CN117629147B - Obstacle detection method, cloud control platform and unmanned vehicle - Google Patents
Obstacle detection method, cloud control platform and unmanned vehicle Download PDFInfo
- Publication number
- CN117629147B CN117629147B CN202410104206.2A CN202410104206A CN117629147B CN 117629147 B CN117629147 B CN 117629147B CN 202410104206 A CN202410104206 A CN 202410104206A CN 117629147 B CN117629147 B CN 117629147B
- Authority
- CN
- China
- Prior art keywords
- height
- grid
- fitting
- determining
- preset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000008447 perception Effects 0.000 claims abstract description 11
- 238000010845 search algorithm Methods 0.000 claims description 15
- 230000007613 environmental effect Effects 0.000 abstract description 4
- 238000004364 calculation method Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012795 verification Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000004888 barrier function Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 238000012876 topography Methods 0.000 description 2
- 238000006563 Carroll rearrangement reaction Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 125000006850 spacer group Chemical group 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Abstract
The embodiment of the disclosure provides a detection method of an obstacle, a cloud control platform and an unmanned vehicle, wherein the method comprises the following steps: under the condition that the sensing system senses that the ground height h0 is larger than a preset threshold value, projecting the sensed ground point cloud data to a planar grid map; determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0; judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; and according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area, fitting the true value of the height of the ground area corresponding to h0, and determining that the obstacle exists in the ground area corresponding to h0 under the condition that the true value is larger than the preset height value. The reliability of environmental perception in the unmanned system is improved, and the driving safety is improved.
Description
Technical Field
The disclosure relates to the technical field of automatic driving, in particular to a detection method of an obstacle, a cloud control platform and an unmanned vehicle.
Background
A grid map is a grid-based map representation method in which a map area is divided into grid cells, and specific attribute information is assigned to each grid cell. Each grid cell may represent different geographic features, types of terrain, altitude, etc.
Grid maps are commonly used in the field of autopilot technology. The automatic driving perception system adopts a preset mode to perform ground fitting, and realizes accurate modeling of ground topology by analyzing interaction between laser beams and the ground. Subsequently, the height of the potential obstacle is calculated by measuring the difference in height between the laser beam and the grid area floor model. These altitude data are further applied to an altitude threshold analysis to determine if an impending obstacle is present to enable a feasibility assessment of the environment. When the height difference exceeds a preset threshold value, the existence of the obstacle in the area can be confirmed, and the area is marked as an impenetrable area, so that the basis for the detection of the obstacle and the path planning is provided.
The ground fitting also shows the deficiency under specific circumstances, especially in soft foundation, non-pavement road surface, such as open mine scene, the ground fitting equation suffers from distortion, the height measurement error is obviously increased, and the situation of false alarm of obstacles is easily caused. In conventional approaches, attempts have been made to solve this problem by reducing the grid size to increase the grid resolution, but this has led to an exponential increase in computational resource requirements, limited to small scale sensing tasks. Another attempt is to use a surface fitting method, however, this strategy is prone to distortion of the ground environment and may even cause missed detection of obstacles due to difficulty in effectively adapting to complex terrain, which constitutes a potential risk to unmanned vehicle safety.
Disclosure of Invention
The embodiment of the disclosure provides a detection method of an obstacle, a cloud control platform and an unmanned vehicle, which are used for solving the problem that the accuracy of obstacle detection by means of a built grid map is low in the prior art.
Based on the above-mentioned problems, in a first aspect, an embodiment of the present disclosure provides a method for detecting an obstacle, including:
under the condition that the sensing system senses that the ground height h0 is larger than a preset threshold value, projecting the sensed ground point cloud data to a planar grid map;
determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0;
judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; and is combined with
And according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area, fitting a true value of the height of the ground area corresponding to h0, and determining that the obstacle exists in the ground area corresponding to h0 under the condition that the true value is larger than the preset height value.
In combination with the first aspect, in one embodiment provided by the present disclosure, determining a fitting height of at least two layers of grid regions includes: determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0 and a second fitting height h16 of a second grid region adjacent to the first grid region; judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including: judging whether h0, h8 and h16 meet the following conditions: h0> h8> h16, and dh8< α1× (dh 8+dh 16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh8 characterizes the height difference between h0 and h 8; dh16 characterizes the height difference between h8 and h16; α1 characterizes a preset value.
In combination with the first aspect, in one embodiment provided by the present disclosure, determining a fitting height of at least two layers of grid regions includes: determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0, a second fitting height h16 of a second grid region adjacent to the first grid region, and a third fitting height h1 of a third grid region adjacent to the second grid region; judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including: judging whether h0, h16 and h1 meet the following conditions: h0> h16> h1, and dh16< α2× (dh1+dh16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh16 characterizes the height difference between h8 and h 16; dh1 characterizes the height difference between h0 and h1; α2 characterizes a preset value.
In combination with the first aspect, in one embodiment provided by the present disclosure, determining a fitting height of at least two layers of grid regions includes: determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0 and a third fitting height h1 of a third grid region spaced one layer of grid from the first grid region; judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including: judging whether h0, h8 and h1 meet the following conditions: h0> h8> h1, and dh8< α3× (dh1+dh8); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh8 characterizes the height difference between h0 and h 8; dh1 characterizes the height difference between h0 and h1; α3 characterizes a preset value.
With reference to the first aspect, in an implementation manner provided in the present disclosure, according to h0 and a fitting height of a preset layer grid region in the multi-layer grid region, the fitting h0 corresponds to a true value of a ground region height, including: determining a difference value between h0 and a second fitting height h16, and determining the difference value as a true value of the fitting height of the ground area corresponding to h 0; the second fitting height h16 is a fitting height of a grid region of a layer of grids spaced from the grid region corresponding to h 0.
With reference to the first aspect, in an embodiment provided in the present disclosure, the preset height value is 30 cm.
With reference to the first aspect, in an implementation manner provided in the present disclosure, according to h0 and a fitting height of a preset layer grid region in the multi-layer grid region, the fitting h0 corresponds to a true value of a ground region height, including: determining a difference value between h0 and a first fitting height h8, and determining the difference value as a true value of the fitting height of the ground area corresponding to h 0; the first fitting height h8 is the fitting height of the grid region adjacent to the grid region corresponding to h 0; the preset height value is less than 30 cm.
In combination with the first aspect, in an embodiment provided by the present disclosure, the h0 corresponding grid region includes a plurality of adjacent grids.
With reference to the first aspect, in an embodiment provided by the present disclosure, the grids included in the grid map have the same size and are 0.2×0.2; or the grids contained in the grid map are different in size.
With reference to the first aspect, in an embodiment provided in the present disclosure, the method further includes: traversing grids of the grid map by adopting a preset search algorithm, and clustering the grids with the preset characteristic targets; merging adjacent grids of the same category according to the clustering result, and determining a grid area corresponding to h0 according to the merged grid area; wherein the preset search algorithm is a breadth-first search algorithm; the preset features include preset size information.
In a second aspect, a cloud control platform is provided, including:
the grid map generation module is used for projecting the perceived ground point cloud data to the planar grid map under the condition that the perceived ground height h0 is larger than a preset threshold value;
the fitting height determining module is used for determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0;
the gentle slope terrain determining module is used for judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
The obstacle determining module is used for determining that the obstacle exists in the ground area corresponding to the h0 according to the h0 and the fitting height of the preset layer grid area in the multi-layer grid area, wherein the fitting h0 corresponds to the real value of the ground area height, and the obstacle exists in the ground area corresponding to the h0 under the condition that the real value is larger than the preset height value.
In a third aspect, there is provided an unmanned vehicle comprising: a perception system and an autopilot system;
the sensing system is used for scanning and detecting the ground and providing a scanning and detecting result to the automatic driving system;
the autopilot system is configured to perform the steps of the first aspect, or the obstacle detection method in combination with any implementation of the first aspect.
The beneficial effects of the embodiment of the disclosure include:
the embodiment of the disclosure provides a detection method of an obstacle, a cloud control platform and an unmanned vehicle, wherein the method comprises the following steps: under the condition that the sensing system senses that the ground height h0 is larger than a preset threshold value, projecting the sensed ground point cloud data to a planar grid map; determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0; judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; and according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area, fitting the true value of the height of the ground area corresponding to h0, and determining that the obstacle exists in the ground area corresponding to h0 under the condition that the true value is larger than the preset height value. According to the obstacle detection method provided by the embodiment of the disclosure, under the condition that the ground height h0 is perceived to be larger than the preset threshold value, fitting heights of at least two layers of grid areas around the grid area corresponding to h0 are detected to determine whether the terrain around the ground area corresponding to h0 is a gentle slope, if so, fitting is further performed on the real height of the ground area corresponding to h0, and whether an obstacle exists in the ground area corresponding to h0 is determined according to the real height. According to the embodiment of the disclosure, the multi-layer height calculation method is adopted to acquire the height information of the terrain at a plurality of abstract levels, so that the change of the terrain features is more comprehensively described, more height information is provided for terrain analysis and application, the judgment of the obstacle is more accurate, and the construction of a high-precision ground model is realized under the complex terrain condition, so that the false alarm rate is reduced, the reliability of environmental perception in an unmanned system is improved, and the driving safety is increased.
Drawings
Fig. 1 is a flowchart of a method for detecting an obstacle according to an embodiment of the disclosure;
FIG. 2 is a schematic diagram of a grid map provided by an embodiment of the present disclosure;
fig. 3 is a schematic diagram of an application scenario of an obstacle detection method according to an embodiment of the disclosure;
FIG. 4 is a schematic diagram of an application scenario of an obstacle detection method according to an embodiment of the disclosure;
fig. 5 is a schematic structural diagram of a cloud control platform according to an embodiment of the present disclosure.
Detailed Description
The embodiments of the present disclosure provide a method for detecting an obstacle, a cloud control platform and an unmanned vehicle, and hereinafter, preferred embodiments of the present disclosure will be described with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present disclosure only, and are not intended to limit the present disclosure. And embodiments and features of embodiments in this application may be combined with each other without conflict.
An embodiment of the present disclosure provides a method for detecting an obstacle, as shown in fig. 1, including:
s101, under the condition that a sensing system senses that the ground height h0 is larger than a preset threshold value, projecting sensed ground point cloud data to a planar grid map;
s102, determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0;
S103, judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, go to step S104;
s104, determining the ground area corresponding to the multi-layer grid area as a gentle slope;
s105, fitting a true value of the height of the ground area corresponding to h0 according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area, and determining that the obstacle exists in the ground area corresponding to h0 under the condition that the true value is larger than the preset height value.
The obstacle detection method provided by the embodiment of the disclosure can be applied to a cloud control platform for performing terrain analysis and also can be applied to a vehicle (unmanned vehicle or manned vehicle) with an automatic driving or auxiliary driving function for performing environment sensing. When the method is applied to a cloud control platform, a vehicle can acquire target ground data through a sensing system (such as a camera, a laser radar, a millimeter wave radar, an ultrasonic radar and the like), and the cloud control platform analyzes terrain according to the acquired data by executing the obstacle detection method provided by the disclosure to determine obstacle information of the target ground. When the method is applied to a vehicle, the surrounding environment can be sensed in real time through a sensing system (such as a camera, a laser radar, a millimeter wave radar, an ultrasonic radar and the like) configured by the vehicle, and an automatic driving system can be used for analyzing the environmental data sensed by the sensing system in real time by executing the obstacle detection method provided by the disclosure so as to determine the obstacle in the current environment.
In the embodiment of the disclosure, ground point cloud data can be obtained by scanning the ground through a laser radar in a vehicle sensing system, and the ground point cloud data is projected to a plane grid area to obtain a grid map under the condition that the ground height h0 is perceived to be larger than a preset threshold value. Due to the high complexity of the topography of soft foundations, non-paved roads (as in surface mine scenes), false detection situations often occur, for example: identifying a soil slope as an obstacle or a portion of an obstacle, or a vehicle sinking into a deeper rut, calculating the depth of the rut as a portion of the height of the obstacle, can result in false detection of the obstacle, reducing the vehicle passing efficiency. In order to reduce false detection of an obstacle, in the embodiment of the present disclosure, under the condition that it is perceived that the ground height h0 is greater than a preset threshold, it is not directly determined that an obstacle exists in a ground area corresponding to h0, but the fitting heights of at least two layers of grid areas around the grid area corresponding to h0 are detected to determine whether the terrain around the ground area corresponding to h0 is a gentle slope, if so, the true height of the ground area corresponding to h0 is calibrated, and whether an obstacle exists in the ground area corresponding to h0 is determined according to the obtained true height; if the gentle slope judgment condition is not satisfied, other processes are entered, and the disclosure is not limited. According to the obstacle detection method provided by the embodiment of the disclosure, the height information is acquired from the plurality of hierarchical areas around the h0 corresponding area, so that the change of the terrain features is more comprehensively described, the judgment of the obstacle is more accurate, the construction of a high-precision ground model is realized under the complex terrain condition, the false alarm rate is reduced, a good foundation is provided for the formulation of the driving strategy, and the driving safety is increased.
Fig. 2 is a schematic diagram of a grid map constructed in an embodiment of the disclosure, and as shown in fig. 2, a grid region denoted by h0 in the middle represents a grid region corresponding to h 0. The one-layer grid region surrounding the h 0-corresponding grid region (the grid region denoted as h8 in the figure, hereinafter referred to as the first grid region) is formed by the grids adjacent to the h 0-corresponding grid region, and further expanded outward, the one-layer grid region surrounding the first grid region (the grid region denoted as h16 in the figure, hereinafter referred to as the second grid region) may be formed by the grids adjacent to the first grid region, and so on, and the one-layer grid region surrounding the second grid region (the grid region denoted as h1 in the figure, hereinafter referred to as the third grid region) may be determined, and the first, second, and third grid regions may be regarded as a multi-layer grid region surrounding the h 0-corresponding grid region.
Further, as shown in fig. 2, for the grid region corresponding to h0, the innermost neighborhood is the first grid region for fitting the height value h8. The second layer neighborhood extends to a second grid region for fitting the height value h16. Empirically, it can be considered that the third layer neighborhood (i.e., the third grid region) is an initial ground fit (here, simply an initial ground, whether it is a real ground or not, and further judgment is needed in a subsequent step) for calculating an initial ground height value h1. It should be noted that, since a certain layer of grid area surrounding the grid area corresponding to h0 may have a certain width and gradient (related to the size of the grid division), the actual height of the layer of grid area cannot be represented by a numerical value, which requires fitting a corresponding height value to each layer of grid area. In this way, the periphery of the grid area corresponding to h0 is divided into a plurality of layers of areas, and the height information can be obtained on different abstraction levels of the topographic data by carrying out multi-layer abstraction neighborhood height calculation.
In order to acquire detailed height information, a height difference between the three layers of heights may be calculated. Firstly, calculating a difference value between h0 and the height h8 of the first grid area, which is denoted as dh8, wherein the calculation mode is h0-h8; then, calculating a difference value between the first grid area height h8 and the second grid area height h16, which is denoted as dh16, and calculating the difference value in a mode of h8-h16; finally, the difference between h0 and the initial ground fitting height h1, denoted dh1, is calculated in a manner of h0-h1. The multi-layer height calculation method can understand the height information of the terrain at a plurality of abstract levels, so that the change of the terrain features is more fully described, and more height information is provided for terrain analysis and application. In the subsequent step, it may be determined whether the multi-layered grid region corresponds to a gentle slope terrain according to the height differences.
The multi-layer grid area is exemplified by three layers, and the number of layers may be determined according to practical situations, and the outermost layer may be determined as the initial ground area, which is not limited herein.
In yet another embodiment provided in the present disclosure, "determining the fitting height of the at least two layers of grid regions" in step S102 may be implemented as the following steps:
Determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0 and a second fitting height h16 of a second grid region adjacent to the first grid region;
step S103' judges whether h0 and the determined at least two fitting heights meet the preset height difference condition; if so, determining that the ground area corresponding to the multi-layer grid area is a gentle slope "may be implemented as:
judging whether h0, h8 and h16 meet the following conditions: h0> h8> h16, and dh8< α1× (dh 8+dh 16); if yes, determining the ground area corresponding to the multi-layer grid area as a gentle slope;
wherein dh8 characterizes the height difference between h0 and h 8; dh16 characterizes the height difference between h8 and h16; α1 characterizes a preset value.
In this embodiment, it is first ensured that h0 is higher than the height of its neighborhood (i.e., the first fitting height h8 of the first grid region), and the first grid region height h8 is higher than the second grid region height h16. In addition, it is also necessary to verify that h0 and the first grid region height difference (h0-h8=dh8) are smaller than the product of the preset value α1 and the sum of dh8 and dh 16. When these conditions are satisfied, it is determined that the ground area corresponding to the above-described multi-layered grid area is a gentle slope terrain.
In yet another embodiment provided in the present disclosure, "determining the fitting height of the at least two layers of grid regions" in step S102 may be implemented as the following steps:
Determining a first fitting height h8 of a first grid region adjacent to the grid region corresponding to h0, a second fitting height h16 of a second grid region adjacent to the first grid region, and a third fitting height h1 of a third grid region adjacent to the second grid region;
step S103' judges whether h0 and the determined at least two fitting heights meet the preset height difference condition; if so, determining that the ground area corresponding to the multi-layer grid area is a gentle slope "may be implemented as:
judging whether h0, h16 and h1 meet the following conditions: h0> h16> h1, and dh16< α2× (dh1+dh16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
wherein dh16 characterizes the height difference between h8 and h 16; dh1 characterizes the height difference between h0 and h1; α2 characterizes a preset value.
In this embodiment, first, h0 is ensured to be higher than a second grid region height h16 of a grid spaced apart from the grid region corresponding to h0 by one layer, and the second grid region height h16 is higher than a third grid region height h1 adjacent thereto. Meanwhile, it is also required to satisfy that the height difference (h8-h16=dh 16) between the first and second grid regions is smaller than the product of the preset value α2 and the sum of the height difference (h0-h1=dh 1) and the height difference (dh 16). When these conditions are satisfied, it is determined that the ground area corresponding to the above-described multi-layered grid area is a gentle slope terrain.
In yet another embodiment provided in the present disclosure, "determining the fitting height of the at least two layers of grid regions" in step S102 may be implemented as the following steps:
determining a first fitting height h8 of a first grid region adjacent to the grid region corresponding to h0 and a third fitting height h1 of a third grid region spaced apart from the first grid region by a layer of grid;
step S103' judges whether h0 and the determined at least two fitting heights meet the preset height difference condition; if so, determining that the ground area corresponding to the multi-layer grid area is a gentle slope "may be implemented as:
judging whether h0, h8 and h1 meet the following conditions: h0> h8> h1, and dh8< α3× (dh1+dh8); if yes, determining the ground area corresponding to the multi-layer grid area as a gentle slope;
wherein dh8 characterizes the height difference between h0 and h 8; dh1 characterizes the height difference between h0 and h1; α3 characterizes a preset value.
In this embodiment, it is first ensured that h0 is higher than the height of its neighborhood (i.e., the first fitting height h8 of the first grid region), and the first grid region height h8 is higher than the third grid region height h1. Furthermore, it is necessary to verify that h0 and the first grid region height difference (h0-h8=dh 8) are smaller than the product of the preset value α3 and the sum of dh1 and dh 8. When these conditions are satisfied, it is determined that the ground area corresponding to the above-described multi-layered grid area is a gentle slope terrain.
The multi-layer grid region conforming to any of the gentle slope terrain judging modes can determine that the corresponding ground region is a gentle slope terrain. In addition, the preset values α1, α2, and α3 in the three modes of determining the gentle slope topography may be set and adjusted according to empirical values, and the three preset values may be the same or different.
In still another embodiment provided by the present disclosure, the step S104 "fitting h0 to the actual value of the ground area height according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area" may be implemented as the following steps:
determining a difference value between the h0 and the second fitting height h16, and determining the difference value as a true value of the fitting h0 corresponding to the ground area height;
the second fitting height h16 is a fitting height of the grid region separated from the grid region corresponding to h0 by a layer of grid.
In this embodiment, after the ground area corresponding to the multi-layer grid area is determined to be a gentle slope terrain, it may be determined that there is a gentle slope around the ground area corresponding to h0, and since the small surmountable obstacle located on the gentle slope may be misjudged as an insurmountable large obstacle located on the flat ground, misjudgment on the obstacle is caused, so that the true height of the ground area corresponding to h0, that is, the true height of the obstacle, may be checked and calculated.
In practice, dh2 may be obtained by subtracting the fitting height h16 to the second grid region of the spacer grid from h 0. This calculation aims to obtain altitude difference information of the terrain for further terrain analysis and environmental awareness to ensure that the vehicle can make appropriate decisions and path planning under different terrain conditions. The elevation verification process provided by the present embodiment is generally used to evaluate the slope and elevation differences of terrain (to avoid elevation errors of identified obstacles caused by rutting or to avoid elevation errors of identified obstacles caused by gentle slopes) to support critical tasks such as navigation, path planning, and decision making.
In yet another embodiment provided by the present disclosure, the preset height value is 30 centimeters.
In this embodiment, for the case where the check height dh2 is determined through h0 and h16, the preset height value (i.e., the actual height threshold of the obstacle) may be set to 30 cm. That is, if the verification height (i.e., the true value of the height of the ground area corresponding to h 0) is greater than 30 cm, it is determined that the obstacle exists in the ground area corresponding to h0, otherwise, it is determined that the obstacle does not exist in the ground area corresponding to h0, or that a smaller obstacle capable of being surmounted exists.
In still another embodiment provided by the present disclosure, the step S104 "fitting h0 to the actual value of the ground area height according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area" may be implemented as the following steps:
determining a difference value between the h0 and the first fitting height h8, and determining the difference value as a true value of the fitting h0 corresponding to the ground area height;
the first fitting height h8 is the fitting height of the grid region adjacent to the grid region corresponding to h 0; the preset height value is less than 30 cm.
In this embodiment, h0, which may be fitted by the height difference between h0 and h8, corresponds to a true value of the ground area height, and the preset height value may be set to be less than 30 cm. The specific setting value can be flexibly adjusted according to the actual situation, and will not be described here again.
Through the correction of the real height of the ground area corresponding to h0, a large number of false alarm obstacle detection results (the corrected height is larger than a preset height value and is marked as an obstacle affecting traffic, otherwise, the corrected height is not considered as an obstacle) can be effectively screened, so that the parking frequency of a vehicle and the frequency of an unmanned vehicle needing manual connection can be obviously reduced. The operation efficiency of the automatic driving system is improved, and obvious economic benefits are brought to operators. By reducing false alarms, the vehicle can run more stably, unnecessary parking and intervention are reduced, and smoothness and continuity of transportation tasks are improved. In addition, reducing the frequency of manual takeover also means reducing the cost of specialized training and scheduling personnel, reducing the management burden on the operator. In a word, by optimizing the obstacle detection and treatment process, more efficient, economical and reliable automatic driving vehicle operation is realized, and remarkable economic benefit is brought.
In yet another embodiment provided by the present disclosure, the h0 corresponding grid region includes a plurality of adjacent grids.
In this embodiment, since the grid size is smaller, in practice, the area of the grid corresponding to h0 corresponding to the ground area is larger, and the area of the grid corresponding to h0 may include a plurality of adjacent grids.
In yet another embodiment provided by the present disclosure, the grids included in the grid map are the same size and 0.2m x 0.2m; or the grid map contains grids of different sizes.
In this embodiment, when dividing the grids of the grid map, the grid sizes may be different. For example, the first grid region and the second grid region may be made to correspond to different actual floor region sizes, and the grid sizes included in the first grid region and the second grid region may be different. In another embodiment, the grid size of the partitions may be made the same for the purpose of conserving computing resources.
In still another embodiment provided by the present disclosure, after the grid map is built, the grid region corresponding to h0 may be determined by:
traversing grids of the grid map by adopting a preset search algorithm, and clustering the grids with the preset characteristic targets;
step two, merging adjacent grids of the same category according to a clustering result, and determining a grid area corresponding to h0 according to the merged grid area;
Wherein the preset search algorithm is breadth-first search algorithm; the preset feature includes preset size information.
In this embodiment, after the grid map is constructed, the obstacle in the grid map may also be identified. The neighbor clustering can be performed on all grids by using a preset search algorithm, optionally, the size of the obstacle can be limited (for example, the obstacle with the length, the width and the height all within 0.5 meter is screened out) during the search, the target objects with preset characteristics are clustered, and a group of well-defined obstacle clusters is formed by gradually expanding and merging adjacent grids, so that the grid area corresponding to h0 is determined. According to the embodiment of the disclosure, the obstacle information with various sizes and shapes can be extracted from a large amount of discrete data with high efficiency, and irrelevant data can be further filtered out by restraining the size of the obstacle, so that the final clustering result has higher interpretability and application value. The preset search algorithm may be a breadth-first search algorithm; the preset feature includes preset size information.
Fig. 3 is one application scenario of the obstacle detection method provided in the embodiment of the present disclosure. As shown in fig. 3, the sensing system of the unmanned vehicle senses that the front ground height h0 is greater than the preset threshold, projects the sensed ground point cloud data to the planar grid map for analysis, and determines the third grid area (i.e., the initial ground) height h1 according to the grid map shown in fig. 1, so that the pre-detected obstacle 301 height is h0—h1=dh 1. If dh1 is directly determined as the height of the obstacle 301, it can be seen that the height of the gentle slope where the obstacle 301 is located is misjudged as a part of the height of the obstacle 301. Therefore, the method for detecting the obstacle provided by the embodiment of the present disclosure needs to further divide the neighborhood of the grid map and analyze the fitting heights (h 0, h8, h16, h 1) of each neighborhood, and finally determine that the ground area corresponding to the multi-layer grid area surrounding the obstacle 301 is the gentle slope 302, and perform the verification calculation on the real height of the ground area corresponding to h0 (i.e., the real height of the obstacle 301), so as to obtain the real height dh2 of the obstacle 301 and the gentle slope plane height h2, so as to determine whether the insurmountable obstacle exists in front according to the verified real height dh 2.
Fig. 4 is a second application scenario of the obstacle detection method according to the embodiment of the disclosure. As shown in fig. 4, since the ground in the mine scene is a non-paved ground, the mine car rolls back and forth to form a very deep rut 401 on the road, and the unmanned mine car running in the rut determines the height of the bottom of the rut as the initial ground height h1, so that the pre-detected obstacle height (or the height of the soil slope formed between the ruts) dh1 contains the depth information of the rut, and the obstacle height is misjudged. Therefore, by adopting the obstacle detection method provided by the embodiment of the disclosure, perceived ground point cloud data is projected to a planar grid map, the grid map shown in fig. 1 is subjected to neighborhood division, fitting heights (h 0, h8, h16 and h 1) of all the neighborhoods are analyzed, finally, initial ground corresponding to a third grid area is determined to be rutting, the corrected true height of h0 corresponding to the ground area is dh2, and the ground height is h2, so that whether an insurmountable obstacle exists in front is determined according to the verified true height dh 2.
Based on the same disclosure concept, the embodiments of the present disclosure further provide a cloud control platform and an unmanned vehicle, and because the principles of the problems solved by the cloud control platform and the unmanned vehicle are similar to the detection method of the obstacle, the implementation of the cloud control platform and the unmanned vehicle can refer to the implementation of the method, and the repetition is omitted.
The embodiment of the disclosure provides a cloud control platform, as shown in fig. 5, including:
the grid map generation module 501 is configured to, when the sensing system senses that the ground height h0 is greater than a preset threshold, project the sensed ground point cloud data to a planar grid map;
a fitting height determining module 502, configured to determine a fitting height of at least two layers of grid areas from a plurality of layers of grid areas surrounding the grid area corresponding to h 0;
a gentle slope terrain determining module 503, configured to determine whether h0 and the determined at least two fitting heights satisfy a preset height difference condition; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
the obstacle determining module 504 is configured to determine, according to h0 and a fitting height of a preset layer grid area in the multi-layer grid area, that h0 corresponds to a real value of a ground area height, and determine that an obstacle exists in the ground area corresponding to h0 if the real value is greater than the preset height value.
In yet another embodiment provided by the present disclosure, a fitting height determining module 502 is configured to determine a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0, and a second fitting height h16 of a second grid region adjacent to the first grid region;
The gentle slope terrain determining module 503 is configured to determine whether h0, h8, and h16 satisfy: h0> h8> h16, and dh8< α1× (dh 8+dh 16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh8 characterizes the height difference between h0 and h 8; dh16 characterizes the height difference between h8 and h 16; α1 characterizes a preset value.
In yet another embodiment provided by the present disclosure, the fitting height determining module 502 is configured to determine a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0, a second fitting height h16 of a second grid region adjacent to the first grid region, and a third fitting height h1 of a third grid region adjacent to the second grid region;
the gentle slope terrain determining module 503 is configured to determine whether h0, h16, and h1 satisfy: h0> h16> h1, and dh16< α2× (dh1+dh16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh16 characterizes the height difference between h8 and h 16; dh1 characterizes the height difference between h0 and h1; α2 characterizes a preset value.
In yet another embodiment provided by the present disclosure, a fitting height determining module 502 is configured to determine a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0, and a third fitting height h1 of a third grid region spaced apart from the first grid region by a layer of grids;
The gentle slope terrain determining module 503 is configured to determine whether h0, h8, and h1 satisfy: h0> h8> h1, and dh8< α3× (dh1+dh8); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; wherein dh8 characterizes the height difference between h0 and h 8; dh1 characterizes the height difference between h0 and h 1; α3 characterizes a preset value.
In yet another embodiment provided by the present disclosure, an obstacle determining module 504 is configured to determine a difference between h0 and the second fitting height h16, and determine the difference as a true value of the fitting h0 corresponding to the ground area height; the second fitting height h16 is a fitting height of a grid region of a layer of grids spaced from the grid region corresponding to h 0.
In yet another embodiment provided by the present disclosure, the preset height value is 30 centimeters.
In yet another embodiment provided by the present disclosure, an obstacle determining module 504 is configured to determine a difference between h0 and the first fitting height h8, and determine the difference as a true value of the fitting h0 corresponding to the ground area height; the first fitting height h8 is the fitting height of the grid region adjacent to the grid region corresponding to h 0; the preset height value is less than 30 cm.
In yet another embodiment provided by the present disclosure, the h0 corresponding grid region includes a plurality of adjacent grids.
In yet another embodiment provided by the present disclosure, the grid map includes grids that are the same size and 0.2m by 0.2m; or the grids contained in the grid map are different in size.
In yet another embodiment provided in the present disclosure, the grid map generating module 501 is further configured to traverse the grids of the grid map by using a preset search algorithm, and cluster the grids with the preset feature objects; merging adjacent grids of the same category according to the clustering result, and determining a grid area corresponding to h0 according to the merged grid area; wherein the preset search algorithm is a breadth-first search algorithm; the preset features include preset size information.
Based on the same technical concept, the embodiment of the application also provides electronic equipment. Including a processor, memory, and a bus. The memory is used for storing execution instructions, and comprises a memory and an external memory; the internal memory is also called an internal memory, and is used for temporarily storing operation data in the processor and data exchanged with an external memory such as a hard disk, the processor exchanges data with the external memory through the internal memory, and when the electronic device operates, the processor and the memory are communicated through a bus, so that the processor executes instructions executed by each module of the cloud control platform.
The disclosed embodiments also provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the obstacle detection method described in the method embodiments above. Wherein the storage medium may be a volatile or nonvolatile computer readable storage medium
The embodiment of the disclosure provides an unmanned vehicle, comprising: a perception system and an autopilot system;
the sensing system is used for scanning and detecting the ground and providing a scanning and detecting result to the automatic driving system;
the automatic driving system is used for executing the steps of the obstacle detection method according to any embodiment of the disclosure.
In the embodiment of the disclosure, the unmanned vehicle can be a mining vehicle with unmanned transportation of a mine, and the drive-by-wire chassis has complete functions and can respond to instructions from an unmanned system, such as driving, braking, steering, lifting and other driving and transportation operation requests.
According to the obstacle detection method, the cloud control platform and the unmanned vehicle, which are provided by the embodiment of the disclosure, for the feature detection of the gentle slope, firstly, detailed information of an actual scene can be provided for a ground fitting model, so that the perception capability of the terrain areas is obviously enhanced. The introduction of such information can be used to make the autopilot system pay more attention to these gentle slope regions by improving the attentiveness mechanism, thereby effectively reducing the probability of false positive obstacles. By detecting gentle slope features, more contextual information is provided, enabling the automated driving system to be more able to distinguish topographical features from potential obstacles. The advanced perception mode is beneficial to the system to more accurately identify the terrain change, reduces the risk of false detection, and improves the safety and reliability of the automatic driving vehicle under the complex terrain condition;
Secondly, aiming at the barrier height verification on soft foundations and uneven roads, the challenge of calculating errors of the barrier height difference caused by a single ground fitting model is overcome, and the accuracy and reliability of barrier perception are obviously improved. The conventional ground fitting model is prone to errors in altitude calculation when faced with uneven road surfaces or soft foundations, which may lead to inaccurate altitude estimation of the obstacle. By adopting the height verification strategy provided by the embodiment of the disclosure, the actual height of the obstacle can be measured more accurately, the accuracy of obstacle perception can be improved no matter how complex the ground condition is, more reliable environment perception capability is provided for the automatic driving vehicle, and the driving safety and feasibility are improved.
From the foregoing description of the embodiments, it will be apparent to those skilled in the art that the embodiments of the present disclosure may be implemented in hardware, or may be implemented in software plus a necessary general purpose hardware platform. Based on such understanding, the technical solution of the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.), and includes several instructions for causing a computer device (may be a personal computer, a server, or a network device, etc.) to perform the method described in the embodiments of the present disclosure.
Those skilled in the art will appreciate that the drawing is merely a schematic illustration of one preferred embodiment and that the modules or flows in the drawing are not necessarily required to practice the present disclosure.
Those skilled in the art will appreciate that modules in an apparatus of an embodiment may be distributed in an apparatus of an embodiment as described in the embodiments, and that corresponding changes may be made in one or more apparatuses different from the present embodiment. The modules of the above embodiments may be combined into one module, or may be further split into a plurality of sub-modules.
The foregoing embodiment numbers of the present disclosure are merely for description and do not represent advantages or disadvantages of the embodiments.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present disclosure without departing from the spirit or scope of the disclosure. Thus, the present disclosure is intended to include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (12)
1. A method for detecting an obstacle, comprising:
under the condition that the sensing system senses that the ground height h0 is larger than a preset threshold value, projecting the sensed ground point cloud data to a planar grid map;
Determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0;
judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope; and is combined with
And according to h0 and the fitting height of the preset layer grid area in the multi-layer grid area, fitting a true value of the height of the ground area corresponding to h0, and determining that the obstacle exists in the ground area corresponding to h0 under the condition that the true value is larger than the preset height value.
2. The method of claim 1, wherein determining a fitting height of at least two layers of grid regions comprises:
determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0 and a second fitting height h16 of a second grid region adjacent to the first grid region;
judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including:
judging whether h0, h8 and h16 meet the following conditions: h0> h8> h16, and dh8< α1× (dh 8+dh 16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
Wherein dh8 characterizes the height difference between h0 and h 8; dh16 characterizes the height difference between h8 and h 16; α1 characterizes a preset value.
3. The method of claim 1, wherein determining a fitting height of at least two layers of grid regions comprises:
determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0, a second fitting height h16 of a second grid region adjacent to the first grid region, and a third fitting height h1 of a third grid region adjacent to the second grid region;
judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including:
judging whether h0, h16 and h1 meet the following conditions: h0> h16> h1, and dh16< α2× (dh1+dh16); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
wherein dh16 characterizes the height difference between h8 and h 16; dh1 characterizes the height difference between h0 and h1; α2 characterizes a preset value.
4. The method of claim 1, wherein determining a fitting height of at least two layers of grid regions comprises:
determining a first fitting height h8 of a first grid region adjacent to a grid region corresponding to h0 and a third fitting height h1 of a third grid region spaced one layer of grid from the first grid region;
Judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope, including:
judging whether h0, h8 and h1 meet the following conditions: h0> h8> h1, and dh8< α3× (dh1+dh8); if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
wherein dh8 characterizes the height difference between h0 and h 8; dh1 characterizes the height difference between h0 and h 1; α3 characterizes a preset value.
5. The method of claim 1, wherein fitting h0 to the true value of the ground area height based on h0 and the fit height of the predetermined one of the multi-layer grid areas, comprises:
determining a difference value between h0 and a second fitting height h16, and determining the difference value as a true value of the fitting height of the ground area corresponding to h 0;
the second fitting height h16 is a fitting height of a grid region of a layer of grids spaced from the grid region corresponding to h 0.
6. The method of any one of claims 1-5, wherein the predetermined height value is 30 cm.
7. The method of claim 1, wherein fitting h0 to the true value of the ground area height based on h0 and the fit height of the predetermined one of the multi-layer grid areas, comprises:
Determining a difference value between h0 and a first fitting height h8, and determining the difference value as a true value of the fitting height of the ground area corresponding to h 0;
the first fitting height h8 is the fitting height of the grid region adjacent to the grid region corresponding to h 0; the preset height value is less than 30 cm.
8. The method of any of claims 1-5, wherein the h0 corresponding grid region comprises a plurality of adjacent grids.
9. The method of any one of claims 1-5, wherein the grid map comprises grids of the same size and 0.2 x 0.2m; or the grids contained in the grid map are different in size.
10. The method of any one of claims 1-5, further comprising:
traversing grids of the grid map by adopting a preset search algorithm, and clustering the grids with the preset characteristic targets;
merging adjacent grids of the same category according to the clustering result, and determining a grid area corresponding to h0 according to the merged grid area;
wherein the preset search algorithm is a breadth-first search algorithm; the preset features include preset size information.
11. The cloud control platform is characterized by comprising:
The grid map generation module is used for projecting the perceived ground point cloud data to the planar grid map under the condition that the perceived ground height h0 is larger than a preset threshold value;
the fitting height determining module is used for determining fitting heights of at least two layers of grid areas from the multi-layer grid areas surrounding the grid area corresponding to h 0;
the gentle slope terrain determining module is used for judging whether h0 and the determined at least two fitting heights meet the preset height difference condition or not; if yes, determining that the ground area corresponding to the multi-layer grid area is a gentle slope;
the obstacle determining module is used for determining that the obstacle exists in the ground area corresponding to the h0 according to the h0 and the fitting height of the preset layer grid area in the multi-layer grid area, wherein the fitting h0 corresponds to the real value of the ground area height, and the obstacle exists in the ground area corresponding to the h0 under the condition that the real value is larger than the preset height value.
12. An unmanned vehicle, comprising: a perception system and an autopilot system;
the sensing system is used for scanning and detecting the ground and providing a scanning and detecting result to the automatic driving system;
the automated driving system for performing the steps of the obstacle detection method according to any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410104206.2A CN117629147B (en) | 2024-01-25 | 2024-01-25 | Obstacle detection method, cloud control platform and unmanned vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410104206.2A CN117629147B (en) | 2024-01-25 | 2024-01-25 | Obstacle detection method, cloud control platform and unmanned vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117629147A CN117629147A (en) | 2024-03-01 |
| CN117629147B true CN117629147B (en) | 2024-03-26 |
Family
ID=90020328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410104206.2A Active CN117629147B (en) | 2024-01-25 | 2024-01-25 | Obstacle detection method, cloud control platform and unmanned vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117629147B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110457771A (en) * | 2019-07-19 | 2019-11-15 | 河海大学 | A kind of DEM water (flow) direction calculation method based on height variation transmitting |
| CN111127646A (en) * | 2019-12-26 | 2020-05-08 | 西南林业大学 | Construction method and system of rasterized elevation curved surface for measuring height difference of landform |
| WO2020212102A1 (en) * | 2019-04-16 | 2020-10-22 | Robert Bosch Gmbh | Method and device for creating an elevation profile for an area to be driven in by a vehicle |
| CN113848944A (en) * | 2021-10-18 | 2021-12-28 | 云鲸智能科技（东莞）有限公司 | Map construction method and device, robot and storage medium |
| CN114035584A (en) * | 2021-11-18 | 2022-02-11 | 上海擎朗智能科技有限公司 | Method for detecting obstacle by robot, robot and robot system |
| CN115143964A (en) * | 2022-07-05 | 2022-10-04 | 中国科学技术大学 | Four-footed robot autonomous navigation method based on 2.5D cost map |
| WO2023000337A1 (en) * | 2021-07-23 | 2023-01-26 | 华为技术有限公司 | Road gradient determination method and apparatus, lane line projection method and apparatus, and lane line display method and apparatus |
| CN116434181A (en) * | 2022-12-16 | 2023-07-14 | 北京百度网讯科技有限公司 | Ground point detection method, device, electronic equipment and medium |
| CN116740670A (en) * | 2022-03-01 | 2023-09-12 | 广东省智能网联汽车创新中心有限公司 | Obstacle segmentation recognition method and system |
| CN116859413A (en) * | 2023-05-31 | 2023-10-10 | 华能伊敏煤电有限责任公司 | Perception model building method for open-air mine car |
| CN117250944A (en) * | 2022-06-09 | 2023-12-19 | 追觅创新科技(苏州)有限公司 | Obstacle processing method and device, storage medium and electronic device |
-
2024
- 2024-01-25 CN CN202410104206.2A patent/CN117629147B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020212102A1 (en) * | 2019-04-16 | 2020-10-22 | Robert Bosch Gmbh | Method and device for creating an elevation profile for an area to be driven in by a vehicle |
| CN110457771A (en) * | 2019-07-19 | 2019-11-15 | 河海大学 | A kind of DEM water (flow) direction calculation method based on height variation transmitting |
| CN111127646A (en) * | 2019-12-26 | 2020-05-08 | 西南林业大学 | Construction method and system of rasterized elevation curved surface for measuring height difference of landform |
| WO2023000337A1 (en) * | 2021-07-23 | 2023-01-26 | 华为技术有限公司 | Road gradient determination method and apparatus, lane line projection method and apparatus, and lane line display method and apparatus |
| CN113848944A (en) * | 2021-10-18 | 2021-12-28 | 云鲸智能科技（东莞）有限公司 | Map construction method and device, robot and storage medium |
| CN114035584A (en) * | 2021-11-18 | 2022-02-11 | 上海擎朗智能科技有限公司 | Method for detecting obstacle by robot, robot and robot system |
| CN116740670A (en) * | 2022-03-01 | 2023-09-12 | 广东省智能网联汽车创新中心有限公司 | Obstacle segmentation recognition method and system |
| CN117250944A (en) * | 2022-06-09 | 2023-12-19 | 追觅创新科技(苏州)有限公司 | Obstacle processing method and device, storage medium and electronic device |
| CN115143964A (en) * | 2022-07-05 | 2022-10-04 | 中国科学技术大学 | Four-footed robot autonomous navigation method based on 2.5D cost map |
| CN116434181A (en) * | 2022-12-16 | 2023-07-14 | 北京百度网讯科技有限公司 | Ground point detection method, device, electronic equipment and medium |
| CN116859413A (en) * | 2023-05-31 | 2023-10-10 | 华能伊敏煤电有限责任公司 | Perception model building method for open-air mine car |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117629147A (en) | 2024-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11709058B2 (en) | Path planning method and device and mobile device | |
| US11680820B2 (en) | Map information system | |
| CN110920611B (en) | Vehicle control method and device based on adjacent vehicles | |
| CN111429520B (en) | Negative obstacle detection method, device, terminal equipment and storage medium | |
| CN104833370A (en) | System and method for mapping, localization and pose correction | |
| US9829575B2 (en) | Method for representing a vehicle environment with position points | |
| EP3875905B1 (en) | Method, device and medium for detecting environmental change | |
| CN114926809A (en) | Passable area detection method and device, moving tool and storage medium | |
| CN114842450A (en) | Driving region detection method, device and equipment | |
| CN115342821A (en) | Unmanned vehicle navigation cost map construction method under complex unknown environment | |
| CN112147638A (en) | Ground information acquisition method, device and system based on laser point cloud reflection intensity | |
| Song et al. | Identifying critical test scenarios for lane keeping assistance system using analytic hierarchy process and hierarchical clustering | |
| CN111951552B (en) | Method and related device for risk management in automatic driving | |
| CN117629147B (en) | Obstacle detection method, cloud control platform and unmanned vehicle | |
| US20230314169A1 (en) | Method and apparatus for generating map data, and non-transitory computer-readable storage medium | |
| US20230182743A1 (en) | Method and system for extracting road data and method and system for controlling self-driving car | |
| US20220404506A1 (en) | Online validation of lidar-to-lidar alignment and lidar-to-vehicle alignment | |
| CN116500643A (en) | Road deformation disease detection method and system based on single-line laser point cloud | |
| US20220196829A1 (en) | Radar Reference Map Generation | |
| US20220196828A1 (en) | Radar Reference Map Generation | |
| CN114638848A (en) | Method, device and equipment for detecting lane edge line change and storage medium | |
| Seeger et al. | 2-d evidential grid mapping with narrow vertical field of view sensors using multiple hypotheses and spatial neighborhoods | |
| CN109753065A (en) | The environmental modeling capability assessment method and system of automatic driving vehicle | |
| CN117233792B (en) | Retaining wall detection method, retaining wall detection device and storage medium | |
| CN114355952B (en) | Unmanned vehicle trafficability assessment method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |