Path planning method and device, electronic equipment and computer readable storage medium
By acquiring and matching route information and real-time images, predicting vehicle locations and determining planned paths, the path planning problems of autonomous vehicles in complex environments are solved and driving safety is improved.
Patent Information
- Application Number
- CN202310652440.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-02
- Publication Date
- 2024-12-03
AI Technical Summary
The prior art is difficult to effectively solve the path planning problem of autonomous vehicles in longitudinal and lateral planning, especially in lane lines or complex environments, and it is difficult to ensure the safety and smoothness of the path.
By obtaining route information and real-time images, matching the target driving path, predicting the next position of the vehicle, and determining the planned path based on the current and next position, generating a safe and smooth driving path.
Improves the driving safety of autonomous vehicles, and can generate appropriate paths in complex environments, reducing the risk of collision with other vehicles or obstacles.
Smart Images

Figure CN119063747A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the field of artificial intelligence technology, and in particular, relates to a path planning method, device, electronic device and computer-readable storage medium. Background Art
[0002] Many existing methods convert route planning into solving optimization problems, using dynamic programming advance rules, quadratic programming, and sequential quadratic programming. When solving these optimization problems, it is also necessary to distinguish different optimization problems, such as longitudinal planning and lateral planning. Lateral planning is planning in the driving direction, which can be seen as planning how to steer the wheel, and it determines the shape of the trajectory. There are usually two solutions to this problem, one is a scenario without lanes, and the other is a scenario with lanes.
[0003] For vertical planning, since the decision-making problem is an NP-hard problem, it is not easy to solve directly and it is difficult to fit people's experience with rules. Currently, some of the available methods to solve this challenge are to establish mathematical models based on various situations and use machine learning methods to make decisions. In addition, the variability and complexity of the scenarios make it difficult to make decisions and will face many challenges.
[0004] As for lateral planning, the moving vehicles often do not drive according to the lane lines, and may cross the lane lines to occupy the space of the lane where the autonomous driving vehicle is located, or in the main and auxiliary road scenarios without lane lines. It is difficult to solve the path planning problem when there are no lane lines or the vehicle does not drive according to the lane lines. Another challenge is the problem of the environment, because the driving environment outside the car changes rapidly, and it is also difficult to predict the surrounding environment. The autonomous driving car is preparing to change lanes to the left, and the motorcycle on the left is driving forward at a high speed. The vehicle quickly makes a decision to cancel the lane change and generates a smooth curve to return to the original driving road. Therefore, in the face of such situations, the trajectory planning module needs to ensure that the planned path is smooth and controllable.
[0005] In other words, there is an urgent need for a method for path planning for autonomous driving vehicles to improve the safety of autonomous driving vehicles. Summary of the invention
[0006] The embodiments of the present application provide a path planning method, device, electronic device and computer-readable storage medium, which can generate a planned path for a vehicle and improve the safety of vehicle driving.
[0007] In a first aspect, an embodiment of the present application provides a path planning method, including:
[0008] Acquiring route information, the route information including a plurality of driving paths and a captured image acquired by the vehicle on each of the driving paths;
[0009] Acquire a real-time image currently collected by the target vehicle, and match the real-time image with the collected image to determine a target driving path from the multiple driving paths;
[0010] Based on the target driving path, predicting the next position of the target vehicle;
[0011] A planned path of the target vehicle is determined according to the current position and the next position of the target vehicle.
[0012] In a second aspect, an embodiment of the present application provides a path planning device, including:
[0013] A collection module, used to obtain route information, the route information includes multiple driving paths, and a collection image obtained by the vehicle on each of the driving paths;
[0014] An acquisition module is used to acquire the real-time image currently collected by the target vehicle;
[0015] A matching module, used for matching the real-time image with the collected image to determine a target driving path from the multiple driving paths;
[0016] A prediction module, used to predict the next position of the target vehicle based on the target driving path;
[0017] The determination module is used to determine the planned path of the target vehicle according to the current position and the next position of the target vehicle.
[0018] In a third aspect, an embodiment of the present application provides an electronic device, the device comprising: a processor and a memory storing computer program instructions;
[0019] When the processor executes the computer program instructions, the method according to the first aspect is implemented.
[0020] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described in the first aspect is implemented.
[0021] In a fifth aspect, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the method described in the first aspect.
[0022] The path planning method, device, electronic device and computer-readable storage medium of the embodiment of the present application, wherein the method includes: obtaining route information, the route information includes multiple driving paths, and the captured images captured by the vehicle on each of the driving paths; obtaining the real-time image currently captured by the target vehicle, and matching the real-time image with the captured image to determine the target driving path from the multiple driving paths; based on the target driving path, predicting the next position of the target vehicle; and determining the planned path of the target vehicle according to the current position and the next position of the target vehicle. Through the above steps, a planned path can be generated for the vehicle, and the vehicle can travel according to the planned path, thereby improving the safety of the vehicle's automatic driving. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0024] Figure 1 This is a flow chart of a path planning method provided by an embodiment of the present application;
[0025] Figure 2 is another flowchart of a path planning method provided by an embodiment of the present application;
[0026] Figure 3 is a schematic diagram of the structure of a path planning device provided by an embodiment of the present application;
[0027] Figure 4 It is a structural diagram of an electronic device provided by yet another embodiment of the present application. DETAILED DESCRIPTION
[0028] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.
[0029] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0030] In order to solve the problems in the prior art, the embodiments of the present application provide a path planning method, device, electronic device, medium and product. The path planning method provided by the embodiments of the present application is first introduced below.
[0031] Figure 1 FIG. 1 is a flow chart of a path planning method provided by an embodiment of the present application. Figure 1 As shown, the path planning method provided in the embodiment of the present application includes the following steps 101 to 104, wherein:
[0032] Step 101 : Acquire route information, wherein the route information includes a plurality of driving paths and a captured image acquired by a vehicle on each of the driving paths.
[0033] Different vehicle conditions around the vehicle can be obtained by installing cameras in different directions on the test vehicle, i.e., the collection vehicle, including a front-view camera and a surround-view camera. When the collection vehicle collects route information, the driver drives the test vehicle to record different scene images of the driving route. In addition, the best and safest decisions made by the driver under different vehicle conditions can also be recorded.
[0034] The multiple driving paths may include driving paths of the vehicle on the same route but under different road conditions, or may include driving paths of the vehicle on different routes, which is not limited here.
[0035] Optionally, step 101, obtaining route information, includes:
[0036] During the driving process of the collection vehicle, a driving path of the collection vehicle under different road conditions and a plurality of collection images during the driving process are obtained;
[0037] Analyze each of the collected images to determine the position of the collected vehicle on the driving path;
[0038] A corresponding relationship between each of the captured images and the position of the captured vehicle on the driving path is established.
[0039] Specifically, collecting the driving paths of vehicles under different road conditions can be understood as: collecting the driving paths of vehicles on the same route but under different vehicle conditions, and / or collecting the driving paths of vehicles on different routes. It should be noted that when collecting the driving paths of vehicles on different routes, the road conditions must be different. The road conditions may refer to information such as pedestrians, other vehicles, traffic lights, and road conditions.
[0040] When the collection vehicle is driving, the camera installed on the collection vehicle collects multiple images. After the images are collected, the images are gray-processed, and the real-world position of the vehicle body is analyzed through post-processing. That is, based on the collected images, the position of the collection vehicle on the driving path is determined, and a corresponding relationship between each of the collected images and the position of the collection vehicle on the driving path is established, that is, the collected images are matched with the specific positions on the driving path, so that a driving path and the collected images on the driving path can be obtained.
[0041] Step 102, obtaining a real-time image currently collected by the target vehicle, and matching the real-time image with the collected image to determine the target driving path from the multiple driving paths.
[0042] The target vehicle in this step can be understood as an autonomous driving vehicle. In step 102, the target vehicle adopts autonomous driving. When the target vehicle is moving, the camera set on the target vehicle collects images in real time, and the collected real-time images are matched with the collected images in the route information, and the target driving path is determined according to the matching results. For example, if the collected real-time image matches the first image in the route information, and the first image is an image corresponding to the first driving path, then the first driving path is the target driving path.
[0043] Optionally, step 102 specifically includes:
[0044] Acquire multiple real-time images of the target vehicle currently collected within a preset time period;
[0045] The multiple real-time images are matched one by one with the collected images to determine a target driving path from the multiple driving paths.
[0046] In the above, when matching the real-time image with the image in the route information, multiple real-time images collected within a preset time period are used for matching. The preset time period can be 10 seconds or 15 seconds, which is not limited here. Since the images collected during the vehicle's driving process are on a section of the driving path within the preset time period, when matching, the images on the section of the driving path are matched one by one, which can improve the matching accuracy, thereby improving the accuracy of determining the target driving path. For the multiple real-time images collected, the multiple real-time images are matched one by one with the images in the route information. For example, if there are two real-time images, and both of the two real-time images match the image corresponding to the first driving path, then the first target driving path is the target driving path, that is, the images that match the multiple real-time images one by one are all collected images corresponding to the target driving path.
[0047] Step 103: predicting the next position of the target vehicle based on the target driving path.
[0048] Determine the current position of the target vehicle in the target driving path. For example, the current position can be obtained from navigation software, or the current position can be determined based on a captured image matched by a real-time image. For example, if the real-time image matches the first image, and the first image corresponds to the first position of the target driving path, the current position of the target vehicle is the first position. The next position of the target vehicle can be understood as the next position of the current position.
[0049] Optionally, step 103 includes:
[0050] Determining the current position of the target vehicle on the target driving path according to the real-time image;
[0051] The next position of the target vehicle is predicted based on the current position and the road condition information of the target vehicle.
[0052] Specifically, the current position is determined according to the captured image matched by the real-time image. For example, if the real-time image matches the first image, and the first image corresponds to the first position of the target driving path, the current position of the target vehicle is the first position. The next position of the target vehicle can be understood as the next position of the current position.
[0053] Route information also includes road condition information and driving decision information corresponding to each driving path. During the human driving process, the collected vehicle will make driving decisions based on the road condition information. For example, if there is a white car parked on the side of the road in front and the white car suddenly starts, the human driver will make the best decision in this situation, such as: 1. Slow down and honk to signal the white car to give way, and then speed up to pass. 2. Slow down, wait for the white car to pass, and then speed up. The driving paths corresponding to these two decisions are different, and the next position of the current position is also different. When predicting the next position of the vehicle, the road condition information of the vehicle is comprehensively considered, so that the predicted next position is more in line with the actual situation, improving the safety of the vehicle's automatic driving.
[0054] In this embodiment, when predicting the next position of the target vehicle based on the current position and the road condition information of the target vehicle, it is necessary to comprehensively consider the road condition information and driving decision information corresponding to the target driving path to make a prediction. The prediction can be made through an algorithm model, and the algorithm model is trained through route information. The trained algorithm model can make a decision based on the actual road condition information, whether to maintain the direction unchanged and continue driving, or to drive around obstacles, and predict the next position based on the decision result.
[0055] Traffic condition information may refer to the distribution of pedestrians and vehicles on the road, traffic lights, road information, etc., which may be set according to actual conditions and are not limited here.
[0056] Step 104, determining a planned path of the target vehicle according to the current position and the next position of the target vehicle.
[0057] After the current position and the next position are determined, the current position and the next position may be connected by a straight line to determine the planned path of the target vehicle. Optionally, step 104 includes:
[0058] Constructing a first path between the current position and the next position according to the current position and the next position of the target vehicle;
[0059] Based on the next position and the road condition information of the target vehicle, predicting a first position of the target vehicle, the first position being a next position adjacent to the next position;
[0060] A second path between the next position and the first position is constructed, and the planned path includes the first path and the second path.
[0061] Specifically, the first path can be a line between the current position and the next position. According to the next position and the road condition information of the target vehicle, the first position of the target vehicle is predicted. The prediction method can refer to the prediction method of the next position above, which will not be repeated here.
[0062] The distance between the next position and the first position can be regarded as a straight-line distance. The two position points are connected by a straight line to obtain a second path. Similarly, the second position can be predicted. The second position is the next position of the first position. The distance between the first position and the second position is regarded as a straight-line distance. The two positions are connected by a straight line to obtain a planned path. The planned path can be displayed on the vehicle's display screen, or the planned path can be transmitted to the control center for the next step of processing.
[0063] The method in this embodiment obtains route information, the route information includes multiple driving paths, and the captured images captured by the vehicle on each of the driving paths; obtains the real-time image currently captured by the target vehicle, and matches the real-time image with the captured image to determine the target driving path from the multiple driving paths; predicts the next position of the target vehicle based on the target driving path; and determines the planned path of the target vehicle based on the current position and the next position of the target vehicle. Through the above steps, a planned path can be generated for the vehicle, and the vehicle can travel according to the planned path, thereby improving the safety of the vehicle's automatic driving.
[0064] In some embodiments of the present application, the method further includes:
[0065] Annotating the collected image to obtain an annotated image;
[0066] Using the labeled images to train a network model to obtain a classification model;
[0067] The step of matching the real-time image with the collected image to determine a target driving path from the multiple driving paths includes:
[0068] Inputting the real-time image into the classification model to obtain the classification of the real-time image;
[0069] The target driving path is determined based on the classification of the real-time image and matching with the collected image.
[0070] Specifically, the collected images in the collected route information are classified and labeled according to the objects in the collected images, for example, cars, billboards, pedestrians, green plants, etc. These classifications are labeled. For example, according to the objects in the collected images, the collected images are classified, including the collected images of cars are put into one category, and the collected images including billboards are put into another category, and so on. It should be noted that if the same collected image includes billboards and pedestrians, the collected image can belong to both the billboard category and the pedestrian category. The classified collected image is called a labeled image. The labeled image is used to train the self-training algorithm model, that is, the network model, to obtain a classification model. The self-training algorithm model can select the CenterNet model. The backbone (feature extraction network) used by the model can be hourglass or ResNet50. If ResNet50 is used, a ResNet50 can be trained through moco, a model for training backbone, and then the parameters of this ResNet50 are replaced with the parameters of ResNet50 in CenterNet for training to try to improve the classification accuracy. After that, some unlabeled images, such as the collected images, are put into the trained CenterNet model to reduce the labeled images, that is, to reduce the number of objects prompted to the machine in the images, and the classification model is further trained by collecting images, and so on, until the trained images have no labels. The classification model obtained by such training can accurately classify any unlabeled image.
[0071] The real-time image is input into the classification model to obtain the classification of the real-time image, and the classification of the real-time image is matched with the collected image. Specifically, the classification of the real-time image is matched with the collected image to determine the target driving path, including:
[0072] Matching the real-time image with a first collected image in the route information, where the first collected image and the real-time image belong to the same category;
[0073] If the target image in the first collected image matches the real-time image, the driving path corresponding to the target image is determined as the target driving path.
[0074] Specifically, the collected images in the route information can be divided into multiple categories according to the objects included in the images. After determining the category of the real-time image, the real-time image can be matched with each image in the first collected image. The first collected image and the real-time image belong to the same category. In this way, for the collected images in the categories different from the real-time image, no matching is required, thereby improving the matching efficiency. If the target image in the first collected image matches the real-time image, the driving path corresponding to the target image is determined as the target driving path.
[0075] The path planning method provided in this application is illustrated below by way of example.
[0076] Step 1: Install cameras in different directions on the test vehicle, including front-view cameras and surround-view cameras, to obtain different vehicle conditions around the vehicle. When driving manually, use OpenCV to record different scene images of the driving route, and make the best and safest decision under different vehicle conditions. After collecting the image, grayscale the image, and analyze the real-world position of the vehicle body through post-processing to match the route of the driving process with the image one by one, and then save the matched route and image as an expert image library. In this way, the challenges in the longitudinal planning mentioned above can be solved. The white car parked on the side of the road suddenly starts, and the human driver will make the best decision in this case. There are many decisions: 1. Slow down and honk to signal the white car to give way, and then speed up to pass. 2. Slow down, wait for the white car to pass, and then speed up. Provide subsequent algorithm models for learning.
[0077] Of course, for the automatic driving of the machine, it is best to choose the best decision. You can choose the second option to ensure safe driving to the greatest extent, or you can combine the two options and give different weights to let the machine learn and choose different options for driving. Then you can collect human-driven routes and environmental images on the routes in other different scenarios like this. Finally, as mentioned above, the human-driven routes and images on the routes will be processed by 3D post-processing, and the images corresponding to the routes will be classified and counted into an expert database, and the route information will be stored in the expert database.
[0078] Step 2: Classify and annotate the collected expert route images according to the physical objects in the images, for example, cars, billboards, pedestrians, green plants, etc. These classifications are annotated to let the machine know what these annotated images represent. Put these annotated images into the self-training algorithm model for training to generate a model. The self-training algorithm model here can choose the CenterNet model. The backbone (feature extraction network) used by the model can be hourglass or ResNet50. If ResNet50 is used, a ResNet50 can be trained through moco, a training backbone model, and then the parameters of this ResNet50 are replaced with the parameters of ResNet50 in CenterNet for training to try to improve the classification accuracy. After that, put some unlabeled images into the trained CenterNet model to reduce the labeled images, that is, to reduce the prompts to the machine about what is in the image, and train the model until there are no labeled images in the training images, and only unlabeled images. After that, the trained mcga model can still accurately classify the objects in the image.
[0079] Step three, use OpenCV to take the images captured by the autonomous vehicle as input into the final mcga model. The model matches the image changes in the expert image library with the images of a series of changes in the surrounding environment it captures. After matching the image with the same changes in the expert image library, it determines the roadmap in the image library that matches the image. Because the images in the expert library and the roadmap have been matched one by one, a roadmap will be obtained quickly. Through the changes in the roadmap, the next position image where the vehicle should travel is predicted.
[0080] Step 4: Analyze the current position of the autonomous driving vehicle and the terminal position predicted in step 3 (the next position to be driven to) through 3D post-processing, and combine the distance between the two positions, which are directly regarded as a straight line. Because the route and image are matched frame by frame, the position change of the vehicle will be very small. After multiple transformations of the driving position, there will be multiple straight lines. Connecting them can get a route. Both straight and curved paths can be obtained through this method. In this way, the planned path of the vehicle is obtained. The planned path can be displayed on the display screen for the driver's reference. The planned path can also be transmitted to the control center to control the car to drive according to the generated path.
[0081] The above steps can be summarized as follows: Figure 2 As shown, the following steps are included:
[0082] Step 201, obtaining an expert database, wherein the expert database stores route information, the route information including a plurality of driving paths, and a captured image captured by a vehicle on each of the driving paths.
[0083] Step 202, using the expert database to train the network model, i.e., the self-training algorithm model;
[0084] Step 203, obtaining a trained model;
[0085] Step 204, acquiring a real-time image;
[0086] Step 205, obtaining the real-time position of the vehicle, ie, the current position, according to the real-time image.
[0087] Step 206, generating a model to predict the classification of the real-time image based on the real-time image;
[0088] Step 207, according to the classification of the real-time image, matching with the collected image in the expert database, determining the target driving path, and predicting the next position according to the target driving path;
[0089] Step 208, determining a planned path based on the real-time position and the predicted next position.
[0090] The path planning method provided in the present application does not use an optimization problem to solve the path planning problem, but converts it into a prediction matching problem. Through the learning ability of the algorithm model, as long as the collected images stored in the expert library are fully learned, the vehicle can learn various road conditions from various decisions made by human driving. This method is a method based on real scenarios rather than a formula-based method. There is no need to consider how to use formulas to represent real paths and road conditions, etc., and learning is directly carried out through the collected real images. It is equivalent to only requiring people to teach the vehicle how to make decisions for the first time, and the vehicle will make decisions by itself after learning, so that the path planning of autonomous driving is more based on real driving conditions, and it is easier to make the best and safest decisions under complex road conditions.
[0091] Figure 3 FIG. 1 shows a structural diagram of a path planning device provided in an embodiment of the present application. Figure 3 As shown, the path planning device 300 includes:
[0092] The acquisition module 301 is used to acquire route information, wherein the route information includes a plurality of driving paths and an acquired image acquired by the vehicle on each of the driving paths;
[0093] The first acquisition module 302 is used to acquire the real-time image currently collected by the target vehicle;
[0094] A matching module 303, used for matching the real-time image with the collected image to determine a target driving path from the multiple driving paths;
[0095] A prediction module 304 is used to predict the next position of the target vehicle based on the target driving path;
[0096] The determination module 305 is used to determine the planned path of the target vehicle according to the current position and the next position of the target vehicle.
[0097] Optionally, the matching module 303 includes:
[0098] A first acquisition submodule is used to acquire a plurality of real-time images of the target vehicle currently collected within a preset time period;
[0099] The first determination submodule is used to match the multiple real-time images with the collected images one by one to determine a target driving path from the multiple driving paths.
[0100] Optionally, the device further comprises:
[0101] A second acquisition module is used to annotate the collected image to obtain an annotated image;
[0102] A third acquisition module is used to train the network model using the labeled image to obtain a classification model;
[0103] The matching module 303 includes:
[0104] A second acquisition submodule is used to input the real-time image into the classification model to obtain the classification of the real-time image;
[0105] The matching submodule is used to match the collected image based on the classification of the real-time image to determine the target driving path.
[0106] Optionally, the matching submodule includes:
[0107] a matching unit, configured to match the real-time image with a first collected image in the route information, wherein the first collected image and the real-time image belong to the same category;
[0108] A determination unit is configured to determine a driving path corresponding to the target image as the target driving path if the target image in the first collected image matches the real-time image.
[0109] Optionally, the prediction module 304 includes:
[0110] A second determination submodule, used to determine the current position of the target vehicle on the target driving path according to the real-time image;
[0111] The first prediction submodule is used to predict the next position of the target vehicle according to the current position and the road condition information of the target vehicle.
[0112] Optionally, the determining module 305 includes:
[0113] A first construction submodule, configured to construct a first path between the current position and the next position according to the current position and the next position of the target vehicle;
[0114] A second prediction submodule, configured to predict a first position of the target vehicle based on the next position and the road condition information of the target vehicle, wherein the first position is a next position adjacent to the next position;
[0115] The second construction submodule is used to construct a second path between the next position and the first position, and the planned path includes the first path and the second path.
[0116] Optionally, the acquisition module 301 includes:
[0117] A collection submodule, used for obtaining, during the driving process of the collection vehicle, a driving path of the collection vehicle under different road conditions and a plurality of collection images during the driving process;
[0118] A third determination submodule is used to analyze each of the collected images to determine the position of the collected vehicle on the driving path;
[0119] The fourth determination submodule is used to establish a corresponding relationship between each of the captured images and the position of the captured vehicle on the driving path.
[0120] The path planning device 300 provided in the embodiment of the present application can implement the various processes implemented in the aforementioned path planning method embodiment and achieve the same technical effect. To avoid repetition, it will not be repeated here.
[0121] Figure 4 A schematic diagram of the hardware structure of the path planning method provided in an embodiment of the present application is shown.
[0122] The electronic device may include a processor 601 and a memory 602 storing computer program instructions.
[0123] Specifically, the processor 601 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.
[0124] The memory 602 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 602 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 602 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 602 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 602 is a non-volatile solid-state memory.
[0125] The memory may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, typically, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to the first aspect of the present disclosure.
[0126] The processor 601 implements any one of the path planning methods in the above embodiments by reading and executing computer program instructions stored in the memory 602 .
[0127] In one example, the electronic device may further include a communication interface 603 and a bus 610. Figure 4 As shown, the processor 601, the memory 602, and the communication interface 603 are connected via a bus 610 and communicate with each other.
[0128] The communication interface 603 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.
[0129] Bus 610 includes hardware, software or both, and the parts of path planning method are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 610 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.
[0130] In addition, in combination with the path planning method in the above embodiments, the present application embodiment can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the path planning methods in the above embodiments is implemented.
[0131] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.
[0132] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.
[0133] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.
[0134] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.
[0135] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.
Claims
1. A path planning method, characterized in that: The method comprises: Acquiring route information, the route information including a plurality of driving paths and a captured image acquired by the vehicle on each of the driving paths; Acquire a real-time image currently collected by the target vehicle, and match the real-time image with the collected image to determine a target driving path from the multiple driving paths; Based on the target driving path, predicting the next position of the target vehicle; A planned path of the target vehicle is determined according to the current position and the next position of the target vehicle.
2. The method according to claim 1, characterized in that: The step of acquiring a real-time image currently collected by the target vehicle and matching the real-time image with the collected image to determine a target driving path from the multiple driving paths includes: Acquire multiple real-time images of the target vehicle currently collected within a preset time period; The multiple real-time images are matched one by one with the collected images to determine a target driving path from the multiple driving paths.
3. The method according to claim 1, characterized in that The method further comprises: Annotating the collected image to obtain an annotated image; Using the labeled images to train a network model to obtain a classification model; The step of matching the real-time image with the collected image to determine a target driving path from the multiple driving paths includes: Inputting the real-time image into the classification model to obtain the classification of the real-time image; The target driving path is determined based on the classification of the real-time image and matching with the collected image.
4. The method according to claim 3, characterized in that The classification based on the real-time image is matched with the collected image to determine the target driving path, including: Matching the real-time image with a first collected image in the route information, where the first collected image and the real-time image belong to the same category; If the target image in the first collected image matches the real-time image, the driving path corresponding to the target image is determined as the target driving path.
5. The method according to claim 1, characterized in that The predicting the next position of the target vehicle based on the target driving path includes: Determining the current position of the target vehicle on the target driving path according to the real-time image; The next position of the target vehicle is predicted based on the current position and the road condition information of the target vehicle.
6. The method according to claim 1, characterized in that Determining the planned path of the target vehicle according to the current position and the next position of the target vehicle includes: Constructing a first path between the current position and the next position according to the current position and the next position of the target vehicle; Based on the next position and the road condition information of the target vehicle, predicting a first position of the target vehicle, the first position being a next position adjacent to the next position; A second path between the next position and the first position is constructed, and the planned path includes the first path and the second path.
7. The method according to claim 1, characterized in that The step of obtaining route information includes: During the driving process of the collection vehicle, a driving path of the collection vehicle under different road conditions and a plurality of collection images during the driving process are obtained; Analyze each of the collected images to determine the position of the collected vehicle on the driving path; A corresponding relationship between each of the captured images and the position of the captured vehicle on the driving path is established.
8. A path planning device, characterized in that: The device comprises: A collection module, used to obtain route information, the route information includes multiple driving paths, and a collection image obtained by the vehicle on each of the driving paths; An acquisition module is used to acquire the real-time image currently collected by the target vehicle; A matching module, used for matching the real-time image with the collected image to determine a target driving path from the multiple driving paths; A prediction module, used to predict the next position of the target vehicle based on the target driving path; The determination module is used to determine the planned path of the target vehicle according to the current position and the next position of the target vehicle.
9. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program instructions.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.
