CN115079115A

CN115079115A - Radar, data processing method and device for radar, and readable storage medium

Info

Publication number: CN115079115A
Application number: CN202110262008.5A
Authority: CN
Inventors: 顾天长; 王重阳; 杨晋; 向少卿
Original assignee: Hesai Technology Co Ltd
Current assignee: Hesai Technology Co Ltd
Priority date: 2021-03-10
Filing date: 2021-03-10
Publication date: 2022-09-20
Also published as: WO2022188495A1

Abstract

The radar, the data processing method and the device for the radar and the readable storage medium are provided, wherein the data processing method for the radar comprises the following steps: 1) acquiring a point cloud of the radar, the point cloud comprising: information points obtained by a plurality of detectors at a plurality of angles; 2) selecting a point to be judged from the point cloud; 3) using the detector corresponding to the point to be judged as a reference detector, using the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and using the obtained information point as a reference point; 4) and judging whether the point to be judged is a noise point or not based on the correlation degree between the point to be judged and the reference point. By adopting the scheme, the accuracy of the noise point identification result can be improved, the misjudgment probability is reduced, and the data quality of the point cloud is guaranteed.

Description

Radar, data processing method and device for radar, and readable storage medium

Technical Field

The embodiment of the specification relates to the technical field of radars, in particular to a radar, a data processing method and device for the radar, and a readable storage medium.

Background

At present, a radar can calculate information points of a three-dimensional space by transmitting a detection signal to an environment where the radar is located and receiving an echo signal reflected by an obstacle, so as to obtain a point cloud of the three-dimensional space. The interference signal will cause noise in the point cloud, thereby reducing the accuracy of subsequent processing.

In some applications, the information points collected by the same detector can be judged by setting a threshold value, so that noise points can be identified and filtered. For example, if a detector in the radar is at time t _x-1 、t _x And t _x+1 Respectively measure information points PC _X-1 、PC _X And PC _x+1 Then three information points PC are obtained _X-1 、PC _X And PC _x+1 And the information point PC is determined based on the parameters such as the distance and the reflectivity _X And a PC _X-1 And information point PC _X And PC _X+1 If the degree of correlation is lower than a preset threshold value, determining the information point PC _X Is a noise point and further filters the information point PC _X 。

However, due to the small obstacle or the low radar resolution, the information points acquired by the same detector at different times may be sparsely distributed and have a weak correlation, and if the threshold is set to be high, useful information points (i.e., information points other than noise points) that are sparsely distributed may be erroneously determined as noise points, and further, the useful information points may be filtered, resulting in a problem of missing points. In order to reserve the sparse useful information points as much as possible, only the threshold value is reduced, the judgment scale is widened, and meanwhile, the risk that the noise points are mistakenly judged as the useful information points is increased. Therefore, the existing solutions are less accurate for noise filtering.

Disclosure of Invention

In view of this, embodiments of the present disclosure provide a radar, a data processing method and device for a radar, and a readable storage medium, which can improve accuracy of a noise point identification result, reduce a false determination probability, and ensure data quality of point cloud.

The present specification provides a data processing method for a radar, comprising the steps of:

1) acquiring a point cloud of the radar, the point cloud comprising: information points obtained by a plurality of detectors at a plurality of angles;

2) selecting a point to be judged from the point cloud;

3) taking the detector corresponding to the point to be judged as a reference detector, taking the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and taking the obtained information point as a reference point;

4) and judging whether the point to be judged is a noise point or not based on the correlation degree between the point to be judged and the reference point.

Optionally, before the step 4), the following steps are further included:

A) and searching detection information of adjacent detectors corresponding to the adjacent angles of the reference angle and the reference detector, and taking the obtained information points as reference points.

Optionally, the adjacent detector is one or more detectors adjacent to the reference detector in the arrangement position;

the adjacent angle is one or more angles of which the angle difference with the reference angle belongs to a preset range.

Optionally, for the detection information of the reference detector, the adjacent angle is an angle whose angle difference from the reference angle belongs to a first range;

for the detection information of the adjacent detector, the adjacent angle is an angle of which the angle difference with the reference angle belongs to a second range;

wherein the span of the first range is greater than the span of the second range.

Optionally, a span difference between the first range and the second range is positively correlated with a field of view angle difference between the detectors.

Optionally, when the adjacent detector includes a plurality of detectors located on one side of the reference detector in the arrangement position, the second range includes a plurality of angular ranges respectively corresponding to the plurality of detectors, and a span of the angular ranges is inversely related to a distance from the adjacent detector to the reference detector in the arrangement position.

Optionally, the step 4) includes:

4-1) determining whether the point to be determined is a noise point according to the point distance between the point to be determined and the reference point.

Optionally, the step 4-1) comprises the following steps:

4-11) respectively calculating the distance between each reference point and the point to be judged to obtain a point distance data set;

4-12) determining a point distance threshold value;

4-13) judging whether point distance data lower than the point distance threshold exist in the point distance data set or not, and if yes, determining that the point to be judged is not a noise point; otherwise, determining the point to be determined as a noise point.

Optionally, the step 4-12) comprises the steps of:

4-121) obtaining the distance of the obstacle corresponding to the point to be determined;

4-122) determining the point distance threshold value of the point to be determined based on the corresponding relation between the obstacle distance and the point distance threshold value.

Optionally, the point-to-point distance threshold is positively correlated with the distance of the obstacle corresponding to the point to be determined, and the point-to-point distance threshold is negatively correlated with the resolution of the radar.

Optionally, when the point loss rate of the radar is higher than a preset point loss rate threshold, the point distance threshold is increased.

Optionally, the step 4) includes:

4-2) determining whether the point to be determined is a noise point according to the number of the points of the reference point.

Optionally, the step 4-2) comprises the following steps:

4-21) judging whether the point number of the reference point is lower than a preset point number threshold value, if so, determining that the point to be judged is a noise point; otherwise, determining that the point to be determined is not a noise point.

Optionally, the point threshold is positively correlated with a resolution of the radar.

Optionally, when the point loss rate of the radar is higher than a preset point loss rate threshold, the point threshold is increased.

Optionally, the data processing method for radar further includes:

5) and if the point to be determined is judged to be a noise point, deleting the point to be determined.

Optionally, the data processing method for radar further includes:

6) and judging whether information points which are not subjected to noise point judgment exist, if so, selecting the point to be judged from the information points which are not subjected to noise point judgment, and continuing to execute the step 3).

The present specification also provides a data processing apparatus for a radar, comprising:

the data storage unit is suitable for caching point cloud of the radar, and the point cloud comprises information points obtained by a plurality of detectors at a plurality of angles;

the noise point identification unit is suitable for selecting a point to be judged from the point cloud; taking the detector corresponding to the point to be judged as a reference detector, taking the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and taking the obtained information point as a reference point; and judging whether the point to be judged is a noise point or not based on the correlation degree of the point to be judged and the reference point.

Optionally, the noise point identification unit is further adapted to find detection information of an adjacent detector corresponding to the reference angle and an adjacent detector corresponding to the reference detector, and use the obtained information point as a reference point.

Optionally, the degree of correlation between the point to be determined and the reference point includes: and the point distance between the point to be determined and the reference point.

Optionally, the degree of correlation between the point to be determined and the reference point includes: the number of points of the reference point.

Optionally, the noise point identifying unit is further adapted to delete a point to be determined which is determined as noise point.

Optionally, the noise point identification unit is further adapted to select the point to be determined from information points for which noise point determination is not performed, and perform noise point determination.

The present specification also provides a radar comprising a plurality of transmitters, a plurality of detectors and a data processing device, wherein:

the emitter and the detector have a corresponding relation; the transmitter is suitable for transmitting the detection signals in multiple angles; the detector is suitable for multi-angle echo signal acquisition;

the data processing device is adapted to perform data processing according to the detection signal of the transmitter and the echo signal of the corresponding detector to obtain a corresponding information point, and to perform the method according to any of the above embodiments.

Optionally, the radar is a lidar.

The present specification also provides a computer readable storage medium having stored thereon computer instructions which, when executed, perform the steps of the method of any of the above embodiments.

By adopting the data processing method for the radar of the embodiment of the specification, the reference points which are related to the point to be determined in the same dimension of space and time sequence are obtained by searching the adjacent angle corresponding to the reference angle and the detection information corresponding to the reference detector and the adjacent detector corresponding to the reference detector and the detection information corresponding to the reference angle, so that the reliability and diversity of the reference points are improved, the accuracy of the identification result can be improved when the noise point is identified according to the correlation between the point to be determined and the reference point, the misjudgment probability is reduced, the noise point can be effectively identified, and the data quality of the point cloud is guaranteed.

Furthermore, the correlation between the point to be determined and the reference point set is represented by the distance between each reference point in the reference point set and the point to be determined, so that the reliability of the correlation can be enhanced, the problem of point loss caused by misjudgment is effectively reduced, the accuracy of a noisy point identification result is improved, and the data quality of point cloud is ensured.

Furthermore, the correlation between the point to be judged and the reference point set is represented by the point number of the reference point, so that the calculated amount can be reduced, and the noise point identification efficiency is improved under the condition of ensuring the accuracy of the noise point identification result.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings used in the embodiments of the present specification or in the description of the prior art will be briefly described below, it is obvious that the drawings described below are only some embodiments of the present specification, and it is also possible for a person having ordinary skill in the art to obtain other drawings based on the drawings without creative efforts.

Fig. 1a is a schematic view of a part of the field of view of a radar in a prior art application.

Fig. 1b is a schematic diagram of arrangement of information points acquired by a part of a detector of a radar corresponding to the field of view of fig. 1a in a three-dimensional space.

Fig. 2 is a flowchart of a data processing method for radar in an embodiment of the present specification.

Fig. 3 is a flowchart of another data processing method for radar in an embodiment of the present specification.

Fig. 4 is a schematic diagram of a distribution of a radar point cloud in an embodiment of the present disclosure.

Fig. 5 to 11 are schematic diagrams of various reference point selections corresponding to fig. 4.

FIG. 12 is a schematic diagram of the arrangement of information points acquired by a portion of the detector of another radar corresponding to the field of view of FIG. 1a in three-dimensional space.

Fig. 13 is a flowchart of a noise determination method in an embodiment of the present disclosure.

Fig. 14 is a flowchart of a method for determining a dot pitch threshold in an embodiment of the present specification.

Fig. 15 is a schematic diagram illustrating a correspondence relationship between a point-to-threshold value and an obstacle distance in an embodiment of the present specification.

Fig. 16 is a schematic diagram illustrating a correspondence relationship between a point-to-threshold value and an obstacle distance in another embodiment of the present specification.

Fig. 17 is a block diagram of a data processing device for radar in an embodiment of the present specification.

Fig. 18 is a block diagram of a radar in the embodiment of the present specification.

Detailed Description

As known from the background art, if there is noise in the point cloud, the accuracy of the subsequent processing will be reduced. In order to make the problems of the prior art more clearly known to those skilled in the art, the following detailed description is given with reference to the accompanying drawings and the prior application.

In one currently available implementation, a vehicle is loaded with a radar that transmits a detection signal to the environment, and the received signal includes an echo signal and an interference signal reflected from an obstacle, and the point cloud obtained by the radar includes noise.

FIG. 1a shows a schematic view of a partial field of view of a radar; fig. 1b shows a schematic diagram of a partial detector of a radar arranged in three-dimensional space corresponding to information points acquired by the field of view of fig. 1 a.

Multiple detectors may be included in the radar, with one detector TC in the radar for ease of description _X Are objects of description.

With combined reference to fig. 1a and 1b, during actual detection of the radar, the detector TC _X At t ₁ ～t ₄ The detection signals reflected by the external wall WA are received at the moment, and because the wall WA is wider, the detection signals are detected at the detector TC _X The horizontal angle span in the visual field is large, and four useful information points with strong relevance, namely the useful information point a in fig. 1b, can be collected ₁ ～a ₄ 。

Detector TC _X At t ₅ The visual field at the moment corresponds to the GA of the telegraph pole, the detection signal reflected by the GA of the telegraph pole is received, and the telegraph pole is thinner, so that the detection signal is detected by a detector TC _X Has small horizontal angle span in the field of view, and the detector TC _X Only one useful information point, namely the useful information point a in fig. 1b, is acquired in the width perpendicular to the axial direction of the utility pole ₅ . At t ₅ Next adjacent detection time of detector TC _X The field of view deviates from the utility pole GA.

Detector TC _X At t ₆ ～t ₉ The detection signals reflected by the external wall WB are respectively received at all times, and the wall WB is wider, so that the detection signals are detected at the detector TC _X Has a large horizontal angle span in the field of view, and obtains four useful information points with strong relevance, namely the useful information point a in fig. 1b ₆ ～a ₉ 。

As can be seen from FIG. 1b, the useful information point a ₅ And useful information points a ₄ Has weak correlation and is provided with useful information points a ₅ And useful information points a ₆ The correlation between them is also weak.

If the scheme of the prior art is adopted, t is obtained ₄ Time t ₅ Time and t ₆ Time of dayRespectively measured useful information points a ₄ Useful information point a ₅ And useful information points a ₆ And according to the useful information point a ₄ Useful information point a ₅ And useful information points a ₆ Will obtain useful information points a ₄ And useful information points a ₅ And useful information point a ₅ And useful information points a ₆ The correlation between the two is lower than a threshold value, so that the useful information point a is positioned ₅ And misjudging the signal as noise and filtering the noise.

Therefore, when the prior art scheme is adopted to judge the noise point, the useful information point with low correlation degree can be mistakenly judged and deleted (for example, the useful information point a in fig. 1 b) ₅ )。

Thus, existing solutions do not allow for a high degree of accuracy in filtering noise.

In order to solve the above problems, the present specification provides a data processing method for a radar, which selects a point to be determined and reference points corresponding to a plurality of detectors in a point cloud in combination with an angle and a detector position, and thereby determines whether the point to be determined is a noise point based on a correlation between the point to be determined and the reference points. Therefore, the accuracy of the noise point identification result is improved, the misjudgment probability is reduced, and the data quality of the point cloud is guaranteed.

For the purpose of promoting a better understanding of the principles of the invention, its embodiments, and advantages, reference should be made to the drawings and to the accompanying drawings.

Referring to fig. 2, a flowchart of a data processing method for a radar is provided for an embodiment of the present specification, and in the embodiment of the present specification, as shown in fig. 2, the method may include the following steps:

s1, acquiring a point cloud of the radar, wherein the point cloud comprises: the information points obtained by the plurality of detectors at the plurality of angles.

In a particular implementation, the radar may include multiple transmitters and multiple detectors. The plurality of transmitters can be arranged according to a designated direction (such as the vertical direction of a radar), and the included angle between the detection signal emitted by each transmitter and the designated direction is different; the detector and the transmitter have a corresponding relationship and are used for receiving signals of the surrounding environment, and after the detector receives signals (such as echo signals, interference signals and the like), information points can be obtained through calculation according to detection signals of the corresponding transmitter and the signals received by the detector.

The radar rotates along the appointed axis at a certain rotating speed in the working process, and information point acquisition is carried out after rotating a certain angle every time according to the set sampling frequency, so that information around the radar is acquired in the rotating process, the sensing of the surrounding environment is realized, and the information points obtained after the radar rotates for a circle form a frame of point cloud.

It should be noted that, during the radar operation, the detection information of the detector may include the calculated information point and the blank result (i.e., the signal is not received at the angle). For example, if the detector is at radar from angle α ₁ Rotated to an angle alpha ₂ The corresponding information point can be calculated by receiving the signal before, and the detector is at the angle alpha ₁ The detection information of (1) is an information point; otherwise, the detector is at angle α ₁ The detected information of (2) is a blank result.

In order to distinguish and know the origin of each information point, each of said information points may comprise detector data and angle data.

Taking a conventional mechanical rotary radar as an example, a plurality of detectors are arranged along the vertical direction of the radar, and different detectors are used for receiving echo signals with different vertical angles, so that information points measured by different detectors can obtain the corresponding vertical angles according to the positions of the detectors.

The vertical angle difference corresponding to two adjacent detectors is the vertical angle resolution of the radar, and thus, the detection information of the adjacent detectors corresponding to the reference detector described in this specification may be understood as: the corresponding vertical angle is separated from the reference detector of the point to be determined by detection information of the detector of one or a plurality of continuous vertical angle resolutions.

The radar can rotate in the horizontal direction by 360 degrees, and after the radar finishes the rotation of one horizontal angle, a plurality of transmitters and detectors are started in turn to detect the outside. After all detectors complete the round trip, detection information corresponding to the radar vertical Field of View (FOV) at that horizontal angle is obtained.

After the detection under one horizontal angle is finished, the radar rotates to another horizontal angle, and the next round of inspection is carried out.

The horizontal angle difference corresponding to two adjacent signal detections of the same detector may be the horizontal angle resolution of the radar, and thus, the detection information corresponding to adjacent angles described in this specification may be understood as meaning: the corresponding horizontal angle is separated from the reference angle of the point to be determined by one or by detection information of a plurality of horizontal angular resolutions in succession.

As can be seen from the above, the detector data can be used to characterize the detector corresponding to the information point, and the angle data can be used to characterize the angle that the detector corresponding to the information point has rotated relative to the reference position.

Wherein the angle is related to a rotational speed and a sampling frequency of the radar. And, the reference position may be an initial position of the radar, and the angle may be an included angle between an initial position of the corresponding detector and the current position.

And S2, selecting a point to be judged from the point cloud.

In specific implementation, the selection mode of the point to be determined can be set according to actual application scenes and requirements. For example, information points may be sequentially selected from the point cloud in a specified order as points to be determined. For another example, a part of information points may be selected from the point cloud as points to be determined. The embodiments of the present specification are not limited thereto.

S3, using the detector corresponding to the point to be judged as a reference detector, using the angle corresponding to the point to be judged as a reference angle, searching the detection information corresponding to the adjacent angle of the reference angle and the reference detector, the detection information corresponding to the adjacent detector of the reference detector and the reference angle, and using the obtained information point as a reference point.

In specific implementation, according to the detector data of the point to be determined, the detector corresponding to the point to be determined is determined, and the detector is used as a reference detector. And determining the corresponding angle according to the angle data of the point to be determined, and taking the angle as a reference angle. Then, based on the reference detector and the reference angle, searching detection information of which the detector and the angle meet the preset adjacent condition from the detection information of each detector, and taking an information point obtained by the detection information meeting the preset adjacent condition as a reference point, thereby determining the information point in the adjacent space domain of the point to be determined.

Wherein the adjacent detector may be one or more detectors adjacent to the reference detector in the arrangement position; the adjacent angle may be one or more angles having an angle difference from the reference angle within a preset range.

In other words, the reference points may include: the method comprises the steps of calculating other information points except a point to be judged when the radar rotates to a reference angle, and calculating information points corresponding to a reference detector when the radar rotates to an adjacent angle of the reference angle. Therefore, reference points from different sources are obtained, the reliability and diversity of the reference points are increased, and spatial and time sequence correlation exists between the reference points and the points to be judged.

In specific implementation, the preset range can be set according to the sampling frequency, the rotating speed and the actual requirement of the radar. Wherein the span of the preset range (i.e. the difference between the maximum boundary value and the minimum boundary value in the range) is at least larger than the angle rotated by a single sample, thereby ensuring that at least one adjacent angle can be determined according to the range. For example, if the radar has a rotational speed of ω and a sampling frequency of f, the span of the range is at least ω/f.

In specific implementation, the determination mode of the adjacent detector may be set according to an actual application scenario and a requirement, and according to different determination modes, the adjacent detector may be a detector located on a designated side of the reference detector, or may be detectors located on both sides of the reference detector. Correspondingly, the adjacent angle determination mode can be set according to the actual application scene and the requirement, according to different determination modes, the adjacent angle can be the angle detected last time corresponding to the reference angle, and the angle difference between the adjacent angle and the reference angle is represented by a negative value; or the angle corresponding to the reference angle detected at the later time, and the angle difference between the adjacent angle and the reference angle is represented by a positive value; the corresponding angle can be detected for a plurality of times before and after the reference angle, and then the corresponding range can be set according to different judging modes. The present specification does not specifically limit the manner of determination of the adjacent detectors and the adjacent angles.

And S4, judging whether the point to be judged is a noise point or not based on the correlation degree of the point to be judged and the reference point.

The reference points comprise information points of adjacent angles (transverse dimensions) and adjacent detectors (longitudinal dimensions), so that reference points which are related to the points to be judged in the same dimensions of space and time sequence are obtained, the reliability and diversity of the reference points can be improved, the accuracy of identification results can be improved, the misjudgment probability can be reduced, the noise points can be effectively identified and the data quality of point clouds can be guaranteed when the noise points are identified according to the correlation between the points to be judged and the reference points.

In order to make the technical effects of the technical solutions provided in the present specification more clearly understood by those skilled in the art, the following description will proceed to specific examples of the corresponding realizable applications in fig. 1a and 1 b.

As can be seen from FIGS. 1a and 1b and the related description, the detector TC _X Collected useful information points a ₅ And a detector TC _X Including the useful information point a ₁ ～a ₄ And useful information points a ₆ ～a ₉ ) The correlation between the two is weak, so that the prior technical scheme can be used for the useful information point a ₅ Misjudgment is noise.

Hypothesis and detector TC _X The detectors that differ by one vertical angular resolution are: detector TC _X-1 And a detector TC _X+1 . Wherein the detector TC _X-1 At t ₅ A detection signal reflected by an external telegraph pole GA is received at the moment to obtain a useful information point c ₁ . Detector TC _X+1 At t ₁ ～t ₄ Respectively receiving detection signals reflected by an external wall WA at any time to obtain a useful information point d ₁ ～d ₄ And at t ₅ Moment obtaining useful information point d ₅ 。

As can be seen from FIG. 1a and FIG. 1b, useful information point c ₁ Useful information point a ₁ And useful information points d ₅ I.e. a list of useful information points corresponding to the axis of the utility pole GA in figure 1 b.

For the technical solution provided in this specification, if the detector TC is used _X At t ₅ Useful information point a collected at any moment ₅ As the point to be determined, the reference points thereof may include: useful information point a ₄ Useful information point a ₆ Useful information point c ₁ And useful information points d ₅ 。

Due to the detector TC _X Useful information point a of ₅ And a detector TC _X Including the useful information point a ₁ ～a ₄ And useful information points a ₆ ～a ₉ ) The correlation degree between the two is low, so that the useful information point a is positioned at ₅ After selecting the point to be determined, the point a to be determined ₅ From a reference detector TC _X Reference point a of ₄ And a reference point a ₆ The correlation between them is low.

However, since the fields of view of the detectors corresponding at the same angle can be considered approximately the same, adjacent detectors TC _X-1 And adjacent detectors TC _X+1 At reference angle (i.e. detector TC) respectively _X At the point a of collecting useful information ₅ Angle of rotation) acquired reference point c ₁ And a reference point d ₅ Are all connected with the point a to be determined ₅ Has higher correlation degree, and improves the point a to be judged ₅ And the overall correlation degree with the reference point of the useful information point, thereby effectively reducing the misjudgment rate of the useful information point.

In a specific implementation, in order to further improve the accuracy of the noise point identification result, as shown in fig. 3, a flowchart of another data processing method for radar provided in the embodiment of the present specification is different from the example shown in fig. 2 in that, before the step S4), the following steps may be further included:

and SA, searching detection information of adjacent detectors corresponding to the adjacent angles of the reference angle and the reference detector, and taking the obtained information points as reference points.

Thus, the reference point may further include: and when the radar rotates to the adjacent angle of the reference angle, calculating the information points corresponding to the adjacent detectors. Therefore, the sources of the reference points are further enriched, and the reliability and diversity of the reference points are improved.

In a specific implementation, for the detection information of the reference detector, the adjacent angle is an angle whose angle difference from the reference angle belongs to a first range; for the detection information of the adjacent detector, the adjacent angle is an angle whose angle difference from the reference angle belongs to a second range. The span of the first range may be equal to the span of the second range, or may not be equal to the span of the second range.

In a specific implementation, since the farther the distance between the detectors is, the larger the difference of the field angle between the two collected information points is, the weaker the correlation between the corresponding information points is, the span of the first range may be larger than that of the second range in order to effectively reduce the data amount.

In an alternative example, the difference in span between the first range and the second range is positively correlated with the difference in angle of field between the detectors.

In a specific implementation, the adjacent detector may be a detector on one side of the reference detector, or may be a detector on both sides of the reference detector. When the adjacent detector includes a plurality of detectors on one side of the reference detector in the arrangement position, that is, when there are a plurality of adjacent detectors on one side of the reference detector, the second range may include a plurality of angular ranges respectively corresponding to the plurality of detectors. Further, since the farther the distance between the detectors is, the larger the angular difference of the field of view between the information points acquired by the detectors is, the weaker the correlation between the corresponding information points is, and the span of the angular range is inversely related to the distance between the adjacent detector and the reference detector at the arrangement position. This can effectively reduce the data amount.

In order to facilitate understanding and implementation of the reference point acquisition process by those skilled in the art, the following description is provided by way of specific embodiments and accompanying drawings.

It should be noted that, for convenience of description and intuitive understanding, some embodiments in the present specification employ a two-dimensional plane view angle, and the intervals between the information points are equal. However, this is not a specific limitation on the dimension and the actual distribution of the information points, and in practical applications, the information points may be three-dimensional data, and the information points may be non-uniformly distributed in a three-dimensional space, that is, the distances between the information points may not be equal. The embodiments of the present specification do not specifically limit this.

In the implementation, as shown in fig. 4, a distribution diagram of a radar point cloud is shown. In the present example, the radar may comprise (m + n +1) vertically arranged detectors, i.e. detectors DE ₁ To the detector DE _m+n+1 And the radar rotates horizontally along the vertical axis with a horizontal angular resolution of theta, i.e. the span between two consecutive angles is theta. Wherein, the detection information of each detector comprises (p + q +1) information points, which respectively correspond to the angle AG ₁ To angle AG _p+q+1 . Angle AG ₁ To angle AG _p+q+1 Any two angles in succession span theta. The (p + q +1) × (m + n +1) information points form a radar point cloud.

Selecting an information point A from the point cloud ₁ (i.e., information points filled with diagonal lines in fig. 4) as points to be determined. With the point A to be determined ₁ Corresponding detector DE _m+1 As a reference detector, the point A to be judged is used ₁ Corresponding angle AG _p+1 As a reference angle.

In some embodiments, in conjunction with FIGS. 2 and 4, at detector DE ₁ To the detector DE _m+n+1 In the probe information of (2), finding the corresponding reference angle AG _p+1 And corresponding to said reference detector DE _m+1 Probe information of (2), correspondingReference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (1).

According to the different judging modes of the adjacent detectors and the different judging modes of the adjacent angles, the detection information of different conditions can be obtained, and then different reference points can be obtained. The following description is given by way of example in several cases, and it should be noted that these examples do not limit other cases that may exist in the embodiments of the present disclosure in practical application.

In some cases, the adjacent detectors may be determined in the following manner: a detector positioned above the reference detector.

For the adjacent angle of the reference detector, the determination method may be as follows: the first angular range spans no less than the horizontal angular resolution theta and spans no more than twice the horizontal angular resolution 2 theta. For example, the first angle range may be (-2 θ, 0), so that an angle corresponding to the previous detection of the reference angle by the reference detector may be obtained according to the first angle range.

For the adjacent angle of the adjacent detector, the determination method may be: the second angular range spans less than twice the horizontal angular resolution 2 theta. For example, the second angle range may be (- θ, θ). Thus, the adjacent angle of the adjacent detector cannot be obtained according to the second angle range, in other words, the detection information of the adjacent detector at the adjacent angle does not need to be searched.

According to the above-mentioned determination method of the adjacent detector and the determination method of the adjacent angle of the reference detector, a reference point selection diagram corresponding to fig. 4 can be obtained as shown in fig. 5, in which:

corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At angle AG _p Is detected (i.e. information point B in fig. 5) ₁₁ )。

Corresponding to the reference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (a) may include: detectorDE _m At an angle AG _p+1 Is detected (i.e. information point C in fig. 5) ₁₁ )。

Then, information point B is added ₁₁ And information point C ₁₁ As a point A to be determined ₁ To the reference point of (c).

In other cases, the adjacent detectors may be determined in the following manner: one on each side of the reference detector.

For the adjacent angle of the reference detector, the determination method may be as follows: the first angular range spans no less than two times the horizontal angular resolution 2 theta and no more than three times the horizontal angular resolution 3 theta. For example, the first angle range may be [ - θ, θ ], such that the angles corresponding to the previous detection and the next detection of the reference angle by the reference detector can be obtained according to the first angle range.

For the adjacent angle of the adjacent detector, the determination method may be as follows: the second angular range spans less than twice the horizontal angular resolution 2 θ. For example, the second angle range may be (- θ, θ). Thus, the adjacent angle of the adjacent detector cannot be obtained according to the second angle range, in other words, the detection information of the adjacent detector at the adjacent angle does not need to be searched.

According to the above-mentioned determination method of the adjacent detector and the determination method of the adjacent angle of the reference detector, another reference point selection diagram corresponding to fig. 4 can be obtained as shown in fig. 6, in which:

corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At an angle AG _p Probe information (i.e. information point B in fig. 6) ₁₁ ) And at angle AG _p+2 Probe information (i.e. information point B in fig. 6) ₂₁ )。

Corresponding to the reference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (a) may include: detector DE _m At an angle AG _p+1 Is detected (i.e. information point C in fig. 6) ₁₁ ) And a detector DE _m+2 At an angle AG _p+1 Is detected (i.e. information point C in fig. 6) ₂₁ )。

Then, information point B is added ₁₁ Information point B ₂₁ Information point C ₁₁ And information point C ₂₁ As a point A to be determined ₁ The reference point of (a).

In other cases, the adjacent detectors may be determined in the following manner: j detectors located on either side of the reference detector. Wherein j is an integer greater than 1.

For the adjacent angle of the reference detector, the determination method may be as follows: the first angular range spans not less than 2i times the horizontal angular resolution 2i × θ and spans not more than 3i times the horizontal angular resolution 3i × θ, where i is an integer greater than 1. For example, the first angle range may be [ -i × θ, i × θ ], so that i angles respectively corresponding to the first i times and the last i times of detection of the reference angle by the reference detector can be obtained according to the first angle range.

According to the judgment method of the adjacent detector and the judgment method of the adjacent angle of the reference detector, the following detection information can be obtained:

corresponding to the reference angle AG _p+1 And corresponds to the reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At an angle AG _p Is detected (i.e. information point B in fig. 7) ₁₁ ) To at an angle AG _p-i+1 And (not shown in fig. 7), and a detector DE _m+1 At an angle AG _p+2 Probe information (i.e. information point B in fig. 7) ₂₁ ) To at an angle AG _p+i+1 Not shown in fig. 7.

Corresponding to the reference detector DE _m+1 Is adjacent toA detector corresponding to the reference angle AG _p+1 The probe information of (a) may include: detector DE _m At an angle AG _p+1 Probe information (i.e. information point C in fig. 7) ₁₁ ) To the detector DE _m-j+1 At an angle AG _p+1 And a detector DE _m+2 At an angle AG _p+1 Is detected (i.e. information point C in fig. 7) ₂₁ ) To the detector DE _m+j+1 At an angle AG _p+1 The probe information of (1). And the obtained information points are taken as reference points.

For example, when i is 2 and j is 2, as shown in fig. 7, a schematic diagram is drawn for another reference point corresponding to fig. 4, and the information point B can be obtained according to the above-described determination manner of the adjacent detector and the determination manner of the adjacent angle of the reference detector ₁₁ And B ₁₂ Information point B ₂₁ And B ₂₂ Information point C ₁₁ And C ₁₂ And information point C ₂₁ And information point C ₂₂ And is taken as a point A to be determined ₁ To the reference point of (c).

In other embodiments, in conjunction with FIGS. 3 and 4, at detector DE ₁ To the detector DE _m+n+1 In the probe information of (2), finding the corresponding reference angle AG _p+1 And corresponds to the reference detector DE _m+1 And corresponding reference detector DE _m+1 And corresponding to the reference angle AG _p+1 And corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The detection information of the adjacent detector.

According to the determination method of different adjacent detectors and the determination method of different adjacent angles, detection information of different situations and further different reference points are obtained, which is described below by exemplifying several situations, and it should be noted that these examples do not limit other situations that may exist when the embodiments of the present specification are actually applied.

In some cases, the adjacent detectors may be determined in the following manner: one on each side of the reference detector.

For the adjacent angle of the reference detector, the determination method may be as follows: the first angular range spans no less than twice the horizontal angular resolution 2 theta and no more than three times the horizontal angular resolution 3 theta. For example, the first angle range may be [ - θ, θ ], so that the angles corresponding to the previous detection and the subsequent detection of the reference angle by the reference detector can be obtained according to the first angle range.

For the adjacent angle of the adjacent detector, the determination method may be as follows: the second angular range spans no less than two times the horizontal angular resolution 2 theta and no more than three times the horizontal angular resolution 3 theta. For example, the second angle range may be [ - θ, θ ], so that the angles corresponding to the previous detection and the subsequent detection of the reference angle corresponding to the adjacent detectors can be obtained according to the second angle range.

According to the above-mentioned determination method of the adjacent detector and the determination method of the adjacent angle of the reference detector, another reference point selection diagram corresponding to fig. 4 can be obtained as shown in fig. 8, in which:

corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At an angle AG _p Is detected (i.e. information point B in fig. 9) ₁₁ ) And at angle AG _p+2 Is detected (i.e. information point B in fig. 3) ₂₁ )。

Corresponding to the reference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (a) may include: detector DE _m At an angle AG _p+1 Is detected (i.e. information point C in fig. 3) ₁₁ ) And a detector DE _m+2 At an angle AG _p+1 Is detected (i.e. information point C in fig. 3) ₂₁ )

Corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The detection information of the adjacent detector may include: detector DE _m At angle AG _p Probe information (i.e. information point D in fig. 8) ₁₁ ) And inAngle AG _p+2 Is detected (i.e. information point D in fig. 8) ₂₁ ) Detector DE _m+2 At an angle AG _p Is detected (i.e. information point D in fig. 8) ₃₁ ) And at angle AG _p+2 Is detected (i.e. information point D in fig. 8) ₄₁ )。

Then, information point B is added ₁₁ Information point B ₂₁ Information point C ₁₁ Information point C ₂₁ Information point D ₁₁ Information point D ₂₁ Information point D ₃₁ And an information point D ₄₁ As a point A to be determined ₁ To the reference point of (c).

For the adjacent angle of the reference detector, the determination method may be as follows: the first angular range spans no less than four times the horizontal angular resolution 4 theta and no more than five times the horizontal angular resolution 5 theta. For example, the first angle range may be [ -2 θ,2 θ ], so that angles respectively corresponding to the first two detections and the last two detections of the reference angle corresponding to the reference detector can be obtained according to the first angle range.

For the adjacent angle of the adjacent detector, the determination method may be as follows: the second angular range spans no less than two times the horizontal angular resolution 2 theta and no more than three times the horizontal angular resolution 3 theta. For example, the second angle range may be [ - θ, θ ], so that the angles corresponding to the previous detection and the next detection of the reference angle corresponding to the adjacent detectors can be obtained according to the second angle range.

According to the above-mentioned determination method of the adjacent detector and the determination method of the adjacent angle of the reference detector, another reference point selection diagram corresponding to fig. 4 can be obtained as shown in fig. 9, in which:

corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At an angle AG _p-1 Is detected (i.e. information point B in fig. 9) ₁₂ ) At an angle AG _p Is detected (i.e. information point B in fig. 9) ₁₁ ) At angle AG _p+2 Is detected (i.e. information point B in fig. 9) ₂₁ ) And at angle AG _p+3 Is detected (i.e. information point B in fig. 9) ₂₂ )。

Corresponding to the reference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (a) may include: detector DE _m At angle AG _p+1 Is detected (i.e. information point C in fig. 9) ₁₁ ) And a detector DE _m+2 At an angle AG _p+1 Is detected (i.e. information point C in fig. 9) ₂₁ )。

Corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The detection information of the adjacent detector may include: detector DE _m At an angle AG _p Is detected (i.e. information point D in fig. 9) ₁₁ ) And at angle AG _p+2 Is detected (i.e. information point D in fig. 9) ₂₁ ) Detector DE _m+2 At an angle AG _p Is detected (i.e. information point D in fig. 9) ₃₁ ) And at angle AG _p+2 Is detected (i.e. information point D in fig. 9) ₄₁ )。

Then, information point B is added ₁₁ Information point B ₁₂ Information point B ₂₁ Information point B ₂₂ Information point C ₁₁ Information point C ₂₁ Information point D ₁₁ Information point D ₂₁ Information point D ₃₁ And an information point D ₄₁ As a point A to be determined ₁ To the reference point of (c).

In other cases, the adjacent detectors may be determined in the following manner: two detectors on either side of the reference detector.

For adjacent detectors which are positioned at the upper side and the lower side of the reference detector and have no other detector in the middle, the adjacent angle can be determined in the following manner: the second angular range spans no less than twice the horizontal angular resolution 2 theta and no more than three times the horizontal angular resolution 3 theta. For example, the angular range of the adjacent detector may be [ - θ, θ ], so that the angles corresponding to the previous detection and the next detection of the reference angle corresponding to the adjacent detector can be obtained according to the angular range of the adjacent detector.

For adjacent detectors which are positioned at the upper side and the lower side of the reference detector and have a detector spaced in the middle, the adjacent angle can be determined in the following manner: the angular range of the adjacent detectors spans less than twice the horizontal angular resolution 2 theta. For example, the angular extent of the adjacent detectors may be (- θ, θ). Therefore, the adjacent angle cannot be obtained according to the angle range of the adjacent detector, in other words, the detection information of the adjacent detector at the adjacent angle does not need to be searched.

According to the above-mentioned determination method of the adjacent detector and the determination method of the adjacent angle of the reference detector, another reference point selection diagram corresponding to fig. 4 can be obtained as shown in fig. 10, in which:

corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The probe information of (a) may include: detector DE _m+1 At an angle AG _p-1 Is detected (i.e., information point B in fig. 10) ₁₂ ) At angle AG _p Is detected (i.e., information point B in fig. 10) ₁₁ ) At angle AG _p+2 Is detected (i.e., information point B in fig. 10) ₂₁ ) And at angle AG _p+3 Probe information (i.e., information point B in fig. 10) ₂₂ )。

Corresponding to the reference detector DE _m+1 And corresponding to the reference angle AG _p+1 The probe information of (a) may include: detector DE _m At an angle AG _p+1 Is detected (i.e. information point C in fig. 10) ₁₁ ) Detector DE _m-1 At an angle AG _p+1 Is detected (i.e. information point C in fig. 10) ₁₂ ) Detector DE _m+2 At an angle AG _p+1 Is detected (i.e. information point C in fig. 10) ₂₁ ) Detector DE _m+3 At an angle AG _p+1 Probe information (i.e., information point C in fig. 10) ₂₂ )。

Corresponding to the reference angle AG _p+1 And corresponding to said reference detector DE _m+1 The detection information of the adjacent detector may include: detector DE _m At an angle AG _p Is detected (i.e. information point D in fig. 9) ₁₁ ) And at angle AG _p+2 Probe information (i.e. information point D in fig. 9) ₂₁ ) Detector DE _m+2 At an angle AG _p Is detected (i.e. information point D in fig. 9) ₃₁ ) And at angle AG _p+2 Probe information (i.e. information point D in fig. 9) ₄₁ )。

Then, information point B is added ₁₁ Information point B ₁₂ Information point B ₂₁ Information point B ₂₂ Information point C ₁₁ Information point C ₁₂ Information point C ₂₁ Information point C ₂₂ Information point D ₁₁ Information point D ₂₁ Information point D ₃₁ And an information point D ₄₁ As a point A to be determined ₁ To the reference point of (c).

It is understood that, in practical applications, since the detection information of the detector includes the information points and the blank result, there may be a case where the detection information that meets the condition is found, and the reference point cannot be obtained. The embodiments of the present specification do not specifically limit this.

For example, as shown in fig. 11, it is another radar point cloud distribution diagram, and compared with the example shown in fig. 10, the difference is that: detector DE _m+1 At an angle AG _p The detected information of (2) is a blank result. Thereby, even if the detector DE _m+1 At an angle AG _p The detection information of (2) meets the adjacent condition, and no information point can be used as a reference point because the detection information is a blank result.

It is understood that the number of adjacent detectors and the number of adjacent angles can be set according to practical application scenarios and requirements. For example, the parameters such as the number of detection signals, the rotating speed, the resolution ratio and the like of the radar can be set, so that the phenomenon that points are lost due to too few obtained reference points is avoided, the phenomenon that the obtained reference points are too many, the data size is too large, too many computing resources are occupied, and the data processing efficiency is influenced are avoided. The embodiments of the present disclosure do not specifically limit the number of adjacent detectors and the number of adjacent angles.

In some embodiments, if the number of adjacent detectors is small, there may still be a misjudgment. For example, if the resolution of the radar is low, the intervals between the information points in the point cloud may be large, thereby resulting in a weak correlation between the information points.

In particular, reference may be made to fig. 12, which shows a schematic diagram of the arrangement of partial detectors of another radar in three-dimensional space corresponding to the information points acquired by the field of view of fig. 1 a.

Referring to fig. 1a and 12 in combination, when corresponding to the field of view of the radar shown in fig. 1a, the resolution of the radar is low, and the angular span of the utility pole GA in the horizontal direction is small, and TC _X Two adjacent detectors TC _X-1 And TC _X+1 The detection signal reflected by the telegraph pole GA is not detected, only the detection signal reflected by the wall WA is received, and the information point d is obtained ₁ ～d ₄ . At this time, if the adjacent detector is judged by the following method: one detector on each side of the reference detector, i.e. two adjacent detectors, may still be set up as a ₅ Points are misjudged as noise.

In order to reduce the false positive probability, the number of adjacent detectors can be increased, thereby enlarging the source range and diversity of the reference points. For example, the adjacent detectors are determined in the following manner: two detectors on either side of the reference detector, i.e. when four adjacent detectors are set, the useful information point a ₅ When the point to be judged is used as the point to be judged, the source range of the reference point is expanded, the reference points are more diversified, and the overall correlation degree between the point to be judged and the reference points can be enhanced. As shown in FIG. 12, adjacent detectors TC in the radar can be obtained _X-2 Useful information point e of ₁ And adjacent detectionDevice TC _X-2 Useful information point f of ₁ And uses it as a reference point. Due to the reference point e ₁ And f ₁ And a point a to be determined ₅ Has stronger correlation between the useful information points a, thereby reducing the useful information points a ₅ Probability of being misinterpreted as noise.

In a specific implementation, the determining whether the point to be determined is a noise point based on the correlation between the point to be determined and the reference point may include: and determining whether the point to be determined is a noise point according to the point distance between the point to be determined and the reference point.

Specifically, as shown in fig. 13, a flowchart of a noise determination method may include the following steps:

and S4-11, respectively calculating the distance between each reference point and the point to be judged to obtain a point distance data set.

As described above, according to the detector and radar rotation data corresponding to the information point, the three-dimensional angle corresponding to the information point can be obtained. In specific implementation, the distance from the obstacle to the radar can be calculated by using a Time Of Flight (TOF) algorithm according to the Time Of the detection signal Of the transmitter and the Time Of the signal received by the corresponding detector. Therefore, the information point may further include: and the obstacle distance data are used for representing the obstacle distance corresponding to the information point.

And S4-12, determining a point distance threshold value.

S4-13, judging whether the point distance data set has point distance data lower than the point distance threshold value, if so, determining that the point to be judged is not a noise point; otherwise, determining the point to be determined as a noise point.

For example, the point distance dataset is { dis ₁ ，dis ₂ ，dis ₃ ，dis ₄ ，dis ₅ }，dis _k And point distance data between the k-th reference point and the point to be determined is shown, wherein k is 1,2,3,4 and 5. And obtaining a point distance threshold value r based on the corresponding relation between the obstacle distance and the point distance threshold value. Thereby, each dot pitch data dis is judged _k And a dot spacing threshold r if there is at least oneDot pitch data dis _k If not, determining the point to be determined as the noise point.

Therefore, the relevance between the point to be judged and the reference point set is represented by the distance between each reference point in the reference point set and the point to be judged, the reliability of the relevance can be enhanced, the problem of point loss caused by misjudgment is effectively reduced, the accuracy of a noisy point identification result is improved, and the data quality of point cloud is ensured.

Based on the point distance, the point distance threshold of the point to be determined can be determined according to the obstacle distance corresponding to the point to be determined. Specifically, as shown in fig. 14, it is a flowchart of a method for determining a dot pitch threshold, which may include the following steps:

4-121) obtaining the distance of the obstacle corresponding to the point to be determined.

Therefore, the point distance threshold value can be changed according to the distance of the obstacle corresponding to the point to be judged, and the adjustability of the point distance threshold value can improve the accuracy of a noise judgment result.

In specific implementation, due to perspective relation of near, far and small, objects with the same size are farther away from the radar, the information points which can be collected by the radar are fewer, and the relevance between the information points is weaker. Based on this, the point distance threshold value is positively correlated with the distance of the obstacle corresponding to the point to be determined, that is, the larger the distance of the obstacle corresponding to the point to be determined is, the larger the point distance threshold value is. Therefore, a more accurate point distance threshold value is obtained, and a noise point identification result is guaranteed.

For example, as shown in fig. 15, a schematic diagram of a corresponding relationship between a point-to-point threshold and an obstacle distance is shown, and in fig. 15, the point-to-point threshold and the obstacle distance are in a linear relationship, which facilitates to determine the point-to-point threshold quickly.

In practical application, in consideration of noise filtering effect and point loss rate, the point distance threshold and the barrier distance can be in a nonlinear relation, and according to the point cloud requirement, a curve segment with the point distance threshold changing rapidly along with the barrier distance is set in a key information part, so that the probability of misjudgment of useful information is reduced, the point loss rate is reduced, and the key information in the lost point cloud is reduced; in the non-key information part, a curve segment with a point-to-threshold value changing slowly along with the distance of the obstacle is set, and the noise filtering effect is improved.

For example, as shown in fig. 16, another corresponding relationship diagram of the point-to-threshold value and the obstacle distance is shown, and in fig. 16, the point-to-threshold value and the obstacle distance are in a nonlinear relationship. Which may include: the curve section with the point distance threshold value changing fast along with the distance of the obstacle, namely the first part (i) and the third part (i), and the curve section with the point distance threshold value changing slow along with the distance of the obstacle, namely the second part (i).

As can be seen from fig. 16, the key information part corresponds to the obstacle existing in a short distance and the obstacle existing in a long distance, and it is ensured that the key information of the two parts can be beneficial to subsequent data processing analysis; and the non-key information part corresponds to an obstacle in a moderate distance, and the radar can obtain more information in a moderate distance section due to the moderate distance from the radar, and the relevance between information points is strong, so that the noise filtering effect can be improved by slowing down the change amplitude of the point distance threshold.

In specific implementation, the line number of the radar (the number of detection signals transmitted in a single acquisition) and the rotation speed of the radar affect the resolution of the radar, wherein the resolution of the radar is higher as the line number of the radar is larger under the condition that other parameters are unchanged; accordingly, the higher the speed of rotation of the radar, the lower the resolution of the radar, while the other parameters remain unchanged. Further, the lower the resolution of the radar is, the more sparse the information points acquired by the radar are, and at this time, if the point distance threshold is set to be smaller, the situation of misjudgment is easy to occur.

Based on this, in order to obtain a more accurate point distance threshold value and improve the accuracy of the noise point identification result, the point distance threshold value may be inversely related to the resolution of the radar, that is, the lower the resolution of the radar is, the larger the point distance threshold value is. In an implementation, there may be a probability of point loss for each frame of point cloud of the radar, i.e., a point loss rate of the radar. The point distance threshold value is too small, the point loss rate of the radar is increased, the point loss rate of the radar can be calculated for considering the point loss rate and the noise point identification effect of the radar, and the point distance threshold value is improved when the point loss rate of the radar is higher than the preset point loss rate threshold value. The method for improving the point distance threshold value may be to modify a corresponding relationship curve between the point distance threshold value and the obstacle distance, or to obtain the increased point distance threshold value by superimposing a correction amount after obtaining the point distance threshold value according to the corresponding relationship curve between the point distance threshold value and the obstacle distance.

It can be understood that, after the data processing method provided by the embodiment of the present specification is executed to perform noise point identification, the radar point loss rate calculation may be executed; or the radar missing point rate calculation may be performed first, and then the data processing method provided in the embodiments of the present specification may be performed. The present specification does not specifically limit the execution order of the two.

In a specific implementation, the determining whether the point to be determined is a noise point based on the correlation between the point to be determined and the reference point may include: and determining whether the point to be determined is a noise point according to the number of the points of the reference point. Specifically, counting the number of points of the reference point, judging whether the number of points of the reference point is lower than a preset point threshold value, and if so, determining that the point to be judged is a noise point; otherwise, determining that the point to be determined is not a noise point. Wherein the point threshold is a positive integer greater than 2.

Therefore, the correlation between the point to be judged and the reference point set is represented by the point number of the reference point, the calculated amount can be reduced, and the noise point identification efficiency is improved under the condition of ensuring the accuracy of the noise point identification result.

As can be seen from the above description, in a specific implementation, the number of lines of the radar and the rotation speed of the radar affect the resolution of the radar, and the lower the resolution of the radar is, the more sparse the information points acquired by the radar are, that is, the fewer the number of the information points is, at this time, if the value of the point threshold is too large, the noise filtering effect will be deteriorated; if the value of the point threshold is too small, the misjudgment probability is increased, and point loss is caused.

Based on this, in order to obtain a more accurate point threshold and improve the accuracy of the noise identification result, the point threshold may be positively correlated with the resolution of the radar, that is, the lower the resolution of the radar is, the smaller the point threshold is.

In concrete implementation, the threshold value undersize of counting, will increase the lost point rate of radar, in order to compromise lost point rate and the noise point recognition effect of radar, can calculate the lost point rate of radar, when the lost point rate of radar is higher than predetermineeing lost point rate threshold value, improve the threshold value of counting.

In specific implementation, the noise point judgment can be performed by adopting at least one of the point number and the point distance according to the actual application scene and the requirement.

Further, in order to achieve both the processing efficiency and the accuracy of the noise identification result, the noise determination based on the point number may be performed first, and then the noise determination based on the point distance may be performed when the correlation degree of the point number is satisfied.

Because the data volume judged by the noise based on the point number is less than the data volume judged by the noise based on the point distance, if the point to be judged is judged as the noise in the noise judgment based on the point number, the noise judgment based on the point distance can be omitted, so that the operation resource is saved, and the processing efficiency is improved.

Because the accuracy of the noise judgment based on the point number is lower than that of the noise judgment based on the point distance, if the point to be judged is not judged as the noise in the noise judgment based on the point number, the noise judgment based on the point distance can be carried out, and the accuracy of the noise identification result is improved.

In a specific implementation, as shown in fig. 2, after the step S4, the data processing method may further include a step S5 of deleting the point to be determined if the point to be determined is determined to be noise.

And, the noise point judgment operation can be continuously executed for the information points which are not subjected to the noise point judgment. Specifically, referring to fig. 2 or 3, the data processing method may further include the steps of:

s6, judging whether there is information point without noise point judgment, if yes, continuing to step S7;

s7, selecting the point to be determined from the information points without noise determination, and continuing to execute the step S3.

Therefore, noise point identification operation can be carried out on all information points in the point cloud, noise points are effectively filtered, and data quality is improved.

In specific implementation, in order to reduce the amount of calculation and improve the processing efficiency, part of information points may be selected from the point cloud according to actual conditions to perform noise point identification. For example, if the point cloud has a densely distributed information point region and a sparsely distributed information point region, noise point identification can be performed on the densely distributed information point region; for another example, if a detector obtains information points at a plurality of adjacent angles, the information points may be selected at intervals according to a set number of intervals as the points to be determined. The embodiment of the present specification does not specifically limit the selection manner of the partial information points.

It should be noted that fig. 2 and fig. 3 only show the processing operation in the case where there is an information point for which noise is not determined, and in practical applications, there may be no information point for which noise is not determined, and corresponding processing operation is executed in this case, for example, when there is no information point for which noise is not determined, the process may be ended, and after the radar obtains more point clouds, the data processing method described in the embodiment of the present specification may be performed again. This specification does not specifically limit this.

It is to be understood that while various embodiments have been provided in the foregoing description, alternatives to those embodiments described above may be combined, cross-referenced, and so forth without conflict, to extend to the various possible embodiments, all of which are deemed to be disclosed and disclosed herein.

The present specification also provides a data processing apparatus corresponding to the above data processing method, which is described in detail below by specific embodiments with reference to the accompanying drawings. It should be noted that the data processing apparatus described below can be regarded as a functional module provided for implementing the data processing method provided in the present specification; the contents of the data processing apparatus described below may be referred to in correspondence with the contents of the data processing method described above.

In a specific implementation, as shown in fig. 17, which is a block diagram of a data processing device for radar in an embodiment of the present specification, the data processing device M1 may include:

a data storage unit M11 adapted to buffer a point cloud of the radar, the point cloud comprising information points obtained by a plurality of detectors at a plurality of angles;

a noise point identification unit M12, adapted to select a point to be determined from the point cloud; using the detector corresponding to the point to be judged as a reference detector, using the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and using the obtained information point as a reference point; and judging whether the point to be judged is a noise point or not based on the correlation degree of the point to be judged and the reference point.

By adopting the scheme, the reference points comprise the information points of adjacent angles (transverse dimensions) and adjacent detectors (longitudinal dimensions), and the reference points which are related to the points to be judged in the same dimensions of space and time sequence are obtained, so that the reliability and diversity of the reference points can be improved, the accuracy of the identification result can be improved, the misjudgment probability can be reduced, the noise points can be effectively identified and the data quality of point cloud can be guaranteed when the noise points are identified according to the correlation between the points to be judged and the reference points.

In a specific implementation, with continued reference to fig. 17, the noise identification unit M12 is further adapted to find detection information of neighboring detectors corresponding to the reference angle and to the reference detector, and to use the obtained information point as a reference point. For details, reference may be made to the above related contents, which are not described herein again.

Thus, the reference point may further include: when the radar rotates to the adjacent angle of the reference angle, the information points corresponding to the adjacent detectors are obtained through calculation, so that the sources of the reference points are further enriched, and the reliability and diversity of the reference points are improved.

In a specific implementation, the adjacent detector is one or more detectors adjacent to the reference detector in the arrangement position. The adjacent angle is one or more angles of which the angle difference with the reference angle belongs to a preset range. For details, reference may be made to the above related contents, which are not described herein again.

Therefore, reference points from different sources are obtained, the reliability and diversity of the reference points are increased, and spatial and temporal correlation exists between the reference points and the points to be judged.

In a specific implementation, for the detection information of the reference detector, the adjacent angle is an angle whose angle difference from the reference angle belongs to a first range; for the detection information of the adjacent detector, the adjacent angle is an angle whose angle difference from the reference angle belongs to a second range.

Alternatively, the first range may have a larger span than the second range in order to effectively reduce the data amount, since the farther the distance between the detectors is, the larger the difference in the angle of the field of view between the information points acquired by the detectors is, the weaker the correlation between the corresponding information points is. Further, a span difference between the first range and the second range is positively correlated with a field of view angle difference between the detectors.

In a specific implementation, when the adjacent detector includes a plurality of detectors located on one side of the reference detector in the arrangement position, the second range may include a plurality of angular ranges respectively corresponding to the plurality of detectors, and a span of the angular ranges is inversely related to a distance of the adjacent detector to the reference detector in the arrangement position.

In a specific implementation, the correlation between the point to be determined and the reference point includes: and the point distance between the point to be determined and the reference point. For details, reference may be made to the above related contents, which are not described herein again.

In a specific implementation, the correlation between the point to be determined and the reference point includes: the number of points of the reference point. For details, reference may be made to the above related contents, which are not described herein again.

In a specific implementation, with continued reference to fig. 17, the noise recognition unit M12 is further adapted to delete the point to be determined that is determined to be noise. Further, the noise point identifying unit M12 is further adapted to select the point to be determined from information points for which noise point determination is not performed, and perform noise point determination.

The present specification embodiment also provides a radar, as shown in fig. 18, in the present specification embodiment, the radar may include a plurality of transmitters L1 to LX, a plurality of detectors D1 to DY, and a data processing device M2, wherein:

the emitters L1-LX and the detectors D1-DY have corresponding relations; the transmitters L1 to LX are suitable for transmitting detection signals in multiple angles; the detectors D1-DY are suitable for collecting echo signals from multiple angles;

the data processing device M2 is adapted to perform data processing according to the detection signals of the transmitters L1 to LX and the echo signals of the corresponding detectors D1 to DY to obtain corresponding information points, and to perform the method according to any of the embodiments described above.

It should be noted that, when the probe acquires an echo signal, the actually acquired echo signal may include a real echo signal and an interference signal. This specification does not specifically limit this.

In summary, by searching the adjacent angle corresponding to the reference angle and the detection information corresponding to the reference detector and the adjacent detector corresponding to the reference detector and the detection information corresponding to the reference angle, a reference point which is dimensionally associated with a point to be determined in space and time sequence is obtained, so that the reliability and diversity of the reference point are improved, and when noise point identification is performed according to the correlation between the point to be determined and the reference point, the accuracy of an identification result can be improved, the misjudgment probability is reduced, noise points can be effectively identified, and the data quality of point cloud is guaranteed.

In particular implementations, the data processing apparatus may include a memory that may store one or more computer-executable instructions that may be invoked by a processor to perform the steps of the methods provided by embodiments of the present specification.

In particular implementations, the radar may be a laser radar, a millimeter wave radar, or the like.

Embodiments of the present specification further provide a computer-readable storage medium, on which computer instructions are stored, and when the computer instructions are executed, the computer instructions may perform the steps of any one of the above-described method embodiments of the present specification. The computer readable storage medium may be various suitable readable storage media such as an optical disc, a mechanical hard disc, a solid state hard disc, and the like. The instructions stored in the computer-readable storage medium may be used to execute the method according to any of the embodiments, which may specifically refer to the embodiments described above and will not be described again.

The computer-readable storage medium may include, for example, any suitable type of memory unit, memory device, memory article, memory medium, storage device, storage article, storage medium and/or storage unit, for example, memory, removable or non-removable media, erasable or non-erasable media, writeable or re-writeable media, digital or analog media, hard disk, floppy disk, compact disk read Only memory (CD-ROM), compact disk recordable (CD-R), compact disk Rewriteable (CD-RW), optical disk, magnetic media, magneto-optical media, removable memory cards or disks, various types of Digital Versatile Disk (DVD), a tape, a cassette, or the like.

The computer instructions may include any suitable type of code, such as source code, compiled code, interpreted code, executable code, static code, dynamic code, encrypted code, and the like, implemented using any suitable high-level, low-level, object-oriented, visual, compiled and/or interpreted programming language.

It is noted that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic may be included in at least one implementation of the specification. Also, in the description of the present specification, the terms "first", "second", and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, a feature defined as "first," "second," etc. may explicitly or implicitly include one or more of the feature. Moreover, the terms "first," "second," and the like are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the specification described herein are, for example, capable of operation in sequences other than those illustrated or otherwise described herein.

Although the embodiments of the present specification are disclosed above, the present specification is not limited thereto. Various changes and modifications can be made by one skilled in the art without departing from the spirit and scope of the present disclosure, and therefore, the scope of the present disclosure should be determined only by the appended claims.

Claims

1. A data processing method for a radar, characterized in that the data processing method comprises the steps of:

2) selecting a point to be judged from the point cloud;

3) using the detector corresponding to the point to be judged as a reference detector, using the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and using the obtained information point as a reference point;

2. The data processing method for radar according to claim 1, further comprising, before the step 4), the steps of:

3. The data processing method for radar according to claim 1 or 2, wherein the adjacent detector is one or more detectors adjacent to the reference detector in the arrangement position;

the adjacent angle is one or more angles of which the angle difference with the reference angle is within a preset range.

4. The data processing method for radar according to claim 1 or 2, wherein the adjacent angle is an angle whose angle difference from the reference angle belongs to a first range, with respect to the detection information of the reference detector;

5. The data processing method for radar according to claim 4, wherein a span difference between the first range and the second range is positively correlated with a field-of-view angle difference between the detectors.

6. The data processing method for radar according to claim 4, wherein when the adjacent detector includes a plurality of detectors located on a side of the reference detector in the arrangement position, the second range includes a plurality of angular ranges respectively corresponding to the plurality of detectors, and a span of the angular ranges is inversely related to a distance of the adjacent detector to the reference detector in the arrangement position.

7. The data processing method for radar according to claim 1 or 2, wherein the step 4) includes:

8. The data processing method for radar according to claim 7, wherein the step 4-1) includes the steps of:

4-12) determining a point distance threshold value;

4-13) judging whether point distance data lower than the point distance threshold exist in the point distance data set, if so, determining that the point to be judged is not a noise point; otherwise, determining the point to be determined as a noise point.

9. The data processing method for radar according to claim 8, wherein the step 4-12) includes the steps of:

10. The data processing method for radar according to claim 8, wherein the threshold value of the point-to-point distance is positively correlated with an obstacle distance corresponding to the point to be determined, and the threshold value of the point-to-point distance is negatively correlated with a resolution of the radar.

11. The data processing method for a radar according to claim 8, wherein the point distance threshold is increased when a point loss rate of the radar is higher than a preset point loss rate threshold.

12. The data processing method for radar according to claim 1 or 2, wherein the step 4) includes:

4-2) determining whether the point to be judged is a noise point according to the point number of the reference point.

13. The data processing method for radar according to claim 12, wherein the step 4-2) includes the steps of:

4-21) judging whether the number of points of the reference point is lower than a preset point threshold value, if so, determining that the point to be judged is a noise point; otherwise, determining that the point to be determined is not a noise point.

14. The data processing method for radar of claim 13, wherein the point threshold is positively correlated with a resolution of the radar.

15. The data processing method for a radar according to claim 13, wherein the point number threshold is increased when the point loss rate of the radar is higher than a preset point loss rate threshold.

16. The data processing method for radar according to claim 1 or 2, characterized by further comprising:

17. The data processing method for radar according to claim 1 or 2, characterized by further comprising:

18. A data processing device for a radar, comprising:

the noise point identification unit is suitable for selecting a point to be judged from the point cloud; using the detector corresponding to the point to be judged as a reference detector, using the angle corresponding to the point to be judged as a reference angle, searching detection information corresponding to the adjacent angle of the reference angle and the reference detector, detection information corresponding to the adjacent detector of the reference detector and the reference angle, and using the obtained information point as a reference point; and judging whether the point to be judged is a noise point or not based on the correlation degree between the point to be judged and the reference point.

19. The data processing apparatus for radar according to claim 18, wherein the noise point identifying unit is further adapted to find detection information of adjacent detectors corresponding to adjacent angles of the reference angle and corresponding to the reference detector, and to use the obtained information point as a reference point.

20. The data processing apparatus for radar according to claim 18 or 19, wherein the adjacent detector is one or more detectors adjacent to the reference detector in the arrangement position;

21. The data processing apparatus for radar according to claim 18 or 19, wherein the adjacent angle is an angle whose angle difference from the reference angle belongs to a first range, with respect to the detection information of the reference detector;

22. The data processing apparatus for radar according to claim 18 or 19, wherein the degree of correlation of the point to be determined and the reference point includes: and the point distance between the point to be determined and the reference point.

23. The data processing apparatus for radar according to claim 18 or 19, wherein the degree of correlation of the point to be determined and the reference point includes: the number of points of the reference point.

24. The data processing apparatus for radar according to claim 18 or 19, wherein the noise point identifying unit is further adapted to delete a point to be determined which is determined as noise point.

25. The data processing apparatus for radar according to claim 18 or 19, wherein the noise point identifying unit is further adapted to select the point to be determined from information points for which noise point determination is not made, and to make noise point determination.

26. A radar comprising a plurality of transmitters, a plurality of detectors and a data processing device, wherein:

the data processing device is adapted to perform data processing according to the detection signal of the transmitter and the echo signal of the corresponding detector to obtain a corresponding information point, and perform the method according to any one of claims 1 to 17.

27. A radar as claimed in claim 26, wherein the radar is a lidar.

28. A computer readable storage medium having computer instructions stored thereon for performing the steps of the method of any one of claims 1 to 17 when the computer instructions are executed.