CN113791409A - Millimeter wave radar-based static target extraction method and system - Google Patents
Millimeter wave radar-based static target extraction method and system Download PDFInfo
- Publication number
- CN113791409A CN113791409A CN202110893540.7A CN202110893540A CN113791409A CN 113791409 A CN113791409 A CN 113791409A CN 202110893540 A CN202110893540 A CN 202110893540A CN 113791409 A CN113791409 A CN 113791409A
- Authority
- CN
- China
- Prior art keywords
- detection
- points
- radar
- extraction method
- wave radar
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G01S13/66—Radar-tracking systems; Analogous systems
- G01S13/72—Radar-tracking systems; Analogous systems for two-dimensional tracking, e.g. combination of angle and range tracking, track-while-scan radar
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention relates to the technical field of automobile radars, in particular to a static target extraction method and system based on a millimeter wave radar. The method comprises the following steps: collecting multi-frame radar detection data in the advancing process of the automobile, and screening detection points which are static to the ground in the detection data; performing motion compensation on multi-frame historical detection points according to the automobile motion information, and integrating multi-frame radar detection point data; performing density clustering in the detection point data to obtain and extract effective clusters; and fitting and calculating the effective clusters to finish the edge extraction of the static target. The invention realizes the edge extraction of the static target through point cloud collection, motion compensation, clustering and fitting, can extract the road edge or fence with any curvature, can self-check the extraction effectiveness, and adaptively correct the extraction result, thereby being beneficial to improving the accuracy of alarm application and reducing the false alarm rate of the false detection target near the static target through the edge extraction of the static target such as the road edge, the fence and the like.
Description
Technical Field
The invention relates to the technical field of automobile radars, in particular to a static target extraction method and system based on a millimeter wave radar.
Background
With the development of the automobile industry, the application of sensors such as radars and the like on automobiles is concerned by users and manufacturers more and more, at present, vehicle-mounted angle radars are widely applied to target identification, alarming and the like, but are limited by detection precision, the probability of missing detection and false detection exists, in addition, the detection distance of the angle radars is limited, so that the detection rate of the edges of static targets such as road edges, fences and the like is not high at high speed, and the edges with larger curvatures, such as ramp guardrails, flower bed edges and the like, are more difficult to identify.
Disclosure of Invention
The invention provides a static target extraction method and system based on a millimeter wave radar, aiming at solving the technical problem that the detection precision of a road static target is low when a current automobile runs.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a millimeter wave radar-based stationary target extraction method, the method comprising:
collecting multi-frame radar detection data in the advancing process of the automobile, and screening detection points which are static to the ground in the detection data;
performing motion compensation on multi-frame historical detection points according to the automobile motion information, and integrating multi-frame radar detection point data;
performing density clustering in the detection point data to obtain and extract effective clusters;
and fitting and calculating the effective clusters to finish the edge extraction of the static target.
Further, performing density clustering on the detection point data to obtain and extract effective clusters specifically includes:
the position wave gate between two detection points in the detection point data is within a preset value, the two detection points are clustered into one class, and all the detection points are clustered;
and judging the number of the points in the cluster, if the number of the points in the cluster meets the requirement of the number of the points, determining the cluster as an effective cluster, and otherwise, discarding the cluster.
Further, the preset value of the wave gate is 2.5 +/-0.5 m, and the required value of the number of the points in the cluster is 15 +/-5.
Further, the fitting and calculating the effective clusters to complete the edge extraction of the stationary target includes:
fitting points in the effective clusters by adopting polynomial fitting, and extracting an expression of a static target;
and judging the validity of the fitting result according to the contact ratio of the expression obtained by fitting and the actual position of the clustering point, if the error meets a preset range value, extracting the expression as an effective expression of the static target, and otherwise, performing density clustering and fitting again.
Further, the preset range value has an error smaller than 1 +/-0.8 m.
Further, the radar detection data includes position information, speed information, and angle information of the detection point.
Further, the screening of the geostationary detection points in the detection data includes:
screening detection points detected by a radar according to the running speed of the vehicle and the relative speed of the detection points to screen out detection points which are relatively static to the ground; wherein, the detection point of the relative rest to the ground is the detection point of the ground speed less than 1 m/s.
Further, the motion compensation is to perform motion compensation on the historical frame according to the vehicle position of the historical frame and the vehicle position of the current frame.
Further, the integrating the multi-frame radar detection point data includes:
and integrating the plurality of historical frames after motion compensation with the current frame to obtain detection point data with more stable target feature identification.
In addition, the invention also provides a millimeter wave radar-based static target tracking system, which comprises an angle radar and a processor, wherein the processor receives the detection information detected by the angle radar and executes the millimeter wave radar-based static target extraction method.
The invention realizes the edge extraction of the static target through point cloud collection, motion compensation, clustering and fitting, can extract the road edge or fence with any curvature, can self-check the extraction effectiveness, and adaptively correct the extraction result, thereby helping the vehicle to identify scenes through the edge extraction of the static target such as the road edge, the fence and the like, being beneficial to improving the accuracy of alarm application and reducing the false alarm rate of false detection targets near the static target.
Drawings
Fig. 1 is a structural flow chart of a millimeter-wave radar-based stationary target extraction method in the embodiment of the present invention.
Fig. 2 is a flowchart of one example of a millimeter-wave radar-based stationary target extraction method in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The same or similar reference numerals in the drawings of the embodiments of the present application correspond to the same or similar components; in the description of the present application, it is to be understood that the terms "upper", "lower", "left", "right", "top", "bottom", "inner", "outer", and the like, if any, are used in the orientations and positional relationships indicated in the drawings only for the convenience of describing the present application and for simplicity of description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and therefore the terms describing the positional relationships in the drawings are for illustrative purposes only and are not to be construed as limiting the present patent.
Furthermore, if the terms "first," "second," and the like are used for descriptive purposes only, they are used for mainly distinguishing different devices, elements or components (the specific types and configurations may be the same or different), and they are not used for indicating or implying relative importance or quantity among the devices, elements or components, but are not to be construed as indicating or implying relative importance.
Fig. 1 shows a structural flow chart of a millimeter wave radar-based stationary target extraction method.
As shown in fig. 1, the present embodiment provides a method for extracting a stationary target based on a millimeter wave radar, which is mainly used for identifying, detecting and extracting geostationary and continuous road buildings, such as railings, fences, or curbs, on a driving road of an automobile by using an angular radar of the automobile, so as to help the automobile to identify a scene and improve the accuracy of automobile alarm. Specifically, the method comprises the following steps:
101. collecting multi-frame radar detection data in the advancing process of the automobile, and screening detection points which are static to the ground in the detection data; the radar detection data, which is an information sum of detection points detected by the radar, specifically includes position information, speed information, and angle information of the detection points.
Preferably, in the above step, the screening of the geostationary detection points in the detection data specifically includes: screening detection points detected by a radar according to the running speed of the vehicle and the relative speed of the detection points to screen out detection points which are relatively static to the ground; wherein, the detection point of the relative rest to the ground is the detection point of the ground speed less than 1 m/s.
In the screening process, the screening of the detection points at ground standstill needs to be corrected by the vehicle speed when the vehicle runs, and in the screening, the detection points at ground speed less than 1m/s are screened according to the vehicle speed when the vehicle runs and the relative speed of the detection points, and are identified as the stationary targets. For example, when the angular radar detects a detection point within 50 ± 10m, and compensates the velocity value of the detection point, and compensates the vehicle velocity, and then compensates the detection point less than 1 ± 0.2m/s, the detection point is considered as a stationary detection point with respect to the ground. Specifically, the ground speed of the detection point is the detection point relative speed + the vehicle speed cos (detection point angle).
102. Performing motion compensation on multi-frame historical detection points according to the automobile motion information, and integrating multi-frame radar detection point data; since the position information of the detection point in the historical frame is the coordinates of the position of the vehicle relative to the historical frame, in the current frame, the historical frame needs to be subjected to trace compensation, and the historical distance of the vehicle between the current frame and the historical frame during the motion of the vehicle is compensated, so as to restore the relative positions of the stationary target at different times. Because the position of the static target is unchanged, the problems of few and uneven detection points obtained in one frame of the radar can be solved after the motion position compensation of multiple frame points, and the target feature identification is more stable.
For better understanding, a concrete compensation example of the embodiment is provided, for example, the automobile angle radar acquires the position coordinate target (x) of the stationary target a relative to the host vehicle at time tt,yt) And the vehicle is moving to the groundThe driving track is (Sx)t,Syt) The running angle is S thetatAt this time, the coordinates of the stationary object A to the ground are target (abs _ x)t,abs_yt)；
Wherein:
abs_xt＝Sxt+yt*sin(Sθt)+xt*cos(Sθt)；
abs_yt＝Syt+yt*cos(Sθt)-xt*sin(Sθt)；
the straight-going running track of the vehicle at the time t +1 is updated to (Sx)t+1,Syt+1) The running angle is S thetat+1Then, the current time of the stationary object A is updated to the target (x) with respect to the position coordinates of the vehiclet+1,yt+1)；
Wherein:
xt+1＝(abs_xt-Sxt+1)*cos(Sθt)-(abs_yt-Syt+1)*sin(Sθt)；
yt+1＝(abs_xt-Sxt+1)*sin(Sθt)+(abs_yt-Syt+1)*cos(Sθt)。
after the motion position compensation, the processor integrates the plurality of history frames subjected to the motion compensation with the current frame, so that detection point data with more stable target feature identification are obtained and integrated into radar detection data, and subsequent clustering and fitting are based on the radar detection data.
Certainly, in the subsequent operation process, in order to better detect stationary targets such as fences and road edges, the radar detection data can be further screened, the screening is mainly used for screening points in an effective range, and the points within the distance of 100 +/-50 m are extracted from the radar detection data.
103. And performing density clustering in the detection point data to obtain and extract effective clusters.
Preferably, the steps specifically include:
the position wave gate between two detection points in the detection point data is within a preset value, the two detection points are clustered into one class, and all the detection points are clustered;
and judging the number of the points in the cluster, if the number of the points in the cluster meets the requirement of the number of the points, determining the cluster as an effective cluster, and otherwise, discarding the cluster.
The density clustering can set a threshold value, if the distance between two detection points is within the threshold value, the two points can be determined to belong to the same cluster, and the detection points in the detection point data are clustered according to the judging mode to obtain a plurality of clusters. It is noted that if points a and B are of the same type and points B and C are of the same type, then points A, B, C are of the same type.
Specifically, in terms of specific clustering numerical values, the detection point position wave gate epsilon 1 of the clustering is set to be 2.5 +/-0.5 m, the detection point speed wave gate epsilon 2 is set to be 1 +/-0.2 m/s, namely, by calculating the linear distance delta d and the speed difference delta v between every two points, if the delta d is less than epsilon 1 and the delta v is less than epsilon 2, the two points are clustered into one class.
After the clustering is finished, effective clustering is carried out for extraction, and static targets such as railings and road edges are targets with long continuous lengths, so that if the number of points of the clustering is less than a certain numerical value, the static targets such as the railings and the road edges can be determined. Specifically, if the number of points in the cluster satisfies 15 ± 5 or more, the following fitting requirements are satisfied; otherwise, the operation is abandoned. The specific number of the clustering points can be reasonably adjusted according to the detection precision requirement.
104. And fitting and calculating the effective clusters to finish the edge extraction of the static target. And finally determining the effective expression of the edge extraction to finish the edge extraction of the static target.
In some embodiments, the step 104 of performing fitting processing and calculation on the valid clusters to complete the edge extraction of the stationary target specifically includes:
fitting points in the effective clusters by adopting polynomial fitting, and extracting an expression of a static target;
and judging the validity of the fitting result according to the contact ratio of the expression obtained by fitting and the actual position of the clustering point, if the error meets a preset range value, extracting the expression as an effective expression of the static target, and otherwise, performing density clustering and fitting again.
And the processor judges the validity of the fitting result according to the contact degree of the expression obtained by fitting and the actual position of the clustering point. In the judgment process, if the error meets the preset range value, the edges of the road edge, the fence and the like are successfully extracted, otherwise, the step 103 is re-executed for clustering fitting operation until all the points are processed. Specifically, the error of the preset range value is less than 1 ± 0.8 m.
For better operation experience, a specific extraction step is provided in this embodiment, please refer to fig. 2, where the processor performs point cloud extraction through the angle radar, that is, extracts detection point information detected by the angle radar, and performs motion compensation for screening. And in the screening process, detecting the number of the detection points, if the number of the detection points is too small, giving up edge extraction, if the number of the detection points meets the requirement, performing density clustering and fitting, finally judging the fitting effect, obtaining a final effective expression, and finishing the extraction of the static target.
The method has the advantages that the method can extract the edge of the static target through point cloud collection, motion compensation, clustering and fitting, can extract the road edge or the fence with any curvature, can self-check the extraction effectiveness, and can adaptively correct the extraction result, so that the method helps a vehicle to identify scenes through the edge extraction of the static target such as the road edge or the fence, is beneficial to improving the accuracy of alarm application, and reduces the false alarm rate of the false detection target near the static target.
In addition, the present embodiment also provides a millimeter wave radar-based stationary target extraction system, which includes an angle radar and a processor, where the processor receives detection information detected by the angle radar and executes the relevant steps of the above millimeter wave radar-based stationary target extraction method, so as to realize the extraction of stationary target edges such as road edges and fences.
It should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.
Claims (10)
1. A static target extraction method based on millimeter wave radar is characterized by comprising the following steps:
collecting multi-frame radar detection data in the advancing process of the automobile, and screening detection points which are static to the ground in the detection data;
performing motion compensation on multi-frame historical detection points according to the automobile motion information, and integrating multi-frame radar detection point data;
performing density clustering in the detection point data to obtain and extract effective clusters;
and fitting and calculating the effective clusters to finish the edge extraction of the static target.
2. The millimeter wave radar-based stationary target extraction method according to claim 1, wherein the performing density clustering on the detection point data to obtain and extract valid clusters specifically comprises:
the position wave gate between two detection points in the detection point data is within a preset value, the two detection points are clustered into one class, and all the detection points are clustered;
and judging the number of the points in the cluster, if the number of the points in the cluster meets the requirement of the number of the points, determining the cluster as an effective cluster, and otherwise, discarding the cluster.
3. The millimeter wave radar-based stationary target extraction method according to claim 2, wherein the preset value of the gate is 2.5 ± 0.5m, and the required value of the number of points in the cluster is 15 ± 5.
4. The millimeter wave radar-based stationary target extraction method according to claim 1, wherein the fitting and calculating the effective clusters to complete the edge extraction of the stationary target comprises:
fitting points in the effective clusters by adopting polynomial fitting, and extracting an expression of a static target;
and judging the validity of the fitting result according to the contact ratio of the expression obtained by fitting and the actual position of the clustering point, if the error meets a preset range value, extracting the expression as an effective expression of the static target, and otherwise, performing density clustering and fitting again.
5. The millimeter wave radar-based stationary target extraction method according to claim 4, wherein the preset range value has an error of less than 1 ± 0.8 m.
6. The millimeter wave radar-based stationary target extraction method according to claim 1, wherein the radar detection data includes position information, speed information, angle information of a detection point.
7. The millimeter wave radar-based stationary target extraction method according to claim 6, wherein the screening of the detection points at which the ground is stationary in the detection data includes:
screening detection points detected by a radar according to the running speed of the vehicle and the relative speed of the detection points to screen out detection points which are relatively static to the ground; wherein, the detection point of the relative rest to the ground is the detection point of the ground speed less than 1 m/s.
8. The millimeter wave radar-based stationary target extraction method according to claim 1, wherein the motion compensation is to perform motion compensation on the history frame according to the vehicle position of the history frame and the vehicle position of the current frame.
9. The millimeter wave radar-based stationary target extraction method according to claim 1, wherein the integrating multi-frame radar detection point data comprises:
and integrating the plurality of historical frames after motion compensation with the current frame to obtain detection point data with more stable target feature identification.
10. A millimeter-wave radar-based stationary target tracking system comprising an angle radar and a processor, the processor receiving detection information detected by the angle radar and executing the millimeter-wave radar-based stationary target extraction method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110893540.7A CN113791409A (en) | 2021-08-04 | 2021-08-04 | Millimeter wave radar-based static target extraction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110893540.7A CN113791409A (en) | 2021-08-04 | 2021-08-04 | Millimeter wave radar-based static target extraction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113791409A true CN113791409A (en) | 2021-12-14 |
Family
ID=78877138
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110893540.7A Pending CN113791409A (en) | 2021-08-04 | 2021-08-04 | Millimeter wave radar-based static target extraction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113791409A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114529858A (en) * | 2022-04-21 | 2022-05-24 | 浙江大华技术股份有限公司 | Vehicle state recognition method, electronic device, and computer-readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110807412A (en) * | 2019-10-30 | 2020-02-18 | 驭势科技（北京）有限公司 | Vehicle laser positioning method, vehicle-mounted equipment and storage medium |
| CN111289980A (en) * | 2020-03-06 | 2020-06-16 | 成都纳雷科技有限公司 | Roadside stationary object detection method and system based on vehicle-mounted millimeter wave radar |
| CN111316128A (en) * | 2018-12-28 | 2020-06-19 | 深圳市大疆创新科技有限公司 | Continuous obstacle detection method, device, system and storage medium |
| CN111699406A (en) * | 2019-03-29 | 2020-09-22 | 深圳市大疆创新科技有限公司 | Tracking detection method of millimeter wave radar, millimeter wave radar and vehicle |
-
2021
- 2021-08-04 CN CN202110893540.7A patent/CN113791409A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111316128A (en) * | 2018-12-28 | 2020-06-19 | 深圳市大疆创新科技有限公司 | Continuous obstacle detection method, device, system and storage medium |
| CN111699406A (en) * | 2019-03-29 | 2020-09-22 | 深圳市大疆创新科技有限公司 | Tracking detection method of millimeter wave radar, millimeter wave radar and vehicle |
| CN110807412A (en) * | 2019-10-30 | 2020-02-18 | 驭势科技（北京）有限公司 | Vehicle laser positioning method, vehicle-mounted equipment and storage medium |
| CN111289980A (en) * | 2020-03-06 | 2020-06-16 | 成都纳雷科技有限公司 | Roadside stationary object detection method and system based on vehicle-mounted millimeter wave radar |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114529858A (en) * | 2022-04-21 | 2022-05-24 | 浙江大华技术股份有限公司 | Vehicle state recognition method, electronic device, and computer-readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109870680B (en) | Target classification method and device | |
| US9292750B2 (en) | Method and apparatus for detecting traffic monitoring video | |
| US6670912B2 (en) | Method for detecting stationary object located above road | |
| US8994823B2 (en) | Object detection apparatus and storage medium storing object detection program | |
| WO2016129403A1 (en) | Object detection device | |
| EP2189944A1 (en) | Image processing device and image processing method | |
| US20110063157A1 (en) | Method for operating a radar system in the event of possible concealment of the target object and radar system for performing the method | |
| JP2003502745A (en) | A method for detecting objects near road vehicles up to a considerable distance | |
| US10481254B2 (en) | Radar device and method for determining targets to be followed | |
| CN112764034B (en) | Method for identifying obstacle at low speed by vehicle-mounted BSD millimeter wave radar | |
| CN112597839B (en) | Road boundary detection method based on vehicle-mounted millimeter wave radar | |
| CN108344997B (en) | Road guardrail rapid detection method based on trace point characteristics | |
| CN112731296A (en) | Automobile millimeter wave radar point trace condensation method and system | |
| CN113791409A (en) | Millimeter wave radar-based static target extraction method and system | |
| CN113740874A (en) | Road edge detection method, electronic equipment and vehicle | |
| CN116089903A (en) | Multi-sensor target tracking method and device | |
| CN113962301B (en) | Multi-source input signal fused pavement quality detection method and system | |
| CN114662600A (en) | Lane line detection method and device and storage medium | |
| CN113496199B (en) | Histogram-based L-shaped detection of target objects | |
| CN112949489A (en) | Road boundary identification method and device, electronic equipment and storage medium | |
| Lee et al. | A cumulative distribution function of edge direction for road-lane detection | |
| JP2001330665A (en) | On-vehicle object detector using radar and image processing | |
| CN115151836A (en) | Method for detecting a moving object in the surroundings of a vehicle and motor vehicle | |
| CN115656962A (en) | Method for identifying height-limited object based on millimeter wave radar | |
| CN110736981B (en) | Radar-based tracking of slowly moving objects |
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 |