Obstacle detection method and device and unmanned vehicle

Abstract

The invention provides an obstacle detection method and device and an unmanned vehicle. The method comprises the steps of collecting point cloud data of a road in front of the unmanned vehicle; according to the point cloud data, determining a scanning point corresponding to a first scanning beam of the ground scanned by the preset single-line laser; according to the scanning position of each scanning point, determining a first scanning point with a step in the scanning position; according to the first scanning point, determining whether an obstacle exists on the front road; and a corresponding driving strategy is determined under the condition that the existence of the obstacle is determined. According to the invention, the point cloud data of the road ahead is collected, the scanning points of the preset single-line laser for scanning the ground are determined, and whether the step phenomenon exists between the scanning points is determined under the view angle of the single-line laser scanning beam, so that the obstacle is accurately detected according to the typical characteristics of the isolated obstacle, and the driving strategy is determined in time; and the transportation operation can be efficiently and safely carried out.

Description

Technical Field

[0001] The present disclosure relates to the field of unmanned driving technology, and in particular to an obstacle detection method, device and unmanned vehicle. Background Art

[0002] In the earthwork freight operation environment of mines and construction, there are often fallen rocks rolling down from high places and scattered materials on the road during transportation. Such isolated obstacles such as fallen rocks are extremely difficult to find and identify in the dusty construction environment. When vehicles run over such obstacles, they may cause safety risks such as tire blowouts and rollovers. In order to ensure the safety of transportation operations, a certain proportion of auxiliary vehicles are usually deployed on site for inspection. Their main task is to promptly detect and clean up these fallen rocks and bulk materials to improve the road transportation environment. However, when the mining area is large, auxiliary vehicles are often unable to patrol such obstacles in time, resulting in delayed cleaning work and low overall efficiency. Summary of the invention

[0003] The embodiments of the present disclosure provide a method and device for detecting an obstacle, and an unmanned vehicle, so as to solve the existing problem that isolated obstacles cannot be accurately detected.

[0004] Based on the above problems, in a first aspect, an embodiment of the present disclosure provides a method for detecting an obstacle, including:

[0005] Collect point cloud data of the road ahead of the unmanned vehicle;

[0006] Determine, based on the point cloud data, a scanning point corresponding to a first scanning line beam of a preset single-line laser scanning on the ground;

[0007] According to the scanning positions of the scanning points, determining the first scanning point where the scanning position has a step;

[0008] Determine whether there is an obstacle on the road ahead according to the first scanning point; and determine a corresponding driving strategy if it is determined that there is an obstacle.

[0009] In combination with the first aspect, in a possible implementation manner, determining, according to the scanning positions of the scanning points, the first scanning point having a step in the scanning position includes:

[0010] Traversing the scanning positions of the scanning points according to a preset step length;

[0011] For each traversal, determining whether the discrete degree of the scanning position of the scanning point of the traversal is greater than a first threshold;

[0012] When the value is greater than the first threshold, for the scan points of the current traversal, the distance between the scan position of each scan point and the scan position of the adjacent previous scan point is determined in sequence;

[0013] A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

[0014] In combination with the first aspect, in a possible implementation manner, determining, according to the scanning positions of the scanning points, the first scanning point having a step in the scanning position includes:

[0015] According to the scanning position of each scanning point, the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in sequence;

[0016] A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

[0017] In combination with the first aspect, in a possible implementation manner, determining whether there is an obstacle on the road ahead according to the first scanning point includes:

[0018] Determine each of two consecutive first scanning points in the traversal order, and determine whether the step directions of the first scanning point pair relative to the adjacent previous scanning point are opposite;

[0019] If the step direction is opposite, it is determined that there is an obstacle between the first scanning point pair on the road ahead.

[0020] In combination with the first aspect, in a possible implementation manner, if the step direction is opposite, the method further includes:

[0021] Determine a second scanning point whose scanning position is located between the scanning positions corresponding to the first scanning point pair;

[0022] The determining that there is an obstacle between the first scanning point pair on the road ahead includes:

[0023] When there is no scanning position step at the second scanning point, it is determined that there is an obstacle on the road ahead between the first scanning point pair.

[0024] In combination with the first aspect, in a possible implementation manner, the second threshold corresponding to the road ahead is determined in the following manner:

[0025] When the relative slope between the road ahead and the road where the unmanned vehicle is located is zero, a second threshold value corresponding to the road ahead is determined according to a geometric relationship between the height of the preset single-line laser emission position relative to the ground, the height of the preset obstacle, the intersection position of the first scanning line beam trajectory extending after passing through the top of the preset obstacle and the ground, and the position of the unmanned vehicle;

[0026] When the relative slope between the road ahead and the road where the unmanned vehicle is located is not zero, the second threshold corresponding to the road ahead is determined based on the geometric relationship between the first scanning beam trajectory, the preset obstacle height, the relative slope of the road ahead relative to the road where the unmanned vehicle is located, and the distance from the unmanned vehicle to the starting point of the relative slope.

[0027] In combination with the first aspect, in a possible implementation manner, when it is determined that an obstacle exists, the method further includes:

[0028] Determine the information of the obstacle, including:

[0029] For a single-line laser radar, mapping the first scanning point pair to a global viewing space to determine the spatial distribution posture of the obstacle; and determining the information of the obstacle;

[0030] For the multi-line laser radar, respectively determine the third scanning point corresponding to the second scanning line beam of at least one other laser line in the multi-line laser except the preset single-line laser for scanning the ground;

[0031] For each laser line of the at least one other laser line, determining a third scanning point pair having a step in the third scanning points corresponding to the laser line;

[0032] Determine, according to the index of the first scanning point pair in the point cloud data, a third scanning point pair corresponding to the position of the first scanning point pair;

[0033] The first scanning point pair and the third scanning point pair corresponding to the position are mapped to the global viewing angle space to determine the spatial distribution posture of the obstacle; and the information of the obstacle is determined.

[0034] In combination with the first aspect, in a possible implementation manner, the method further includes:

[0035] Determine a preset number of second point cloud frames after the first point cloud frame to which the scanning point corresponding to the first scanning line beam belongs;

[0036] For each second point cloud frame, determining a fourth scanning point corresponding to the preset single-line laser in the second point cloud frame; and based on the scanning position of the corresponding fourth scanning point in the second point cloud frame, determining whether an obstacle corresponding to the position of the first scanning point pair is determined based on the second point cloud frame;

[0037] When obstacles corresponding to the positions of the first scanning point pairs are determined based on each second point cloud frame, it is determined that there are obstacles on the road ahead.

[0038] In combination with the first aspect, in a possible implementation manner, when it is determined that an obstacle exists, the method further includes:

[0039] Notify upstream vehicles to re-plan their routes, including:

[0040] Notify the vehicles within a first preset range upstream to re-plan their driving routes based on the information of the obstacle; and / or

[0041] If an obstacle is identified, this also includes:

[0042] Dispatching auxiliary vehicles to clear the obstacles includes:

[0043] If there is an auxiliary vehicle within a second preset distance range of the obstacle, dispatching the auxiliary vehicle to clear the obstacle;

[0044] When there is no auxiliary vehicle within the second preset distance range of the obstacle, the cloud platform is notified to dispatch an auxiliary vehicle to clear the obstacle.

[0045] In combination with the first aspect, in a possible implementation manner, when it is determined that there is an obstacle, determining a corresponding driving strategy includes:

[0046] When it is determined that there is an obstacle, a driving strategy is determined to control the unmanned vehicle to ride or detour to avoid the obstacle.

[0047] In a second aspect, a device for detecting an obstacle is provided, comprising:

[0048] LiDAR module, used to collect point cloud data of the road ahead of the unmanned vehicle;

[0049] The obstacle detection module is used to determine the scanning point corresponding to the first scanning line beam of the preset single-line laser scanning on the ground according to the point cloud data; determine the first scanning point where there is a step in the scanning position according to the scanning position of each scanning point; determine whether there is an obstacle on the road ahead according to the first scanning point; and determine the corresponding driving strategy when it is determined that there is an obstacle.

[0050] In a third aspect, an unmanned vehicle is provided, comprising: an obstacle detection device as described in the second aspect.

[0051] The beneficial effects of the embodiments of the present disclosure include:

[0052] The present disclosure provides an obstacle detection method, device and unmanned vehicle, comprising: collecting point cloud data of a road ahead of the unmanned vehicle; determining, based on the point cloud data, a scanning point corresponding to a first scanning line beam of a preset single-line laser scanning the ground; determining, based on the scanning positions of each scanning point, a first scanning point where a step exists in the scanning position; determining, based on the first scanning point, whether there is an obstacle on the road ahead; and determining a corresponding driving strategy when it is determined that there is an obstacle. An obstacle detection method provided by the embodiment of the present disclosure utilizes the principle of rectilinear propagation of light during laser scanning, and determines the scanning points corresponding to the scanning beam of a preset single-line laser scanning the ground by collecting point cloud data of the road ahead. When there are isolated obstacles such as fallen rocks and bulk materials on the road ahead of the unmanned vehicle, the propagation path of the laser cannot reach the ground due to the obstruction of the obstacle. From the perspective of the single-line laser scanning beam, the position point where the laser reaches the ground and the position point where the laser reaches the obstacle after being blocked by the obstacle will not be flush, and the position point blocked by the obstacle will have a step compared to the position point that normally reaches the ground. Based on the typical characteristics of isolated obstacles, the embodiment of the present disclosure accurately detects the obstacle by determining whether there is a step phenomenon between each scanning point, and promptly determines the driving strategy to ensure that the transportation operation can be carried out efficiently and safely. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flowchart of a method for detecting an obstacle provided in an embodiment of the present disclosure;

[0054] Figure 2 One of the schematic diagrams of scanning points corresponding to the first scanning line beam provided in an embodiment of the present disclosure;

[0055] Figure 3 A second schematic diagram of scanning points corresponding to the first scanning line beam provided in an embodiment of the present disclosure;

[0056] Figure 4 One of the schematic diagrams of the road features ahead of the unmanned vehicle provided in the embodiment of the present disclosure;

[0057] Figure 5 A second schematic diagram of the road features ahead of the unmanned vehicle provided in an embodiment of the present disclosure;

[0058] Figure 6 A flowchart of determining obstacle information provided by an embodiment of the present disclosure;

[0059] Figure 7 A schematic diagram of a vehicle dispatching strategy provided by an embodiment of the present disclosure;

[0060] Figure 8 A schematic diagram of the structure of an obstacle detection device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0061] The embodiments of the present disclosure provide an obstacle detection method, device and unmanned vehicle. The preferred embodiments of the present disclosure are described below in conjunction with the drawings of the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure, and are not used to limit the present disclosure. In addition, the embodiments and features in the embodiments of the present application can be combined with each other if there is no conflict.

[0062] The present disclosure provides a method for detecting obstacles. Figure 1 As shown, the following steps are included:

[0063] S101, collecting point cloud data of the road ahead of the unmanned vehicle;

[0064] S102, determining, according to the point cloud data, a scanning point corresponding to a first scanning line beam of a preset single-line laser scanning the ground;

[0065] S103, determining, according to the scanning positions of the scanning points, a first scanning point having a step in the scanning position;

[0066] S104: Determine whether there is an obstacle on the road ahead according to the first scanning point; and if it is determined that there is an obstacle, determine a corresponding driving strategy.

[0067] The disclosed embodiments are applicable to both urban road scenarios and non-urban road scenarios, such as mines, construction sites, etc. In particular, in earthwork freight operation scenarios in mines and construction, there are often falling rocks and bulk materials. Such isolated obstacles such as falling rocks are extremely difficult to detect and identify in a dusty construction environment. If vehicles run over them while ignoring the obstacles, they may cause safety risks such as tire blowouts and rollovers. In actual operation scenarios, a certain proportion of auxiliary vehicles are usually equipped on site, and their main task is to clean up these fallen rocks and bulk materials to improve the quality of the project. However, when the operating area of ​​the mining area is large and there are many operating vehicles, the auxiliary vehicles are often unable to obtain obstacle information in a timely manner, resulting in a delay in the cleaning work and low overall efficiency. In order to accurately detect such isolated obstacles, LiDAR (Light Detection and Ranging) can be used. LiDAR is widely used in the autonomous driving industry. The working principle of automotive LiDAR draws on the TOF (Time of Flight) measurement method used by bats for ranging. It outputs point cloud data by analyzing the reflected energy size, amplitude, frequency, and phase of the reflected spectrum on the surface of the target object. Point cloud data can be a massive set of points that express the spatial distribution of the target object and the surface characteristics of the target object in the same spatial reference system. The point cloud data collected based on the working principle of LiDAR can include information such as the three-dimensional coordinates of the target object and the laser reflection intensity. In order to accurately detect and identify isolated obstacles such as falling rocks, how to analyze and process these point cloud data is the key.

[0068] In the disclosed embodiment, first, a laser radar can be used to collect point cloud data of the road ahead of the unmanned vehicle. The laser radar can be a single-line laser radar or a multi-line laser radar. For a single-line laser radar, it includes a single laser emitter inside, and the single laser emitter can rotate inside the laser radar. Each time it rotates a small angle, the laser is emitted once. After patrolling the preset angle, the point cloud data scanned by the single-line laser radar can be regarded as a row of dot matrix within a certain height range, which is used to describe linear information. For a multi-line laser radar, it includes multiple laser emitters inside (for ease of understanding, it can be regarded as multiple single-line laser radar laser emitters placed vertically), and multiple laser emitters can be rotated. After each laser emitter rotates the preset angle, it obtains a row of dot matrixes at its corresponding height. Multiple rows of dot matrixes at different heights can form planar information within a certain height range. Therefore, compared with a single-line laser radar, the point cloud data of the target object obtained by using a multi-line laser radar is usually richer and the accuracy is usually higher. Taking a 32-line laser radar as an example, each time the 32 laser emitters rotate a preset angle, it can be considered that a scan is completed. Each laser emitter emits n lasers in one scan, and each laser emission will generate a laser scanning line beam. The laser scanning line beam propagates from the emission point to the emission direction and scans the object in front (the ground or obstacles, etc.) to form a scanning point. Each laser emitter generates n scanning points in one scan, so 32 laser emitters can generate 32n scanning points in one scan. These scanning point information (for example: index information (index in 32n points), coordinate information, etc.) will be stored as point cloud data. In one possible implementation, the scanning position information of the scanning point can be obtained by projecting the point cloud coordinates, for example: projecting the point cloud coordinates onto the ground to obtain the scanning position information of the scanning point; in another possible implementation, the scanning position information of the scanning point can be represented by the three-dimensional coordinates of the scanning point.

[0069] In the disclosed embodiment, after the laser radar is used to collect the point cloud data of the road ahead of the unmanned vehicle, the point cloud data is processed to determine the scanning point corresponding to the first scanning line beam of the preset single-line laser scanning the ground. The preset single-line laser can be a laser emitted by any laser emitter in the laser radar. For example, for a single-line laser radar, the preset single-line laser can be a laser emitted by a single laser emitter, and for a multi-line laser radar, the preset single-line laser can be a laser emitted by any one of the multiple laser emitters, or the laser emitted by the topmost laser emitter among the multiple laser emitters can be used as the preset single-line laser, so that a longer distance and a larger range can be scanned.

[0070] Furthermore, the first scanning line beam may be a scanning line beam formed by a laser emitted when a preset single-line laser scans the ground, and the scanning point corresponding to the first scanning line beam may be determined according to the stored scanning point information corresponding to the preset single-line laser. Figure 2 As shown, the black square represents a scanning point corresponding to the first scanning line beam of a laser emitted by a laser transmitter and scanning the ground. Figure 2 is a top view of the scanning point corresponding to the first scanning line beam, Figure 2 The scanning position of the middle scanning point can be obtained by projecting the scanning point coordinates onto the ground. Among the scanning points corresponding to the first scanning beam, the scanning positions of the scanning points on the ground are usually arranged evenly and regularly. However, for isolated obstacles such as fallen rocks and bulk materials, the scanning points of the first scanning beam cannot reach the ground due to the blocking of the laser propagation path by these isolated obstacles. Instead, the scanning points are generated when they reach the isolated obstacles. This makes it so that under the perspective of the single-line laser scanning beam, the position point where the first scanning beam reaches the ground and the position point where it reaches the obstacle after being blocked by the obstacle are not flush. The position point blocked by the obstacle will have a step compared to the position point that normally reaches the ground (that is, its position will be retreated to a certain extent compared to the position point that reaches the ground). Based on this feature, it is possible to detect whether there are isolated obstacles such as fallen rocks on the road ahead.

[0071] Furthermore, the presence of a step in the scanning position of a scanning point may refer to a sudden change or discontinuity in the scanning point corresponding to the first scanning line beam, and this change appears as an obvious jump in several dimensions of the three-dimensional coordinates (such as the x, y, and z axes). A step phenomenon may occur when encountering different objects or different surfaces during the laser radar scanning process. For example, when the laser scans a plane (such as the ground), the coordinates of the scanning points obtained will show a continuous and relatively gentle characteristic, but when the laser scans from one plane (such as the ground) to another different plane (such as an obstacle), one or more dimensions of the coordinates will show an obvious step change. Then, based on the scanning position of each scanning point, determine the first scanning point where there is a step in the scanning position. For example, if Figure 2 As shown, position A and position D represent the scanning positions when the laser scans the ground, position B and position C represent the scanning positions when the laser scans an obstacle, and position A and position D are located on both sides of the obstacle. If there is a step between the position of point B and the position of point A, and there is a step between the position of point D and the position of point C, then point B and point D can be determined as the first scanning point.

[0072] Further, according to the first scanning point, it is determined whether there is an obstacle on the road ahead. If it is determined that there is an obstacle, a corresponding driving strategy is determined. For example, the vehicle avoids the obstacle by riding or detouring, and notifies the upstream vehicle to avoid it in time or notifies the auxiliary vehicle to clear the obstacle in time.

[0073] The embodiment of the present application analyzes and processes the point cloud data of the road ahead in combination with the typical characteristics of isolated obstacles such as falling rocks, and determines whether there are steps between each scanning point from the perspective of a single-line laser scanning beam, and then determines whether there are obstacles. When it is determined that there are obstacles, the driving strategy is determined in a timely manner to ensure that the transportation operation can be carried out efficiently and safely.

[0074] In another embodiment of the present disclosure, in the above step S103, determining the first scanning point where the scanning position has a step according to the scanning positions of each scanning point includes the following steps:

[0075] Step 1: traverse the scanning position of each scanning point according to the preset step length;

[0076] Step 2: for each traversal, determine whether the discrete degree of the scanning position of the scanning point of the traversal is greater than a first threshold;

[0077] Step 3: When the value is greater than the first threshold, for the scan points traversed this time, the distance between the scan position of each scan point and the scan position of the adjacent previous scan point is determined in sequence;

[0078] Step 4: Determine a scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold as the first scanning point having a step.

[0079] In the embodiment of the present disclosure, in order to make the determination of the first scanning point more efficient, the scanning points that may have steps in a small range may be determined first, and then the first scanning point may be further determined for the scanning points in the small range. For the above step 1, the scanning positions of each scanning point are traversed according to the preset step length. When implemented, a sliding window detector may be constructed to traverse the scanning positions of each scanning point according to the preset step length within the detection window of the sliding window detector. For example, Figure 3 As shown, a sliding window detector 300 is constructed, and within the detection window of the sliding window detector 300, the scanning position of each scanning point is traversed along the x direction according to a preset step size.

[0080] For the above step 2, for each traversal, determine whether the discrete degree of the scanning position of the scanning point of the traversal is greater than the first threshold value, for example, traverse the scanning positions of the scanning points in the window of the above sliding window detector 300, perform preliminary screening on the scanning points, and determine whether the discrete degree of the scanning positions of these scanning points is greater than the first threshold value. The calculation method for characterizing the discrete degree of the scanning position of the scanning point may include: calculating the average value and / or variance of the scanning position coordinates. The discrete degree may be calculated by the average value and / or variance of the scanning point position coordinates, or by other formulas characterizing the discrete degree. In a possible implementation, the average value may be calculated as the average value of one (one dimension, such as the coordinate dimension that may produce a step) or multiple dimensions (X, Y, Z), respectively, and the average value may be regarded as the "center point" or "center of mass" of these scanning points, and the variance measures the average value of the square of the difference between the scanning point position coordinates and their average value, and the variance reflects the degree of dispersion of the scanning point position in space. The larger the variance, the more dispersed the scanning point position is in the corresponding dimension. By determining whether the discrete degree of the scanning positions of the scanning points traversed this time is greater than the first threshold, the scanning points can be screened more quickly, thereby improving the efficiency of obstacle detection.

[0081] For the above step 3, after the initial screening, the scanning points with a discrete degree greater than the first threshold are obtained, and then the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in turn. Figure 3 As shown, the detection window of the sliding window detector 300 includes five scanning points E, F, A, B, and G. Then, the distance between point E and the scanning position of the previous scanning point, the distance between point F and point E, the distance between point A and point F, and so on can be determined in sequence. For the above step 4, the scanning point whose distance to the scanning position of the adjacent previous scanning point is greater than the second threshold is determined as the first scanning point with a step. For example, when traversing to point B, the distance between point B and point A is greater than the second threshold, then point B is determined as the first scanning point with a step. When traversing to point D, the distance between point D and point C is greater than the second threshold, then point D is determined as the first scanning point with a step.

[0082] In another embodiment of the present disclosure, in the above step S103, determining the first scanning point where the scanning position has a step according to the scanning positions of each scanning point includes the following steps:

[0083] Step 1: According to the scanning position of each scanning point, the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in sequence;

[0084] Step 2: Determine a scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold as the first scanning point with a step.

[0085] In the embodiment of the present disclosure, the first scanning point can be determined directly by comparing the scanning positions of adjacent scanning points. For the above step 1, among the scanning points corresponding to the first scanning line beam of the preset single-line laser scanning on the ground, the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in sequence, for example, Figure 2 As shown, the distance between point F and point E is determined in turn, the distance between point A and point F is determined in turn, and so on. For the above step 2, the scanning point whose distance from the scanning position of the adjacent previous scanning point is greater than the second threshold is determined as the first scanning point with a step. For example, if the distance between point B and point A is greater than the second threshold and the distance between point D and point C is greater than the second threshold, then point B and point D can be determined as the first scanning point with a step.

[0086] In another embodiment of the present disclosure, in the above step S104, determining whether there is an obstacle on the road ahead according to the first scanning point includes the following steps:

[0087] Step 1: Determine two consecutive first scanning points in the traversal order, and determine whether the step directions of the first scanning point pair relative to the adjacent previous scanning point are opposite;

[0088] Step 2: If the step direction is opposite, it is determined that there is an obstacle between the first scanning point pair on the road ahead.

[0089] In the embodiment of the present disclosure, for the above step 1, two consecutive first scanning points are determined in the traversal order, and it is determined whether the step directions of the first scanning point pair composed of the two first scanning points are opposite to the adjacent previous scanning points. The step direction can be the vector direction from the position coordinates of the previous scanning point adjacent to the first scanning point to the position coordinates of the first scanning point. Whether the step directions are opposite can be determined by whether the signs of the above vectors are opposite. For example, Figure 2 As shown, two consecutive first scanning points B and D are determined in the traversal order, and it is determined whether the step directions of the first scanning point pair composed of points B and D are opposite to their adjacent previous scanning points A and C respectively.

[0090] Regarding the above step 2, for isolated obstacles such as falling rocks, during the scanning process, the laser scans from the ground to the obstacle and from the obstacle to the ground, and the position of the scanning point will have two steps in opposite directions. Therefore, if the step directions are opposite, it is determined that there is an obstacle between the first scanning point pair on the road ahead. For example, Figure 2 As shown, the step directions of the first scanning point pair composed of point B and point D relative to their adjacent previous scanning points A and C are AB and CD respectively. At this time, the step directions are opposite, so it is determined that there is an obstacle between point B and point D.

[0091] In addition, the traversal order in the embodiment of the present disclosure can be the traversal order of the above-mentioned "step one, traverse the scanning position of each scanning point according to the preset step size", and the step direction is judged in the process of determining the first scanning point. A separate process can also be used to determine the first scanning point pair in the traversal order corresponding to the process.

[0092] In another embodiment of the present disclosure, if the step direction is opposite, the method further includes:

[0093] Determine a second scanning point whose scanning position is located between the scanning positions corresponding to the first scanning point pair;

[0094] In the above step 2, determining that there is an obstacle between the first scanning point pair on the road ahead includes:

[0095] When there is no scanning position step at the second scanning point, it is determined that there is an obstacle on the road ahead between the first scanning point pair.

[0096] In the embodiment of the present disclosure, although the shape of the obstacle may be irregular, the obstacle itself usually does not produce a step, but a step will be produced between the obstacle and the ground. In order to further confirm that there is an obstacle between the first scanning point pair, the step condition of the second scanning point between the first scanning point pair can be further judged. During implementation, if the step direction is opposite, it is determined that the scanning position is located at the second scanning point between the scanning positions corresponding to the first scanning point pair, for example, Figure 2 As shown, the scanning point between the first scanning point pair B and D is determined as the second scanning point. There is no scanning position step between the second scanning points. At this time, the second scanning point represents the surface of the obstacle scanned by the laser, and it is determined that there is an obstacle between the first scanning point pair on the road ahead.

[0097] In another embodiment of the present disclosure, the second threshold corresponding to the road ahead is determined in the following manner:

[0098] Step 1: When the relative slope between the road ahead and the road where the unmanned vehicle is located is zero, determine the second threshold corresponding to the road ahead according to the geometric relationship between the height of the preset single-line laser emission position relative to the ground, the preset obstacle height, the intersection position of the first scanning line beam trajectory after passing through the top of the preset obstacle and the ground, and the position of the unmanned vehicle;

[0099] Step 2: When the relative slope between the road ahead and the road where the unmanned vehicle is located is not zero, determine the second threshold corresponding to the road ahead based on the geometric relationship among the first scanning beam trajectory, the preset obstacle height, the relative slope of the road ahead relative to the road where the unmanned vehicle is located, and the distance from the unmanned vehicle to the starting point of the relative slope.

[0100] In the embodiment of the present disclosure, when determining the first scanning point, the distance between the scanning position of the current scanning point and the scanning position of the adjacent previous scanning point is compared with the second threshold value, and whether there is a step between the two scanning points is determined according to the second threshold value, thereby determining whether there is an obstacle. It can be seen that the second threshold value is related to the predefined obstacle height, and because the road terrain is complex, especially for mine roads, obstacles of the same height may have different conclusions on whether there is a step between the two scanning points when the terrain of the road ahead is flat and when the terrain of the road ahead is uphill, that is, the conclusion on whether there is an obstacle on the road ahead may be different. Therefore, when determining the second threshold value, it is necessary to consider the preset obstacle height information and the terrain information of the road ahead.

[0101] For step 1 above, if Figure 4 As shown, the vehicle is moving forward. When the relative slope between the road ahead and the road where the unmanned vehicle is located is zero, a rectangular coordinate system is established with the position of the unmanned vehicle as the coordinate origin. The height of the single-line laser emission position F relative to the ground is preset as h, the preset obstacle height is CE, and the first scanning beam trajectory FC, after FC passes through the preset obstacle top point C, extends to the intersection position with the ground as point A. According to the geometric relationship of the above parameters, the length of the second threshold corresponding to the road ahead is determined to be AC, where AC = AF × (CE / h).

[0102] For step 2, Figure 5 As shown, the vehicle is traveling forward. When the relative slope between the road ahead and the road where the unmanned vehicle is located is not zero, a rectangular coordinate system is established with the position of the unmanned vehicle as the coordinate origin. According to the first scanning beam trajectory, the vertical angle of the beam trajectory is determined to be pitch, and the slope of the beam trajectory is k1=tan(pitch). The first scanning beam trajectory FC, FC passes through the preset obstacle top point C and extends to the intersection with the ground as point A, the preset obstacle height is CE, the relative slope of the road ahead relative to the road where the unmanned vehicle is located is alpha, the slope of the slope is k2=tan(alpha), and the distance from the unmanned vehicle to the starting point D of the relative slope is d. According to the geometric relationship of the above parameters, the length of the second threshold corresponding to the road ahead is determined to be AC. For the convenience of calculation, point B is determined as the equivalent projection point of the top C of the obstacle CE at the bottom of the slope. Then AC=AB×(CE / BD), where the coordinates of point A are A(x_A,y_A), and the coordinates of point B are B(x_B,y_B),

[0103] x_A = (k2×d+h) / (k2-k1);

[0104] y_A=k1×(k2×d+h) / (k2-k1)+h;

[0105] x_B = (d-k2×h) / (k1×k2+1);

[0106] y_B=k1×(d-k2×h) / (k1×k2+1)+h.

[0107] In another embodiment of the present disclosure, when it is determined that there is an obstacle, the method further includes:

[0108] Information to identify obstacles, including:

[0109] Step 1: For a single-line laser radar, map the first scanning point pair to the global viewing space to determine the spatial distribution posture of the obstacle; and determine the information of the obstacle;

[0110] Step 2: for the multi-line laser radar, respectively determine the third scanning point corresponding to the second scanning line beam of at least one other laser line in the multi-line laser except the preset single-line laser scanning the ground;

[0111] Step 3: for each laser line of at least one other laser line, determine a third scanning point pair having a step in the third scanning point corresponding to the laser line;

[0112] Step 4: Determine a third scanning point pair corresponding to the position of the first scanning point pair according to the index of the first scanning point pair in the point cloud data;

[0113] Step 5: Map the first scanning point pair and the third scanning point pair corresponding to the position to the global viewing space to determine the spatial distribution posture of the obstacle; and determine the information of the obstacle.

[0114] In the disclosed embodiment, the main difference between a single-line laser radar and a multi-line laser radar is the number of laser beams. A single-line laser radar only emits one laser beam at a time, while a multi-line laser radar emits multiple laser beams at the same time. Compared with a single-line laser radar, a multi-line laser radar has a higher scanning accuracy. Therefore, there is no conflict in the process of determining obstacle information for a single-line laser radar and a multi-line laser radar. The following describes the information for determining obstacles for a single-line laser radar and the information for determining obstacles for a multi-line laser radar respectively:

[0115] For the above step 1, for a single-line laser radar, first determine the scanning point corresponding to the first scanning line beam of the preset single-line laser scanning on the ground, then determine the first scanning point pair with a step in the scanning point, and if there is an obstacle between the first scanning point pair, map the first scanning point pair to the global perspective space. In a possible implementation, the mapping to the global perspective space can be based on the point cloud bird's eye view (BEV) principle. The principle of the point cloud bird's eye view is to project the three-dimensional point cloud data onto a two-dimensional plane to obtain a bird's eye view of the environment, and visualize the projected two-dimensional point cloud data to determine the spatial distribution posture of the obstacle. Finally, convert these point cloud data representing the obstacle into a standard obstacle information format to determine the obstacle information, and the obstacle information includes the location information of the obstacle and the spatial distribution posture information of the obstacle, such as the height and width of the obstacle.

[0116] For the above steps 2 to 5, for the multi-line laser radar, first determine the scanning point corresponding to the first scanning line beam of the preset single-line laser scanning the ground, then determine the first scanning point pair with a step in the scanning point, and if it is determined that there is an obstacle between the first scanning point pair, store the index of the first scanning point pair in the point cloud data. After that, check whether there is a scanning point pair corresponding to the first scanning point pair in at least one other laser line in the multi-line laser except the preset single-line laser, so as to increase the probability of the existence of obstacles or improve the obstacle information (the number and position of at least one other laser line can be determined according to actual needs and are not limited here).

[0117] like Figure 6 As shown, in one possible implementation, each laser line in at least one other laser line is traversed to determine the third scanning point corresponding to the second scanning line beam of the laser scanning the ground, and to determine whether there is a third scanning point pair with a step in the third scanning point. When it is determined that there is a third scanning point pair with a step, the index of the third scanning point pair in the point cloud data can be stored until the third scanning point corresponding to a laser line in at least one other laser line is traversed to determine whether there is no third scanning point pair with a step. In another possible implementation, it is not necessary to traverse the at least one other laser line to determine whether there is a corresponding third scanning point pair, but to manually specify at least one laser line for which it is necessary to determine whether there is a corresponding third scanning point pair.

[0118] In order to ensure that the first scanning point pair and the third scanning point pair represent the same obstacle, the third scanning point pair corresponding to the position of the first scanning point pair can be determined according to the index of the first scanning point pair in the point cloud data, and summarized. The first scanning point pair and the third scanning point pair corresponding to the position are mapped to the global perspective space to determine the spatial distribution posture of the obstacle and the information of the obstacle. The method here is the same as the method of determining obstacle information by a single-line laser radar, and will not be repeated here.

[0119] It should be particularly noted that, in a possible implementation, for lasers other than the preset single-line laser, if there is no step third scanning point in the third scanning point corresponding to the second scanning line beam of the laser scanning the ground, it does not affect the determination of the spatial distribution posture of the obstacle and the obstacle information. The third scanning points corresponding to the second scanning line beam of the laser scanning the ground can assist in determining whether there is an obstacle, or can only participate in the determination of the obstacle information.

[0120] In another embodiment of the present disclosure, the following steps are also included:

[0121] Step 1, determining a preset number of second point cloud frames after the first point cloud frame to which the scanning point corresponding to the first scanning line beam belongs;

[0122] Step 2: for each second point cloud frame, determine the fourth scanning point corresponding to the preset single-line laser in the second point cloud frame; and based on the scanning position of the corresponding fourth scanning point in the second point cloud frame, determine whether an obstacle corresponding to the position of the first scanning point pair is determined based on the second point cloud frame;

[0123] Step 3: When obstacles corresponding to the positions of the first scanning point pairs are determined based on each second point cloud frame, it is determined that there are obstacles on the road ahead.

[0124] In the disclosed embodiment, in order to avoid false detection due to noise, the presence of obstacles on the road ahead can be confirmed by comparing multiple point cloud frames. When the first scanning point pair is determined in the first point cloud frame and there is an obstacle before the first scanning point pair, a preset number of second point cloud frames after each first point cloud frame are used to determine whether there is an obstacle corresponding to the position of the first scanning point pair. When obstacles corresponding to the position of the first scanning point pair are determined based on each second point cloud frame, it is determined that the same obstacle exists on the road ahead. In this way, the information of the same obstacle determined by multiple point cloud frames is more accurate.

[0125] In another embodiment of the present disclosure, when it is determined that there is an obstacle, the method further includes:

[0126] Step 1: notifying upstream vehicles to re-plan their driving routes, including: notifying upstream vehicles within a first preset range to re-plan their driving routes based on information about the obstacle; and / or

[0127] If an obstacle is identified, this also includes:

[0128] Step 2, dispatching an auxiliary vehicle to clear the obstacle, including: when there is an auxiliary vehicle within a second preset distance range of the obstacle, dispatching the auxiliary vehicle to clear the obstacle; when there is no auxiliary vehicle within the second preset distance range of the obstacle, notifying the cloud platform to dispatch an auxiliary vehicle to clear the obstacle.

[0129] In the embodiment of the present disclosure, for the above step 1, as Figure 7 As shown, when the unmanned vehicle detects an obstacle, the unmanned vehicle can notify the upstream vehicles within the first preset range to re-plan the driving route through vehicle-to-vehicle communication technology (V2V) to avoid the obstacle and avoid being stuck or collided with the obstacle. Vehicle-to-vehicle communication technology is a technology that allows vehicles to exchange information wirelessly to improve road safety and traffic efficiency.

[0130] For the above step 2, if there is an auxiliary vehicle within the second preset distance range of the obstacle, the unmanned vehicle can dispatch the auxiliary vehicle to clear the obstacle through vehicle-to-vehicle communication technology. If there is no auxiliary vehicle within the second preset distance range of the obstacle, the unmanned vehicle can notify the cloud platform to dispatch the auxiliary vehicle to clear the obstacle through 4G / 5G communication technology. The cloud platform can perform a global assessment to determine the relationship between the position of the auxiliary vehicle and the position of the obstacle to improve the efficiency of obstacle clearing.

[0131] In another embodiment of the present disclosure, in the above step S104, when it is determined that there is an obstacle, determining a corresponding driving strategy includes:

[0132] When it is determined that there is an obstacle, a driving strategy is determined to control the unmanned vehicle to ride or detour to avoid the obstacle.

[0133] In the disclosed embodiment, when the unmanned vehicle detects an isolated obstacle such as a falling rock, in order to ensure traffic safety, if there is enough drivable space, the unmanned vehicle's braking system can slow down and take riding or detour actions to avoid the obstacle. If the road is narrow and there is no alternative drivable space, the unmanned vehicle can stop and notify the auxiliary vehicle to clear the obstacle in time.

[0134] Based on the same inventive concept, the embodiments of the present disclosure also provide an obstacle detection device and an unmanned vehicle. Since the principles of the problems solved by these devices and the unmanned vehicle are similar to the aforementioned detection method applied to obstacles, the implementation of the device and the unmanned vehicle can refer to the implementation of the aforementioned method, and the repeated parts will not be repeated.

[0135] An obstacle detection device provided by the embodiment of the present disclosure is as follows: Figure 8 As shown, including:

[0136] The laser radar module 801 is used to collect point cloud data of the road ahead of the unmanned vehicle;

[0137] The obstacle detection module 802 is used to determine the scanning point corresponding to the first scanning line beam of the preset single-line laser scanning on the ground based on the point cloud data; determine the first scanning point where there is a step in the scanning position based on the scanning position of each scanning point; determine whether there is an obstacle on the road ahead based on the first scanning point; and determine the corresponding driving strategy when it is determined that there is an obstacle.

[0138] In another embodiment of the present disclosure, the obstacle detection module 802 is used to determine, according to the scanning positions of each scanning point, a first scanning point where a step exists in the scanning position, including:

[0139] Traverse the scanning position of each scanning point according to the preset step length;

[0140] For each traversal, determining whether the discrete degree of the scanning position of the scanning point of the traversal is greater than a first threshold;

[0141] When the value is greater than the first threshold, for the scan points of the current traversal, the distance between the scan position of each scan point and the scan position of the adjacent previous scan point is determined in sequence;

[0142] A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

[0143] In another embodiment of the present disclosure, the obstacle detection module 802 is used to determine, according to the scanning positions of each scanning point, a first scanning point where a step exists in the scanning position, including:

[0144] According to the scanning position of each scanning point, the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in sequence;

[0145] A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

[0146] In another embodiment of the present disclosure, the obstacle detection module 802 is used to determine whether there is an obstacle on the road ahead according to the first scanning point, including:

[0147] Determine each of two consecutive first scanning points in the traversal order, and determine whether the step directions of the first scanning point pair relative to the adjacent previous scanning point are opposite;

[0148] If the step direction is opposite, it is determined that there is an obstacle between the first scanning point pair on the road ahead.

[0149] In another embodiment of the present disclosure, if the step direction is opposite, the obstacle detection module 802 is further configured to:

[0150] Determine a second scanning point whose scanning position is located between the scanning positions corresponding to the first scanning point pair;

[0151] The obstacle detection module 802 is used to determine whether there is an obstacle between the first scanning point pair on the road ahead, including:

[0152] When there is no scanning position step at the second scanning point, it is determined that there is an obstacle on the road ahead between the first scanning point pair.

[0153] In another embodiment of the present disclosure, the obstacle detection module 802 is used to determine the second threshold corresponding to the road ahead in the following manner:

[0154] When the relative slope between the road ahead and the road where the unmanned vehicle is located is zero, a second threshold value corresponding to the road ahead is determined according to a geometric relationship between the height of a preset single-line laser emission position relative to the ground, a preset obstacle height, a position of an intersection of the first scanning line beam trajectory extending from the top of the preset obstacle and the ground, and the position of the unmanned vehicle;

[0155] When the relative slope between the road ahead and the road where the unmanned vehicle is located is not zero, the second threshold corresponding to the road ahead is determined based on the geometric relationship between the first scanning beam trajectory, the preset obstacle height, the relative slope of the road ahead relative to the road where the unmanned vehicle is located, and the distance from the unmanned vehicle to the starting point of the relative slope.

[0156] In another embodiment of the present disclosure, when it is determined that there is an obstacle, the obstacle detection module 802 is further configured to:

[0157] Information to identify obstacles, including:

[0158] For a single-line laser radar, the first scanning point pair is mapped to the global viewing space to determine the spatial distribution posture of the obstacle; and the information of the obstacle is determined;

[0159] For the multi-line laser radar, respectively determine the third scanning point corresponding to the second scanning line beam of at least one other laser line in the multi-line laser except the preset single-line laser for scanning the ground;

[0160] For each laser line of at least one other laser line, determining a third scanning point pair having a step in the third scanning points corresponding to the laser line;

[0161] Determine, according to the index of the first scanning point pair in the point cloud data, a third scanning point pair corresponding to the position of the first scanning point pair;

[0162] The first scanning point pair and the third scanning point pair corresponding to the position are mapped to the global viewing space to determine the spatial distribution posture of the obstacle; and the information of the obstacle is determined.

[0163] In another embodiment of the present disclosure, the obstacle detection module 802 is further configured to:

[0164] Determine a preset number of second point cloud frames after the first point cloud frame to which the scanning point corresponding to the first scanning line beam belongs;

[0165] For each second point cloud frame, determine a fourth scanning point corresponding to the preset single-line laser in the second point cloud frame; and based on the scanning position of the corresponding fourth scanning point in the second point cloud frame, determine whether an obstacle corresponding to the position of the first scanning point pair is determined based on the second point cloud frame;

[0166] When obstacles corresponding to the positions of the first scanning point pairs are determined based on each of the second point cloud frames, it is determined that there are obstacles on the road ahead.

[0167] In another embodiment of the present disclosure, when it is determined that there is an obstacle, the obstacle detection module 802 is further configured to:

[0168] Notifying upstream vehicles to re-plan their driving routes, including: notifying upstream vehicles within a first preset range to re-plan their driving routes based on information about the obstacle; and / or

[0169] When it is determined that there is an obstacle, the obstacle detection module 802 is further configured to:

[0170] Dispatching an auxiliary vehicle to clear an obstacle includes: when there is an auxiliary vehicle within a second preset distance range of the obstacle, dispatching the auxiliary vehicle to clear the obstacle; when there is no auxiliary vehicle within the second preset distance range of the obstacle, notifying a cloud platform to dispatch an auxiliary vehicle to clear the obstacle.

[0171] In another embodiment of the present disclosure, the obstacle detection module 802 is used to determine a corresponding driving strategy when it is determined that an obstacle exists, including:

[0172] When it is determined that there is an obstacle, a driving strategy is determined to control the unmanned vehicle to ride or detour to avoid the obstacle.

[0173] An unmanned vehicle provided by an embodiment of the present disclosure includes: an obstacle detection device as described in any of the above embodiments.

[0174] Through the description of the above implementation methods, those skilled in the art can clearly understand that the embodiments of the present disclosure can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present disclosure.

[0175] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present disclosure.

[0176] Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description of the embodiment, or can be changed accordingly and located in one or more devices different from the present embodiment. The modules in the above embodiment can be combined into one module, or can be further divided into multiple sub-modules.

[0177] The serial numbers of the above-mentioned embodiments of the present disclosure are only for description and do not represent the advantages or disadvantages of the embodiments.

[0178] Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A method for detecting an obstacle, characterized in that: include: Collect point cloud data of the road ahead of the unmanned vehicle; Determine, based on the point cloud data, a scanning point corresponding to a first scanning line beam of a preset single-line laser scanning on the ground; According to the scanning positions of the scanning points, determining the first scanning point where the scanning position has a step; Determining whether there is an obstacle on the road ahead according to the first scanning point; And when it is determined that there are obstacles, the corresponding driving strategy is determined.

2. The method according to claim 1, characterized in that The step of determining the first scanning point having a step in the scanning position according to the scanning positions of the scanning points comprises: Traversing the scanning positions of the scanning points according to a preset step length; For each traversal, determining whether the discrete degree of the scanning position of the scanning point of the traversal is greater than a first threshold; When the value is greater than the first threshold, for the scan points of the current traversal, the distance between the scan position of each scan point and the scan position of the adjacent previous scan point is determined in sequence; A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

3. The method according to claim 1, characterized in that The step of determining the first scanning point having a step in the scanning position according to the scanning positions of the scanning points comprises: According to the scanning position of each scanning point, the distance between the scanning position of each scanning point and the scanning position of the adjacent previous scanning point is determined in sequence; A scanning point whose distance from a scanning position of an adjacent previous scanning point is greater than a second threshold is determined as a first scanning point with a step.

4. The method according to any one of claims 1 to 3, characterized in that: The determining, based on the first scanning point, whether there is an obstacle on the road ahead includes: Determine each of two consecutive first scanning points in the traversal order, and determine whether the step directions of the first scanning point pair relative to the adjacent previous scanning point are opposite; If the step direction is opposite, it is determined that there is an obstacle between the first scanning point pair on the road ahead.

5. The method according to claim 4, characterized in that If the step direction is opposite, the method further includes: Determine a second scanning point whose scanning position is located between the scanning positions corresponding to the first scanning point pair; The determining that there is an obstacle between the first scanning point pair on the road ahead includes: When there is no scanning position step at the second scanning point, it is determined that there is an obstacle on the road ahead between the first scanning point pair.

6. The method according to claim 2 or 3, characterized in that: The second threshold corresponding to the road ahead is determined in the following manner: When the relative slope between the road ahead and the road where the unmanned vehicle is located is zero, a second threshold value corresponding to the road ahead is determined according to a geometric relationship between the height of the preset single-line laser emission position relative to the ground, the height of the preset obstacle, the intersection position of the first scanning line beam trajectory extending after passing through the top of the preset obstacle and the ground, and the position of the unmanned vehicle; When the relative slope between the road ahead and the road where the unmanned vehicle is located is not zero, the second threshold corresponding to the road ahead is determined based on the geometric relationship between the first scanning beam trajectory, the preset obstacle height, the relative slope of the road ahead relative to the road where the unmanned vehicle is located, and the distance from the unmanned vehicle to the starting point of the relative slope.

7. The method according to claim 4, characterized in that If an obstacle is identified, this also includes: Determine the information of the obstacle, including: For a single-line laser radar, mapping the first scanning point pair to a global viewing space to determine the spatial distribution posture of the obstacle; and determining the information of the obstacle; For the multi-line laser radar, respectively determine the third scanning point corresponding to the second scanning line beam of at least one other laser line in the multi-line laser except the preset single-line laser for scanning the ground; For each laser line of the at least one other laser line, determining a third scanning point pair having a step in the third scanning points corresponding to the laser line; Determine, according to the index of the first scanning point pair in the point cloud data, a third scanning point pair corresponding to the position of the first scanning point pair; The first scanning point pair and the third scanning point pair corresponding to the position are mapped to the global viewing angle space to determine the spatial distribution posture of the obstacle; and the information of the obstacle is determined.

8. The method according to claim 4, characterized in that Also includes: Determine a preset number of second point cloud frames after the first point cloud frame to which the scanning point corresponding to the first scanning line beam belongs; For each second point cloud frame, determining a fourth scanning point corresponding to the preset single-line laser in the second point cloud frame; and based on the scanning position of the corresponding fourth scanning point in the second point cloud frame, determining whether an obstacle corresponding to the position of the first scanning point pair is determined based on the second point cloud frame; When obstacles corresponding to the positions of the first scanning point pairs are determined based on each second point cloud frame, it is determined that there are obstacles on the road ahead.

9. The method according to claim 1, characterized in that If an obstacle is identified, this also includes: Notify upstream vehicles to re-plan their routes, including: Notify the vehicles within a first preset range upstream to re-plan their driving routes based on the information of the obstacle; and / or If an obstacle is identified, this also includes: Dispatching auxiliary vehicles to clear the obstacles includes: If there is an auxiliary vehicle within a second preset distance range of the obstacle, dispatching the auxiliary vehicle to clear the obstacle; When there is no auxiliary vehicle within the second preset distance range of the obstacle, the cloud platform is notified to dispatch an auxiliary vehicle to clear the obstacle.

10. The method according to claim 1, characterized in that The step of determining a corresponding driving strategy when determining that an obstacle exists includes: When it is determined that there is an obstacle, a driving strategy is determined to control the unmanned vehicle to ride or detour to avoid the obstacle.

11. An obstacle detection device, characterized in that: include: LiDAR module, used to collect point cloud data of the road ahead of the unmanned vehicle; The obstacle detection module is used to determine, based on the point cloud data, a scanning point corresponding to a first scanning line beam of a preset single-line laser scanning the ground; determine, based on the scanning positions of each scanning point, a first scanning point where a step exists in the scanning position; and determine, based on the first scanning point, whether there is an obstacle on the road ahead; And when it is determined that there are obstacles, the corresponding driving strategy is determined.

12. An unmanned vehicle, characterized in that: A device for detecting obstacles comprising the obstacle as claimed in claim 11.

Images