A method, device, equipment and storage medium for determining road boundary lines

Abstract

This disclosure provides a method, apparatus, device, and storage medium for determining road boundaries, relating to the fields of artificial intelligence technology, particularly autonomous driving and deep learning. A specific implementation scheme comprises: obtaining an image of the environment ahead of a vehicle while driving; performing semantic segmentation on the image to determine the road outline and background outline within the image; determining at least two key points on the road boundary from the road outline; and determining the road boundary based on the at least two key points on the road boundary. This technical solution enables the rapid and accurate determination of road boundaries.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, in particular to the field of autonomous driving and deep learning technology, and specifically to a method, apparatus, device, and storage medium for determining a road boundary line. Background Art

[0002] Autonomous driving is a current research hotspot in computer vision. In autonomous driving, onboard sensors collect data in real time and need to identify road boundaries. Boundaries help vehicles perceive their location, assisting in autonomous driving decisions and avoiding dangerous accidents. Therefore, how to quickly and accurately determine road boundaries is crucial for autonomous driving. Summary of the Invention

[0003] The present disclosure provides a road boundary line determination method, apparatus, device, and storage medium.

[0004] According to one aspect of the present disclosure, a method for determining a road boundary line is provided, the method comprising:

[0005] Acquire the environment image ahead of the vehicle collected during driving;

[0006] Performing semantic segmentation on the environment image to determine a road contour and a background contour in the environment image;

[0007] determining at least two key points on a road boundary from the road profile;

[0008] Determine the boundary line of the road based on at least two key points on the road boundary.

[0009] According to one aspect of the present disclosure, a road boundary line determination device is provided, the device comprising:

[0010] The environmental image acquisition module is used to acquire the environmental image in front of the vehicle during driving;

[0011] a contour determination module, configured to perform semantic segmentation on the environment image to determine a road contour and a background contour in the environment image;

[0012] a key point determination module, configured to determine at least two key points on a road boundary from the road profile;

[0013] The boundary line determination module is used to determine the boundary line of the road according to at least two key points on the road boundary.

[0014] According to another aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the road boundary line determination method described in any embodiment of the present disclosure.

[0018] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the road boundary line determination method described in any embodiment of the present disclosure.

[0019] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the road boundary line determination method according to any embodiment of the present disclosure.

[0020] According to the technology disclosed herein, road boundary lines can be determined quickly and accurately.

[0021] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0023] Figure 1 is a flowchart of a road boundary line determination method provided according to an embodiment of the present disclosure;

[0024] Figure 2A is a flowchart of another road boundary line determination method provided according to an embodiment of the present disclosure;

[0025] Figure 2B is a schematic diagram of an environment image and a contour image provided according to an embodiment of the present disclosure;

[0026] Figure 3A is a flowchart of another road boundary line determination method provided according to an embodiment of the present disclosure;

[0027] Figure 3B is a schematic diagram of key point determination provided according to an embodiment of the present disclosure;

[0028] Figure 3C This is a boundary line fitting effect diagram provided according to an embodiment of the present disclosure;

[0029] Figure 4 1 is a schematic structural diagram of a road boundary line determination device provided according to an embodiment of the present disclosure;

[0030] Figure 5 It is a block diagram of an electronic device used to implement the road boundary line determination method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0032] Figure 1 This is a flow chart of a road boundary line determination method provided according to an embodiment of the present disclosure. This embodiment is applicable to situations where road boundary lines are determined, and is particularly applicable to situations where road boundary lines are determined in autonomous driving scenarios. The method can be executed by a road boundary line determination device, which can be implemented in software and / or hardware and can be integrated into an electronic device that carries the road boundary line determination function, such as a vehicle controller. Figure 1 As shown, the road boundary line determination method of this embodiment may include:

[0033] S101, obtaining an image of the environment ahead of the vehicle collected by the vehicle during the vehicle's driving process.

[0034] In this embodiment, the collection vehicle can be an autonomous vehicle. Optionally, an image acquisition device can be installed directly above the front windshield of the collection vehicle to capture images of the environment in front of the vehicle while the vehicle is in motion. The so-called environmental image refers to an image of the environment in front of the collection vehicle and can be a color image or a grayscale image.

[0035] Specifically, the environment image ahead of the vehicle collected during driving can be acquired in real time and periodically. For example, the environment image ahead of the vehicle collected during driving can be acquired at set intervals.

[0036] S102 , performing semantic segmentation on the environment image to determine a road outline and a background outline in the environment image.

[0037] In this embodiment, the road contour refers to the overall contour composed of the road and moving objects on the road, while the background contour refers to the overall contour composed of other objects other than the road.

[0038] Optionally, semantic segmentation can be performed on the environment image based on a semantic segmentation model to obtain road and background contours in the environment image. The semantic segmentation model can be trained based on a deep learning algorithm using pre-labeled sample environment images; the sample environment images include two categories: road and background.

[0039] S103: Determine at least two key points on the road boundary from the road contour.

[0040] In this embodiment, a key point refers to a point that has a high probability of being located at a road boundary.

[0041] Specifically, at least two key points on the road boundary can be determined from the road contour based on certain key point extraction rules. For example, coordinate points of all contours can be extracted from the road contour, and then at least two key points can be determined from all contour coordinate points.

[0042] S104: Determine a road boundary line based on at least two key points on the road boundary.

[0043] In this embodiment, the boundary line of the road is a location where a medium change occurs in the road, where the medium can be a green belt, a curb, a guardrail, a hard barrier, etc.

[0044] Optionally, the boundary line of the road may be determined based on a boundary line extraction model and according to at least two key points on the road boundary.

[0045] The technical solution of the embodiment of the present disclosure obtains an environmental image of the vehicle ahead of the vehicle while the vehicle is driving, and then performs semantic segmentation on the environmental image to determine the road contour and background contour in the environmental image, and then determines at least two key points on the road boundary from the road contour, and determines the road boundary line based on the at least two key points on the road boundary. Compared with the existing technology that relies on three-dimensional spatial information obtained by a vehicle-mounted laser scanning system to extract the road boundary line, the technical solution of the present disclosure processes the environmental image collected by the vehicle to determine the road boundary line, thereby reducing the cost of extracting the boundary line while also being able to quickly and accurately determine the road boundary line. At the same time, the road boundary line can be identified from the environmental image in real time, and can also be identified from the environmental image offline.

[0046] Figure 2A This is a flow chart of another method for determining road boundary lines according to an embodiment of the present disclosure. Based on the above embodiment, this embodiment further optimizes "semantic segmentation of the environment image to determine the road outline and background outline in the environment image" and provides an optional solution. Figure 2A As shown, the road boundary line determination method of this embodiment may include:

[0047] S201, obtaining an image of the environment ahead of the vehicle collected during the vehicle's driving process.

[0048] S202 , performing semantic segmentation on the environment image to obtain object contours of at least one type of object.

[0049] Specifically, a semantic segmentation network model can be used to perform semantic segmentation on the environmental image to obtain the object contours of at least one type of object. Since each object in the environmental image has distinct characteristics, the semantic segmentation network model can accurately learn the corresponding object characteristics, improving precision and recall, and facilitating subsequent contour reassembly to obtain an accurate road contour.

[0050] The semantic segmentation network model can include different segmentation networks, such as the FCN series, UNET series, Deeplab series, and other segmentation networks. Optionally, the semantic segmentation network model can be trained based on object category annotations, enabling the network model to output the outlines of objects in the image, clearly depicting each object and providing a basis for boundary line extraction strategies.

[0051] The object category may include at least one of the following: background, road, ground bus, ground car, bicycle, curb, guardrail, green belt, fence, and collection vehicle. It should be noted that in this embodiment, smaller objects such as pedestrians on the road are directly regarded as roads.

[0052] S203 , merging the object contours belonging to the road and the moving object category on the road according to the object category, so as to distinguish and obtain the road contour and the background contour.

[0053] In this embodiment, the types of moving objects on the road may be ground buses, ground cars, and bicycles.

[0054] Optionally, based on certain merging rules and object categories, the object contours belonging to the road and the moving object category on the road can be merged to distinguish the road contour from the background contour. For example, in order to ensure the integrity and accuracy of the contour, the contours of the road and the ground bus, ground car, and bicycle can be combined to generate a new contour as the road contour; the object contours of other objects can be merged into a new contour as the background contour. For example, Figure 2B The upper middle image is the environment image, and the lower image is the processed contour image including the road and background.

[0055] S204: Determine at least two key points on the road boundary from the road contour.

[0056] S205: Determine the boundary line of the road based on at least two key points on the road boundary.

[0057] The technical solution of the disclosed embodiment obtains an image of the environment ahead of the vehicle while it is in motion, then performs semantic segmentation on the image to obtain the object contours of at least one category of objects. Based on the object category, the object contours belonging to the road and moving objects on the road are merged to distinguish the road contour from the background contour. At least two key points on the road boundary are then determined from the road contour, and the road boundary line is determined based on the at least two key points on the road boundary. By merging the obtained fine-category object contours to determine the road contour, the above technical solution ensures the accuracy of the road contour, thereby providing a guarantee for the subsequent determination of the road boundary line.

[0058] Based on the above embodiment, as an optional method of the embodiment of the present disclosure, before determining at least two key points on the road boundary from the road profile, coordinate points outside a set distance from the collection vehicle can be removed from the coordinate points of the road profile.

[0059] Specifically, coordinate points outside the set distance from the collection vehicle can be removed from the road profile. That is, the road profile coordinate points outside the set distance from the collection vehicle are assigned a background type. This means that, based on the set distance, an image at a certain height (h) can be intercepted from the profile image and assigned as the background to obtain a new profile image. The set distance can be set by those skilled in the art based on actual circumstances. For example, an actual distance of 20m or 30m can be used.

[0060] It is understandable that since the roads in the environmental images collected by the collection vehicle that are far away from the collection vehicle may not be very accurate, and according to the actual requirements of the autonomous driving scenario, removing the parts of the road contour that are far away from the collection vehicle can make the road contour more accurate and meet the scenario requirements.

[0061] Based on the above embodiment, as an optional method of the embodiment of the present disclosure, before determining at least two key points on the road boundary from the road profile, the coordinate points adjacent to the collection vehicle side may be removed from the coordinate points of the road profile.

[0062] Specifically, the coordinate points closest to the vehicle can be removed from the road contour, leaving only the leftmost and rightmost points on the vehicle's side. It should be noted that due to the vehicle's shooting angle, in addition to the leftmost and rightmost points on the vehicle's side, there are also captured boundary points, i.e., those belonging to the road boundary in the contour image.

[0063] It can be understood that by removing the coordinate points close to the side of the acquisition vehicle, the key points on the boundary of the road can be determined more accurately.

[0064] Figure 3A This is a flow chart of another method for determining road boundary lines according to an embodiment of the present disclosure. Based on the above embodiment, this embodiment further optimizes "determining at least two key points on the road boundary from the road contour" and provides an optional implementation scheme. Figure 3A As shown, the road boundary line determination method of this embodiment may include:

[0065] S301, obtaining an image of the environment ahead of the vehicle collected during the vehicle's driving process.

[0066] S302: Perform semantic segmentation on the environment image to determine the road contour and background contour in the environment image.

[0067] S303 : Select the coordinate points closest to and farthest from the collection vehicle from each side boundary of the road contour as two key points on the side boundary.

[0068] In this embodiment, the key points on each side of the boundary include an upper key point and a lower key point.

[0069] Optionally, from each side boundary of the road profile, the coordinate point farthest from the collection vehicle is selected as the upper key point on the side boundary. Figure 3B Key point 1 and key point 2 in.

[0070] Optionally, when the acquisition vehicle side includes the leftmost coordinate point and the rightmost coordinate point, in each side boundary of the road profile, the sidemost coordinate point is directly selected as the lower key point on the side boundary.

[0071] Furthermore, when the collection vehicle side includes the leftmost coordinate point or the rightmost coordinate point, as well as the shooting boundary point of the shooting road, the shooting boundary point farthest from the collection vehicle on the side boundary including the shooting boundary point is used as the lower key point on the side boundary, such as Figure 3B The key point 4 in the side boundary including the leftmost coordinate point or the rightmost coordinate point is used as the next key point of the side, such as Figure 3B Key point 3.

[0072] S304: Determine a road boundary line based on at least two key points on the road boundary.

[0073] The technical solution of the disclosed embodiments obtains an image of the environment ahead of the vehicle while it is in motion, then performs semantic segmentation on the image to determine the road and background contours within the image. The coordinate points closest and farthest from the vehicle on each side of the road contour are then selected as the two key points on that side of the road contour. Based on these at least two key points on the road boundary, the road boundary is determined. This technical solution, by determining the key points of the boundary using the coordinate points closest and farthest from the vehicle, can improve the efficiency of subsequent road boundary determination.

[0074] On the basis of the above embodiments, as an optional method of the present disclosure, before determining the boundary line of the road based on at least two key points on the road boundary, the probability that the key point belongs to the boundary line can be verified based on the key point and the object category of the coordinate points within the distance range set by the key point.

[0075] Specifically, the object category of the coordinate points within the set distance range of the key point is determined to be a road. If so, the key point is determined to be a boundary line, i.e., the key point is valid. If more than a set number of pixels in the key point and its surrounding pixels are not roads, such as cars or buses on the road, then the true boundary line is obscured and the currently extracted key point is not a point on the true boundary line. The set number and set distance can be set by those skilled in the art based on actual circumstances.

[0076] If the key point fails after verification, the recognition result of the road boundary line determined based on the collected environmental image in this cycle is invalid. You can wait for the next cycle to test again.

[0077] It can be understood that by verifying the probability that a key point belongs to a boundary line, the accuracy of determining the boundary line of the road can be guaranteed.

[0078] Based on the above embodiments, as an optional method of the present disclosure, determining the boundary line of the road according to at least two key points on the road boundary can be to perform linear fitting based on at least two key points on the road boundary to determine the boundary line of the road.

[0079] Specifically, for each side boundary of the road profile, a linear fit is performed based on the upper key point and the lower key point on the side road boundary to obtain the boundary line of the side road. For example, the fitting effect of the boundary line is as follows: Figure 3C shown.

[0080] It can be understood that by performing linear fitting on key points to determine the boundary line of the road, the boundary line of the road can be obtained quickly and accurately.

[0081] Figure 4This is a schematic diagram of the structure of a road boundary line determination device provided according to an embodiment of the present disclosure. This embodiment is applicable to situations where road boundary lines are determined, and is particularly applicable to situations in autonomous driving scenarios, such as situations where road boundary lines are determined. The device can be implemented in software and / or hardware, and can be integrated into electronic devices that carry road boundary line determination functions. Figure 4 As shown, the road boundary line determination device 400 of this embodiment may include:

[0082] The environment image acquisition module 401 is used to acquire the environment image in front of the vehicle during the driving process;

[0083] A contour determination module 402 is configured to perform semantic segmentation on the environment image to determine a road contour and a background contour in the environment image;

[0084] A key point determination module 403 is used to determine at least two key points on the road boundary from the road contour;

[0085] The boundary line determination module 404 is configured to determine the boundary line of the road based on at least two key points on the road boundary.

[0086] The technical solution of the embodiment of the present disclosure obtains an environmental image of the vehicle ahead of the vehicle while the vehicle is driving, and then performs semantic segmentation on the environmental image to determine the road contour and background contour in the environmental image, and then determines at least two key points on the road boundary from the road contour, and determines the road boundary line based on the at least two key points on the road boundary. Compared with the existing technology that relies on three-dimensional spatial information obtained by a vehicle-mounted laser scanning system to extract the road boundary line, the technical solution of the present disclosure processes the environmental image collected by the vehicle to determine the road boundary line, thereby reducing the cost of extracting the boundary line while also being able to quickly and accurately determine the road boundary line. At the same time, the road boundary line can be identified from the environmental image in real time, and can also be identified from the environmental image offline.

[0087] Furthermore, the device also includes:

[0088] The contour correction module is used to remove coordinate points that are beyond a set distance from the collection vehicle from the coordinate points of the road contour before determining at least two key points on the boundary of the road from the road contour.

[0089] Furthermore, the device also includes:

[0090] The contour correction module is further configured to remove coordinate points adjacent to the acquisition vehicle from the coordinate points of the road contour before determining at least two key points on the road boundary from the road contour.

[0091] Furthermore, the key point determination module 403 is specifically configured to:

[0092] From each side boundary of the road contour, select the coordinate points closest and farthest from the collection vehicle as the two key points on the side boundary.

[0093] Furthermore, the device also includes:

[0094] The key point verification module is used to verify the probability that a key point belongs to the boundary line before determining the boundary line of the road based on at least two key points on the road boundary and the object category of the coordinate points within the set distance range of the key points.

[0095] Furthermore, the contour determination module 402 is specifically configured to:

[0096] Perform semantic segmentation on the environment image to obtain the object contours of at least one type of object;

[0097] According to the object category, the object contours belonging to the road and the moving object category on the road are merged to distinguish the road contour and the background contour.

[0098] Furthermore, the object categories include: background, road, ground bus, ground car, bicycle, curb, guardrail, green belt, fence or collection vehicle.

[0099] Furthermore, the boundary line determination module 404 is specifically configured to:

[0100] A linear fit is performed based on at least two key points on the road boundary to determine the road boundary line.

[0101] It should be noted that the collection, storage, use, processing, transmission, provision and disclosure of environmental images involved in the technical solution of the present disclosure are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0102] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0103] Figure 5 It is a block diagram of an electronic device used to implement the road boundary line determination method according to an embodiment of the present disclosure. Figure 5A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0104] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. Various programs and data required for the operation of the electronic device 500 can also be stored in the RAM 503. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0105] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a magnetic disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0106] The computing unit 501 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the road boundary line determination method. For example, in some embodiments, the road boundary line determination method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the road boundary line determination method described above can be performed. Alternatively, in other embodiments, the computing unit 501 can be configured to perform the road boundary line determination method by any other suitable means (e.g., via firmware).

[0107] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0108] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0109] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0110] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0111] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0112] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0113] Artificial intelligence (AI) is the study of how computers can simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, specialized AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily encompass computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graphs.

[0114] Cloud computing refers to a technology system that provides network access to elastically scalable shared pools of physical or virtual resources. These resources can include servers, operating systems, networks, software, applications, and storage devices, and can be deployed and managed on-demand in a self-service manner. Cloud computing technology provides efficient and powerful data processing capabilities for the application of technologies such as artificial intelligence and blockchain, as well as for model training.

[0115] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not a limitation herein.

[0116] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A method for determining a road boundary line, comprising: Acquire the environment image ahead of the vehicle collected during driving; Performing semantic segmentation on the environment image to determine a road outline and a background outline in the environment image; the road outline is an overall outline consisting of a road and moving objects on the road; The moving objects on the road are ground buses, ground cars and bicycles; Determining at least two key points on a road boundary from the road profile includes: selecting coordinate points closest to and farthest from the collection vehicle from each side boundary of the road profile as two key points on the side boundary; The probability that the key point belongs to the boundary line is verified based on the object category of the key point and the coordinate points within the set distance range of the key point, including: if the object category of the coordinate points within the set distance range of the key point is the road category, the key point belongs to the boundary line; if the object category of the key point and the pixel points around it exceeds a preset number and none of them belongs to the road category, the key point does not belong to the boundary line; Determining a boundary line of the road based on at least two key points on the road boundary; Before determining at least two key points on the boundary of the road from the road profile, the method further includes: Removing coordinate points outside the set distance from the collection vehicle from the coordinate points of the road profile includes: assigning object categories of the coordinate points outside the set distance from the collection vehicle in the coordinate points of the road profile to background types to obtain a new road profile.

2. The method according to claim 1, before determining at least two key points on the boundary of the road from the road profile, further comprising: The coordinate points adjacent to the collection vehicle are removed from the coordinate points of the road profile.

3. The method according to claim 1, wherein Performing semantic segmentation on the environment image to determine a road contour and a background contour in the environment image includes: Performing semantic segmentation on the environment image to obtain object contours of at least one type of object; According to the object category, the object contours belonging to the road and the moving object category on the road are merged to distinguish the road contour and the background contour.

4. The method according to claim 3, wherein: The object categories include: background, road, ground bus, ground car, bicycle, curb, guardrail, green belt, fence or collection vehicle.

5. The method according to claim 1, wherein Determining a boundary line of the road according to at least two key points on the road boundary includes: A linear fit is performed based on at least two key points on the road boundary to determine the boundary line of the road.

6. A road boundary determination device, comprising: The environmental image acquisition module is used to acquire the environmental image in front of the vehicle during driving; a contour determination module, configured to perform semantic segmentation on the environment image to determine a road contour and a background contour in the environment image; the road contour is the overall contour of the road and the moving objects on the road; the moving objects on the road are ground buses, ground cars, and bicycles; a key point determination module, configured to determine at least two key points on a road boundary from the road profile; The key point determination module is specifically configured to select the coordinate points closest to and farthest from the collection vehicle from each side boundary of the road profile as two key points on the side boundary; A key point verification module, configured to verify the probability that the key point belongs to the boundary line based on the object category to which the key point and the coordinate points within the set distance range of the key point belong; The key point verification module is specifically configured to determine that if the object category of the coordinate points within the set distance range of the key point is the road category, the key point belongs to the boundary line; if the object category of the key point and the pixel points around it exceeds a preset number and none of them belongs to the road category, the key point does not belong to the boundary line; a boundary line determination module, configured to determine a boundary line of the road based on at least two key points on the road boundary; A contour correction module, configured to remove coordinate points that are outside a set distance from the collection vehicle from the coordinate points of the road contour before determining at least two key points on the boundary of the road from the road contour; The contour correction module is specifically used to assign the object category of the coordinate points of the road contour that are beyond the set distance from the collection vehicle to the background type to obtain a new road contour.

7. The apparatus according to claim 6, further comprising: The contour correction module is further configured to remove coordinate points adjacent to the collection vehicle from the coordinate points of the road contour before determining at least two key points on the road boundary from the road contour.

8. The device according to claim 6, wherein The contour determination module is specifically used to: Performing semantic segmentation on the environment image to obtain object contours of at least one type of object; According to the object category, the object contours belonging to the road and the moving object category on the road are merged to distinguish the road contour and the background contour.

9. The device according to claim 8, wherein The object categories include: background, road, ground bus, ground car, bicycle, curb, guardrail, green belt, fence or collection vehicle.

10. The device according to claim 6, wherein The boundary line determination module is specifically used to: A linear fit is performed based on at least two key points on the road boundary to determine the boundary line of the road.

11. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the road boundary line determination method according to any one of claims 1 to 5.

12. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable a computer to execute the road boundary line determination method according to any one of claims 1 to 5. 13 . A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the road boundary line determination method according to claim 1 .

Images