CN114359869A - Method and device for detecting boundary on vehicle driving area - Google Patents

Abstract

The invention discloses a method and a device for detecting a boundary on a vehicle driving area. Wherein, the method comprises the following steps: obtaining a segmentation result of the pavement map, wherein the segmentation result of the pavement map at least comprises the following steps: a travelable area, at least one obstacle located adjacent to the travelable area; acquiring image coordinates and categories of a plurality of boundary points in a travelable area based on a segmentation result of a road surface full map; optimizing a plurality of boundary points in a travelable area by using a detection result of the obstacle to obtain an optimized result, wherein the obstacle is detected by using a 2D obstacle detection model and/or a 3D obstacle detection model, and the detection result comprises: position information of at least one obstacle located adjacent to the travelable area. The invention solves the technical problem that the road surface of the vehicle driving area can not be accurately detected.

Description

Method and apparatus for detecting boundaries on vehicle travel area

technical field

The present invention relates to the field of automatic driving development, and in particular, to a method and device for detecting a boundary on a driving area of a vehicle.

Background technique

At present, the detection of drivable areas for autonomous driving is mainly to provide path planning assistance for autonomous driving, but the existing boundary points of drivable areas do not further optimize the segmentation points that intersect the drivable area and obstacles, which leads to the development of drivable areas. The road surface detection in the area is inaccurate.

In view of the above-mentioned problem that the road surface in the driving area of the vehicle cannot be accurately detected, no effective solution has been proposed so far.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide a method and device for detecting a boundary on a vehicle driving area, so as to at least solve the technical problem that the road surface of the vehicle driving area cannot be accurately detected.

According to an aspect of the embodiments of the present invention, a method for detecting a boundary on a driving area of a vehicle is provided, including: obtaining a segmentation result of a full road map, wherein the segmentation result of the full road map at least includes: a drivable area, a drivable area located in the drivable area At least one obstacle in an adjacent position; based on the segmentation result of the full road map, obtain the image coordinates and categories of multiple boundary points in the drivable area; Perform optimization processing at points to obtain an optimization result, wherein the obstacle is detected by using a 2D obstacle detection model and/or a 3D obstacle detection model, and the detection result includes: position information of at least one obstacle located adjacent to the drivable area .

Optionally, acquiring the segmentation result of the entire road surface map includes: segmenting the entire road surface map based on multiple segmentation categories, and acquiring a segmented area formed for each segmentation category in the entire road surface map, wherein the segmentation categories include at least the following: One: roads, vehicles and pedestrians.

Optionally, based on the segmentation result of the whole road map, obtaining the image coordinates and categories of the boundary points in the drivable area, including: obtaining the area edge of each segmented area obtained by segmentation; using the boundary points adjacent to the area edge as The boundary point of the drivable area is obtained, and the coordinates of the boundary point of the drivable area are obtained; the category of at least one obstacle located above the drivable area is taken as the category of the boundary point of the drivable area.

Optionally, using the detection result of the obstacle to perform optimization processing on multiple boundary points in the drivable area to obtain the optimization result, including: using a 2D obstacle detection model to detect at least one obstacle located adjacent to the drivable area. The first detection result is obtained, and the first detection result is used to perform the first optimization process on multiple boundary points in the drivable area; the 3D obstacle detection model is used to perform the first optimization process on the vehicles located in the adjacent positions of the drivable area. The detection is performed to obtain a second detection result, and a second optimization process is performed on a plurality of boundary points in the drivable area by using the second detection result.

Optionally, the first optimization process includes at least one of the following: if the 2D obstacle detection model detects that any two obstacles have overlapping parts, the coordinates of the non-overlapping parts are taken as the coordinates of the boundary points of the drivable area; The object detection model detects that the boundary distance of any obstacle close to the drivable area is less than a predetermined value, and the coordinates of the point where the obstacle is close to the side of the drivable area are taken as the coordinates of the boundary point of the drivable area.

Optionally, the second optimization process includes at least one of the following: if the 3D obstacle detection model detects that the angle at which the angular coordinates of any obstacle deviates from any direction exceeds a predetermined angle value, the coordinates of the boundary point of the drivable area are Adjustment in predetermined steps.

According to another aspect of the embodiments of the present invention, there is also provided an apparatus for detecting a boundary on a driving area of a vehicle, which is characterized by comprising: a first obtaining module, configured to obtain a segmentation result of a full road map, wherein the road The segmentation result of the map at least includes: a drivable area and at least one obstacle located adjacent to the drivable area; a second acquisition module, configured to obtain the segmentation results of multiple boundary points in the drivable area based on the segmentation result of the entire road map Image coordinates and categories; an optimization module for optimizing multiple boundary points in the drivable area using the detection results of obstacles to obtain optimization results, wherein 2D obstacle detection models and/or 3D obstacles are used The detection model detects obstacles, and the detection result includes: position information of at least one obstacle located adjacent to the drivable area.

Optionally, the first obtaining module includes: a segmentation module, configured to segment the entire road map based on multiple segmentation categories, and obtain the segmentation area formed for each segmentation category in the entire road map, wherein the segmentation categories include the following: At least one of: road, vehicle, and pedestrian.

Optionally, the second acquisition module includes: a reading module for reading the region edge of each segmented region obtained by segmentation; a processing module for using the boundary point adjacent to the region edge as the boundary point of the drivable region , and obtains the coordinates of the boundary point of the drivable area; the assignment module is used for taking the category of at least one obstacle above the drivable area as the category of the boundary point of the drivable area.

In the embodiment of the present invention, the segmentation result of the full road map is obtained, wherein the segmentation result of the full road map at least includes: a drivable area and at least one obstacle located adjacent to the drivable area; segmentation based on the full road map As a result, the image coordinates and categories of multiple boundary points in the drivable area are obtained; the detection results of obstacles are used to optimize the multiple boundary points in the drivable area, and the optimization results are obtained. The detection model and/or the 3D obstacle detection model detects obstacles, and the detection results include: position information of at least one obstacle located adjacent to the drivable area, so as to optimize multiple boundary points in the drivable area , in order to ensure the accuracy and pertinence of the detection of the road surface in the drivable area of the vehicle, thereby achieving the technical effect of accurately detecting the road surface in the drivable area of the vehicle, and solving the technology that cannot accurately detect the road surface in the drivable area of the vehicle. question.

Description of drawings

The accompanying drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the attached image:

FIG. 1 is a flowchart of a method for detecting a boundary on a driving area of a vehicle according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a device for detecting a boundary on a driving area of a vehicle according to an embodiment of the present invention;

Fig. 3 is a kind of effect comparison diagram before and after optimization of drivable area boundary points based on 2D Boundingbox processing according to an embodiment of the present invention;

Fig. 4 is a kind of effect comparison diagram before and after optimization of the boundary point of the drivable area based on 3D Boundingbox processing according to an embodiment of the present invention;

5 is a comparison diagram of the effect before and after the optimization of the boundary points of the drivable area based on the segmentation of the whole image according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of a device for detecting a boundary on a driving area of a vehicle according to an embodiment of the present invention.

Detailed ways

In order to make those skilled in the art better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only Embodiments are part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

It should be noted that the terms "first", "second" and the like in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used may be interchanged under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those expressly listed Rather, those steps or units may include other steps or units not expressly listed or inherent to these processes, methods, products or devices.

Example 1

According to an embodiment of the present invention, an embodiment of a method for detecting a boundary on a driving area of a vehicle is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be implemented in a computer system such as a set of computer-executable instructions. and, although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in an order different from that herein.

FIG. 1 is a flowchart of a method for detecting a boundary on a driving area of a vehicle according to an embodiment of the present invention. As shown in FIG. 1 , the method may include the following steps:

Step S102: Obtain a segmentation result of the full road map, wherein the segmentation result of the full road map at least includes: a drivable area and at least one obstacle located adjacent to the drivable area.

In the technical solution provided by the above step S102 of the present invention, the full image segmentation result may include: segmentation categories and results, wherein the segmentation categories include: roads, vehicles, pedestrians, cones, etc., and the rest are backgrounds, and the results include each category. 2D pixels and the corresponding category of each pixel.

Step S104, based on the segmentation result of the entire road surface map, obtain image coordinates and categories of multiple boundary points in the drivable area.

In the technical solution provided by the above step S104 of the present invention, the image coordinates and categories of the multiple boundary points in the drivable area are the area edges of each segmented area obtained by segmenting the entire road surface.

Optionally, the boundary point adjacent to the edge of the area is used as the boundary point of the drivable area to obtain the image coordinates of the boundary point of the drivable area.

Optionally, the boundary point image category of the drivable area is the category of at least one obstacle located above the drivable area.

Step S106, using the detection result of the obstacle to perform optimization processing on multiple boundary points in the drivable area to obtain the optimization result, wherein the obstacle is detected by using the 2D obstacle detection model and/or the 3D obstacle detection model , and the detection result includes: position information of at least one obstacle located adjacent to the drivable area.

In the technical solution of the above step S106 of the present invention, the detection results may include detection categories and detection results, wherein the detection categories include: vehicles, pedestrians, cyclists, etc. The vehicles include a full vehicle frame and a rear frame, and the detection results include: each 2D Boundingbox coordinates and 3D Boundingbox coordinates of obstacles for each category.

In the above steps S102 to S106 of the present application, the segmentation results of the full road map are obtained, wherein the segmentation results of the full road map include at least: a drivable area and at least one obstacle located adjacent to the drivable area; The segmentation results are used to obtain the image coordinates and categories of multiple boundary points in the drivable area; the detection results of obstacles are used to optimize the multiple boundary points in the drivable area, and the optimization results are obtained. Among them, 2D obstacles are used. The object detection model and/or the 3D obstacle detection model detects obstacles, and the detection results include: position information of at least one obstacle located adjacent to the drivable area, so as to optimize multiple boundary points in the drivable area processing to ensure the accuracy and pertinence of the detection of the road surface in the drivable area of the vehicle, thereby realizing the technical effect of accurately detecting the road surface in the drivable area of the vehicle, and solving the problem that the road surface in the drivable area of the vehicle cannot be accurately detected. technical problem.

The above method of this embodiment will be further described below.

As an optional embodiment, in step S102, obtaining the segmentation result of the entire road surface map, further comprising: segmenting the entire road surface map based on multiple segmentation categories, and obtaining the segmentation results formed for each segmentation category in the entire road surface map. A segmented area, wherein the segmented category includes at least one of the following: road, vehicle and pedestrian.

Optionally, the segmentation result includes the 2D pixel points of each category and the category corresponding to each pixel point.

As an optional embodiment, in step S104, based on the segmentation result of the entire road map, acquiring the image coordinates and categories of boundary points in the drivable area, including: acquiring the area edge of each segmented area obtained by segmentation; The boundary point adjacent to the edge of the area is used as the boundary point of the drivable area, and the coordinates of the boundary point of the drivable area are obtained; the category of at least one obstacle above the drivable area is taken as the category of the boundary point of the drivable area.

In this embodiment, according to the segmentation result based on the whole road map, the area frame formed by each segmentation category is obtained, and then the coordinates of the pixel points are divided according to each category to obtain the area frame coordinates of the segmented area of this category, and the area frame is divided into Sort from small to large, and finally scan the points in each area frame row by row according to the image height from bottom to top, and use the points in the row that belong to the area frame and the previous row is determined as the boundary point of the drivable area as the drivable area. Boundary point, use the obstacle class above the drivable area as the class of the updated drivable area boundary point.

As an optional embodiment, in step S106, using the detection result of the obstacle, performing optimization processing on a plurality of boundary points in the drivable area to obtain the optimization result, including: using a 2D obstacle detection model to detect the obstacles located in the drivable area. At least one obstacle in the adjacent position of the driving area is detected to obtain a first detection result, and the first detection result is used to perform a first optimization process on a plurality of boundary points in the drivable area; the 3D obstacle detection model is used to detect Vehicles at adjacent positions in the drivable area are detected to obtain a second detection result, and a second optimization process is performed on a plurality of boundary points in the drivable area by using the second detection result.

In this embodiment, 2D Boundingbox is used to detect obstacles in the drivable area, and 3DBoundingbox is used to detect other contents of the drivable area, wherein, using the 2D Boundingbox detection result to optimize the pixel-level segmentation result of the drivable area is for the rear of the vehicle Box, using the 3D Boundingbox detection results to optimize the pixel-level segmentation results of the drivable area is for vehicles.

As an optional embodiment, in step S106, the first optimization process includes at least one of the following: if the 2D obstacle detection model detects that any two obstacles have overlapping parts, take the coordinates of the non-overlapping parts as the drivable area If the 2D obstacle detection model detects that the boundary distance of any obstacle close to the drivable area is less than the predetermined value, the coordinates of the point on the side of the obstacle close to the drivable area are used as the boundary point of the drivable area. coordinate.

In this embodiment, the first optimization process detects obstacles in the drivable area through 2D obstacle detection model detection, and determines the coordinates of the boundary points of the drivable area.

For example, if it is detected that there are partial overlapping of multiple obstacles, the coordinates of the boundary point of the drivable area are the coordinates of the non-overlapping part; The coordinates of the boundary point of the driving area are the coordinates of the point where the obstacle is close to the side of the driving area.

As an optional embodiment, in step S106, the second optimization process includes at least one of the following: if the angle at which the angular coordinates of any obstacle deviates from any direction exceeds a predetermined angle value detected by the 3D obstacle detection model, the The coordinates of the boundary points of the drivable area are adjusted in predetermined steps.

In this embodiment, the second optimization process detects obstacles in the drivable area through 3D obstacle detection model detection, and updates and adjusts the coordinates of the boundary points of the drivable area according to a predetermined step size.

In this embodiment, the segmentation result of the full road map is obtained, wherein the segmentation result of the full road map at least includes: a drivable area and at least one obstacle located adjacent to the drivable area; Image coordinates and categories of multiple boundary points in the driving area; use the detection results of obstacles to optimize multiple boundary points in the drivable area to obtain the optimization results, in which the 2D obstacle detection model and/or Or the 3D obstacle detection model detects obstacles, and the detection results include: position information of at least one obstacle located adjacent to the drivable area, so as to optimize the multiple boundary points in the drivable area to ensure that the The accuracy and pertinence of the detection of the road surface in the drivable area of the vehicle, thereby realizing the technical effect of accurately detecting the road surface in the drivable area of the vehicle, and solving the technical problem that the road surface in the drivable area of the vehicle cannot be accurately detected.

Example 2

The technical solutions of the embodiments of the present invention are illustrated below with reference to the preferred embodiments.

At present, the detection of the drivable area for autonomous driving is mainly to provide path planning assistance for autonomous driving, which can realize the detection of the entire road surface, or only extract part of the road information. However, if there are obstacles above the drivable area, such as vehicles, pedestrians, etc., the segmentation result of the drivable area category based on the segmentation of the whole image is not enough. If the boundary points between obstacles and drivable areas are roughly divided, this will affect road detection and path planning for autonomous driving based on drivable areas. Therefore, in order to obtain more realistic boundary point information of the drivable area, it is very important to further optimize the boundary points of the drivable area.

Existing drivable area boundary point acquisition is mainly based on full-image semantic segmentation results, including traditional machine learning algorithms such as texture extraction, edge extraction, vanishing point extraction, and support vector machine detection classifiers, as well as deep learning-based roads. Driving area detection algorithms, such as DeepLabV3+ and other deep learning models. However, the segmentation points that intersect the drivable area and obstacles are not further optimized, resulting in inaccurate road detection in the drivable area.

Therefore, in order to overcome the above problems, in a related art, an optimization method for automatic driving image semantic segmentation is proposed, which is characterized in that it is divided into the following steps: 1) The method designs an AAM that uses labels to assist activation module, the features extracted by the network are modified by segmentation labels, so that the features extracted from similar objects have approximately the same value; 2) The AAM module is integrated into the encoder and decoder of the segmentation model, and a performance ratio benchmark is obtained through training A model with a better model is called a teacher network; 3) Through knowledge transfer, the knowledge learned by the teacher network based on the AAM module is transferred to the segmentation model to improve its segmentation performance, thereby solving the technology that cannot effectively mine the information of segmentation labels problem, improve the performance of the segmentation model, and do not need to modify the network structure, which has strong application value.

However, the present application proposes a drivable area boundary point optimization method based on full image segmentation, which can make the drivable area boundary points further fit the obstacle target, so as to accurately perform the road surface detection and path based on the drivable area planning, and the optimization algorithm has strong adaptability, low algorithm complexity and easy implementation.

In this embodiment, a flowchart of a method for optimizing a boundary point of a drivable area is proposed, as shown in FIG. 2 , which is a flowchart of a method for optimizing a boundary point of a drivable area according to an embodiment of the present invention.

S202 , obtain 2D image coordinates and categories of boundary points of the drivable area according to the full image segmentation result.

According to the full-image semantic segmentation result of each category, obtain the region frame SEG_BBox formed by each segmentation category, and set the coordinates of the segmentation pixels of this category as (x _i , y _i ), then the SEG_BBox coordinates of the category segmentation area are :

SEG_BBox _xmin =min(x _i ) SEG_BBox _xmax =max(x _i )

SEG_BBox _ymin =min(y _i ) SEG_BBox _ymax =max(y _i )

Sort the SEG_BBox according to the SEG_BBox _ymin from small to large, scan the points in each SEG_BBox line by line from bottom to top according to the image height, and use the points in the line that belong to the SEG_BBox and the previous line is determined as the drivable area boundary point as the drivable area boundary point.

Use the class of obstacles above the drivable area as the class of the updated drivable area boundary points.

S204 , based on the 2D obstacle target detection, use the 2D Boundingbox detection result to optimize the pixel-level segmentation result of the drivable area, where the vehicle frame is the rear frame, and the boundary points before and after optimization are shown in FIG. 3 .

Optionally, if there is an overlapping part between two obstacles, assign the coordinates of the non-overlapping part with the smaller 2D_BBox _ymin to the corresponding boundary point of the drivable area, and judge that the two obstacles do not overlap as follows:

box ₁ .1>box ₂ .r||box ₁ .r<box ₂ .1||box ₁ .t>box ₂ .b

Among them, box1 and box2 are 2D Boundingbox of two obstacles, l is 2D_BBox _xmin , r is 2D_BBox _xmax , t is 2D_BBox _ymin , and b is 2D_BBox _ymax .

Optionally, when the distance between the bottom edge point of the obstacle 2D_BBox _ymax and the boundary point of the drivable area is less than 10 pixels, use the bottom edge point of the obstacle 2D_BBox _ymax to assign the corresponding drivable area boundary point.

S206, for the above-mentioned drivable area boundary points that have not been optimized, based on 3D obstacle target detection, use the 3DBoundingbox detection result to optimize the drivable area pixel-level segmentation result (for vehicles), as shown in FIG. 4 .

Optionally, if the yaw angle of the 3D BoundingBox obstacle is to the left, that is, if the following conditions are met:

box3d.lower.1t.x＜box3d.lower.1b.x&&box3d.lower.1t.x＜left

Among them, lower is the bottom 2D_BBox of the 3D Boundingbox, lt is the upper left point, and lb is the lower left point. Set the coordinates of the rear frame 2D_BBox _xmin as left, and the drivable area boundary point corresponding to left as DRI[left], and use the corresponding DRI[box3d.lower.lt.x] point to pair (box3d.lower.lt.x) with a certain step size. The boundary points of the drivable area within the range of lt.x, left) are optimized, and the step size formula is as follows:

step=(DRI[left]-box3d.lower.1t.y)/(left-box3d.lower.1t.x)

Then the optimization formula for the boundary point of the drivable area within the range of (box3d.lower.lt.x, left) is:

DRI[t+box3d.lower.1r.x]=DRI[box3d.lower.1r.x]+t*step,

t∈[box3d.lower.1r.x, left]

Optionally, if the yaw angle of the 3D BoundingBox obstacle is to the right, that is, if the following conditions are met:

box3d.lower.rt.x>box3d.lower.rb.x&&box3d.lower.rt.x>right

Among them, lower is the bottom 2D_BBox of the 3D Boundingbox, rt is the upper right point, and rb is the lower right point. Set the coordinate of the rear frame 2D_BBox _xmax as right, and the drivable area boundary point corresponding to right is DRI[right], and use the optimized DRI[right] point of the corresponding rear frame to pair (right, box3d. lower, rt.x) within the range of the drivable area boundary points for optimization, the step size formula is as follows:

step=(box3d.lower.rt.y-DRI[right])/(box3d.lower.rt.x-right)

Then the optimization formula of the boundary point of the drivable area within the range of (right, box3d.lower.rt.x) is:

DRI[t+right]=DRI[right]+t*step, t∈[right, box3d.lower.rt.x]

Based on the full image segmentation results, the 2D image coordinates and categories of the boundary points of the drivable area are obtained, then based on the 2D obstacle target detection, the 2D Boundingbox detection results are used to optimize the pixel-level segmentation results of the drivable area, and finally based on the 3D obstacle target detection, use 3DBouundingbox detection result optimizes the pixel-level segmentation result of the drivable area, as shown in Figure 5. Figure 5 is a comparison diagram of the effect before and after optimization of the boundary point of the drivable area based on full-image segmentation according to an embodiment of the present invention, wherein the black line Before optimization, the red line is after optimization.

In this embodiment, by obtaining the segmentation result of the whole road map, and then obtaining the image coordinates and categories of multiple boundary points in the drivable area based on the segmentation result of the whole road map, and finally using the detection results of obstacles, Multiple boundary points in the drivable area are optimized to obtain the optimization result, and then the technical effect of accurate detection of the road surface in the drivable area of the vehicle is realized, and the technical problem that the road surface in the drivable area of the vehicle cannot be accurately detected is solved. .

Example 3

According to an embodiment of the present invention, a device for detecting a boundary on a driving area of a vehicle is also provided. It should be noted that the apparatus for detecting a boundary on a vehicle driving area can be used to execute the method for detecting a boundary on a vehicle driving area in Embodiment 1.

FIG. 6 is a schematic diagram of a device for detecting a boundary on a driving area of a vehicle according to an embodiment of the present invention. As shown in FIG. 6 , the device 600 for detecting the boundary on the driving area of the vehicle may include: a first acquisition module 601 , a second acquisition module 602 , and an optimization module 603 .

The first obtaining module 601 is configured to obtain a segmentation result of the full road map, wherein the segmentation result of the full road map includes at least a drivable area and at least one obstacle located adjacent to the drivable area.

The second obtaining module 602 is configured to obtain image coordinates and categories of multiple boundary points in the drivable area based on the segmentation result of the full road map.

The optimization module 603 is configured to perform optimization processing on a plurality of boundary points in the drivable area by using the detection result of the obstacle, and obtain the optimization result, wherein, the obstacle detection model and/or the 3D obstacle detection model are used to analyze the obstacles. The detection result includes: position information of at least one obstacle located adjacent to the drivable area.

Optionally, the first obtaining module includes: a sub-segmentation module, configured to segment the entire road map based on a plurality of segmentation categories, and obtain a segmentation area formed for each segmentation category in the entire road map, wherein the segmentation categories include: At least one of the following: road, vehicle, and pedestrian.

Optionally, the second acquisition module includes: a sub-reading module for reading the region edge of each segmented region obtained by segmentation; a sub-processing module for using the boundary point adjacent to the region edge as the drivable region. The boundary point is obtained, and the coordinates of the boundary point of the drivable area are obtained; the sub-assignment module is used to take the category of at least one obstacle located above the drivable area as the category of the boundary point of the drivable area.

In this embodiment, the segmentation result of the full road map is obtained by the first acquisition module, and then based on the segmentation result of the full road map, the second acquisition module acquires the image coordinates and categories of multiple boundary points in the drivable area, and finally optimizes the The module uses the detection results of obstacles to optimize multiple boundary points in the drivable area, and obtain the optimization results, thereby realizing the technical effect of accurate detection of the road surface in the drivable area of the vehicle, and solving the problem of inability to detect the road surface in the drivable area of the vehicle. The technical problem of accurate detection of the road surface in the driving area.

The above-mentioned serial numbers of the embodiments of the present invention are only for description, and do not represent the advantages or disadvantages of the embodiments.

In the above-mentioned embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are only illustrative, for example, the division of units may be a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or integrated into Another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of units or modules, and may be in electrical or other forms.

Units described as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented as a software functional unit and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , which includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes .

The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, some improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be regarded as It is the protection scope of the present invention.

Claims (10)

1. A method for detecting a boundary on a vehicle driving area, comprising: Obtaining a segmentation result of the full road map, wherein the segmentation result of the full road map includes at least: a drivable area and at least one obstacle located adjacent to the drivable area; obtaining image coordinates and categories of multiple boundary points in the drivable area based on the segmentation result of the full road surface map; Using the detection result of the obstacle, an optimization process is performed on multiple boundary points in the drivable area to obtain an optimization result, wherein the obstacle is detected by using a 2D obstacle detection model and/or a 3D obstacle detection model. The detection result includes: position information of at least one obstacle located adjacent to the drivable area. 2 . The method according to claim 1 , wherein obtaining the segmentation result of the full road surface map comprises: segmenting the full road surface map based on a plurality of segmentation categories, and obtaining the segmentation results for each road surface map in the full road surface map. 3 . A segmented area formed by segmented categories, wherein the segmented categories include at least one of the following: road, vehicle and pedestrian. 3. The method according to claim 2, wherein, based on the segmentation result of the full road surface map, acquiring the image coordinates and categories of the boundary points in the drivable area, comprising: Obtain the region edge of each segmented region obtained by segmentation; Taking the boundary point adjacent to the edge of the area as the boundary point of the drivable area, and obtaining the coordinates of the boundary point of the drivable area; The category of the at least one obstacle located above the drivable area is used as the category of the boundary point of the drivable area. 4. The method according to any one of claims 1-3, characterized in that, using the detection result of the obstacle, an optimization process is performed on a plurality of boundary points in the drivable area to obtain an optimization result ,include: Use the 2D obstacle detection model to detect at least one obstacle located adjacent to the drivable area to obtain a first detection result, and use the first detection result to detect the obstacles in the drivable area Perform the first optimization process on multiple boundary points of ; The 3D obstacle detection model is used to detect the vehicle located adjacent to the drivable area to obtain a second detection result, and the second detection result is used to detect multiple vehicles in the drivable area. The boundary points are subjected to the second optimization process. 5. The method according to claim 4, wherein the first optimization process comprises at least one of the following: If the 2D obstacle detection model detects that any two obstacles have overlapping parts, the coordinates of the non-overlapping parts are used as the coordinates of the boundary point of the drivable area; If the 2D obstacle detection model detects that the distance of any obstacle close to the boundary of the drivable area is less than a predetermined value, the point coordinates of the obstacle close to one side of the drivable area are taken as the boundary of the drivable area the coordinates of the point. 6. The method according to claim 4, wherein the second optimization process comprises at least one of the following: If the 3D obstacle detection model detects that the angle that the angular coordinates of any obstacle deviate from any direction exceeds a predetermined angle value, the coordinates of the boundary points of the drivable area are adjusted according to a predetermined step size. 7. A device for detecting a boundary on a driving area of a vehicle, comprising: a first obtaining module, configured to obtain a segmentation result of the full road map, wherein the segmentation result of the full road map includes at least: a drivable area and at least one obstacle located adjacent to the drivable area; a second acquisition module, configured to acquire image coordinates and categories of multiple boundary points in the drivable area based on the segmentation result of the full road map; An optimization module, configured to perform optimization processing on a plurality of boundary points in the drivable area by using the detection results of the obstacles to obtain optimization results, wherein a 2D obstacle detection model and/or a 3D obstacle detection model are used. The model detects the obstacles, and the detection result includes: position information of at least one obstacle located adjacent to the drivable area. 8 . The apparatus according to claim 7 , wherein the first obtaining module comprises: a sub-segmentation module, configured to segment the full road map based on a plurality of segmentation categories to obtain the full road map. 9 . The segmented area formed for each segmented category in , wherein the segmented category includes at least one of the following: road, vehicle and pedestrian. 9. The apparatus according to claim 8, wherein the second acquiring module comprises: The sub-reading module is used to read the region edge of each segmented region obtained by segmentation; a sub-processing module, configured to use the boundary point adjacent to the edge of the area as the boundary point of the drivable area, and obtain the boundary point coordinates of the drivable area; The sub-assignment module is configured to use the category of the at least one obstacle located above the drivable area as the category of the boundary point of the drivable area. 10. A vehicle, characterized by comprising the method for detecting a boundary on a driving area of a vehicle according to any one of claims 1-6.

Images