A method and device for acquiring a geographic fence confidence and a storage medium

Abstract

The embodiment of the application discloses a kind of acquisition method and device of geographic fence confidence, the method includes: obtaining driving image and driving parameter, and according to driving image and driving parameter, construct geographic fence parameter;Wherein, driving parameter includes at least one in road curvature, road slope, weather type and illumination intensity;According to geographic fence parameter, through the confidence evaluation model of pre-training completion, the model evaluation confidence of geographic fence is acquired.The technical scheme provided in the embodiment of the application realizes objective evaluation based on confidence evaluation model, avoids the exhaustion of artificial evaluation to specific driving scene, reduces the human cost of evaluation consumption, also avoids the influence of subjective factor on confidence evaluation result, greatly improves the evaluation accuracy of geographic fence confidence.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to automatic driving technology, and in particular to a method and device for obtaining a geo-fence confidence level, an electronic device and a storage medium. BACKGROUND

[0002] A geo-fence is a prerequisite for a vehicle to start an automatic driving function and is a core condition for ensuring the safety of automatic driving. A geo-fence confidence level is used to evaluate the reliability of a geo-fence, and a high confidence level can ensure the accuracy of geo-fence data.

[0003] A traditional geo-fence is evaluated for a confidence level by a subjective evaluation method of a vehicle engineer, that is, the vehicle engineer pre-sets different confidence level values according to different driving scenarios, and the vehicle obtains a matching confidence level value according to a current driving scenario.

[0004] However, such an evaluation method has a large number of and complex automatic driving scenarios, cannot enumerate all scenarios, and is difficult to adapt to all road conditions. In addition, the subjective evaluation of the vehicle engineer lacks a unified standard and has a strong subjective factor, and an accurate confidence level value cannot be obtained. SUMMARY

[0005] Embodiments of the present application provide a method and device for obtaining a geo-fence confidence level, an electronic device and a storage medium to obtain a confidence level evaluation result of a geo-fence by using a pre-trained confidence level evaluation model.

[0006] In a first aspect, embodiments of the present application provide a method for obtaining a geo-fence confidence level, comprising:

[0007] obtaining driving images and driving parameters, and constructing a geo-fence parameter according to the driving images and the driving parameters; wherein the driving parameters include at least one of a road curvature, a road slope, a weather type and an illumination intensity;

[0008] obtaining a model evaluation confidence level of a geo-fence by using a pre-trained confidence level evaluation model according to the geo-fence parameter.

[0009] In a second aspect, embodiments of the present application provide a device for obtaining a geo-fence confidence level, comprising:

[0010] a geo-fence parameter obtaining module configured to obtain driving images and driving parameters, and construct a geo-fence parameter according to the driving images and the driving parameters; wherein the driving parameters include at least one of a road curvature, a road slope, a weather type and an illumination intensity;

[0011] The model evaluation confidence obtaining module is configured to obtain a model evaluation confidence of the geofence according to the geofence parameters and by using a pre-trained confidence evaluation model.

[0012] In a third aspect, an electronic device is provided, and the electronic device includes:

[0013] one or more processors;

[0014] a memory configured to store one or more programs;

[0015] When the one or more programs are executed by the one or more processors, the one or more processors implement the method for obtaining the geofence confidence according to any of the embodiments of the present application.

[0016] In a fourth aspect, a storage medium containing computer executable instructions is provided, and the computer executable instructions implement the method for obtaining the geofence confidence according to any of the embodiments of the present application when executed by a computer processor.

[0017] The technical solution provided by the embodiments of the present application obtains driving images and driving parameters, constructs geofence parameters according to the driving images and the driving parameters, obtains a model evaluation confidence of the geofence according to the geofence parameters and by using a pre-trained confidence evaluation model, and thus objective evaluation based on the confidence evaluation model is achieved, the exhaustive evaluation of specific driving scenarios by manual evaluation is avoided, the human cost of evaluation is reduced, the influence of subjective factors on the confidence evaluation result is avoided, and the evaluation accuracy of the geofence confidence is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of a method for obtaining a geofence confidence provided by the first embodiment of the present application;

[0019] Figure 2 is a flowchart of a method for obtaining a geofence confidence provided by the second embodiment of the present application;

[0020] Figure 3 is a structural block diagram of a device for obtaining a geofence confidence provided by the third embodiment of the present application;

[0021] Figure 4 is a structural block diagram of an electronic device provided by the fourth embodiment of the present application. DETAILED DESCRIPTION

[0022] The application will be described in further detail below with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are intended to be illustrative only and not limiting of the application. It is also to be understood that the terminology used herein is for the purpose of describing the specific embodiments only and is not intended to be limiting.

[0023] Embodiment one

[0024] Figure 1 A flowchart of a method for obtaining a geographic fence confidence provided by the first embodiment of the application. The embodiment can be applied to obtaining the confidence of a geographic fence based on a pre-trained confidence evaluation model. The method can be executed by the device for obtaining a geographic fence confidence in the embodiments of the application. The device can be realized by software and / or hardware and integrated on an electronic device, typically, a server or a vehicle terminal device. The method specifically includes the following steps:

[0025] S110, obtaining a driving image and driving parameters, and constructing a geographic fence parameter according to the driving image and the driving parameters. The driving parameters include at least one of road curvature, road slope, weather type and illumination intensity.

[0026] The driving image is image information captured in the driving direction of the vehicle. It can be obtained by a vehicle-mounted camera installed at the front end of the vehicle for capturing images in front of the vehicle on the route. It can also be obtained by an external camera installed in the cab. The driving image directly reflects the road surface conditions and the density of vehicles in front of the vehicle and other driving information. The driving parameters are parameter information related to the creation of a geographic fence during the driving of the vehicle. The driving parameters can be obtained by recognizing the vehicle-mounted sensor, for example, the illumination intensity of the current driving environment obtained by a camera sensor. The driving parameters can also be obtained by a third-party functional software accessed by an electronic device, for example, an electronic map and navigation software.

[0027] The road curvature (Curvature) represents the degree of bending of the road, and the higher the curvature value, the greater the degree of bending. On a road with a lower curvature value, the reliability of the constructed geofence is higher. The road slope (Slope) is the ratio of the vertical height of the slope to the horizontal distance, which represents the steepness of the ground unit. The greater the slope, the steeper the slope. On a road with a lower slope value, the reliability of the constructed geofence is higher. The weather type can include sunny, rainy, snowy, foggy, hazy and other weather. According to the severity, it can be specifically divided into heavy rain, heavy rain, moderate rain and light rain, as well as heavy snow, heavy snow, moderate snow and light snow. When the weather is good, the reliability of the constructed geofence is higher, and when the weather is bad, the reliability of the constructed geofence is lower. The light intensity reflects the effective recognition distance and recognition accuracy of some vehicle-mounted sensors (for example, a camera). Therefore, when the light intensity is low, that is, when the light condition is poor, the reliability of the constructed geofence is lower.

[0028] The electronic device stamps a time stamp on the obtained driving image and driving parameter according to the system time, and then associates the driving image and the driving parameter under the same time stamp, so as to ensure that the driving image and the driving parameter under each group of geofence parameters have the same time stamp, that is, to ensure that the data in each group of geofence parameters constitutes the same in the time dimension.

[0029] S120, according to the geofence parameter, through the pre-trained confidence evaluation model, the model evaluation confidence of the geofence is obtained.

[0030] The confidence evaluation model can be pre-constructed based on various algorithms to form a basic model (i.e., an initial model), for example, based on a collaborative filtering algorithm or an artificial neural network (ANN) algorithm to form the initial model; the historical sample information includes historical driving images captured by a high-resolution camera and historical driving parameters with the same timestamp as the historical driving images; the acquired historical sample information can be respectively generated into a training sample set, a verification sample set and a test sample set according to a certain proportion (for example, 6:2:2), and the training sample set, the verification sample set and the test sample set are given corresponding confidence labels according to the artificial marking results; then, the initial model is trained according to the training sample set, and the training result is verified through the verification sample set, so that the initial model adjusts the parameters inside the model according to the loss function and the recognition accuracy, and then the test sample set is used to test the recognition accuracy of the confidence evaluation model, and finally the pre-training completed confidence evaluation model is obtained; wherein, whether the pre-training of the confidence evaluation model is completed can be determined according to a preset number threshold, that is, when the number of iterations reaches the preset number threshold, it is determined that the pre-training is completed, or it can be determined according to a preset recognition threshold, that is, when the recognition accuracy reaches the preset recognition threshold, it is determined that the pre-training is completed.

[0031] Optionally, in the embodiments of the present application, the confidence evaluation model comprises at least one of a Transformer neural network model, a convolutional neural network model and a recurrent neural network model. The recurrent neural network (RNN) is a recursive neural network with sequence data as input, recursion in the evolution direction of the sequence and all nodes connected in a chain. The characteristic is that the continuous information of the previous and subsequent inputs is taken as the associated information to ensure the content coherence of the image information. The convolutional neural network (CNN) is a feedforward neural network with convolution calculation and deep structure in deep learning. The characteristic is that the application of convolution operation improves the extraction accuracy of image features, and the application of the pooling layer reduces the calculation complexity of image features, that is, improves the acquisition efficiency of image features. The Transformer neural network model is a neural network model based on the Transformer architecture. The Transformer architecture is an Encoder-Decoder structure of Attention mechanism, which includes multiple stacked Encoder layers and multiple stacked Decoder layers, and outputs the results through the output layer connected with the last Decoder layer. The characteristic is to ensure the independence of different things in the image. The confidence evaluation models under different architectures give the evaluation results of the confidence through different recognition methods.

[0032] Optionally, in the embodiment of the present application, the model evaluation confidence of the geofence is obtained according to the geofence parameters and the pre-trained confidence evaluation model, comprising: a first reference confidence is obtained according to the geofence parameters and the Transformer neural network model, a second reference confidence is obtained according to the geofence parameters and the convolutional neural network model, and a third reference confidence is obtained according to the geofence parameters and the recurrent neural network model; and the model evaluation confidence of the geofence is obtained according to the first reference confidence, the second reference confidence and the third reference confidence. Since the confidence evaluation models under different architectures give the evaluation results of the confidence through different identification methods, the average of the first reference confidence, the second reference confidence and the third reference confidence can be used as the obtained model evaluation confidence after the confidence evaluation results of the geofence are obtained through the neural network models of the above three architectures. In particular, in order to ensure the absolute safety when the automatic driving function is turned on, the minimum value of the first reference confidence, the second reference confidence and the third reference confidence can also be used as the model evaluation confidence, so as to ensure that the minimum value of the above three reference confidences is greater than the preset opening threshold when the automatic driving function is turned on.

[0033] Optionally, in the embodiment of the present application, the model evaluation confidence of the geofence is obtained according to the geofence parameters and the pre-trained confidence evaluation model, comprising: the image features of the driving image are obtained through the confidence evaluation model; wherein the image features include color features, texture features, shape features and spatial relationship features; the driving parameters are added to the color features, the texture features, the shape features and the spatial relationship features respectively, and the model evaluation confidence of the geofence is obtained according to the color features, the texture features, the shape features and the spatial relationship features after the driving parameters are added.

[0034] The confidence evaluation model is used to acquire image features by extracting feature vectors from a feature map; wherein, color features describe the surface properties of the objects corresponding to the image or image region, and are pixel-based features; texture features describe the surface properties of the objects corresponding to the image or image region, and need to be calculated by statistics in a region containing multiple pixels; shape features describe the contour features of the object outer boundary and the overall region features; spatial relationship features are the mutual spatial positions or relative direction relationships between multiple targets segmented from the image, such as connection relationship, overlapping relationship and containing relationship, etc.; the driving parameters are added into the feature vectors of the color features, texture features, shape features and spatial relationship features respectively, so that each image feature vector of the geofence parameters includes complete driving parameters, which not only ensures the relevance of the driving parameters and the driving image, but also increases the proportion of the driving parameters in the geofence parameters.

[0035] The technical scheme provided by the embodiment of the present application, after acquiring the driving image and the driving parameters and constructing the geofence parameters according to the driving image and the driving parameters, acquires the model evaluation confidence of the geofence according to the above geofence parameters through the pre-trained confidence evaluation model, realizes the objective evaluation based on the confidence evaluation model, avoids the exhaustive enumeration of specific driving scenes by manual evaluation, reduces the human cost of evaluation consumption, avoids the influence of subjective factors on the confidence evaluation result, and greatly improves the evaluation accuracy of the geofence confidence.

[0036] Embodiment two

[0037] Figure 2 The flowchart of the method for acquiring the geofence confidence provided by the second embodiment of the present application is based on the above technical scheme and is specific, specifically, after acquiring the driving image and the driving parameters, the method further includes acquiring the sensor evaluation confidence of the geofence, and the method includes the following steps:

[0038] S210, acquire a driving image and a driving parameter, and construct a geofence parameter according to the driving image and the driving parameter; wherein, the driving parameter includes at least one of road curvature, road slope, weather type and illumination intensity.

[0039] S220, according to the geofence parameter, through the pre-trained confidence evaluation model, acquire the model evaluation confidence of the geofence.

[0040] S230, acquire a vehicle sensor list; wherein, the vehicle sensor list includes at least one vehicle sensor related to the geofence, and target driving parameters related to the working performance of the at least one vehicle sensor.

[0041] The construction of the geofence relies on the cooperation of multiple sensors in the vehicle; all vehicle sensors related to the geofence are recorded in the vehicle sensor list, and each vehicle sensor can be in normal working performance under certain target driving parameters. For example, the camera sensor can be in normal working state under the target driving parameters of good light conditions (i.e. large light intensity) and good weather conditions (i.e. sunny day), therefore, the vehicle sensor list records the driving parameters that affect the working performance of each vehicle sensor; different vehicle sensors correspond to different target driving parameters.

[0042] S240, according to the driving parameters and the target driving parameters related to the working performance of the at least one vehicle sensor, obtaining the reliability of the at least one vehicle sensor.

[0043] According to the target driving parameters recorded in the vehicle sensor list that affect the working performance of each vehicle sensor, after obtaining the driving parameters in the current driving environment, it can be determined whether the vehicle sensor is in normal working state. If it is in normal working state, the reliability of the vehicle sensor is high, if it is in abnormal working state, the reliability of the vehicle sensor is low.

[0044] S250, according to the reliability of the at least one vehicle sensor, obtaining the sensor evaluation confidence of the geofence.

[0045] The proportion of vehicle sensors with high reliability in the total number of vehicle sensors can be used as the sensor evaluation confidence.

[0046] S260, according to the model evaluation confidence and the sensor evaluation confidence of the geofence, obtaining the actual evaluation confidence of the geofence.

[0047] The average of the model evaluation confidence and the sensor evaluation confidence can be used as the actual evaluation confidence, or in order to ensure the absolute safety of the automatic driving function, the smaller value of the model evaluation confidence and the sensor evaluation confidence can be used as the actual evaluation confidence, so as to determine whether to start the automatic driving function according to the obtained actual evaluation confidence.

[0048] Optionally, in the embodiment of the present application, the acquiring the actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence of the geofence comprises: acquiring the actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence of the geofence, and a weight value of the model evaluation confidence and a weight value of the sensor evaluation confidence. Different weight values can be assigned to the model evaluation confidence and the sensor evaluation confidence in advance, for example, the weight value of the model evaluation confidence is set to be greater than the weight value of the sensor evaluation confidence, so as to increase the proportion of the model evaluation confidence in the actual evaluation confidence, and then the product of the model evaluation confidence and the corresponding weight value is summed with the product of the sensor evaluation confidence and the corresponding weight value, and the sum is the actual evaluation confidence.

[0049] Optionally, in the embodiment of the present application, the weight value of the model evaluation confidence and the weight value of the sensor evaluation confidence are both related to the driving parameter. The evaluation accuracy of the confidence evaluation model may be different under different driving parameters, for example, the evaluation accuracy of the confidence evaluation model is lower under the driving scene of high road curvature, high road slope, poor weather condition and weak light intensity, and therefore, the weight value of the model evaluation confidence is set to be less than the weight value of the sensor evaluation confidence in the driving scene of the above driving parameters. The evaluation accuracy of the confidence evaluation model is higher under the driving scene of low road curvature, low road slope, good weather condition and strong light intensity, and therefore, the weight value of the model evaluation confidence is set to be greater than the weight value of the sensor evaluation confidence in the driving scene of the above driving parameters, so as to assign the weight values to the model evaluation confidence and the sensor evaluation confidence according to different driving parameters.

[0050] The technical scheme provided by the embodiment of the present application can acquire the model evaluation confidence of the geofence through the pre-trained confidence evaluation model after constructing the geofence parameter according to the driving image and the driving parameter, determine the reliability of each vehicle sensor according to the vehicle sensor list and the current driving parameter, and then acquire the sensor evaluation confidence of the geofence according to the reliability of each vehicle sensor, and finally acquire the actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence, so as to realize the confidence evaluation based on the confidence evaluation model and the confidence evaluation based on the working state of the vehicle sensor, and further improve the evaluation accuracy of the geofence confidence.

[0051] Embodiment three

[0052] Figure 3is a structural block diagram of a geographic fence confidence acquisition device provided by Embodiment Three of the present application, and the device specifically comprises: a geographic fence parameter acquisition module 310 and a model evaluation confidence acquisition module 320.

[0053] The geographic fence parameter acquisition module 310 is used for acquiring a driving image and driving parameters, and constructing a geographic fence parameter according to the driving image and the driving parameters; wherein the driving parameters comprise at least one of a road curvature, a road slope, a weather type and an illumination intensity.

[0054] The model evaluation confidence acquisition module 320 is used for acquiring a model evaluation confidence of a geographic fence according to the geographic fence parameter, through a pre-trained confidence evaluation model.

[0055] The technical scheme provided by the present application in the embodiments, after acquiring a driving image and driving parameters and constructing a geographic fence parameter according to the driving image and the driving parameters, acquires a model evaluation confidence of a geographic fence according to the above-mentioned geographic fence parameter, through a pre-trained confidence evaluation model, realizes objective evaluation based on the confidence evaluation model, avoids exhaustive enumeration of specific driving scenes by artificial evaluation, reduces the human cost of evaluation consumption, also avoids the influence of subjective factors on the confidence evaluation result, and greatly improves the evaluation accuracy of the geographic fence confidence.

[0056] Optionally, on the basis of the above technical scheme, the confidence evaluation model comprises at least one of a Transformer neural network model, a convolutional neural network model and a recurrent neural network model.

[0057] Optionally, on the basis of the above technical scheme, the model evaluation confidence acquisition module 320 comprises:

[0058] The reference confidence acquisition unit is used for acquiring a first reference confidence through the Transformer neural network model, a second reference confidence through the convolutional neural network model and a third reference confidence through the recurrent neural network model according to the geographic fence parameter;

[0059] The model evaluation confidence acquisition unit is used for acquiring a model evaluation confidence of a geographic fence according to the first reference confidence, the second reference confidence and the third reference confidence.

[0060] Optionally, on the basis of the above technical scheme, the model evaluation confidence acquisition module 320 further comprises:

[0061] An image feature acquisition unit is configured to acquire image features of the driving image by using the confidence evaluation model, wherein the image features include color features, texture features, shape features, and spatial relationship features.

[0062] A driving parameter adding unit is configured to add the driving parameters to the color features, the texture features, the shape features, and the spatial relationship features, respectively, and acquire a model evaluation confidence of the geofence according to the color features, the texture features, the shape features, and the spatial relationship features after the driving parameters are added.

[0063] Optionally, based on the above technical solutions, the device for acquiring the confidence of the geofence further includes:

[0064] A vehicle sensor list acquisition module is configured to acquire a vehicle sensor list, wherein the vehicle sensor list includes at least one vehicle sensor related to the geofence and driving conditions related to the working performance of the at least one vehicle sensor.

[0065] A reliability degree acquisition module is configured to acquire the reliability degree of the at least one vehicle sensor according to the driving parameters and the driving conditions of the at least one vehicle sensor.

[0066] A sensor evaluation confidence acquisition module is configured to acquire a sensor evaluation confidence of the geofence according to the reliability degree of the at least one vehicle sensor.

[0067] An actual evaluation confidence acquisition module is configured to acquire an actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence of the geofence.

[0068] Optionally, based on the above technical solutions, the actual evaluation confidence acquisition module is specifically configured to acquire the actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence of the geofence, and a weight value of the model evaluation confidence and a weight value of the sensor evaluation confidence.

[0069] Optionally, based on the above technical solutions, the weight value of the model evaluation confidence and the weight value of the sensor evaluation confidence are both related to the driving parameters.

[0070] The above device can execute the method for acquiring the confidence of the geofence provided by any embodiment of the present application, has the function modules and beneficial effects corresponding to the execution method. Technical details not described in detail in the present embodiment can be referred to the method for acquiring the confidence of the geofence provided by any embodiment of the present application.

[0071] Embodiment Four

[0072] Figure 4 Figure 4 is a schematic diagram of an electronic device according to an embodiment of the application. Figure 4 Figure 1 shows a block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the application. Figure 4 The electronic device 12 shown is merely one example. It should be understood, however, that the functionality of the various embodiments of the application could be implemented in different environments without departing from the scope of the application.

[0073] As shown in Figure 4 Figure 1, the electronic device 12 is in the form of a general- purpose computer device. The components of the electronic device 12 can include, but are not limited to, one or more processors or processing units 16, a memory 28, and a bus 18 that couples various system components, including the memory 28 and the processing unit 16.

[0074] The bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics bus (e.g., an Accelerated Graphics Port, or AGP bus) and a processor or local bus using any of a variety of bus architectures. By way of example, these architectures include Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0075] The electronic device 12 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by the electronic device 12 and includes both volatile and non- volatile media, removable and non-removable media.

[0076] The memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 can further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 34 can be provided for reading from and writing to non-removable, non-volatile magnetic media (e.g., a "hard drive"). Figure 4 not shown, a magnetic hard disk drive for reading from and writing to non-removable, non-volatile magnetic media (e.g., a "hard drive"). Although not specifically shown, such Figure 4 a removable, non-volatile flash memory module. A storage controller can interface with the bus 18 and the memory 28, and can perform reading from and writing to the storage system 34. The memory 28 can include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the application.

[0077] Program / utility 40 having a set of program modules 42 can be stored in memory 28 by way of example, such program modules 42 include an operating system, one or more application programs, other program modules, and program data, each or some combination thereof, which may

[0078] Electronic device 12 can also communicate with one or more external devices 14 such as a keyboard or pointing device, a display 24, etc.; other devices such as portable memory storage devices (e.g., floppy disks, USB flash drives, portable memory cards, etc.); and user interfaces 26 or peripheral devices (e.g., a joystick, a game pad, a satellite dish, a scanner, or the like).

[0079] Processing unit(s) 16 can execute instructions and manipulate data to perform a variety of functions and operations, including the methods of the embodiments of the present application. For example, processing unit(s) 16 can execute instructions to implement the method of acquiring the confidence of a geofence according to the embodiments of the present application. That is, acquiring a driving image and a driving parameter, and constructing a geofence parameter according to the driving image and the driving parameter; wherein the driving parameter comprises at least one of a road curvature, a road slope, a weather type, and an illumination intensity; and acquiring a model evaluation confidence of the geofence according to the geofence parameter through a pre-trained confidence evaluation model.

[0080] Embodiment five

[0081] The embodiment five of the present application further provides a computer readable storage medium, which has stored thereon a computer program, and the computer program is executed by a processor to implement the method of acquiring the confidence of a geofence according to any of the embodiments of the present application; the method comprises:

[0082] acquiring a driving image and a driving parameter, and constructing a geofence parameter according to the driving image and the driving parameter; wherein the driving parameter comprises at least one of a road curvature, a road slope, a weather type, and an illumination intensity;

[0083] According to the geographic fence parameter, a model evaluation confidence of the geographic fence is acquired through a pre-trained confidence evaluation model.

[0084] The computer storage medium of the embodiments of the present application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples (non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, device or apparatus.

[0085] The computer readable signal medium can include a data signal propagating in baseband or propagating as a carrier wave in a propagated signal, in which computer readable program code is embodied. Such propagated signal can take a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a storage medium and that can communicate, propagate or transport program for use by or in connection with an instruction execution system, apparatus, or device.

[0086] The program code contained on the computer readable medium can be transmitted using any suitable medium, including, but not limited to, wireless, wire line, optical fiber, RF, etc., or any suitable combination thereof.

[0087] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0088] It is to be understood that the above description is merely a preferred embodiment of the application and the applied technical principles. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, modifications and substitutions can be made without departing from the scope of the present application. Therefore, although the present application has been described in detail through the above embodiments, the present application is not limited to the above embodiments, and can include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims

1. A method for obtaining geofencing confidence, characterized in that, The method comprises: acquiring driving images and driving parameters, and constructing a geofence parameter according to the driving images and the driving parameters; wherein the driving parameters include at least one of road curvature, road slope, weather type and illumination intensity; acquiring a model evaluation confidence of the geofence by a pre-trained confidence evaluation model according to the geofence parameter; wherein the pre-trained confidence evaluation model is acquired in the following way: generate training sample set, validation sample set and test sample set respectively according to a set proportion from the acquired historical sample information; wherein the historical sample information includes historical driving images captured by a high-resolution camera and historical driving parameters with the same timestamp as the historical driving images; assign corresponding confidence labels to the training sample set, the validation sample set and the test sample set according to artificial marking results; perform sample training on an initial model according to the training sample set, verify the training result by the validation sample set, test the recognition accuracy of the confidence evaluation model by the test sample set, and acquire the pre-trained confidence evaluation model; the method of acquiring a model evaluation confidence of the geofence by a pre-trained confidence evaluation model according to the geofence parameter comprises: acquire image features of the driving images by the confidence evaluation model; wherein the image features include color features, texture features, shape features and spatial relationship features; add the driving parameters to the color features, the texture features, the shape features and the spatial relationship features respectively, and acquire a model evaluation confidence of the geofence according to the color features, the texture features, the shape features and the spatial relationship features after adding the driving parameters; after acquiring the driving images and the driving parameters, the method further comprises: acquire a vehicle sensor list; wherein the vehicle sensor list includes at least one vehicle sensor related to the geofence and target driving parameters related to the working performance of the at least one vehicle sensor; acquire the reliability of the at least one vehicle sensor according to the driving parameters and the target driving parameters related to the working performance of the at least one vehicle sensor; acquire a sensor evaluation confidence of the geofence according to the reliability of the at least one vehicle sensor; wherein the sensor evaluation confidence is the proportion of vehicle sensors with high reliability in the total number of vehicle sensors; acquire an actual evaluation confidence of the geofence according to the model evaluation confidence and the sensor evaluation confidence of the geofence.

2. The method of claim 1, wherein, The confidence evaluation model includes at least one of a Transformer neural network model, a convolutional neural network model and a recurrent neural network model.

3. The method of claim 2, wherein, the method of acquiring a model evaluation confidence of the geofence by a pre-trained confidence evaluation model according to the geofence parameter comprises: According to the geographic fence parameters, a first reference confidence is obtained through the Transformer neural network model, a second reference confidence is obtained through the convolutional neural network model, and a third reference confidence is obtained through the recurrent neural network model; According to the first reference confidence, the second reference confidence, and the third reference confidence, a model evaluation confidence of the geographic fence is obtained.

4. The method of claim 1, wherein, The model evaluation confidence of the geographic fence and the sensor evaluation confidence are used to obtain an actual evaluation confidence of the geographic fence. The model evaluation confidence of the geographic fence and the sensor evaluation confidence, as well as a weight value of the model evaluation confidence and a weight value of the sensor evaluation confidence, are used to obtain an actual evaluation confidence of the geographic fence.

5. The method of claim 4, wherein, The weight value of the model evaluation confidence and the weight value of the sensor evaluation confidence are both related to the driving parameters.

6. An apparatus for acquiring a geo-fence confidence, the apparatus comprising: It includes: A geographic fence parameter acquisition module is configured to acquire driving images and driving parameters, and construct geographic fence parameters according to the driving images and the driving parameters; wherein the driving parameters include at least one of road curvature, road slope, weather type, and light intensity; A model evaluation confidence acquisition module is configured to obtain a model evaluation confidence of the geographic fence according to the geographic fence parameters through a pre-trained confidence evaluation model; The device further includes: A confidence evaluation model acquisition module is configured to generate training sample sets, validation sample sets, and test sample sets according to a set proportion from the obtained historical sample information; wherein the historical sample information includes historical driving images captured by a high-resolution camera and historical driving parameters with the same timestamp as the historical driving images; according to artificial labeling results, the training sample sets, the validation sample sets, and the test sample sets are respectively assigned corresponding confidence labels; an initial model is sample-trained according to the training sample sets, the training results are verified through the validation sample sets, and the recognition accuracy of the confidence evaluation model is tested through the test sample sets to obtain a pre-trained confidence evaluation model; The model evaluation confidence acquisition module includes: An image feature acquisition unit is configured to obtain image features of the driving images through the confidence evaluation model; wherein the image features include color features, texture features, shape features, and spatial relationship features; A driving parameter adding unit is configured to add the driving parameters to the color features, the texture features, the shape features, and the spatial relationship features, respectively, and obtain a model evaluation confidence of the geographic fence according to the color features, the texture features, the shape features, and the spatial relationship features after the driving parameters are added; The geographic fence confidence acquisition device further includes: A vehicle sensor list acquisition module is configured to acquire a vehicle sensor list; wherein the vehicle sensor list includes at least one vehicle sensor related to the geographic fence and target driving parameters related to the working performance of the at least one vehicle sensor; a reliability degree obtaining module, configured to obtain a reliability degree of the at least one vehicle sensor according to the driving parameter and a target driving parameter related to a working performance of the at least one vehicle sensor; a sensor evaluation confidence degree obtaining module, configured to obtain a sensor evaluation confidence degree of the geofence according to the reliability degree of the at least one vehicle sensor; wherein the sensor evaluation confidence degree is a proportion of vehicle sensors with high reliability degree in a total number of vehicle sensors; an actual evaluation confidence degree obtaining module, configured to obtain an actual evaluation confidence degree of the geofence according to the model evaluation confidence degree and the sensor evaluation confidence degree of the geofence.

7. An electronic device, comprising: The electronic device includes: one or more processors; a storage device configured to store one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method for obtaining the geofence confidence degree according to any one of claims 1-5.

8. A storage medium containing computer executable instructions for performing the method for obtaining the geofence confidence degree according to any one of claims 1-5 when executed by a computer processor.

Images