CN116299319B - Synchronous scanning and point cloud data processing method of multiple laser radars and radar system - Google Patents
Synchronous scanning and point cloud data processing method of multiple laser radars and radar system Download PDFInfo
- Publication number
- CN116299319B CN116299319B CN202310601244.4A CN202310601244A CN116299319B CN 116299319 B CN116299319 B CN 116299319B CN 202310601244 A CN202310601244 A CN 202310601244A CN 116299319 B CN116299319 B CN 116299319B
- Authority
- CN
- China
- Prior art keywords
- radar
- point
- point cloud
- cloud data
- radars
- 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
- 230000001360 synchronised effect Effects 0.000 title claims abstract description 15
- 238000003672 processing method Methods 0.000 title claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims abstract description 73
- 238000001914 filtration Methods 0.000 claims abstract description 47
- 230000009466 transformation Effects 0.000 claims abstract description 39
- 238000005259 measurement Methods 0.000 claims abstract description 37
- 230000004927 fusion Effects 0.000 claims abstract description 15
- 238000000034 method Methods 0.000 claims description 33
- 238000012937 correction Methods 0.000 claims description 25
- 238000012545 processing Methods 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 12
- 238000013519 translation Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000003287 optical effect Effects 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000004132 cross linking Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/4802—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The application relates to a synchronous scanning and point cloud data processing method of a multi-laser radar and a radar system; determining space fixed positions and a basic coordinate system of a plurality of cooperative laser radars, and determining a coordinate transformation matrix between each radar coordinate and the basic coordinate based on the mutual position relation between the origin of the basic coordinate system and the radars; correcting each radar coordinate transformation matrix by measuring a standard measured object arranged at a space fixed position; after synchronizing the laser radar clocks, data acquisition is carried out; coordinate transformation is carried out on the collected measurement data according to the corrected coordinate transformation matrix, and a time stamp is given to the measurement data to form multi-radar point cloud data; and performing mutual interference filtering on the multi-radar point cloud data, comparing and checking point by point, and finishing data splicing and fusion after the integral checking of the point cloud data. The application eliminates the interference among internal components, expands the measurement range and measurement precision of a single radar, and improves the overall measurement capability.
Description
Technical Field
The application relates to the technical field of laser radars, in particular to a synchronous scanning and point cloud data processing method of a multi-laser radar and a radar system.
Background
The lidar is a radar system that detects a characteristic quantity such as a position, a speed, etc. of a target by emitting a laser beam. The laser radar can obtain relevant information of the target, such as parameters of the distance, the azimuth, the height, the speed, the gesture, the even the shape and the like of the target by transmitting a detection signal (laser beam) to the target, then comparing the received signal (target echo) reflected from the target with the transmission signal, and properly processing the signal, thereby realizing detection, tracking and identification of the target.
In practical application, the more information is collected, the more real is space restoration, and the more accurate is the pose information, the peripheral contour information, the travel path information and the like obtained by the method. The method is limited by a light detection mechanism of the laser radars, and a single laser radar has certain limitations in the aspects of view field range, detection distance, scanning resolution, scanning precision, reliability and the like, and in order to obtain more complete, more accurate and reliable measurement information, a plurality of laser radars are usually required to be matched for use so as to make up for the defect of single radar measurement. Multiple radars work simultaneously and bring new problems: (1) the measurement distortion caused by mutual interference among radars with the same optical wave band is particularly serious when a plurality of radars are placed on the same horizontal plane; (2) the multi-radar is placed in a scattered mode, coordinate references of output point cloud data are not uniform, and calculation pressure of data splicing and fusion operation of a follow-up central control system is high; (3) the problem of data matching obstacle caused by the asynchronous system time of a plurality of radars is particularly remarkable when aiming at moving a measured target, and the problems are particularly represented by point cloud image staggering, target contour blurring and the like of the measured target.
For radar mutual interference, the common processing method is staggered-layer arrangement or phase control, and the staggered-layer arrangement only solves the problem of radar correlation interference, and the interference problem caused by overlapping of receiving and transmitting view fields at specific moments among different radars still exists. The phase control is mainly used for uniformly regulating and controlling the scanning phases of different radars to realize the locking of scanning phase differences, but the phase monitoring and feedback control locking phase units are required to be added, so that the complexity of cross-linking control of a hardware system of the radar and a multi-radar system is increased.
Disclosure of Invention
In view of the above analysis, the present application aims to disclose a method for synchronous scanning and point cloud data processing of a multi-laser radar and a radar system; the method solves the problems of synchronous scanning and point cloud data fusion of the multi-laser radar.
The application discloses a synchronous scanning and point cloud data processing method of a multi-laser radar, which comprises the following steps:
step S1, determining space fixed positions and a basic coordinate system of a plurality of cooperative laser radars, and determining a coordinate transformation matrix between each radar coordinate and the basic coordinate based on the mutual position relation between the origin of the basic coordinate system and the radars;
s2, correcting each radar coordinate transformation matrix by measuring a standard measured object arranged at a space fixed position;
step S3, after synchronizing the laser radar clocks, data acquisition is carried out; performing coordinate transformation on the acquired measurement data according to the corrected coordinate transformation matrix and giving a time stamp to form multi-radar point cloud data;
and S4, performing mutual interference filtering on the multi-radar point cloud data, comparing the point by point to perform verification, and completing data splicing and fusion after the integral verification of the point cloud data.
Further, in step S2, correction of the transformation matrix is performed by using the ICP algorithm; the method for correcting the transformation matrix by adopting the ICP algorithm comprises the following steps:
step S201, setting a standard object to be measured at a fixed position in a measurement space, enabling all the laser radars to work one by one, ensuring that only one radar works at the same time, and obtaining point cloud data of all the laser radars;
step S202, extracting key point data of point cloud data of each laser radar to obtain a key point data set;
step S203, taking one laser radar as a reference radar A, and the rest laser radars as calibrated radars B; according to the key point data of the radar point clouds of the reference radar A and each calibrated radar B, an ICP algorithm is adopted to obtain a correction matrix, and the correction matrix is overlapped on the coordinate transformation matrix stored by each calibrated radar B, so that the correction of the radar coordinate transformation matrix of the calibrated radar B is realized.
Further, in step S203, the radar coordinate transformation matrix correction process of the calibrated radar B with respect to the reference radar a includes:
1) The reference radar A and the calibrated radar B respectively scan a fixed standard measured object, and respectively take one frame of point cloud data in the measurement result to extract key point data so as to obtain key point data sets a and B;
wherein a= (a) 1 ,a 2 ,…,a i ,…,a n ）；b=（b 1 ,b 2 ,…,b i ,…,b n ）；
2) For the key point data sets a and b, a corresponding relation between the key point data sets a and b is found by adopting a characteristic matching mode, and a correction matrix comprising a rotation matrix R and a translation matrix t is solved by minimizing a matching error; the solved rotation matrix R and translation matrix t are correction matrices of the calibrated radar B relative to the reference radar A;
wherein the matrix is rotatedThe method comprises the steps of carrying out a first treatment on the surface of the Translation matrix->；
U, V SVD decomposition is performed on the matched features to obtain a left singular vector matrix and a right singular vector matrix;centroid of the data of the key point data set a; />Centroid of the b data of the keypoint dataset;
3) And (3) superposing a correction matrix on the coordinate transformation matrix of the calibrated radar B to finish the correction of the calibrated radar B transformation matrix.
Further, in step S4, based on the interference cause, a corresponding point cloud data filtering mode is adopted to perform mutual interference filtering;
in the mutual interference filtering, fixed-point mutual interference and non-fixed-point mutual interference filtering are carried out according to the difference of the scanning phase difference alpha between the two radars.
Further, when the scanning phase difference alpha=0° or 180 degrees is smaller than alpha < 360 degrees of the reference radar A and the calibrated radar B, judging that fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the reference radar A generates optical interference to the calibrated radar B as fixed point mutual interference at only one moment in the time range of realizing one period 360 scanning;
when the scanning phase difference of the reference radar A and the calibrated radar B is 0 degrees less than alpha less than 180 degrees, judging that non-fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the two radar emergent rays intersect at an intersection point P, and the position of the intersection point P in space continuously changes along with the time passing of the scanning process to form a mutual interference curve; the interference caused by points on the mutual interference curve is non-fixed point mutual interference.
Further, for the fixed point mutual interference, filtering is completed by screening and filtering the data of radar measurement points at specific scanning angle moments and adjacent set angle scanning areas which cause single point mutual interference.
Further, when non-fixed point mutual interference filtering is performed, the method comprises the following steps:
1) Establishing a coordinate system for calculating the mutual interference positions of the two radars, and determining the coordinate of an intersection point P where the emergent laser rays of the two radars intersect; when the measured object is coincident with the intersection point P, the step 2) is entered, otherwise, whether the measured object is coincident with the intersection point P is repeatedly judged in the radar measurement process;
2) Judging whether the scanning phase difference alpha approaches to a fixed value in a set time interval in the radar scanning process, if yes, entering the step 3), otherwise, entering the step 4);
3) According to the triangle cosine theorem, calculating an interference curve where the intersection point P is located, and screening and filtering out adjacent points which are on the interference curve and are out of line and accord with the condition of the adjacent threshold value as interference points by setting the threshold value of the adjacent points;
4) Performing point-by-point position calculation on the point cloud data of the reference radars A and B received at the same time to obtain an intersection point P coordinate value of intersection of the two radar emergent laser rays; calculating the distance L from the intersection point P to the origin of a calculated coordinate system according to the coordinate value of the intersection point P;
5) The calculated distance L is substituted for the measured distance value.
Further, in step S4, the point cloud data point-by-point comparison check includes a check of a fixed measured object and a check of an unfixed measured object;
fixing the checking of the measured target, adopting two to three times of radar ranging errors as the reference of mutual checking, and filtering or reducing weight for the point cloud data exceeding the reference;
checking a non-fixed measured target, setting a threshold value by adopting an abnormal quantity, and comparing and processing the same measured position point cloud data among the measured frame data at the same moment of different radars; and filtering or reducing weight processing is carried out on the point cloud data exceeding the distortion amount set threshold value.
Further, in step S4, in the overall comparison and verification of the point cloud data, the key points of the radar point cloud frame data at the same time are extracted, whether the key points of the point cloud frame data are coincident or not is judged, if yes, the radars are judged to work normally, if not, the laser radars with unmatched key points are filtered or weight-reduced to all the point cloud data.
The application also discloses a radar system, which comprises a plurality of laser radars and a central control system, wherein the central control system is connected with the plurality of laser radars one by one to form a star-shaped connection network, and the laser radars are not directly connected with each other physically;
after the spatial position fixing of the plurality of laser radars and the physical connection of the central control system and the plurality of laser radars are finished;
the central control system and each laser radar realize synchronous scanning and point cloud data processing of the multi-laser radar through communication and cooperative work.
The application can realize the following beneficial effects:
1. according to the method, synchronous scanning of the multi-laser radar is realized through position calibration and clock synchronization, filtering, nuclear calibration and data fusion are carried out on point cloud data of the multi-laser radar, interference among internal components is eliminated, the measurement range and measurement precision of a single radar are expanded, and the overall measurement capability is improved;
2. the application preposes multi-radar point cloud splicing fusion, achieves coarse calibration based on position relation and fine calibration by measuring the fixed targets one by the radar, and completes pre-fusion of radar data. Links adopted by fusion after downsampling, key point extraction, point cloud registration, image splicing and the like are omitted, radar data can be directly integrated and output after simple filtering, the system real-time performance is good, and the requirement on the data processing capability of a central control system is low;
3. the application has the advantages that the occupation ratio of the point cloud data of the problems generated by mutual interference in all the point cloud data commonly measured by multiple radars is extremely low, and the integral measurement effect is not influenced by filtering;
4. according to the scheme, under the condition that phase locking is not performed, a corresponding point cloud data filtering mode is adopted based on an interference reason, so that the problem that the profile information of a measured object is omitted due to a conventional mode is avoided;
5. for a fixed measured target, the point cloud data measured by different radars at the same time and the same position can ensure that the corresponding distance difference value is smaller than a threshold value after mutual interference filtering, and for the point cloud data measured by different radars at the same position and the different times, the theoretical distance difference value is smaller than a fixed value, so that the data mutual checking is performed on the basis of filtering, noise can be further filtered, and the accuracy of the point cloud data is improved;
6. the overall comparison of the radar point cloud data can be used to identify the operational failure of a single radar.
Drawings
The drawings are only for purposes of illustrating particular embodiments and are not to be construed as limiting the application, like reference numerals being used to designate like parts throughout the drawings;
FIG. 1 is a flow chart of a method for processing point cloud data in synchronization scanning of multiple lidars according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a coordinate system for calculating the mutual interference position of two radars in an embodiment of the present application;
fig. 3 is a schematic diagram of the connection of the radar system according to an embodiment of the present application.
Detailed Description
Preferred embodiments of the present application are described in detail below with reference to the attached drawing figures, which form a part of the present application and are used in conjunction with embodiments of the present application to illustrate the principles of the present application.
An embodiment of the application discloses a method for synchronously scanning and processing point cloud data of a multi-laser radar, which is shown in fig. 1 and comprises the following steps:
step S1, determining space fixed positions and a basic coordinate system of a plurality of cooperative laser radars, and determining a coordinate transformation matrix between each radar coordinate and the basic coordinate based on the mutual position relation between the origin of the basic coordinate system and the radars; writing the coordinate transformation matrix into each corresponding laser radar one by one, so that the space coordinates of each radar are unified;
s2, correcting each radar coordinate transformation matrix by measuring a standard measured object arranged at a space fixed position;
step S3, after synchronizing the laser radar clocks, data acquisition is carried out; performing coordinate transformation on the acquired measurement data according to the corrected coordinate transformation matrix and giving a time stamp to form multi-radar point cloud data;
and S4, performing mutual interference filtering on the multi-radar point cloud data, comparing the point by point to perform verification, and completing data splicing and fusion after the integral verification of the point cloud data.
Specifically, in step S1, the base coordinate system is selected to have flexibility, and may be coincident with the original coordinate system of a single radar in the radars, or may be independent of each radar, depending on practical application requirements.
In step S1, the coordinate transformation matrix corresponding to the radar includes a rotation matrix and a translation matrix between the radar coordinates and the basic coordinates. After the position of the radar is fixed and a basic coordinate system is selected, based on the self structure of the radar and the radar installation azimuth drawing, each parameter in the rotation matrix and the translation matrix can be obtained through simple calculation.
In step S2, correction of the transformation matrix is performed by adopting an ICP algorithm;
specifically, the method for correcting the transformation matrix by adopting the ICP algorithm comprises the following steps:
step S201, setting a standard object to be measured at a fixed position in a measurement space, enabling all the laser radars to work one by one, ensuring that only one radar works at the same time, and obtaining point cloud data of all the laser radars;
in the measuring space, each radar can irradiate a standard measured object without shielding.
Step S202, extracting key point data of point cloud data of each laser radar to obtain a key point data set;
optionally, one kind of key point data can be selected from multiple kinds of key point data such as ISS, harris3D, NARF, sift3D, uniform sampling or voxel sampling, and the like, and the selection of each radar on the kind of key point should be consistent.
Step S203, taking one laser radar as a reference radar A, and the rest laser radars as calibrated radars B; according to the key point data of the radar point clouds of the reference radar A and each calibrated radar B, an ICP algorithm is adopted to obtain a correction matrix, and the correction matrix is overlapped on the coordinate transformation matrix stored by each calibrated radar B, so that the correction of the radar coordinate transformation matrix of the calibrated radar B is realized.
Specifically, in step S203, the process of correcting the radar coordinate transformation matrix of the calibrated radar B with respect to the reference radar a includes:
1) The reference radar A and the calibrated radar B respectively scan a fixed standard measured object, and respectively take one frame of point cloud data in the measurement result to extract key point data so as to obtain key point data sets a and B;
wherein a= (a) 1 ,a 2 ,…,a i ,…,a n ）；b=（b 1 ,b 2 ,…,b i ,…,b n ）。
2) For the key point data sets a and b, the corresponding relation of the key point data sets a and b is found out by adopting a characteristic matching mode, and the matching error is causedMinimum, solving a correction matrix comprising a rotation matrix R and a translation matrix t;
the solved rotation matrix R and translation matrix t are correction matrices of the calibrated radar B relative to the reference radar A.
The characteristic matching comprises the following specific steps:
(1) Calculating the mass center of the two frames of point cloud data:
；
；
(2) Establishing coordinate positions for key points in two frames of point clouds by taking the mass center as an origin:
；
；
(3) Calculating the characteristic w and performing SVD (singular value decomposition) on the characteristic w to obtain a left singular vector matrixAnd right singular vector matrix->；
；
(4) Obtaining a rotation matrix R and a translation matrix t according to SVD decomposition results;
wherein, the liquid crystal display device comprises a liquid crystal display device,；/>。
3) And (3) superposing a correction matrix on the coordinate transformation matrix calibrated by the calibrated radar B to finish the correction of the calibrated radar B transformation matrix.
Specifically, in step S3, in clock synchronization of each lidar, one device connected to all the radars may be used as a master node for clock providing, and all the radars may be used as slave nodes. Configuring a time server in a master node for providing a time synchronization signal; the same time synchronization protocol is configured on all radars, such as IEEE 1588 Precision Time Protocol (PTP).
The master node transmits a time synchronization signal to all the slave nodes in a periodical broadcasting mode, and the slave nodes receive and update own clocks. When the radar performs package transmission of measurement data, a time stamp based on a synchronized clock is added in a data packet.
In step S4 of the present embodiment, under the condition that no phase locking is performed, the corresponding point cloud data filtering mode is adopted to perform the mutual interference filtering based on the interference reason.
Specifically, in the mutual interference filtering, fixed-point mutual interference and non-fixed-point mutual interference filtering are performed according to the difference of the scanning phase difference alpha between two radars;
when the scanning phase difference alpha=0 DEG or 180 DEG is less than or equal to alpha < 360 DEG, judging that fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the reference radar a generates optical interference to the calibrated radar B as fixed point mutual interference at only one moment in the time range in which the scanning of one period 360 is achieved. The moment of generating the optical interference corresponds to a specific scanning angle alpha+beta; where β is the angle of placement of the calibrated radar B within the scan field of view of the reference radar a.
When the fixed point is mutually disturbed and filtered, the reference radar A emits laser to irradiate the optical window of the calibrated radar B, and the reference radar A emits laser light waves to enter the calibrated radar B without barriers because the working wave bands of the two radars are the same, so that the calibrated radar B receives incident light of the reference radar A and interferes the current ranging echo of the calibrated radar B on the premise that two conditions are simultaneously met. The two preconditions are that the laser travelling path of the reference radar A is not blocked, the laser travelling path can finally irradiate the surface of the detector of the calibrated radar B in a reflection, refraction, scattering and other modes, and the moment when the detector of the calibrated radar B receives the incident light wave of the reference radar A is in the receiving detection time window of the calibrated radar B.
For the fixed point mutual interference, the filtering is completed by screening and filtering the data of radar measurement points at specific scanning angle moments causing single point mutual interference and the adjacent set angle scanning area, wherein the adjacent set angle scanning area is a small angle scanning area set according to actual needs.
When the scanning phase difference of the reference radar A and the calibrated radar B is 0 degrees less than alpha less than 180 degrees, judging that non-fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the two radar emergent rays intersect at an intersection point P, and the position of the intersection point P in space continuously changes along with the time passing of the scanning process to form a mutual interference curve; the interference caused by points on the mutual interference curve is non-fixed point mutual interference.
When non-fixed point mutual interference filtering is carried out, the method comprises the following steps:
1) Establishing a coordinate system for calculating the mutual interference positions of the two radars, and determining the coordinate of an intersection point P where the emergent laser rays of the two radars intersect; when the measured object is coincident with the intersection point P, the step 2) is entered, otherwise, whether the measured object is coincident with the intersection point P is repeatedly judged in the radar measurement process;
specifically, as shown in fig. 2, a rectangular coordinate system is constructed by taking the center point of the A, B two-radar connection line as a calculation coordinate origin and taking the A, B two-radar connection line as an x-axis (assuming that the calculation coordinate system is consistent with the basic coordinate system in step S1), and the scanning planes of the two-radar are coincident with the coordinate system plane. The reference radar A is located at (-x) position in the coordinate system 0 0), the position of the calibrated radar B in the coordinate system is (x) 0 ,0). At a certain moment when A, B two radars are simultaneously operated, the B radar scans the phase value theta 2 Greater than A radar scan phase value θ 1 The phase difference value of the two is alpha.
The two radar emergent laser rays intersect at an intersection point P, so that a triangle delta APB is constructed. If the measured object coincides with the intersection point P in position at the moment, based on the scattering characteristic of the surface of the measured object, the A radar emergent laser can be received by the B radar after being scattered by the intersection point P, and the B radar emergent laser can be received by the A radar after being scattered by the intersection point P. Whereby A, B the measurements of the two radars at this location interfere with each other.
For the same radar, the calculated value of the intersection point P under the condition that no interference occurs is smaller than the distance difference value between the measured value of the intersection point P under the condition that the intersection point P is interfered and the adjacent point cloud data, and the correlation degree is higher.
And for different radars, when no mutual interference exists, the corresponding distance values of the point cloud data measured by the radars at the same position at the same moment are equal or close.
And judging whether the corresponding target point is interfered according to whether the distance value of the adjacent measuring point given by the same radar is continuous or not and whether the distance value given by different radars at the same moment is the same or not.
2) Judging whether the scanning phase difference alpha approaches to a fixed value in a set time interval in the radar scanning process, if yes, entering the step 3), otherwise, entering the step 4);
the set time interval is set according to the actual application scene, for example, 1-2 s. The approach to the fixed value in the embodiment means that the scanning phase difference alpha is constant or slowly changed, and the change range within 1-2S is not more than the set change range value, wherein the change range value can be specifically set according to engineering requirements;
3) According to the theorem of triangular cosineCalculating an interference curve where the intersection point P is located, and screening and filtering out the interference curve and the adjacent points outside the line meeting the condition of the adjacent threshold value as interference points by setting the adjacent point threshold value;
4) Performing point-by-point position calculation on the point cloud data of the reference radars A and B received at the same time to obtain an intersection point P coordinate value of intersection of the two radar emergent laser rays; calculating the distance L from the intersection point P to the origin of a calculated coordinate system according to the coordinate value of the intersection point P;
specifically, a straight line tg (θ 1 )=y/(x+x 0 ) And a straight line tg (θ 2 )=y/(x-x 0 ) And performing intersection calculation to calculate coordinate values of the intersection point P. Wherein θ 1 For the inclination angle, theta, of the straight line formed by the reference radar A and the target point in the calculation coordinate system 2 The inclination angle in the coordinate system is calculated for the straight line formed by the calibrated radar B and the target point.
From the calculated intersection point P coordinates (x, y) by the formulaThe distance L of the intersection point P from the origin of coordinates is calculated.
5) The calculated distance L is quickly replaced by the measured distance value.
In step S4, the point cloud data point-by-point comparison check includes a check of a fixed measured object and a check of an unfixed measured object;
wherein, for a fixed measured target (the relative positions of the measured target and a plurality of laser radars are fixed);
on the premise of multi-radar accurate calibration, the distance measurement results of different radars at the same target position are completely the same in theory, and the density degree of the distribution of the radar scanning points is different for the same target position in view of the fact that the relative distances between each radar and the measured target are different. The scanning angle values in the point cloud data after unifying the coordinate system cannot be completely consistent, meanwhile, the existence of the ranging errors of all the radars is considered, based on the fact that the ranging results of different radars at the same target position are very close, and the difference value between the ranging values given by all the radars is in a very small range.
In this embodiment, for the calibration of the fixed measured target, two to three times of radar ranging errors are used as the reference for mutual calibration, and filtering or weight reducing processing is performed on the point cloud data exceeding the reference.
Wherein, for a non-stationary measured object (the measured object is displaced relative to the multi-lidar system during the measurement process);
the point cloud data distortion caused by the relative motion of the measured object is the same for all radars, the maximum distortion is determined by the time required for scanning one frame and the motion speed of the measured object for single frame data, and the specific numerical value or the range of the interval between the two parameters is a known quantity for a specific application scene;
in this embodiment, for the verification of the non-fixed measured target, an abnormal amount is used to set a threshold value, and the comparison and processing of the same measured position point cloud data between the frame data of the same time measurement of different radars are performed. And filtering or reducing weight processing is carried out on the point cloud data exceeding the distortion amount set threshold value.
Because of the overlapping area of the scan fields of view for each radar, the point cloud contours should tend to be consistent; the most rapid way of representing the outline features of the point cloud is to extract key points; because the positions of the key points are stable under different visual angles and can be repeatedly detected, the judgment basis of whether the radar in the system works normally can be based on whether the key points in the radar point cloud are overlapped.
Based on this, in step S4, in the overall comparison and verification of the point cloud data, the key points of the radar point cloud frame data at the same moment are extracted, whether the key points of the point cloud frame data overlap is determined, if yes, the radars are determined to work normally, if no, filtering or weight reducing processing is performed on all the point cloud data of the laser radars with unmatched key points.
The key points of the point cloud data are various in ISS, harris3D, NARF, sift D, uniform sampling and voxel sampling.
In a specific scheme of the embodiment, acquiring key point data by taking NARF (Normal Aligned Radial Feature) data as key point data of point cloud data; the method comprises the following specific steps:
1) And traversing each point cloud data point, and performing edge detection by searching a position with depth mutation in a neighboring area.
2) Traversing each depth image point, determining a coefficient of measuring the surface variation according to the surface variation of the neighboring area, and determining the main direction of variation.
3) Calculating an interest value according to the main direction found in the second step, and representing the difference between the direction and other directions and the change condition of the surface at the position, namely how stable the point is.
4) And smoothing and filtering the interest value.
5) And carrying out maximum value-free compression to find a final key point, namely the NARF key point.
Specifically, in the point cloud data stitching and fusion of step S4,
an alternative way is: the method comprises the steps of directly packaging and outputting residual point cloud data after finishing cross interference filtering, point-by-point comparison and verification and point cloud data integral verification without processing;
an alternative way is: and (4) performing adjacent data integration (point frequency reduction) or simultaneous data integration according to actual requirements, and finally packaging and outputting the point cloud data by splicing and fusing.
In summary, the embodiment of the application has the following effects:
1. according to the method, synchronous scanning of the multi-laser radar is realized through position calibration and clock synchronization, filtering, nuclear calibration and data fusion are carried out on point cloud data of the multi-laser radar, interference among internal components is eliminated, the measurement range and measurement precision of a single radar are expanded, and the overall measurement capability is improved;
2. the application preposes multi-radar point cloud splicing fusion, achieves coarse calibration based on position relation and fine calibration by measuring the fixed targets one by the radar, and completes pre-fusion of radar data. Links adopted by fusion after downsampling, key point extraction, point cloud registration, image splicing and the like are omitted, radar data can be directly integrated and output after simple filtering, the system real-time performance is good, and the requirement on the data processing capability of a central control system is low;
3. the application has the advantages that the occupation ratio of the point cloud data of the problems generated by mutual interference in all the point cloud data commonly measured by multiple radars is extremely low, and the integral measurement effect is not influenced by filtering;
4. according to the scheme, under the condition that phase locking is not performed, a corresponding point cloud data filtering mode is adopted based on an interference reason, so that the problem that the profile information of a measured object is omitted due to a conventional mode is avoided;
5. for a fixed measured target, the point cloud data measured by different radars at the same time and the same position can ensure that the corresponding distance difference value is smaller than a threshold value after mutual interference filtering, and for the point cloud data measured by different radars at the same position and the different times, the theoretical distance difference value is smaller than a fixed value, so that the data mutual checking is performed on the basis of filtering, noise can be further filtered, and the accuracy of the point cloud data is improved;
6. the overall comparison of the radar point cloud data can be used to identify the operational failure of a single radar.
The other embodiment of the application also discloses a radar system, as shown in fig. 3, comprising a plurality of laser radars and a central control system, wherein the central control system is connected with the plurality of laser radars one by one to form a star-shaped connection network, and the laser radars are not directly connected with each other physically;
after the spatial position fixing of the plurality of laser radars and the physical connection of the central control system and the plurality of laser radars are finished;
the central control system and each laser radar realize synchronous scanning and point cloud data processing of the multi-laser radar in the embodiment through communication and cooperative work.
The specific technical details and advantages of the synchronous scanning and the point cloud data processing related to the multiple lidars in this embodiment are the same as those in the previous embodiment, and specific reference is made to the content of the previous embodiment, and details are not described here.
The present application is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present application are intended to be included in the scope of the present application.
Claims (4)
1. The synchronous scanning and point cloud data processing method of the multi-laser radar is characterized by comprising the following steps of:
step S1, determining space fixed positions and a basic coordinate system of a plurality of cooperative laser radars, and determining a coordinate transformation matrix between each radar coordinate and the basic coordinate based on the mutual position relation between the origin of the basic coordinate system and the radars;
s2, correcting each radar coordinate transformation matrix by measuring a standard measured object arranged at a space fixed position;
step S3, after synchronizing the laser radar clocks, data acquisition is carried out; performing coordinate transformation on the acquired measurement data according to the corrected coordinate transformation matrix and giving a time stamp to form multi-radar point cloud data;
s4, performing mutual interference filtering on the multi-radar point cloud data, comparing the point by point to perform verification, and completing data splicing and fusion after the integral verification of the point cloud data;
in the mutual interference filtering, according to the scanning phase difference between two radarsαPerforming fixed-point mutual interference and non-fixed-point mutual interference filtering on different ranges;
when the reference radar A and the calibrated radar B scan for phase differencesα=0° or 180+.αWhen the angle is less than 360 degrees, judging that fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the reference radar A generates optical interference to the calibrated radar B as fixed point mutual interference at only one moment in the time range of realizing one period 360 scanning;
when the scanning phase difference of the reference radar A and the calibrated radar B is 0 DEG <αWhen the angle is less than 180 degrees, judging that non-fixed point mutual interference occurs between the reference radar A and the calibrated radar B; the two radar emergent rays intersect at an intersection point P, and the position of the intersection point P in space continuously changes along with the time passing of the scanning process to form a mutual interference curve; the interference caused by the points on the mutual interference curve is non-fixed point mutual interference;
for the fixed point mutual interference, filtering is completed by screening and filtering the data of radar measurement points at specific scanning angle moments and adjacent scanning areas with set angles, wherein the radar measurement points cause single point mutual interference;
when non-fixed point mutual interference filtering is carried out, the method comprises the following steps:
1) Establishing a coordinate system for calculating the mutual interference positions of the two radars, and determining the coordinate of an intersection point P where the emergent laser rays of the two radars intersect; when the measured object is coincident with the intersection point P, the step 2) is entered, otherwise, whether the measured object is coincident with the intersection point P is repeatedly judged in the radar measurement process;
2) Judging scanning phase difference in radar scanning processαIf the set time interval is approaching to the fixed value, the step 3) is entered, and if the set time interval is not approaching to the fixed value, the step 4) is entered;
3) According to the triangle cosine theorem, calculating an interference curve where the intersection point P is located, and screening and filtering out adjacent points which are on the interference curve and are out of line and accord with the condition of the adjacent threshold value as interference points by setting the threshold value of the adjacent points;
4) Performing point-by-point position calculation on the point cloud data of the reference radars A and B received at the same time to obtain an intersection point P coordinate value of intersection of the two radar emergent laser rays; calculating the distance from the intersection point P to the origin of the calculated coordinate system according to the coordinate value of the intersection point PL；
5) Distance to be calculatedLReplacing the measured distance value;
in step S4, the point cloud data point-by-point comparison check includes a check of a fixed measured object and a check of an unfixed measured object;
fixing the checking of the measured target, adopting two to three times of radar ranging errors as the reference of mutual checking, and filtering or reducing weight for the point cloud data exceeding the reference;
checking a non-fixed measured target, setting a threshold value by adopting an abnormal quantity, and comparing and processing the same measured position point cloud data among the measured frame data at the same moment of different radars; filtering or reducing weight processing is carried out on the point cloud data exceeding the distortion quantity set threshold value;
in step S4, in the overall comparison and verification of the point cloud data, the key points of the point cloud data of each radar at the same moment are extracted, whether the key points of the point cloud data are coincident or not is judged, if yes, the radars are judged to work normally, if not, the laser radars with unmatched key points are filtered or weight-reduced to all the point cloud data.
2. The method for simultaneous scanning and point cloud data processing of multiple lidars according to claim 1, wherein the multiple lidars are configured to perform the simultaneous scanning,
in step S2, correction of the transformation matrix is performed by adopting an ICP algorithm; the method for correcting the transformation matrix by adopting the ICP algorithm comprises the following steps:
step S201, setting a standard object to be measured at a fixed position in a measurement space, enabling all the laser radars to work one by one, ensuring that only one radar works at the same time, and obtaining point cloud data of all the laser radars;
step S202, extracting key point data of point cloud data of each laser radar to obtain a key point data set;
step S203, taking one laser radar as a reference radar A, and the rest laser radars as calibrated radars B; according to the key point data of the radar point clouds of the reference radar A and each calibrated radar B, an ICP algorithm is adopted to obtain a correction matrix, and the correction matrix is overlapped on the coordinate transformation matrix stored by each calibrated radar B, so that the correction of the radar coordinate transformation matrix of the calibrated radar B is realized.
3. The method for simultaneous scanning and point cloud data processing of multiple lidars according to claim 2, wherein the data processing device comprises,
in step S203, the radar coordinate transformation matrix correction process of the calibrated radar B with respect to the reference radar a includes:
1) The reference radar A and the calibrated radar B respectively scan a fixed standard measured object, and respectively take one frame of point cloud data in the measurement result to extract key point data so as to obtain key point data sets a and B;
wherein a= (a) 1 ,a 2 ,…,a i ,…,a n ）；b=（b 1 ,b 2 ,…,b i ,…,b n ）；
2) For the key point data sets a and b, the corresponding relation of the key point data sets a and b is found by adopting a characteristic matching mode, and the matching error is minimized to solve the rotation matrixRTranslation matrixtIs used for correcting the matrix; solved rotation matrixRTranslation matrixtA correction matrix for the calibrated radar B relative to the reference radar a;
wherein the matrix is rotatedThe method comprises the steps of carrying out a first treatment on the surface of the Translation matrix->；
U、VSVD decomposition is carried out on the matched features to obtain a left singular vector matrix and a right singular vector matrix;centroid of the data of the key point data set a; />Centroid of the b data of the keypoint dataset;
3) And (3) superposing a correction matrix on the coordinate transformation matrix of the calibrated radar B to finish the correction of the calibrated radar B transformation matrix.
4. The radar system is characterized by comprising a plurality of laser radars and a central control system, wherein the central control system is connected with the plurality of laser radars one by one to form a star-shaped connection network, and the laser radars are not directly connected with each other physically;
after the spatial position fixing of the plurality of laser radars and the physical connection of the central control system and the plurality of laser radars are finished;
the central control system and each laser radar realize synchronous scanning and point cloud data processing of the multi-laser radar according to any one of claims 1-3 through communication and cooperative work.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310601244.4A CN116299319B (en) | 2023-05-26 | 2023-05-26 | Synchronous scanning and point cloud data processing method of multiple laser radars and radar system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310601244.4A CN116299319B (en) | 2023-05-26 | 2023-05-26 | Synchronous scanning and point cloud data processing method of multiple laser radars and radar system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116299319A CN116299319A (en) | 2023-06-23 |
| CN116299319B true CN116299319B (en) | 2023-08-15 |
Family
ID=86794605
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310601244.4A Active CN116299319B (en) | 2023-05-26 | 2023-05-26 | Synchronous scanning and point cloud data processing method of multiple laser radars and radar system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116299319B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117214875A (en) * | 2023-11-08 | 2023-12-12 | 山东富锐光学科技有限公司 | Zero point calibration method and structure for laser radar incremental coding |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106773802A (en) * | 2015-11-25 | 2017-05-31 | 北京万集科技股份有限公司 | Many laser radar synchronous scanning control systems and method |
| CN112924955A (en) * | 2021-01-29 | 2021-06-08 | 同济大学 | Dynamic correction method for point cloud coordinates of roadside laser radar |
| CN113552559A (en) * | 2021-07-23 | 2021-10-26 | 北京轻舟智航科技有限公司 | Laser radar external parameter correction method in unmanned driving |
| WO2021253193A1 (en) * | 2020-06-15 | 2021-12-23 | 深圳市大疆创新科技有限公司 | Calibration method and calibration apparatus for external parameters of multiple groups of laser radars, and computer storage medium |
| CN113888649A (en) * | 2021-10-18 | 2022-01-04 | 上海振华重工(集团)股份有限公司 | Multi-laser-radar external parameter calibration method, device, equipment and storage medium |
| CN113902860A (en) * | 2021-10-10 | 2022-01-07 | 北京工业大学 | Multi-scale static map construction method based on multi-line laser radar point cloud |
| WO2022075356A1 (en) * | 2020-10-07 | 2022-04-14 | パイオニア株式会社 | Information processing apparatus, control method, program, and storage medium |
| WO2022179094A1 (en) * | 2021-02-24 | 2022-09-01 | 长沙行深智能科技有限公司 | Vehicle-mounted lidar external parameter joint calibration method and system, medium and device |
| CN115542338A (en) * | 2022-11-30 | 2022-12-30 | 湖南大学 | Laser radar data learning method based on point cloud space distribution mapping |
-
2023
- 2023-05-26 CN CN202310601244.4A patent/CN116299319B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106773802A (en) * | 2015-11-25 | 2017-05-31 | 北京万集科技股份有限公司 | Many laser radar synchronous scanning control systems and method |
| WO2021253193A1 (en) * | 2020-06-15 | 2021-12-23 | 深圳市大疆创新科技有限公司 | Calibration method and calibration apparatus for external parameters of multiple groups of laser radars, and computer storage medium |
| WO2022075356A1 (en) * | 2020-10-07 | 2022-04-14 | パイオニア株式会社 | Information processing apparatus, control method, program, and storage medium |
| CN112924955A (en) * | 2021-01-29 | 2021-06-08 | 同济大学 | Dynamic correction method for point cloud coordinates of roadside laser radar |
| WO2022179094A1 (en) * | 2021-02-24 | 2022-09-01 | 长沙行深智能科技有限公司 | Vehicle-mounted lidar external parameter joint calibration method and system, medium and device |
| CN113552559A (en) * | 2021-07-23 | 2021-10-26 | 北京轻舟智航科技有限公司 | Laser radar external parameter correction method in unmanned driving |
| CN113902860A (en) * | 2021-10-10 | 2022-01-07 | 北京工业大学 | Multi-scale static map construction method based on multi-line laser radar point cloud |
| CN113888649A (en) * | 2021-10-18 | 2022-01-04 | 上海振华重工(集团)股份有限公司 | Multi-laser-radar external parameter calibration method, device, equipment and storage medium |
| CN115542338A (en) * | 2022-11-30 | 2022-12-30 | 湖南大学 | Laser radar data learning method based on point cloud space distribution mapping |
Non-Patent Citations (1)
| Title |
|---|
| 基于手眼模型的三维激光雷达外参数标定;韩栋斌;徐友春;李华;谢德胜;陈文;;光电工程(第08期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116299319A (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lenz et al. | Techniques for calibration of the scale factor and image center for high accuracy 3-D machine vision metrology | |
| CN106526593B (en) | Sub-pixel-level corner reflector automatic positioning method based on the tight imaging model of SAR | |
| CN116299319B (en) | Synchronous scanning and point cloud data processing method of multiple laser radars and radar system | |
| CN110208771B (en) | Point cloud intensity correction method of mobile two-dimensional laser radar | |
| CN107339935B (en) | Target space intersection measuring method for full-view scanning measuring system | |
| JP2012189445A (en) | Object detection device and object detection method | |
| Reulke | Combination of distance data with high resolution images | |
| US11692812B2 (en) | System and method for measuring three-dimensional coordinates | |
| WO2022001325A1 (en) | Point cloud data fusion method and apparatus, electronic device, storage medium, and computer program | |
| CN113160328A (en) | External reference calibration method, system, robot and storage medium | |
| Wang et al. | Modelling and calibration of the laser beam-scanning triangulation measurement system | |
| CN116802688A (en) | Apparatus and method for correspondence analysis within an image | |
| Usamentiaga | Static calibration for line-scan cameras based on a novel calibration target | |
| CN114089373B (en) | Radar system scanning method and device based on image fusion laser | |
| Hasheminasab et al. | Linear Feature-based image/LiDAR integration for a stockpile monitoring and reporting technology | |
| Lambertus et al. | Empirical investigation of a stochastic model based on intensity values for terrestrial laser scanning | |
| EP4086850A1 (en) | Calibrating system for colorizing point-clouds | |
| US20220196386A1 (en) | Three-dimensional scanner with event camera | |
| Long et al. | Portable visual metrology without traditional self-calibration measurement model | |
| WO2021049490A1 (en) | Image registration device, image generation system, image registration method and image registration program | |
| CN114219908A (en) | Map construction method and device | |
| CN116184427B (en) | Distance detection system based on laser distance sensing technology | |
| CN110967693B (en) | Robust and efficient fast decomposition projection automatic focusing method and system | |
| WO2021102980A1 (en) | System error calibration method and apparatus, and computer-readable storage medium | |
| Altuntas | Measurement Precision of Average Frame for Solid-State Time-of-Flight Camera |
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 |