar 102 and terminal 104 may be a wired connection or a wireless connection. Taking the example that the laser radar 102 and the terminal 104 are installed in a running vehicle, the laser radar 102 emits a laser beam to scan the target environment in a 360-degree surrounding manner, and returns an echo signal. The terminal 104 receives the returned echo signal, analyzes and processes the echo signal, and acquires the road edge information. Laser radar 102 may be, but is not limited to, a pulsed laser radar and a continuous wave laser radar, among others. The terminals 104 may be, but are not limited to, various personal computers, laptops, smartphones, tablets, and navigators.
In one embodiment, as shown in fig. 2, a method for detecting a road edge is provided, which is described by taking the method as an example for being applied to the terminal in fig. 1, and includes the following steps:
step 202, scanning a target environment by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scanning points; the echo signal carries the three-dimensional coordinate information of the scanning point.
In this embodiment, a laser radar is used to emit a laser beam to scan the target environment in a 360-degree surrounding manner, and when the laser beam irradiates a real object in the target environment, an echo signal is returned. The target environment refers to a three-dimensional space in a certain range which is diffused outwards by taking a laser radar as a center, and the real object can be a road surface, a road edge, a plant, a building, other vehicles and the like. Specifically, the lidar emits a laser beam at an inclination and frequency while scanning a target environment. When the laser beam is irradiated on the real object, an echo signal is returned. The echo signal carries the three-dimensional coordinate information of the scanning point reflected by the laser beam irradiated on the real object in the target environment. When the laser radar emits a laser beam every time and correspondingly returns an echo signal, the terminal receives the echo signal and analyzes the echo signal to obtain a scanning point and three-dimensional coordinate information corresponding to the scanning point. The target environment comprises a plurality of scanning points, the target environment is scanned at a certain frequency through the laser radar within a preset time, and the terminal can obtain the plurality of scanning points in the target environment and three-dimensional coordinate information corresponding to the plurality of scanning points.
In one embodiment, a target environment is scanned with a plurality of laser beams. Specifically, a multiline laser radar is adopted to simultaneously emit a plurality of laser beams at different inclination angles to scan a target environment. In the preset time, a plurality of laser beams are used for scanning the target environment, and the terminal can obtain more scanning points, so that the finally obtained road edge information is more accurate.
And 204, carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points.
The plurality of scanning points can form a three-dimensional point cloud, and the three-dimensional point cloud can form a three-dimensional image, wherein the three-dimensional image comprises a plurality of points, lines and surfaces. The scanning points in the three-dimensional image can be discrete points, can also form lines with other scanning points, and can also form surfaces with other scanning points. The line characteristics of a scanning point include the line formed by the scanning point and other scanning points and the significance of the formed line. The normal characteristic of the scanning point refers to the normal characteristic of the surface of the scanning point at the point. The surface of the scanning point can be described by the surface characteristics of the scanning point, and the surface characteristics of the scanning point comprise the surfaces formed by the scanning point and other scanning points and the significance of the surfaces formed by the scanning point and other scanning points.
After the terminal obtains a plurality of scanning points according to a scanning target environment of the laser radar in a preset time, tensor voting is respectively carried out on the plurality of scanning points according to three-dimensional coordinate information of the plurality of scanning points, and a plurality of tensor voting results corresponding to the plurality of scanning points are obtained. And analyzing the tensor voting result by the terminal to obtain the line characteristics and normal characteristics of a plurality of scanning points.
In one embodiment, the tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain the line features and normal features of the plurality of scanning points includes: encoding the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points; tensor voting is carried out on a plurality of tensors; carrying out tensor decomposition according to the result of tensor voting to obtain the rod tensor and the plate tensor of a plurality of scanning points; and extracting features according to the rod tensor and the plate tensor of the plurality of scanning points to obtain line features and normal features of the plurality of scanning points.
And the terminal encodes the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points. Specifically, the terminal encodes the scan points as a sparse tensor. The sparse tensor can be encoded by using an identity matrix, and in this embodiment, the identity matrix obtained according to the three-dimensional coordinate information of the scanning point is a third-order identity matrix. Further, the terminal tensor votes for the plurality of sparse tensors. Sparse tensor voting can be expressed by double integration, which is expressed as:
where p is the sparse tensor corresponding to the scan point, RγαIs a transformation matrix of a three-dimensional sphere in 360-degree direction, TSAnd (4) performing sparse tensor voting on the scanning points in the 360-degree direction of the three-dimensional sphere. Transformation matrix RγαFor coupling TSRotated into the corresponding direction. And the terminal conducts sparse tensor voting on the sparse tensor of each scanning point in the 360-degree direction of the three-dimensional sphere to obtain sparse tensor voting results in each direction, and accumulates the sparse tensor voting results in each direction to obtain a refined tensor corresponding to the scanning point. And the terminal can obtain the estimated normal features of the scanning points by analyzing the refinement tensor.
Further, the terminal conducts dense tensor voting on the scanning points according to the refined tensor of the scanning points. Specifically, the terminal performs dense tensor voting on a plurality of scanning points aiming at spatial grid points to form a dense tensor field. The dense tensor field is formed by dense tensors of all scanning points obtained after the terminal conducts dense tensor voting on the refined tensors of the scanning points. Specifically, the dense tensor voting includes a stick tensor voting and a slab tensor voting. The expression of the wand tensor vote is:
wherein, VpFor the estimated normal feature of the scan point, DF (S, k, δ) is the decay function, which is defined as follows:
wherein, S is the arc length between the current scanning point and the adjacent scanning point, k is the curvature of the arc between the current scanning point and the adjacent scanning point, δ is a parameter for controlling the voting decay speed, and C is a parameter for balancing the distance and the curvature between the current scanning point and the adjacent scanning point.
The expression of the slab tensor vote is:
wherein theta is an included angle formed by a connecting line of the scanning point and the adjacent scanning point and a tangent line of the surface where the scanning point is located, and R isθIs the long axis of the plane of the scanning point.
Further, the terminal performs tensor decomposition according to the result of dense tensor voting, and the expression obtained by tensor decomposition is as follows:
wherein e is1、e2、e3For the decomposed eigenvectors, lambda1、λ2、λ3Is the eigenvalue.
Is the rod tensor of the scanning point, the terminal rootAnd extracting features according to the rod tensor of the scanning point to obtain the line features of the scanning point. Lambda [ alpha ]1-λ2Is the saliency of the rod tensor of a scan point, i.e. the saliency of the line features of a scan point. When the line feature is more significant, the line feature of a scanning point is more significant, that is, a line formed by the scanning point and another scanning point is more significant.
And the terminal extracts features according to the plate tensor of the scanning point to obtain the surface features of the scanning point. Lambda [ alpha ]2-λ3Is the saliency of the sheet tensor, i.e., the saliency of the surface features of the scan point. The more significant the surface feature is, the more significant the surface feature of the scanning point is, that is, the more significant the surface formed by the scanning point and the other scanning points is. Further, the terminal obtains the normal characteristic of the surface where the scanning point is located at the scanning point according to the surface characteristic of the scanning point. Specifically, the terminal calculates the tangent plane of the plane where the scanning point is located at the scanning point, and further calculates the normal feature of the tangent plane at the point according to the given tangent plane.
And step 206, extracting the road surface scanning points from the plurality of scanning points by using the normal features, and calculating the road surface height corresponding to the road surface scanning points.
In this embodiment, the terminal needs to acquire a scanning point reflected back on the road surface irradiated by the laser beam emitted by the laser radar, so as to further analyze the scanning point reflected back on the road surface to acquire the road edge information. The road surface scanning points comprise scanning points reflected back by the road surface. The vehicle chassis and the road surface which currently bears the vehicle are in parallel relation at any time when the vehicle provided with the laser radar runs. And the terminal extracts the road surface scanning points from the scanning points according to the geometric position relation between the vehicle chassis and the road surface and the three-dimensional coordinate information of the scanning points. Further, the terminal calculates the road height of the road surface where the road surface scanning point is located according to the normal characteristic of the road surface scanning point. If the road surface of the current bearing vehicle is taken as a reference object, the height of the road surface is the distance from the road surface scanning point to the plane where the road surface of the current bearing vehicle is located, and if the chassis of the vehicle is taken as a reference object, the height of the road surface is the distance from the road surface scanning point to the plane where the chassis of the vehicle is located.
In one embodiment, the road edge detection method further comprises the steps of establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; the method comprises the following steps of extracting road surface scanning points from a plurality of scanning points by utilizing normal characteristics, and calculating the road surface height corresponding to the road surface scanning points, wherein the steps comprise: acquiring an included angle between the normal characteristic and the Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; and acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis.
And the terminal establishes a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis. Specifically, the terminal can establish a three-dimensional coordinate system by taking any plane parallel to the road surface as an X-Y plane and taking the intersection point of the X-Y plane and the Z axis as an origin. For convenience of calculation, a three-dimensional coordinate system can be established by taking a projection point of the laser radar on the road surface as an origin, the road surface as an X-Y plane and the direction vertical to the road surface and upward as a Z axis. And a three-dimensional coordinate system can be established by taking the laser radar as an origin, taking a plane where the laser radar is positioned and parallel to the road surface as an X-Y plane, and taking a direction vertical to the road surface and upward as a Z axis.
In one embodiment, the terminal establishes a three-dimensional coordinate system by taking a plane where a road surface bearing vehicles currently is located as an X-Y plane and an upward direction perpendicular to the road surface as a Z axis and taking a projection point of a laser radar on the road surface as an origin. And the terminal acquires an included angle between the normal characteristic and the Z axis according to the normal characteristic of the scanning point. Because the road surface scanning points are the scanning points reflected back by the road surface in a certain range which takes the projection of the laser radar on the road surface as the center and spreads outwards, the included angle between the normal characteristic of the road surface scanning points and the Z axis is in a certain included angle threshold range. And the terminal determines an included angle threshold in advance, and extracts the pavement scanning points from the multiple scanning points by using the included angle threshold and the included angle between the normal characteristic of the scanning points and the Z axis. Specifically, the terminal compares the normal characteristic of the scanning point with the included angle of the Z axis and the included angle threshold, and determines whether the included angle between the normal characteristic of the scanning point and the Z axis is greater than the included angle threshold. And when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a road surface scanning point. Thereby extracting the road surface scanning point from the plurality of scanning points.
Further, the terminal obtains the road height corresponding to the plurality of road surface scanning points according to the component of the normal features of the plurality of road surface scanning points in the Z axis. In this embodiment, the road height is a distance from the road scanning point to a plane where the road currently carrying the vehicle is located. Because the Z axis is perpendicular to the road surface of the current bearing vehicle, the terminal can acquire the corresponding road surface height according to the component of the normal characteristic of the road surface scanning point on the Z axis. The terminal can also directly acquire the corresponding road height according to the three-dimensional coordinate information of the road scanning point. And the Z coordinate in the three-dimensional coordinate information of the road scanning point is the road height corresponding to the road scanning point.
In one embodiment, the terminal takes the laser radar as an origin, a plane where the laser radar is located and is parallel to the road surface of the bearing vehicle is an X-Y plane, and a direction perpendicular to the road surface of the bearing vehicle and facing upwards is a Z axis, so that a three-dimensional coordinate system is established. More specifically, the direction in which the vehicle travels is taken as the Y-axis.
And step 208, acquiring a height threshold corresponding to the road height and a threshold of line characteristics of the road scanning points.
And step 210, filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
Because the scanning points, in which the included angle between the normal characteristic of the scanning point and the Z axis is within the threshold value of the included angle, of the plurality of scanning points obtained by the terminal according to the scanning target environment of the laser radar also include the scanning points reflected by other objects parallel to the road surface, and the other objects may be, for example, the roof of other vehicles. That is, the road surface scanning points obtained according to step 206 also include the scanning points reflected back by other objects, and since the scanning points reflected back by other objects are not required by this embodiment, the terminal needs to filter the extracted road surface scanning points to obtain more accurate road surface scanning points reflected back by the road surface. Specifically, the terminal may filter the road surface scanning points according to a certain filtering condition. The filtering conditions may be selected based on the distinguishing characteristics of the road surface from other objects. In this embodiment, the selected filtering condition is the road height. Because the embodiment finally needs to acquire the road edge information, the road edge information is a road edge line with remarkable line characteristics, and the road surface scanning points extracted by the terminal also include scanning points with inconspicuous line characteristics, that is, scanning points reflected by a road surface far away from the road edge line. Therefore, the terminal also needs to filter the road scanning points according to the significance of the line features.
The terminal obtains a height threshold corresponding to the road height and a threshold of line characteristics of the road scanning points. The height threshold and the line feature threshold may be preset. The road scanning points are stored in the terminal in the form of echo signals. Further, the terminal filters the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics. Specifically, the terminal filters out the road surface scanning points with the road surface height greater than the height threshold value to obtain the accurate road surface scanning points with the road surface height less than or equal to the height threshold value. And the terminal filters the echo signal corresponding to the accurate road surface scanning point according to the threshold value of the line characteristic. Specifically, the terminal compares the saliency of the line characteristic of the accurate road surface scanning point with a threshold value of the line characteristic, filters the accurate road surface scanning point of which the saliency of the line characteristic is smaller than the threshold value of the line characteristic, obtains the accurate road surface scanning point of which the saliency of the line characteristic is greater than or equal to the threshold value of the line characteristic, and obtains the road edge information.
Further, the terminal projects the obtained road edge information to the spatial grid points to obtain a road edge projection image for storage, and the road edge projection image is displayed on a display screen of the terminal.
In this embodiment, the terminal scans the target environment for multiple times by using the laser beam to obtain multiple scanning points reflected back by the target environment, and performs tensor voting on the scanning points to extract line features and normal features of the scanning points, so as to further analyze the line features and the normal features of the scanning points to obtain road edge information.
In an embodiment, the echo signal also carries a relative displacement of the scanning point, and the road edge detection method provided in this embodiment further includes the following steps:
the relative displacement of the scanning spot between the multiple scans of the laser beam is calculated, step 212.
In this embodiment, a three-dimensional coordinate system established by using a laser radar to emit a laser beam, and using a plane where a road surface currently bearing a vehicle is located as an X-Y plane, a vehicle traveling direction as a Y axis, an upward direction perpendicular to the road surface as a Z axis, and a projection point of the laser radar on the road surface as an origin is described as an example. The scanning point obtained by the terminal is obtained by scanning a target environment by the laser radar installed in a running vehicle, namely, the laser radar is in a motion state, namely, a three-dimensional coordinate system established by taking a projection point of the laser radar on a road surface as an origin is in a motion state. And because the scanning point obtained by the terminal is obtained by multiple times of scanning within the preset time, and the position of the laser radar in the target environment is changed between different two times of scanning, namely the original point of the three-dimensional coordinate system is changed, the relative position of the scanning point in the target environment cannot be accurately determined according to the three-dimensional coordinate information of the scanning point, so that the relative displacement of the scanning point needs to be obtained, and the position of the scanning point in the target environment is determined by combining the relative displacement with the three-dimensional coordinate information of the scanning point. Wherein the relative displacement can be obtained by calculating the displacement of the vehicle running between two scans within a preset time.
The terminal calculates the relative displacement of the scanning spot between two scans of the laser beam. Specifically, the terminal may estimate the relative displacement of the vehicle from the scan point transition between two scans of the lidar measurement laser beam. Further, the laser radar estimates the speed of the vehicle by measuring the relative displacement of the vehicle, and the speed of the vehicle is obtained according to the interval time of two scans.
And step 214, determining the relative position of the scanning point obtained by the plurality of times of scanning according to the relative displacement.
Because the laser radar is in motion in the scanning process of transmitting the laser beam by the laser radar within the preset time, that is, the three-dimensional coordinate system is dynamically changed, the three-dimensional coordinate information of the scanning point obtained by the terminal according to any two times of scanning of the laser beam transmitted by the laser radar within the preset time is not in the same three-dimensional coordinate system, and therefore the relative positions of the two scanning points corresponding to the target environment need to be determined, so that the road edge information of the target environment can be more accurately determined.
And the terminal determines the relative position of the scanning point obtained by two times of scanning according to the relative displacement. Specifically, the terminal adjusts the three-dimensional coordinate information of the scanning point according to the relative displacement. Specifically, the terminal selects the three-dimensional coordinate system during one scanning as a standard coordinate system, and correspondingly adjusts the three-dimensional coordinate information of the scanning point obtained by other scanning within the preset time to the three-dimensional coordinate information under the standard coordinate system. For example, the terminal selects the three-dimensional coordinate system at the first scanning within the preset time as the standard coordinate system. Further, the terminal calculates three-dimensional coordinate information of the scanning point obtained by the current scanning in the standard coordinate system according to relative displacement of the laser radar in the other scanning relative to the laser radar in the first scanning and the corresponding relative displacement, and obtains relative positions of the scanning point obtained by the other scanning relative to the scanning point obtained by the first scanning. The relative displacement is a vector, and the calculation process is to add the three-dimensional coordinate information of the scanning point and the relative displacement to obtain the three-dimensional coordinate information under the standard coordinate system.
And step 216, correcting the road edge information according to the relative position of the scanning point.
The terminal corrects the road edge information obtained in the previous embodiment according to the relative positions of the scanning points obtained by the plurality of times of scanning. And the terminal adjusts each scanning point in the spatial grid point according to the three-dimensional coordinate information of the scanning point in the standard coordinate system to obtain a corrected road edge projection image.
In this embodiment, the relative displacement of the vehicle between the multiple scans of the laser beam is measured, the relative position of the scanning point obtained by the multiple scans is determined according to the relative displacement, and the road edge information is corrected according to the relative position of the scanning point, so that the obtained road edge information is more accurate.
In one embodiment, the road edge detection method further comprises: measuring linear acceleration and angular acceleration of the vehicle; when the linear acceleration is larger than the linear acceleration threshold, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than the angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
During the running of the vehicle, the acceleration of the vehicle may change, for example, the linear acceleration of the vehicle changes when the speed of the vehicle is increased or decreased, and the angular acceleration of the vehicle may also change due to the change of the road curvature. When the linear acceleration and the angular acceleration of the vehicle change, the relative displacement between the multiple scans, which is obtained by the lidar only based on the velocity and time measurements, is not accurate enough, so it is necessary to estimate the relative displacement between the multiple scans based on the acceleration of the vehicle to make it more accurate.
In this embodiment, the linear acceleration and the angular acceleration of the laser radar are measured by using the inertial measurement unit. Before the terminal calculates the relative displacement between two scans, whether the linear acceleration is greater than a linear acceleration threshold value or not and whether the angular acceleration is greater than an angular acceleration threshold value or not are judged. Specifically, the terminal calculates the linear acceleration average value and the angular acceleration average value between two scans, and judges whether the linear acceleration average value is greater than the linear acceleration threshold value and whether the angular acceleration average value is greater than the angular acceleration threshold value. Further, when the linear acceleration average value is larger than the linear acceleration threshold value, the relative displacement is calculated according to the linear acceleration average value. Specifically, the terminal performs secondary integration on the linear acceleration for the interval time of two scanning to obtain the relative displacement of the vehicle between the two scanning, namely the relative displacement of the laser radar. And when the angular acceleration average value is larger than the angular acceleration threshold value, correcting the relative displacement according to the angular acceleration. Specifically, the terminal performs a second integration of the angular acceleration with respect to the interval time between two scans to obtain the rotation angle of the vehicle between two scans, and the terminal corrects the relative displacement obtained in step 212 in the previous embodiment according to the rotation angle. Wherein the relative displacement obtained in step 212 is substantially the length of an arc line of the vehicle rotation, and further, the terminal calculates the relative displacement of the vehicle on the straight line in two scans according to the length of the arc line and the rotation angle.
In this embodiment, the linear acceleration and the angular acceleration of the vehicle are measured, and the relative displacement of the vehicle measured in the previous embodiment is corrected by combining the linear acceleration and the angular acceleration, so that the obtained relative displacement is more accurate, and the road edge information corrected according to the relative displacement is more accurate.
In one embodiment, the road edge detection method further comprises: counting the road height corresponding to the road scanning points, and dividing a plurality of road scanning points into a plurality of intervals; calculating the average road height corresponding to the road scanning points in the multiple sections; and correcting the road edge information according to the average road surface height.
In an actual scene, a road surface in a target environment generally has certain height fluctuation, the road surface heights of a plurality of corresponding road surface scanning points have a plurality of different height values, the road surface scanning points are discrete points, and because the road surface heights obtained by terminal calculation have certain errors, the road surface scanning points in the obtained road edge information are not coherent and cannot clearly reflect a coherent and smooth road surface, and in order to more clearly and coherently reflect the height change of the road surface, namely the height change of the road edge, the road surface heights of the plurality of road surface scanning points need to be further processed.
The terminal counts the road height corresponding to the road scanning points and divides the road scanning points into a plurality of intervals. The terminal may equally divide the road surface scanning points according to a certain road surface height range to obtain a plurality of sections, for example, the terminal may divide the plurality of scanning points into a plurality of sections according to the road surface height of the equal difference. The terminal can also divide the interval according to the number of road surface scanning point in certain altitude range, and specifically, the terminal is arranged a plurality of scanning points according to road surface height, intercepts equivalent scanning point in proper order and divides into an interval to divide a plurality of scanning points into a plurality of intervals. Further, the terminal calculates average road heights corresponding to the road scanning points in the multiple sections, and corrects the road edge information according to the average road heights. Specifically, the terminal corrects the three-dimensional coordinate information of the road scanning points in the interval according to the average road height of the road scanning points in the interval, and obtains corrected road edge information. Further, the terminal projects the corrected road edge information to the spatial grid points to obtain road edge images, and the road edge images are displayed on a display screen of the terminal.
In this embodiment, the road surface scanning points are divided into a plurality of sections according to the road surface height, and the average road surface height corresponding to the road surface scanning points in the plurality of sections is calculated respectively, so that the road edge information is corrected, the continuity of the road edge information is improved, the height change of the road edge can be reflected more clearly and more consistently, and the road edge information is made more clear and more definite.
It should be understood that although the various steps in the flow charts of fig. 2-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 4, there is provided a road edge detection apparatus including: an acquisition module 410, a tensor voting module 420, a data processing module 430, and a filtering module 440, wherein:
the acquisition module 410 is configured to scan a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scanning points; the echo signals carry three-dimensional coordinate information of the scanning points;
the tensor voting module 420 is configured to perform tensor voting according to the three-dimensional coordinate information of the multiple scanning points to obtain line features and normal features of the multiple scanning points;
the data processing module 430 is configured to extract road surface scanning points from the multiple scanning points by using the normal features, and calculate a road surface height corresponding to the road surface scanning points;
the filtering module 440 is configured to obtain a height threshold corresponding to the road height and a threshold of a line characteristic of a road scanning point; and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
In one embodiment, the tensor voting module is further used for encoding the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points; tensor voting is carried out on a plurality of tensors; carrying out tensor decomposition according to the result of tensor voting to obtain the rod tensor and the plate tensor of a plurality of scanning points; and extracting features according to the rod tensor and the plate tensor of the plurality of scanning points to obtain line features and normal features of the plurality of scanning points.
In one embodiment, the data processing module is further used for establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; acquiring an included angle between the normal characteristic and the Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; and acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis.
In one embodiment, the filtering module is further configured to filter out the corresponding road surface scanning point when the road surface height is greater than the height threshold or the line characteristic value of the road surface scanning point is greater than the line characteristic threshold.
In one embodiment, the echo signals also carry the relative displacement of the scanning point; the data processing module is also used for measuring the relative displacement of the vehicle among a plurality of scans of the laser beam; determining the relative position of a scanning point obtained by multiple times of scanning according to the relative displacement; and correcting the road edge information according to the relative position of the scanning point.
In one embodiment, the data processing module is further configured to measure linear and angular accelerations of the vehicle; when the linear acceleration is larger than the linear acceleration threshold, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than the angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
In one embodiment, the data processing module is further configured to count road heights corresponding to the road surface scanning points, and divide the plurality of road surface scanning points into a plurality of intervals; calculating the average road height corresponding to the road scanning points in the multiple sections; and correcting the road edge information according to the average road surface height.
For the specific definition of the road edge detection device, reference may be made to the above definition of the road edge detection method, which is not described herein again. The modules in the above-mentioned road edge detection device can be wholly or partially implemented by software, hardware and their combination. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 5. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a road edge detection method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scanning points; the echo signals carry three-dimensional coordinate information of the scanning points;
carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
extracting pavement scanning points from the multiple scanning points by using normal characteristics, and calculating the pavement height corresponding to the pavement scanning points;
acquiring a height threshold corresponding to the height of the road surface and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
In one embodiment, the processor, when executing the computer program, further performs the steps of: encoding the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points; tensor voting is carried out on a plurality of tensors; carrying out tensor decomposition according to the result of tensor voting to obtain the rod tensor and the plate tensor of a plurality of scanning points; and extracting features according to the rod tensor and the plate tensor of the plurality of scanning points to obtain line features and normal features of the plurality of scanning points.
In one embodiment, the processor, when executing the computer program, further performs the steps of: establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; acquiring an included angle between the normal characteristic and the Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; and acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis.
In one embodiment, the processor, when executing the computer program, further performs the steps of: and when the road height is greater than the height threshold value or the line characteristic value of the road scanning point is greater than the line characteristic threshold value, filtering out the corresponding road scanning point.
In one embodiment, the echo signals also carry the relative displacement of the scanning point, and the processor executes the computer program to further implement the following steps: measuring relative displacement of the vehicle between multiple scans of the laser beam; determining the relative position of a scanning point obtained by multiple times of scanning according to the relative displacement; and correcting the road edge information according to the relative position of the scanning point.
In one embodiment, the processor, when executing the computer program, further performs the steps of: measuring linear acceleration and angular acceleration of the vehicle; when the linear acceleration is larger than the linear acceleration threshold, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than the angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
In one embodiment, the processor, when executing the computer program, further performs the steps of: counting the road height corresponding to the road scanning points, and dividing a plurality of road scanning points into a plurality of intervals; calculating the average road height corresponding to the road scanning points in the multiple sections; and correcting the road edge information according to the average road surface height.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scanning points; the echo signals carry three-dimensional coordinate information of the scanning points;
carrying out tensor voting according to the three-dimensional coordinate information of the plurality of scanning points to obtain line characteristics and normal characteristics of the plurality of scanning points;
extracting pavement scanning points from the multiple scanning points by using normal characteristics, and calculating the pavement height corresponding to the pavement scanning points;
acquiring a height threshold corresponding to the height of the road surface and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
In one embodiment, the computer program when executed by the processor further performs the steps of: encoding the plurality of scanning points into a plurality of tensors according to the three-dimensional coordinate information of the plurality of scanning points; tensor voting is carried out on a plurality of tensors; carrying out tensor decomposition according to the result of tensor voting to obtain the rod tensor and the plate tensor of a plurality of scanning points; and extracting features according to the rod tensor and the plate tensor of the plurality of scanning points to obtain line features and normal features of the plurality of scanning points.
In one embodiment, the computer program when executed by the processor further performs the steps of: establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; acquiring an included angle between the normal characteristic and the Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; and acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis.
In one embodiment, the computer program when executed by the processor further performs the steps of: and when the road height is greater than the height threshold value or the line characteristic value of the road scanning point is greater than the line characteristic threshold value, filtering out the corresponding road scanning point.
In an embodiment, the echo signals also carry the relative displacement of the scanning point, and the computer program, when executed by the processor, further performs the steps of: measuring relative displacement of the vehicle between multiple scans of the laser beam; determining the relative position of a scanning point obtained by multiple times of scanning according to the relative displacement; and correcting the road edge information according to the relative position of the scanning point.
In one embodiment, the computer program when executed by the processor further performs the steps of: measuring linear acceleration and angular acceleration of the vehicle; when the linear acceleration is larger than the linear acceleration threshold, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than the angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
In one embodiment, the computer program when executed by the processor further performs the steps of: counting the road height corresponding to the road scanning points, and dividing a plurality of road scanning points into a plurality of intervals; calculating the average road height corresponding to the road scanning points in the multiple sections; and correcting the road edge information according to the average road surface height.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method of road edge detection, the method comprising:
scanning a target environment for multiple times by using a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
respectively coding each scanning point into a sparse tensor according to the three-dimensional coordinate information of the plurality of scanning points; performing sparse tensor voting on the sparse tensor to obtain a refined tensor corresponding to each scanning point; performing dense tensor voting on the scanning points according to the refinement tensor corresponding to each scanning point; obtaining line features and normal features of corresponding scanning points according to the result of the dense tensor voting;
establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; acquiring an included angle between the normal characteristic and a Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis;
acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points;
and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
2. The method of claim 1, wherein obtaining line features and normal features of corresponding scan points from the result of the dense tensor voting comprises:
carrying out tensor decomposition on the result of the dense tensor voting to obtain a rod tensor and a plate tensor of each scanning point;
and extracting features according to the rod tensor and the plate tensor of each scanning point, and respectively obtaining the line features and the normal features of each scanning point.
3. The method of claim 1, wherein the dense tensor voting comprises:
a wand tensor vote and a board tensor vote.
4. The method of claim 1, wherein the step of filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of line features comprises:
and when the road surface height is greater than the height threshold value or the line characteristic value of the road surface scanning point is greater than the line characteristic threshold value, filtering out the corresponding road surface scanning point.
5. The method of claim 1, wherein the echo signals also carry relative displacements of the scanning point, the method further comprising: measuring relative displacement of the vehicle between the multiple scans of the laser beam; determining the relative position of a scanning point obtained by multiple times of scanning according to the relative displacement; and correcting the road edge information according to the relative position of the scanning point.
6. The method of claim 5, further comprising: measuring linear acceleration and angular acceleration of the vehicle; when the linear acceleration is larger than a linear acceleration threshold value, calculating the relative displacement according to the linear acceleration; and when the angular acceleration is larger than an angular acceleration threshold value, correcting the relative displacement according to the angular acceleration.
7. The method of claim 1, further comprising:
counting the road height corresponding to the road scanning points, and dividing a plurality of road scanning points into a plurality of intervals;
calculating the average road height corresponding to the road scanning points in the multiple sections;
and correcting the road edge information according to the average road surface height.
8. A road edge detection device, the device comprising:
the acquisition module is used for scanning a target environment for multiple times by utilizing a laser beam to obtain an echo signal reflected within a preset time; the target environment comprises a plurality of scan points; the echo signal carries three-dimensional coordinate information of the scanning point;
the encoding module is used for encoding each scanning point into a sparse tensor according to the three-dimensional coordinate information of the plurality of scanning points;
the sparse tensor voting module is used for carrying out sparse tensor voting on the sparse tensor to obtain a thinned tensor corresponding to each scanning point;
the dense tensor voting module is used for performing dense tensor voting on the scanning points according to the refinement tensor corresponding to each scanning point;
the obtaining module is used for obtaining line characteristics and normal characteristics of corresponding scanning points according to the result of the dense tensor voting;
the data processing module is used for establishing a three-dimensional coordinate system by taking the direction vertical to the road surface and upward as a Z axis; acquiring an included angle between the normal characteristic and a Z axis according to the normal characteristics of the plurality of scanning points; when the included angle is smaller than or equal to the included angle threshold value, recording the corresponding scanning point as a pavement scanning point; acquiring the road heights corresponding to the plurality of road scanning points according to the components of the normal features of the plurality of road scanning points in the Z axis;
the filtering module is used for acquiring a height threshold corresponding to the road surface height and a threshold of line characteristics of the road surface scanning points; and filtering the echo signals corresponding to the road surface scanning points according to the height threshold and the threshold of the line characteristics to obtain road edge information corresponding to the road surface scanning points.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810311877.0A CN108957432B (en) | 2018-04-09 | 2018-04-09 | Road edge detection method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810311877.0A CN108957432B (en) | 2018-04-09 | 2018-04-09 | Road edge detection method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108957432A CN108957432A (en) | 2018-12-07 |
| CN108957432B true CN108957432B (en) | 2021-01-12 |
Family
ID=64498794
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810311877.0A Active CN108957432B (en) | 2018-04-09 | 2018-04-09 | Road edge detection method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108957432B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111562592B (en) * | 2019-01-29 | 2024-04-05 | 北京京东乾石科技有限公司 | Object edge detection method, device and storage medium |
| CN115015964A (en) * | 2022-05-13 | 2022-09-06 | 深圳优地科技有限公司 | Road edge detection method and device, terminal equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766058A (en) * | 2015-03-31 | 2015-07-08 | 百度在线网络技术（北京）有限公司 | Method and device for obtaining lane line |
| CN105184852A (en) * | 2015-08-04 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Laser-point-cloud-based urban road identification method and apparatus |
| JP2017009378A (en) * | 2015-06-19 | 2017-01-12 | 株式会社トヨタマップマスター | Point group data processor, point group data processing method, program, and recording medium |
| CN107169464A (en) * | 2017-05-25 | 2017-09-15 | 中国农业科学院农业资源与农业区划研究所 | A kind of Method for Road Boundary Detection based on laser point cloud |
-
2018
- 2018-04-09 CN CN201810311877.0A patent/CN108957432B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104766058A (en) * | 2015-03-31 | 2015-07-08 | 百度在线网络技术（北京）有限公司 | Method and device for obtaining lane line |
| JP2017009378A (en) * | 2015-06-19 | 2017-01-12 | 株式会社トヨタマップマスター | Point group data processor, point group data processing method, program, and recording medium |
| CN105184852A (en) * | 2015-08-04 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Laser-point-cloud-based urban road identification method and apparatus |
| CN107169464A (en) * | 2017-05-25 | 2017-09-15 | 中国农业科学院农业资源与农业区划研究所 | A kind of Method for Road Boundary Detection based on laser point cloud |
Non-Patent Citations (2)
| Title |
|---|
| 基于激光雷达的路沿检测与跟踪研究;陆恒 等;《计算机光盘软件与应用》;20141215(第23期);18-19 * |
| 高分辨率三维点云目标识别技术研究;汪汉云;《中国博士学位论文全文数据库（电子期刊）》;20170228(第2期);63-69 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108957432A (en) | 2018-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110443786B (en) | Laser radar point cloud filtering method and device, computer equipment and storage medium | |
| CN109975773B (en) | Millimeter wave radar calibration method, device, equipment and storage medium | |
| US9576375B1 (en) | Methods and systems for detecting moving objects in a sequence of image frames produced by sensors with inconsistent gain, offset, and dead pixels | |
| CN111563450B (en) | Data processing method, device, equipment and storage medium | |
| JP6487283B2 (en) | Point cloud data processing device, point cloud data processing method, program, and recording medium | |
| CN111895989A (en) | Robot positioning method and device and electronic equipment | |
| CN109635816B (en) | Lane line generation method, apparatus, device, and storage medium | |
| CN114270410A (en) | Point cloud fusion method and system for moving object and computer storage medium | |
| CN105335955A (en) | Object detection method and object detection apparatus | |
| CN110008941A (en) | Drivable region detection method, device, computer equipment and storage medium | |
| CN111179274B (en) | Map ground segmentation method, device, computer equipment and storage medium | |
| CN114359848A (en) | Target detection method and device, vehicle-end sensing equipment and storage medium | |
| CN111932537A (en) | Object deformation detection method and device, computer equipment and storage medium | |
| CN110782465A (en) | Ground segmentation method and device based on laser radar and storage medium | |
| CN111488812A (en) | Obstacle position recognition method and device, computer equipment and storage medium | |
| JP6552448B2 (en) | Vehicle position detection device, vehicle position detection method, and computer program for vehicle position detection | |
| CN108957432B (en) | Road edge detection method and device, computer equipment and storage medium | |
| EP4213128A1 (en) | Obstacle detection device, obstacle detection system, and obstacle detection method | |
| CN111292275A (en) | Point cloud data filtering method and device based on complex ground and computer equipment | |
| CN114359334A (en) | Target tracking method and device, computer equipment and storage medium | |
| CN114758504A (en) | Online vehicle overspeed early warning method and system based on filtering correction | |
| CN112102375A (en) | Method and device for detecting reliability of point cloud registration and mobile intelligent equipment | |
| CN114663598A (en) | Three-dimensional modeling method, device and storage medium | |
| CN112985772B (en) | Depth map detection apparatus, depth map detection method, electronic device, and computer-readable storage medium | |
| KR102462230B1 (en) | Unmmaned aerial vehicle and operating method thereof |
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 |