Traffic information determination method and device, electronic equipment and vehicle

Abstract

The invention relates to a traffic information determination method and device, electronic equipment and a vehicle, and relates to the technical field of data processing, the method comprises the following steps: obtaining a visual identification result and a map calibration result, the visual identification result is obtained by processing a multi-frame detection image comprising a target signboard through a target identification model, and the map calibration result is obtained by processing a multi-frame detection image comprising a target signboard; and according to the type of the target signboard and the environmental information, determining the confidence coefficient of the visual identification result, and then according to the visual identification result, the map calibration result and the confidence coefficient of the visual identification result, obtaining the target traffic information. The confidence of a visual identification result is determined through the type of the target signboard and the environment information, the visual identification result is obtained by processing multiple frames of detection images including the target signboard through the target identification model, the accuracy can be improved, and the accuracy of the visual identification result is improved based on the visual identification result, the map calibration result and the confidence of the visual identification result. And more accurate target traffic information can be obtained.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular, to a method, device, electronic device, and vehicle for determining traffic information. Background Art

[0002] An intelligent speed limit system refers to a system that obtains speed limit information on the current road through different methods and issues warnings based on the vehicle's own speed. In related technologies, traffic signs can be identified through cameras installed on the vehicle. However, the visual recognition of a single camera is greatly affected by external conditions such as weather, resulting in insufficient recognition accuracy. Summary of the Invention

[0003] The purpose of the present disclosure is to provide a traffic information determination method, device, electronic device and vehicle, which determine the confidence of the visual recognition result by the type of the target sign and environmental information, wherein the visual recognition result is obtained by processing multiple frames of detection images including the target sign by a target recognition model, which can improve the accuracy, and based on the visual recognition result, the map marking result and the confidence of the visual recognition result, more accurate target traffic information can be obtained.

[0004] To achieve the above objectives, according to a first aspect of an embodiment of the present disclosure, a method for determining traffic information is provided, comprising: Obtaining a visual recognition result and a map marking result, wherein the visual recognition result is obtained by processing a plurality of detection frames including the target identification plate through a target recognition model; Determining the confidence level of the visual recognition result based on the type of the target sign and environmental information; Target traffic information is obtained according to the visual recognition result, the map marking result, and the confidence of the visual recognition result, where the target traffic information is one of the visual recognition result and the map marking result.

[0005] Optionally, determining the confidence level of the visual recognition result according to the type of the target sign and environmental information includes: Determining a first confidence weight corresponding to the type of the target sign according to a correspondence between the type of the sign and the confidence weight; Determining a second confidence weight corresponding to the environmental information according to a correspondence between the environment and the confidence weight; The confidence of the visual recognition result is determined according to the first confidence weight and the second confidence weight.

[0006] Optionally, obtaining target traffic information according to the visual recognition result, the map marking result, and the confidence level of the visual recognition result includes: In a case where the visual recognition result is inconsistent with the map calibration result, and the confidence of the visual recognition result is greater than or equal to a confidence threshold, the visual recognition result is determined as the target traffic information.

[0007] Optionally, the obtaining of the target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result comprises: In a case where the visual recognition result is inconsistent with the map calibration result, and the confidence of the visual recognition result is less than a confidence threshold, the map calibration result is determined as the target traffic information.

[0008] Optionally, the obtaining of the target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result comprises: In a case where the visual recognition result is inconsistent with the map calibration result, and the confidence of the visual recognition result is greater than or equal to a confidence threshold, the smaller one of the visual recognition result and the map calibration result is determined as the target traffic information.

[0009] Optionally, before the obtaining of the visual recognition result and the map calibration result, the method further comprises: obtaining a plurality of detection images for the same target sign; processing each detection image by using a target recognition model to obtain a detection result corresponding to each detection image; in a case where detection results of more than or equal to a preset number of detection images are consistent, determining the consistent detection result as the visual recognition result.

[0010] Optionally, the target recognition model comprises a detection model and a classification model. The processing of each detection image by using the target recognition model to obtain a detection result corresponding to each detection image comprises: for any detection image, pre-processing the detection image to obtain a plurality of cropped images; processing each cropped image by using the detection model to obtain a detection result corresponding to each cropped image, the detection model being obtained by training a first base model based on a plurality of first sample images, the first sample images comprising labels for labeling traffic signs in the first sample images; obtaining at least one traffic sign image according to the detection result; The classification model is obtained by training a second base model based on a plurality of second sample images, the second sample images include labels for labeling traffic information in the second sample images, and the plurality of second sample images are images in different environments.

[0011] According to a second aspect of the embodiments of the present disclosure, a traffic information determination apparatus is provided, comprising: The first obtaining module is configured to obtain a visual recognition result and a map calibration result, the visual recognition result being obtained by processing a plurality of detection images including a target signboard through a target recognition model; The first determining module is configured to determine a confidence of the visual recognition result according to a type of the target signboard and environment information; The first obtaining module is configured to obtain target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the target traffic information being one of the visual recognition result and the map calibration result.

[0012] According to a third aspect of the embodiments of the present disclosure, an electronic device is provided, comprising: A memory having a computer program stored thereon; A processor configured to execute the computer program in the memory to implement the steps of the traffic information determination method provided in the first aspect of the present disclosure.

[0013] According to a fourth aspect of the embodiments of the present disclosure, a vehicle is provided, comprising the electronic device provided in the third aspect of the present disclosure.

[0014] Through the above technical solution, the visual recognition result and the map calibration result are obtained, the visual recognition result being obtained by processing a plurality of detection images including a target signboard through a target recognition model, which can improve the accuracy of the visual recognition result, and the confidence of the visual recognition result is determined according to the type of the target signboard and the environment information, and more accurate target traffic information can be obtained based on the visual recognition result, the map calibration result, and the confidence of the visual recognition result.

[0015] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0016] The accompanying drawings are included to provide a further understanding of the present disclosure and constitute a part of the specification, and are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation of the present disclosure. In the drawings: Figure 1is a flowchart of a traffic information determination method according to an embodiment of the present disclosure.

[0017] Figure 2 is a flowchart of a traffic information determination method according to another embodiment of the present disclosure.

[0018] Figure 3 is a flowchart of a method of obtaining a detection result according to an embodiment of the present disclosure.

[0019] Figure 4 is a schematic diagram of a traffic information determination apparatus according to an embodiment of the present disclosure.

[0020] Figure 5 is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0021] The specific embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely intended to illustrate and explain the present disclosure, and are not intended to limit the present disclosure.

[0022] It should be noted that all actions of obtaining signals, information or data in the present disclosure are performed in compliance with the corresponding data protection regulations and policies of the country where the device is located, and with the authorization of the corresponding device owner.

[0023] Figure 1 is a flowchart of a traffic information determination method according to an embodiment of the present disclosure, as shown in Figure 1 The method comprises the following steps. In step S101, a visual recognition result and a map calibration result are obtained, the visual recognition result being obtained by processing a plurality of frames of detection images including a target sign by a target recognition model.

[0024] In the present embodiment, considering the stability of the detection algorithm, there may be missed detection or false detection, in order to improve the accuracy of the visual recognition result, the plurality of frames of detection images including the target sign can be processed by the target recognition model, and the visual recognition result can be obtained by synthesizing the plurality of frames of recognition results. The map calibration result can be obtained by calibration information on the map software. Specifically, the position of the vehicle can be obtained, and based on the position of the vehicle, the calibration information of the corresponding position on the map software is determined, so as to determine the corresponding map calibration result according to the calibration information. The visual recognition result and the map calibration result can both be specific speed limit values.

[0025] In step S102, the confidence of the visual recognition result is determined according to the type of the target sign and the environmental information.

[0026] In this embodiment, the confidence of the visual recognition result is determined by determining the type of the target signboard and the environmental information, considering the influence of environmental factors and the type of the signboard on visual recognition, so as to determine the accuracy of the visual recognition result according to the confidence. The type of the target signboard can include a fixed value displayed by an electronic screen, a variable value displayed by an electronic screen, a fixed value printed on a signboard, and the like. Different types of signboards can correspond to different confidence weights. The environmental information can include weather, light intensity, and the like. For example, the weather can include rainy days, cloudy days, and sunny days, and the light intensity can include different light intensity levels. Different weather and light intensity can correspond to different confidence weights. The confidence of the visual recognition result can be determined by the confidence weight corresponding to the type of the target signboard and the confidence weight corresponding to the environmental information.

[0027] In step S103, the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result. The target traffic information is one of the visual recognition result and the map calibration result.

[0028] In this embodiment, one of the visual recognition result and the map calibration result can be determined as the target traffic information according to the specific values of the visual recognition result and the map calibration result, combined with the confidence of the visual recognition result, so as to improve the accuracy of the obtained target traffic information. The vehicle can perform pre-warning based on the target traffic information and the speed of the vehicle.

[0029] The confidence of the visual recognition result is determined by the type of the target signboard and the environmental information, wherein the visual recognition result is obtained by processing a plurality of detection images including the target signboard by a target recognition model, which can improve the accuracy. Based on the visual recognition result, the map calibration result, and the confidence of the visual recognition result, more accurate target traffic information can be obtained.

[0030] In a possible implementation, determining the confidence of the visual recognition result according to the type of the target signboard and the environmental information can include the following steps: According to the correspondence between the type of the signboard and the confidence weight, the first confidence weight corresponding to the type of the target signboard is determined. According to the correspondence between the environment and the confidence weight, the second confidence weight corresponding to the environmental information is determined. The confidence of the visual recognition result is determined according to the first confidence weight and the second confidence weight.

[0031] In the embodiment, different types of signs and different environments correspond to different confidence weights. The first confidence weight corresponding to the type of the target sign can be determined according to the correspondence between the type of the sign and the confidence weight, and the second confidence weight corresponding to the environment information can be determined according to the correspondence between the environment and the confidence weight. Then, the confidence of the visual recognition result is determined according to the reference value and the first confidence weight and the second confidence weight. The reference value can be 1 or 100, and the value range of the first confidence weight and the second confidence weight can be 0 to 1. The greater the confidence of the visual recognition result, the more accurate the visual recognition result.

[0032] In a possible implementation, the method for obtaining the target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result can be as follows: In the case that the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is greater than or equal to the confidence threshold, the visual recognition result is determined as the target traffic information.

[0033] In the embodiment, it is first determined whether the visual recognition result is consistent with the map calibration result. If the visual recognition result is inconsistent with the map calibration result, the accuracy of the visual recognition result can be determined. If the confidence of the visual recognition result is greater than or equal to the confidence threshold, the visual recognition result is considered to be more accurate, and the visual recognition result is determined as the target traffic information. The confidence threshold can be determined according to the type of the target sign and the influence degree of the environment information on the visual recognition, and can also be determined in combination with the accuracy of the map calibration result. In this way, a more accurate value can be selected from the visual recognition result and the map calibration result as the target traffic information.

[0034] In a possible implementation, the method for obtaining the target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result can be as follows: In the case that the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is less than the confidence threshold, the map calibration result is determined as the target traffic information.

[0035] In the embodiment, it is first determined whether the visual recognition result is consistent with the map calibration result. If the visual recognition result is inconsistent with the map calibration result, the accuracy of the visual recognition result can be determined. If the confidence of the visual recognition result is less than the confidence threshold, the visual recognition result is considered to be inaccurate, and the map calibration result is determined as the target traffic information. The confidence threshold can be determined according to the type of the target sign and the influence degree of the environment information on the visual recognition, and can also be determined in combination with the accuracy of the map calibration result. In this way, a more accurate value can be selected from the visual recognition result and the map calibration result as the target traffic information.

[0036] In a possible implementation, according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the method for obtaining the target traffic information can be as follows: In a case where the visual recognition result is inconsistent with the map calibration result, and the confidence of the visual recognition result is greater than or equal to the confidence threshold, the smaller one of the visual recognition result and the map calibration result is determined as the target traffic information.

[0037] In this embodiment, it is first determined whether the visual recognition result is consistent with the map calibration result. If the visual recognition result is inconsistent with the map calibration result, the accuracy of the visual recognition result can be determined. If the confidence of the visual recognition result is greater than or equal to the confidence threshold, it is considered that the visual recognition result is more accurate. In order to ensure that the vehicle can more meet the speed limit requirement, the smaller one of the visual recognition result and the map calibration result can be further determined as the target traffic information. The confidence threshold can be determined according to the type of the target signboard and the influence degree of the environmental information on the visual recognition, and can also be determined in combination with the accuracy of the map calibration result. In this way, a value that more meets the speed limit requirement can be selected from the visual recognition result and the map calibration result as the target traffic information.

[0038] In a possible implementation, before the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the validity of the visual recognition result and the map calibration result can be first determined. It can be determined whether the timestamp of the obtained map calibration result is overdue. If the timestamp of the map calibration result is overdue, it is determined that the map calibration result is invalid. When the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the map calibration result can be deleted, and the visual recognition result can be directly determined as the target traffic information. If the distance between the position corresponding to the visual recognition result and the current position exceeds the effective distance, it is determined that the visual recognition result is invalid. When the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the visual recognition result can be deleted, and the map calibration result can be directly determined as the target traffic information.

[0039] In a possible implementation, before the visual recognition result and the map calibration result are obtained, the method further includes: a plurality of detection images for the same target signboard are obtained; each detection image is processed by a target recognition model to obtain a detection result corresponding to each detection image; in a case where detection results of more than or equal to a preset number of detection images are consistent, the consistent detection result is determined as the visual recognition result.

[0040] In this embodiment, multiple image acquisition operations can be performed on the same target sign to obtain multiple frames of detection images, so as to continuously track and detect the same target sign and ensure the stability of the output result. Each frame of detection image is processed by the target recognition model, and the detection result corresponding to each frame of detection image is output. The multiple detection results corresponding to the multiple frames of detection images are analyzed to determine whether the multiple detection results are consistent. If the detection results of more than or equal to a preset number of frames of detection images are consistent, the consistent detection result is determined as the visual recognition result, so as to avoid false detection, improve the stability of the algorithm output, and obtain a more accurate detection result. The preset number of frames can be set according to actual conditions, for example, the preset number of frames can be 3 frames.

[0041] Figure 2 is a flowchart of a traffic information determination method according to another embodiment of the present disclosure, as shown in Figure 2 In one possible embodiment, considering that there are various types of traffic signs in practice, it is difficult to collect training data. If the detection result is directly output by the detector, the operability is relatively low, and the algorithm accuracy is not high due to serious class imbalance in the training sample. Therefore, considering the subsequent scalability, the algorithm design adopts a front-end detector + back-end independent classifier mode, that is, the target recognition model includes a detection model and a classification model. The traffic sign recognition algorithm principle mainly includes two steps. The first step is to detect the traffic sign by the detection model to identify the traffic sign in the field of view. The second step is to accurately identify the class of the first step detection result by the classification model to determine the specific speed limit value corresponding to the sign.

[0042] Figure 3 is a flowchart of a method for obtaining a detection result according to an embodiment of the present disclosure, as shown in Figure 3 Each frame of detection image is processed by the target recognition model to obtain the detection result corresponding to each frame of detection image, which can include: In step S301, for any frame of detection image, the frame of detection image is preprocessed to obtain a plurality of cropped images.

[0043] In this embodiment, each frame of detection image can be cropped according to a preset parameter to obtain a plurality of image blocks, and the plurality of image blocks are scaled to adjust the size of the plurality of image blocks to a preset size to obtain a plurality of cropped images.

[0044] In step S302, each cropped image is processed by a detection model to obtain a detection result corresponding to each cropped image. The detection model is obtained by training a first base model based on a plurality of first sample images. The first sample image includes a label for labeling the traffic sign in the first sample image.

[0045] In the embodiment, each cropped image is input into the detection model, and the detection model outputs the detection result corresponding to the cropped image after processing the cropped image. The detection model can be obtained by training a first base model based on a plurality of first sample images. Each first sample image includes a traffic sign, and the traffic sign in each first sample image is labeled as a label of the first sample image for calculating a loss function in the training process. The detection result can be empty, i.e., there is no traffic sign in the cropped image, or the detection result can be a detection box corresponding to the traffic sign in the cropped image, which is used to enclose the traffic sign in the cropped image.

[0046] In a possible implementation, the detection model can be obtained by the following steps: performing multi-round training on the first base model based on a plurality of first sample images; obtaining a predicted detection result output by each round of training; obtaining a first loss function based on the predicted detection result corresponding to the round of training and the label of the first sample image corresponding to the round of training; optimizing the first base model based on the first loss function; and stopping the training and obtaining the detection model when the first base model meets a first training condition. The first training condition can be that the number of training rounds meets a preset number of rounds, or the first base model converges.

[0047] In step S303, at least one traffic sign image is obtained based on the detection result.

[0048] In the embodiment, the traffic sign in the cropped image is cropped based on the detection result. The detection box can be enlarged by a preset multiple, for example, 1.3 times, and then the cropped image is scaled and resized to a preset size to obtain at least one traffic sign image. When the detection result is empty, there is no corresponding traffic sign image.

[0049] In step S304, each traffic sign image is processed by a classification model to obtain a detection result corresponding to the frame of detection image. The classification model is obtained by training a second base model based on a plurality of second sample images. The second sample image includes a label for labeling traffic information in the second sample image, and the plurality of second sample images are images in different environments.

[0050] In the embodiment, images in multiple different environments can be obtained, and traffic information in the multiple images is labeled to obtain multiple second sample images, and the second base model is trained through the multiple second sample images, so that the classification model is obtained. Each traffic sign image is input into the classification model, and the detection result corresponding to each traffic sign image is output through the classification model. The detection result can be a specific value of the traffic sign in the frame detection image.

[0051] In a possible implementation, the classification model can be obtained through the following steps: the second base model is trained through multiple rounds of training through multiple second sample images; after each round of training, a prediction classification result output by the current round of training is obtained; a second loss function is obtained according to the prediction classification result corresponding to the current round of training and the label of the second sample image corresponding to the current round of training; the second base model is optimized according to the second loss function; and the training is stopped when the second base model meets a second training condition, and the classification model is obtained. The second training condition can be that the number of training rounds meets a preset number of rounds, or the second base model converges. Different environments can be different weathers or different light intensities.

[0052] Through the above method, the first step only needs to determine whether it is a traffic sign, which greatly simplifies the collection difficulty of the detection model training sample, and is convenient for integration with other detection tasks. The second step realizes fine classification of the traffic sign, that is, obtaining a specific value, which maximizes the collection difficulty of the classification model training sample, and overcomes the problem of uneven sample distribution, so that a very lightweight classification model can achieve very high precision.

[0053] Figure 4 is a schematic diagram of a traffic information determination apparatus provided according to an embodiment of the present disclosure, as shown in Figure 4 The traffic information determination apparatus 400 includes a first obtaining module 401, a first determining module 402, and a first obtaining module 403.

[0054] The first obtaining module is configured to obtain a visual recognition result and a map calibration result, wherein the visual recognition result is obtained by processing multiple frames of detection images including a target signboard through a target recognition model; The first determining module is configured to determine a confidence of the visual recognition result according to a type of the target signboard and environment information; The first obtaining module is configured to obtain target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, wherein the target traffic information is one of the visual recognition result and the map calibration result.

[0055] Optionally, the first determining module 402 includes: The first determining sub-module is configured to determine a first confidence weight corresponding to the type of the target sign according to a correspondence between the type of the sign and the confidence weight; The second determining sub-module is configured to determine a second confidence weight corresponding to the environment information according to a correspondence between the environment and the confidence weight; The third determining sub-module is configured to determine the confidence of the visual recognition result according to the first confidence weight and the second confidence weight.

[0056] Optionally, the first obtaining module 403 comprises: The fourth determining sub-module is configured to determine the visual recognition result as the target traffic information in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is greater than or equal to a confidence threshold.

[0057] Optionally, the first obtaining module 403 comprises: The fifth determining sub-module is configured to determine the map calibration result as the target traffic information in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is less than a confidence threshold.

[0058] Optionally, the first obtaining module 403 comprises: The sixth determining sub-module is configured to determine the smaller one of the visual recognition result and the map calibration result as the target traffic information in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is greater than or equal to a confidence threshold.

[0059] Optionally, the traffic information determining apparatus 400 further comprises: The second acquiring module is configured to acquire a plurality of frames of detection images for the same target sign; The second obtaining module is configured to process each frame of detection image through a target recognition model to obtain a detection result corresponding to each frame of detection image; The second determining module is configured to determine the consistent detection result as the visual recognition result in a case where there are detection results of greater than or equal to a preset number of frames of detection images consistent.

[0060] Optionally, the target recognition model comprises a detection model and a classification model; The second obtaining module comprises: The first obtaining sub-module is configured to, for any frame of detection image, pre-process the frame of detection image to obtain a plurality of cropped images; The second obtaining sub-module is configured to process each cropped image respectively by using the detection model to obtain a detection result corresponding to each cropped image, the detection model being obtained by training a first base model based on a plurality of first sample images, and the first sample images including labels for labeling traffic signs in the first sample images; The third obtaining sub-module is configured to obtain at least one traffic sign image according to the detection result. The fourth obtaining sub-module is configured to process each traffic sign image by using the classification model to obtain a detection result corresponding to the frame detection image, the classification model being obtained by training a second base model based on a plurality of second sample images, the second sample images including labels for labeling traffic information in the second sample images, and the plurality of second sample images being images in different environments.

[0061] As to the traffic information determination apparatus 400 in the above-described embodiments, the specific manners in which various modules perform operations have been described in detail in the embodiments of the method, and thus will not be described in detail here.

[0062] Figure 5 is a block diagram of an electronic device according to an example embodiment. As shown in Figure 5 The electronic device 500 can include a processor 501, a memory 502. The electronic device 500 can further include one or more of a multimedia component 503, an input / output (I / O) interface 504, and a communication component 505.

[0063] The processor 501 is configured to control overall operations of the electronic device 500 to complete all or part of the steps of the traffic information determination method described above. The memory 502 is configured to store various types of data to support operations of the electronic device 500, which can include, for example, instructions for operating any application or method on the electronic device 500, and application-related data, such as contact data, sent and received messages, pictures, audio, video, and the like. The memory 502 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk, or an optical disk. The multimedia component 503 can include a screen and an audio component. The screen can be, for example, a touch screen, and the audio component is configured to output and / or input audio signals. For example, the audio component can include a microphone configured to receive external audio signals. The received audio signals can be further stored in the memory 502 or transmitted through the communication component 505. The audio component also includes at least one speaker configured to output audio signals. The I / O interface 504 provides an interface between the processor 501 and other interface modules, which can be a keyboard, a mouse, a button, and the like. The buttons can be virtual buttons or physical buttons. The communication component 505 is configured to perform wired or wireless communication between the electronic device 500 and other devices. The wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G, 4G, NB-IOT, eMTC, or other 5G, and the like, or a combination of one or more of them, is not limited herein. Therefore, the corresponding communication component 505 can include a Wi-Fi module, a Bluetooth module, an NFC module, and the like.

[0064] In an exemplary embodiment, the electronic device 500 can be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors, or other electronic elements for performing the traffic information determination method described above.

[0065] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the traffic information determination method described above. For example, the computer-readable storage medium can be the memory 502 described above including program instructions, which can be executed by the processor 501 of the electronic device 500 to complete the traffic information determination method described above.

[0066] In another exemplary embodiment, a computer program product is also provided, which contains a computer program capable of being executed by a programmable device, and the computer program has code portions for executing the traffic information determination method described above when executed by the programmable device.

[0067] In another exemplary embodiment, a vehicle is also provided, which includes the electronic device described above.

[0068] The preferred embodiments of the present disclosure are described in detail above with reference to the accompanying drawings, but the present disclosure is not limited to the specific details in the above-described embodiments. Within the technical concept of the present disclosure, various simple modifications can be made to the technical solutions of the present disclosure, and these simple modifications all belong to the protection scope of the present disclosure.

[0069] In addition, it should be noted that each specific technical feature described in the above-described specific embodiments can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, various possible combinations are not described again in the present disclosure.

[0070] Furthermore, any combination of the various different embodiments of the present disclosure can also be made, as long as it does not deviate from the idea of the present disclosure, and it should also be considered as the disclosed content of the present disclosure.

Claims

1. A traffic information determination method characterized by comprising: The method comprises: obtaining a visual recognition result and a map calibration result, the visual recognition result being obtained by processing, by a target recognition model, a plurality of frames of detection images including a target signboard; determining a confidence of the visual recognition result according to a type of the target signboard and environment information; obtaining target traffic information according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, the target traffic information being one of the visual recognition result and the map calibration result.

2. The method of claim 1, wherein the confidence of the visual recognition result is determined according to the type of the target signboard and the environment information, comprising: determining a first confidence weight corresponding to the type of the target signboard according to a corresponding relationship between the type of the signboard and the confidence weight; determining a second confidence weight corresponding to the environment information according to a corresponding relationship between the environment and the confidence weight; determining the confidence of the visual recognition result according to the first confidence weight and the second confidence weight.

3. The method of claim 1, wherein the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, comprising: in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is greater than or equal to a confidence threshold, determining the visual recognition result as the target traffic information.

4. The method of claim 1, wherein the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, comprising: in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is less than a confidence threshold, determining the map calibration result as the target traffic information.

5. The method of claim 1, wherein the target traffic information is obtained according to the visual recognition result, the map calibration result, and the confidence of the visual recognition result, comprising: in a case where the visual recognition result is inconsistent with the map calibration result and the confidence of the visual recognition result is greater than or equal to a confidence threshold, determining a smaller one of the visual recognition result and the map calibration result as the target traffic information. Before obtaining the visual recognition result and the map calibration result, the method further comprises: obtaining a plurality of frames of detection images for the same target signboard; processing each frame of detection image by a target recognition model to obtain a detection result corresponding to each frame of detection image; in a case where detection results of more than or equal to a preset number of frames of detection images are consistent, determining the consistent detection result as the visual recognition result.

6. The traffic information determination method according to any one of claims 1 to 5, characterized by, The target recognition model comprises a detection model and a classification model. ​ ​ ​ 7. The traffic information determination method according to claim 6, characterized by, ​ The target recognition model is used to process each frame of detection image respectively to obtain a detection result corresponding to each frame of detection image, including: For any frame of detection image, the frame of detection image is preprocessed to obtain a plurality of cropped images; Each cropped image is processed by the detection model to obtain a detection result corresponding to each cropped image, the detection model being obtained by training a first base model based on a plurality of first sample images, the first sample images including labels for labeling traffic signs in the first sample images; At least one traffic sign image is obtained according to the detection result; Each traffic sign image is processed by the classification model to obtain a detection result corresponding to the frame of detection image, the classification model being obtained by training a second base model based on a plurality of second sample images, the second sample images including labels for labeling traffic information in the second sample images, and the plurality of second sample images being images under different environments.

8. A traffic information determination device characterized by comprising: Comprising: The first acquisition module is configured to acquire a visual recognition result and a map calibration result, the visual recognition result being obtained by processing a plurality of frames of detection image including a target signboard through a target recognition model; The first determination module is configured to determine a confidence degree of the visual recognition result according to a type of the target signboard and environmental information; The first obtaining module is configured to obtain target traffic information according to the visual recognition result, the map calibration result, and the confidence degree of the visual recognition result, the target traffic information being one of the visual recognition result and the map calibration result.

9. An electronic device, comprising: Comprising: A memory having a computer program stored thereon; A processor configured to execute the computer program in the memory to implement the steps of the traffic information determination method in any one of claims 1-7.

10. A vehicle characterized by comprising: The electronic device of claim 9 is included.

Images