A method and system for determining whether a vehicle is on an elevated road

Abstract

The embodiments of this specification disclose a method and system for determining whether a vehicle is on an elevated highway. The method comprises: obtaining the vehicle's location; determining, based on the location, whether the vehicle meets a trigger condition; in response to the trigger condition being met, obtaining an image of the vehicle's surroundings; and, based on the image of the surroundings, determining, using an elevated highway recognition model, a confidence level that the vehicle is on and / or not on the elevated highway.

Description

Technical Field

[0001] This specification relates to the field of navigation, and in particular to a method and system for determining whether a vehicle is on an elevated road. Background Art

[0002] With the development of mobile internet and the popularization of automobiles, electronic maps provide navigation through global positioning technology and a range of navigation methods, and real-time route determination and modification are made through the mobile internet. Drift occurs when a vehicle's actual route deviates from its originally planned route. Elevated roads, as a unique road scenario, are more challenging to identify yaw on than conventional routes due to their elevated, three-dimensional form. When yaw occurs on an elevated road, failure to promptly identify the elevated road's upper and lower levels can lead to a series of issues with subsequent guidance and broadcasting functions, resulting in a poor driving experience, navigation detours, and numerous other issues related to driving safety.

[0003] Based on this, there is an urgent need for an effective method to determine whether a vehicle is on an elevated road. Summary of the Invention

[0004] One embodiment of this specification provides a method for determining whether a vehicle is on an elevated highway. The method comprises: obtaining the vehicle's location; determining, based on the location, whether the vehicle meets a trigger condition; in response to the trigger condition being met, obtaining an image of the vehicle's environment; and, based on the image of the environment, determining, using an elevated highway recognition model, a confidence level that the vehicle is on and / or not on the elevated highway.

[0005] One of the embodiments of this specification provides a system for determining whether a vehicle is on an elevated road, the system comprising: a position acquisition module for acquiring the position of the vehicle; a judgment module for determining whether the vehicle satisfies a trigger condition based on the position; an image acquisition module for acquiring an environmental image of the vehicle in response to satisfying the trigger condition; and a confidence determination module for determining, based on the environmental image, the confidence of whether the vehicle is on the elevated road and / or not on the elevated road through an elevated road recognition model.

[0006] One embodiment of the present specification provides a device for determining whether a vehicle is on an elevated road, comprising a processor configured to execute a method for determining whether a vehicle is on an elevated road.

[0007] One embodiment of this specification provides a computer-readable storage medium, which stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes a method for determining whether a vehicle is on an elevated road. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein:

[0009] Figure 1 This is a schematic diagram of an application scenario of a system for determining whether a vehicle is on an elevated road according to some embodiments of this specification;

[0010] Figure 2 is a block diagram of a system for determining whether a vehicle is on an elevated road according to some embodiments of this specification;

[0011] Figure 3 is an exemplary flow chart of a method for determining whether a vehicle is on an elevated road according to some embodiments of this specification;

[0012] Figure 4 is a schematic diagram of an elevated recognition model according to some embodiments of this specification;

[0013] Figure 5 is a schematic diagram of a training method for an elevated recognition model according to some embodiments of this specification. DETAILED DESCRIPTION

[0014] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0015] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0016] As used in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but also include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0017] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0018] Existing methods use the horizontal latitude and longitude and altitude provided by GPS, but there are three main problems: first, in elevated road networks, the intersection distance between roads at different altitudes is close, and the longitude and latitude errors are very small, which are outside the recognition range and cannot be recognized; second, the altitude provided by GPS is unstable and the error is larger, making it impossible to accurately locate the main and auxiliary roads in the elevated bridge area; and all existing methods have lags and cannot make pre-judgments, making it difficult to provide users with advance warnings.

[0019] The method and system for determining whether a vehicle is on an elevated road in one or more embodiments of this specification can be applied to various scenarios, such as navigation correction, route planning, etc.

[0020] This method and system for determining whether a vehicle is on an elevated highway can implement one or more functions, such as using a machine learning model to determine whether a vehicle is on an elevated highway. This method and system can more accurately determine the vehicle's actual driving position, effectively correct positioning and yaw, and achieve one or more beneficial effects.

[0021] It should be understood that the application scenarios of the method and system for determining whether a vehicle is on an elevated road in this specification are merely some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios without creative work.

[0022] Figure 1 It is a schematic diagram of an application scenario of a system for determining whether a vehicle is on an elevated road according to some embodiments of this specification.

[0023] like Figure 1 As shown, the system 100 for determining whether a vehicle is on an elevated platform may include a server 110 , a processor 112 , a terminal 120 , a vehicle 130 , a storage device 140 , and a network 150 .

[0024] In some embodiments, server 110 can be used to process information and / or data related to system 100, for example, to acquire environmental images and determine the confidence level of a vehicle on and / or under an elevated platform. In some embodiments, server 110 can be a single server or a server group. The server group can be centralized or distributed (for example, server 110 can be a distributed system). In some embodiments, server 110 can be local or remote. For example, server 110 can access information and / or data stored in terminal 120, vehicle 130, or storage device 140 via network 150. For another example, server 110 can be directly connected to terminal 120, vehicle 130, and / or storage device 140 to access stored information and / or data. In some embodiments, server 110 can be implemented on a cloud platform or provided virtually. By way of example only, the cloud platform can include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, or any combination thereof. In some embodiments, server 110 can be deployed on terminal 120 (for example, the processor of terminal 120 can serve as server 110 or a portion thereof). In some embodiments, the server 110 may be deployed on a processing device of an online service platform (e.g., the processing device of the online service platform may serve as the server 110 or a portion thereof). In some embodiments, the server 110 may be deployed on a vehicle 130 (e.g., the processor of the vehicle 130 may serve as the server 110 or a portion thereof).

[0025] In some embodiments, server 110 may include a processor 112. Processor 112 may process information and / or data related to system 100 to perform one or more of the functions described herein. For example, processor 112 may obtain a vehicle's location, obtain an image of the vehicle's surroundings, and / or determine whether the vehicle is on or off an elevated highway. In some embodiments, processor 112 may include one or more processing engines (e.g., a single-chip processing engine or a multi-chip processing engine).

[0026] Terminal 120 refers to one or more terminal devices or software used by a user. In some embodiments, terminal 120 may be used by one or more users, including passengers, drivers, back-office maintenance personnel, etc. In some embodiments, the acquisition of environmental images can be triggered by terminal 120. In some embodiments, navigation information can be presented to the user through terminal 120, for example, through the user's mobile phone or through a display. In some embodiments, terminal 120 can be one of other devices with input and / or output functions, such as mobile device 120-1, tablet computer 120-2, laptop computer 120-3, desktop computer 120-4, or any combination thereof. In some embodiments, mobile device 120-1 may include a mobile phone, a smart phone, a personal digital assistant (PDA), a navigation device, a handheld terminal (POS), etc., or any combination thereof. In some embodiments, desktop computer 120-4 may be an in-vehicle computer, an in-vehicle TV, etc.

[0027] Vehicle 130 may include a taxi, public bus, sports car, sedan, minivan, bus, school bus, RV, truck, van, tanker truck, etc. In some embodiments, vehicle 130 may be equipped with a camera, such as a dashcam or external camera. In some embodiments, vehicle 130 may be equipped with a positioning device, such as a GPS positioning device, a Bluetooth positioning device, or a Wi-Fi positioning device. In some embodiments, the camera may transmit captured images to server 110 via network 150, and the positioning device may transmit positioning information to server 110 via network 150.

[0028] The storage device 140 can be used to store data and / or instructions related to the system 100. In some embodiments, the storage device 140 can store data obtained from the terminal 120 and / or the vehicle 130. In some embodiments, the storage device 140 can store historical data, image data, training samples, etc. In some embodiments, the storage device 140 can store data and / or instructions used by the server 110 to execute or use to complete the exemplary methods described in this specification. In some embodiments, the storage device 140 may include one or a combination of large-capacity memory, removable memory, volatile read-write memory, read-only memory (ROM), etc. In some embodiments, the storage device 140 can be implemented through the cloud platform described in this specification. For example, the cloud platform may include one or a combination of private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, cross-cloud, multi-cloud, etc. In some embodiments, the storage device 140 can be deployed on the server 110, the terminal 120, and / or the vehicle 130. In some embodiments, the storage device 140 may be deployed independently, and the server 110 , the terminal 120 , and / or the vehicle 130 may access the storage device 140 directly or through the network 150 to obtain relevant data and / or instructions.

[0029] Network 150 can facilitate the exchange of information and / or data. In some embodiments, one or more components of system 100 (e.g., server 110, vehicle 130, storage device 140) can transmit information and / or data to other components of system 100 via network 150. For example, vehicle 130 can transmit vehicle location, environmental images, etc. to server 110 via network 150. In some embodiments, system 100 can include one or more network access points, such as base stations and / or wireless access points 150-1, 150-2, etc. One or more components of system 100 can connect to network 150 to exchange data and / or information.

[0030] It should be noted that system 100 is provided for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art will readily appreciate that various modifications and variations can be made based on the description of this specification. For example, system 100 may further include a database. For another example, system 100 may implement similar or different functions on other devices. However, such variations and modifications do not deviate from the scope of this specification.

[0031] Figure 2 It is a block diagram of a system for determining whether a vehicle is on an elevated road according to some embodiments of this specification.

[0032] like Figure 2As shown, the system 200 for determining whether a vehicle is on an elevated road may include a position acquisition module 210 , a determination module 220 , an image acquisition module 230 , and a confidence determination module 240 .

[0033] The location acquisition module 210 may be used to acquire the location of the vehicle.

[0034] The determination module 220 may be configured to determine whether the vehicle meets a trigger condition based on the location.

[0035] In some embodiments, the triggering condition may include at least one of the vehicle yaw or suspected yaw, the vehicle is uphill, the vehicle is downhill, the vehicle is approaching or arriving at an elevated entrance, the vehicle is approaching or arriving at an elevated exit, etc.

[0036] The image acquisition module 230 may be configured to acquire an image of the vehicle's environment in response to a trigger condition being met.

[0037] In some embodiments, the image acquisition module 230 can be used to capture environmental images through the vehicle's driving recorder or the vehicle's external camera; or retrieve environmental images captured by the vehicle's driving recorder or the vehicle's external camera from a memory.

[0038] The confidence determination module 240 may be configured to determine the confidence level of whether the vehicle is on an elevated highway and / or is not on an elevated highway using an elevated highway recognition model based on the environment image.

[0039] In some embodiments, the overhead recognition model may include a neural network model.

[0040] In some embodiments, the confidence determination module 240 can be used to calibrate the vehicle's positioning, determine whether the vehicle is yawed, correct the vehicle's yaw, provide a yaw warning to the vehicle, and provide induction broadcasts to the vehicle based on the confidence.

[0041] In some embodiments, the confidence determination module 240 can be used to indicate that the vehicle has veered off course; and / or to indicate that the vehicle has corrected the veer off course.

[0042] In some embodiments, system 200 may include a training module 250. Training module 250 may be configured to input training samples from a training sample set into an elevated highway recognition model, and adjust parameters of the elevated highway recognition model based on the difference between the predicted labels output by the elevated highway recognition model and the labeled labels of the training samples, thereby obtaining a trained elevated highway recognition model. The labeled labels of the training samples may be determined based on the following steps: constructing a training sample set, the training sample set may include at least one of a sample environment image of a sample vehicle traveling on an uphill section of an elevated highway, a sample environment image of a sample vehicle traveling on a mid-section of an elevated highway, and a sample environment image of a sample vehicle traveling on a downhill section of an elevated highway; and for each training sample in the training sample set, labeling the training sample based on the section of the highway in which the sample vehicle is traveling, thereby determining a label corresponding to each training sample. In some embodiments, training module 250 and confidence determination module 240 (and / or location acquisition module 210, judgment module 220, and image acquisition module 230) may be deployed on different processing devices or processors. For example, the training module 250 can train the elevated highway recognition model offline on one or more processing devices; and the confidence determination module 240 can apply the elevated highway recognition model online on another processing device to determine the confidence that the vehicle is on the elevated highway and / or not on the elevated highway.

[0043] In some embodiments, the driving section of the sample vehicle in each training sample can be obtained through road network information.

[0044] It should be noted that the above description of the system 200 and its modules is for convenience only and does not limit this specification to the scope of the embodiments. It is understood that those skilled in the art, after understanding the principles of the system, may arbitrarily combine the modules or form subsystems connected to other modules without departing from the principles. In some embodiments, Figure 1 The location acquisition module 210, judgment module 220, image acquisition module 230, and confidence determination module 240 disclosed herein may be separate modules within a single system, or a single module may implement the functions of two or more of the aforementioned modules. For example, each module may share a storage module, or each module may have its own storage module. Such variations are within the scope of protection of this specification.

[0045] Figure 3 This is an exemplary flow chart of a method for determining whether a vehicle is on an elevated road according to some embodiments of this specification.

[0046] Step 310: Acquire the location of the vehicle.

[0047] In some embodiments, the location acquisition module 210 can acquire the vehicle's location in a variety of ways, such as GPS positioning, base station positioning, Bluetooth positioning, etc. For another example, the vehicle's location can be estimated based on the planned route and / or the elapsed travel time of the vehicle.

[0048] Step 320: Determine whether the vehicle meets the triggering condition based on the location.

[0049] The trigger condition refers to a preset condition that can trigger a corresponding operation (for example, identifying whether a vehicle is on or off an elevated road). The trigger condition can be set by the judgment module 220 or customized by the user.

[0050] In some embodiments, the triggering conditions may include at least one of the vehicle yaw or suspected yaw, the vehicle is uphill, the vehicle is downhill, the vehicle is approaching or arriving at an elevated entrance, the vehicle is approaching or arriving at an elevated exit, the vehicle is approaching or arriving at an elevated preset range, etc.

[0051] Vehicle deviation refers to a situation where the vehicle has deviated from the planned route. For example, the closest distance between the vehicle's position and the planned route is greater than a first preset threshold, which can be set based on experience or needs.

[0052] Suspected deviation refers to situations where the vehicle may have deviated from the planned route. For example, the closest distance between the vehicle's position and the planned route is within a critical value, or between the critical value and a first preset threshold. The critical value can be set based on experience or needs. Another example is when the vehicle's position cannot be obtained due to weather conditions, building obstructions, or other factors, which can be considered suspected deviation.

[0053] In some embodiments, the determination module 220 can determine (e.g., using map data) whether the vehicle is on an uphill and / or downhill slope. For example, if the vehicle's position is on an uphill section indicated by the map data, the vehicle is on an uphill slope. For example, if the vehicle's position is on a downhill section indicated by the map data, the vehicle is on a downhill slope. In some embodiments, the determination module 220 can determine whether the vehicle is on an uphill and / or downhill slope based on the height difference between the front and rear ends of the vehicle. In some embodiments, the determination module 220 can determine whether the vehicle is on an uphill and / or downhill slope using a sensor, such as a slope sensor.

[0054] A vehicle approaching or arriving at an elevated entrance means that the distance between the vehicle's location and the nearest elevated entrance is less than a second preset threshold. A vehicle approaching or arriving at an elevated exit means that the distance between the vehicle's location and the nearest elevated exit is less than a second preset threshold. The second preset threshold can be set based on experience or needs.

[0055] The preset range of the elevated highway can be a range of a preset width extending to both sides of the elevated highway. The preset width can be set based on experience or needs, for example, 5 meters, 10 meters, etc. A vehicle approaching or reaching the preset range of the elevated highway can include various situations, such as when the vehicle is traveling on an elevated auxiliary road, when the vehicle is traveling on a road section below the elevated highway that is perpendicular to the direction of travel of the elevated highway, etc.

[0056] In some embodiments, the triggering conditions may also include timed triggering, for example, every five minutes, at the quarter point, half point, or three-quarter point in the vehicle's journey. In some embodiments, the triggering conditions may also include manual triggering, for example, by the driver, passenger, user requesting a ride on behalf of another, or backend staff.

[0057] In some embodiments, when the determination module 220 determines that the vehicle meets the above triggering conditions, steps 330 - 340 are executed to determine whether the vehicle is on an elevated road.

[0058] Step 330: Acquire an environmental image of the vehicle.

[0059] An environmental image refers to an image of the vehicle's surroundings. The environmental image can be one or more of a grayscale image, an RGB image, an infrared image, an optical flow image, and the like. The field of view of the environmental image can be one or more of the following: directly in front of the vehicle, to the left front of the vehicle, to the right front of the vehicle, diagonally below the vehicle, diagonally above the vehicle, to the side of the vehicle, behind the vehicle, or to the side and rear of the vehicle. In some embodiments, the acquired environmental image can be an image obtained with the user's consent or an image reported by the user.

[0060] In some embodiments, the image acquisition module 230 can capture the environment image through the vehicle's driving recorder or the vehicle's external camera. For example, the image acquisition module 230 can start or trigger the vehicle's driving recorder or the vehicle's external camera to capture the environment image and acquire the environment image.

[0061] In some embodiments, the image acquisition module 230 can retrieve environmental images captured by the vehicle's driving recorder or an external camera of the vehicle. For example, the vehicle's driving recorder or the vehicle's external camera can capture environmental images at a predetermined shooting frequency (e.g., every 5 seconds) or record videos while the vehicle is in motion. In some embodiments, the image acquisition module 230 can retrieve the environmental image corresponding to the moment when the trigger condition is met from multiple environmental images or videos captured by the vehicle's driving recorder or the vehicle's external camera.

[0062] In some embodiments, the image acquisition module 230 can acquire environmental images from a driving recorder or an external camera of the vehicle via a network. Alternatively, the image acquisition module 230 can be integrated with the driving recorder or the external camera of the vehicle. In some embodiments, the image acquisition module 230 can acquire environmental images from the driving recorder or the external camera of the vehicle via a bus. In some embodiments, the environmental images can be acquired through an interface, which can include but is not limited to a program interface, a data interface, a transmission interface, etc. For example, when determining whether the vehicle is on an elevated system, the environmental images can be automatically extracted from the interface.

[0063] Step 340 : Based on the environment image, determine the confidence level of whether the vehicle is on the elevated road and / or not on the elevated road by using the elevated road recognition model.

[0064] In some embodiments, the confidence determination module 240 may input the environment image into the elevated road recognition model to determine the confidence that the vehicle is on the elevated road and / or not on the elevated road. The confidence that the vehicle is not on the elevated road may include the vehicle being under the elevated road, next to the elevated road, or far away from the elevated road.

[0065] For a description of the overhead recognition model and confidence level, see Figure 4 For the training of the overhead recognition model, see Figure 5 .

[0066] In some embodiments, the confidence determination module 240 may also calibrate the positioning of the vehicle, determine whether the vehicle is yawed, correct the yaw of the vehicle, provide a yaw warning to the vehicle, and / or make an inductive announcement to the vehicle based on the confidence.

[0067] Guidance announcements can refer to announcements of guidance point information upon arrival at or approaching a guidance point. For example, announcements such as "Traffic violation photo taken here" or "Speed limit 50, you are speeding" can be made. In some embodiments, the confidence determination module 240 can make guidance announcements based on confidence levels. For example, if the confidence level of a vehicle on an elevated highway is 0.81, the announcement "You have entered the xx elevated highway" can be made.

[0068] In some embodiments, the confidence determination module 240 can indicate that the vehicle has yawed based on the confidence level and / or indicate that the vehicle has corrected its yaw. For example, if, based on the planned route, the vehicle is not on the elevated road at the time the trigger condition is met after measuring the vehicle's actual speed for a certain period of time, and the elevated road recognition model identifies that the trigger condition is met, and the confidence level (e.g., 0.81) of the vehicle being on the elevated road is greater than a third preset threshold (e.g., 0.8), the system 200 indicates that the vehicle has yawed. After determining that the vehicle has yawed, the system 200 can determine that the vehicle is on the elevated road based on the confidence level (e.g., 0.81). Based on the vehicle's position on the elevated road, the system 200 can replan or update the planned route to correct the yaw and indicate (e.g., to the terminal 120 or vehicle 130) that the vehicle has corrected its yaw. For another example, if the elevated road recognition model identifies that the trigger condition is met, the confidence level (e.g., 0.19) of the vehicle being on the elevated road is less than a fourth preset threshold (e.g., 0.2), and the vehicle's position has returned to or is close to the planned route, the system 200 indicates that the vehicle has corrected its yaw.

[0069] It should be noted that the above description of process 300 is for illustration and purpose only and does not limit the scope of application of this specification. Those skilled in the art may make various modifications and alterations to process 300 under the guidance of this specification. However, such modifications and alterations are still within the scope of this specification.

[0070] Figure 4 is a schematic diagram of an elevated recognition model according to some embodiments of this specification.

[0071] An elevated highway recognition model may refer to a machine learning model that identifies whether a vehicle is on an elevated highway.

[0072] The input of the elevated overpass recognition model may include an image of the environment, for example, an image directly in front of the vehicle. The output of the elevated overpass recognition model may include the confidence that the vehicle is on the elevated overpass and / or not on the elevated overpass (or under the elevated overpass), for example, the confidence that the vehicle is on the elevated overpass is 0.81. In some embodiments, the elevated overpass recognition model may be a binary classification model, and its output may include the confidence that the vehicle is on the elevated overpass and the confidence that the vehicle is not on the elevated overpass. In some embodiments, the sum of the confidence that the vehicle is on the elevated overpass and the confidence that the vehicle is not on the elevated overpass may be 1. For example, the confidence that the vehicle is on the elevated overpass is 0.81, and the confidence that the vehicle is not on the elevated overpass is 0.19.

[0073] The overhead recognition model may include at least a neural network model, such as a standard neural network (NN), a deep neural network (DNN), etc., or any combination thereof.

[0074] For example, Figure 4As shown, the elevated highway recognition model may include a deep neural network and a fully connected layer, wherein the deep neural network processes the environment image and extracts features of the environment image, and the fully connected layer processes the features of the environment image to determine the confidence that the vehicle is on the elevated highway and / or not on the elevated highway.

[0075] Figure 5 is a schematic diagram of a training method for an elevated recognition model according to some embodiments of this specification.

[0076] like Figure 5 As shown, the training module 250 can input the training samples in the training sample set into the elevated highway recognition model respectively, and adjust the parameters of the elevated highway recognition model according to the difference between the predicted label output by the elevated highway recognition model (for example, the confidence of the sample vehicle predicted by the elevated highway recognition model that it is on the elevated highway and / or the confidence of the vehicle not on the elevated highway) and the labeled label of the training sample, so as to obtain a trained elevated highway recognition model. Specifically, in some embodiments, it can be determined whether the value of the loss function between the predicted label and the labeled label of the training sample meets the training termination condition (for example, the value of the loss function converges or is less than the fifth preset threshold, and / or whether the number of iterations reaches the preset number of rounds); if so, the training is terminated to obtain a trained elevated highway recognition model; if not, the parameters of the elevated highway recognition model are adjusted, and the elevated highway recognition model after the adjusted parameters is further trained until the training termination condition is met, thereby obtaining a trained elevated highway recognition model.

[0077] In some embodiments, the training module can construct a training sample set based on a large number of labeled training samples. The training sample set may include at least one of a sample environment image of a sample vehicle traveling on an uphill section of an elevated highway (e.g., sample environment image 511), a sample environment image of a sample vehicle traveling on a middle section of an elevated highway (e.g., sample environment image 513), and a sample environment image of a sample vehicle traveling on a downhill section of an elevated highway (e.g., sample environment image 515). The labels of the training samples may include, for example: 1 (on the elevated highway), 0 (not on the elevated highway). In some embodiments, the training sample set may also include sample environment images of a sample vehicle traveling under the elevated highway (e.g., under the elevated highway (such as the main road or auxiliary road under the elevated highway), next to the elevated highway (such as the auxiliary road next to the elevated highway)). The sample vehicle may be the same or a different type of vehicle as the vehicle in step 310.

[0078] In some embodiments, for each training sample in the training sample set, the training module can annotate the training sample based on the road section where the sample vehicle in the corresponding training sample is located to determine the label corresponding to the training sample. Specifically, in some embodiments, the training module can infer whether the sample vehicle was on an elevated road at that time based on the road section where the sample vehicle is located and / or the trajectory of the sample vehicle, as well as the time point when the sample environment image was captured, and annotate the sample environment image. In some embodiments, the annotation of training samples can be automatically implemented by the training module.

[0079] In some embodiments, the road segment on which the sample vehicle is traveling and / or the trajectory of the sample vehicle can be obtained through road network information. Road network information can refer to information about a network of roads at all levels. In some embodiments, the road network information can be constructed and maintained by online ride-hailing service providers, navigation service providers, etc.

[0080] In some embodiments, the training module may clean the training sample set in one or more ways. For example, manual cleaning may be employed. In another example, the training module may perform object detection on the sample environment images and clean the training sample set based on the object detection results. In some embodiments, sample environment images where the object detection results do not include elevated elements (e.g., elevated entrances, elevated exits, elevated bridge piers, elevated road signs, elevated road sections, etc.) may be cleaned.

[0081] In some embodiments, the training module may input the training sample set into the initial overhead recognition model, and adjust the parameters of the overhead recognition model through the labels to obtain a trained overhead recognition model.

[0082] The initial overhead recognition model refers to an untrained, merely initialized neural network model. Initialization methods include minimum initialization and random initialization. The initial overhead recognition model has the same structure as the trained overhead recognition model, but has not yet learned the patterns of the training samples.

[0083] In some embodiments, the training module can randomly select a training set from the training samples and perform data augmentation on the training set; perform several iterative training on the initial elevated recognition model on the training set to obtain a trained elevated recognition model. The iterative training method may include: calculating the gradient of the loss function, and iteratively updating the parameters of the elevated recognition model by the gradient descent method. The loss function may include a cross entropy loss function, a mean square error loss function, an exponential loss function, a logarithmic loss function and / or a square loss function, etc. The gradient descent method may include a standard gradient descent method and / or a stochastic gradient descent method, etc. A variety of learning rate decay strategies may be used in iterative training, for example, piecewise decay, reverse time decay, exponential decay and / or adaptive decay, etc. When the loss function converges or is less than a fifth preset threshold, the iterative training may be terminated. Alternatively, when the number of iterations reaches a preset number of rounds, the iterative training may be terminated.

[0084] The beneficial effects that may be brought about by the embodiments of this specification may include but are not limited to: (1) setting trigger conditions and starting the judgment process after the trigger conditions are met, which can avoid redundant calculations; (2) using a trained neural network model to implement the judgment process, thereby improving the accuracy of yaw recognition; (3) automatically labeling training samples, thereby improving the training efficiency of the neural network model; (4) prompting users in real time based on the judgment results, thereby improving navigation efficiency and user experience.

[0085] While the basic concepts have been described above, it will be apparent to those skilled in the art that the detailed disclosure is merely illustrative and does not limit this specification. Although not explicitly stated herein, various modifications, improvements, and revisions to this specification may be made by those skilled in the art. Such modifications, improvements, and revisions are suggested in this specification and remain within the spirit and scope of the exemplary embodiments of this specification.

[0086] This specification also uses specific terms to describe the embodiments of this specification. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different locations in this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of this specification may be appropriately combined.

[0087] In addition, unless expressly stated in the claims, the order of the processing elements and sequences, the use of alphanumeric characters, or the use of other names described in this specification are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some of the invention embodiments currently considered useful through various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the spirit and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0088] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0089] In some embodiments, numbers are used to describe the quantity of components and attributes. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise stated, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the description and claims are approximate values, which may change according to the required characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the settings of such numerical values are as accurate as possible within the feasible range.

[0090] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, and documents, cited in this specification is hereby incorporated by reference in its entirety. This includes application history documents that are inconsistent with or conflict with the content of this specification, as well as documents (currently or subsequently attached to this specification) that limit the broadest scope of the claims of this specification. It should be noted that if the descriptions, definitions, and / or terminology used in the accompanying materials are inconsistent or conflicting with the content of this specification, the descriptions, definitions, and / or terminology used in this specification will control.

[0091] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A method for determining whether a vehicle is on an elevated road, characterized in that: The method comprises: obtaining the position of the vehicle; determining, based on the position, whether the vehicle meets a trigger condition; In response to satisfying the trigger condition, acquiring an image of the environment of the vehicle; Based on the environment image, determining the confidence level of whether the vehicle is on an elevated road and / or not on an elevated road by using an elevated road recognition model; Based on the confidence level, prompting that the vehicle has deviated; and / or prompting that the vehicle has corrected the deviating; The prompting that the vehicle has deviated includes: prompting that the vehicle has deviated in response to the vehicle not being on the elevated road at the time when the trigger condition is met according to the planned route, and the confidence level that the vehicle is on the elevated road is greater than a third preset threshold; Prompting that the vehicle has corrected the yaw includes: after determining that the vehicle has yawed, replanning the route based on the position of the vehicle on the elevated road, and prompting that the vehicle has corrected the yaw; or, when the triggering condition is met, the confidence level of the vehicle on the elevated road is less than a fourth preset threshold, and the vehicle position has returned to the planned route, prompting that the vehicle has corrected the yaw.

2. The method according to claim 1, characterized in that The triggering conditions include at least one of the vehicle yaw or suspected yaw, the vehicle is on an uphill slope, the vehicle is on a downhill slope, the vehicle is approaching or arriving at an elevated entrance, the vehicle is approaching or arriving at an elevated exit, and the vehicle is approaching or arriving at a preset range of an elevated highway.

3. The method according to claim 1, characterized in that The overhead recognition model includes a neural network model.

4. The method according to claim 3, characterized in that The elevated recognition model is obtained through training, including: Inputting the training samples in the training sample set into the overhead recognition model respectively, and adjusting the parameters of the overhead recognition model according to the difference between the predicted labels output by the overhead recognition model and the labels marked with the training samples, so as to obtain a trained overhead recognition model; The labeled labels of the training samples are determined based on the following steps: constructing a training sample set, the training sample set including at least one of a sample environment image of a sample vehicle traveling on an uphill section of an elevated road, a sample environment image of a sample vehicle traveling on a middle section of an elevated road, and a sample environment image of a sample vehicle traveling on a downhill section of an elevated road; for each training sample in the training sample set, labeling each training sample according to the traveling section of the sample vehicle in each training sample to determine the label corresponding to each training sample.

5. The method according to claim 4, characterized in that The driving section of the sample vehicle in each training sample is obtained through road network information.

6. The method according to claim 1, characterized in that The method further comprises: Based on the confidence level, at least one of calibrating the positioning of the vehicle, determining whether the vehicle is yawed, correcting the yaw of the vehicle, providing a yaw warning to the vehicle, and providing an inductive announcement to the vehicle is performed.

7. A system for determining whether a vehicle is on an elevated road, characterized in that: The system comprises: A position acquisition module, configured to acquire the position of the vehicle; a judgment module, configured to determine whether the vehicle meets a trigger condition based on the position; an image acquisition module, configured to acquire an image of the environment of the vehicle in response to satisfying the trigger condition; A confidence determination module, configured to determine, based on the environment image and using an elevated road recognition model, a confidence level that the vehicle is on an elevated road and / or is not on an elevated road; a confidence processing module, which prompts, based on the confidence, that the vehicle has deviated; and / or prompts that the vehicle has corrected the deviating course; The prompting that the vehicle has deviated includes: prompting that the vehicle has deviated in response to the vehicle not being on the elevated road at the time when the trigger condition is met according to the planned route, and the confidence level that the vehicle is on the elevated road is greater than a third preset threshold; Prompting that the vehicle has corrected the yaw includes: after determining that the vehicle has yawed, replanning the route based on the position of the vehicle on the elevated road, and prompting that the vehicle has corrected the yaw; or, when the triggering condition is met, the confidence level of the vehicle on the elevated road is less than a fourth preset threshold, and the vehicle position has returned to the planned route, prompting that the vehicle has corrected the yaw.

8. A device for determining whether a vehicle is on an elevated road, characterized in that: The apparatus comprises at least one processor and at least one memory; The at least one memory is for storing computer instructions; The at least one processor is configured to execute at least part of the computer instructions to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The storage medium stores computer instructions, and when at least part of the computer instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images