A method and apparatus for processing perception data, an electronic device, and a storage medium

Abstract

The application discloses a kind of processing method, device, electronic equipment and storage medium of perception data, it is related to vehicle control technical field, the method comprises: obtaining initial perception data, wherein, initial perception data includes initial identification information, at least one initial attribute label and the initial information corresponding to each initial attribute label;Determine whether initial perception data meets evaluation requirements based on initial identification information and preset identification information set;If it does not meet evaluation requirements, then classify initial perception data based on initial identification information;If it meets evaluation requirements, then determine target perception data based on initial identification information, each initial attribute label and each initial information;Classify target perception data based on each target attribute label, each target information and the reference information corresponding to each target attribute label.The application can improve the test efficiency and classification accuracy of perception data, reduce the probability of perception data missed test and misclassification, improve the processing quality of perception data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle control, and in particular to a perception data processing method and device, an electronic device and a storage medium. BACKGROUND

[0002] With the development of artificial intelligence, the automatic driving technology is also constantly breaking through, and the application of automatic driving vehicles is gradually widespread. The performance of automatic driving is one of the problems that automobile users focus on.

[0003] Automatic driving needs to use perception data, and therefore, the performance of automatic driving is related to the processing quality of the perception data. At present, the processing process of the perception data is manually completed by technical personnel. For example, the technical personnel analyzes the perception data based on evaluation indexes, and determines the perception data classification result based on the analysis conclusion. However, the manual analysis and classification of each perception data is time-consuming and laborious, and the processing efficiency and evaluation accuracy of the perception data are limited. In addition, the data volume of the perception data is generally large, and manual classification is easy to cause data omission and data misclassification, which affects the processing quality of the perception data. SUMMARY

[0004] The present application provides a perception data processing method and device, an electronic device and a storage medium, which aims to improve the test efficiency and classification accuracy of the perception data, reduce the probability of missing test and misclassification of the perception data, improve the processing quality of the perception data, and save the classification resources of the perception data without manual classification of each perception data.

[0005] According to an aspect of the present application, a perception data processing method is provided, which comprises:

[0006] obtaining initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label;

[0007] determining whether the initial perception data meets the evaluation requirement based on the initial identification information and a preset identification information set;

[0008] if the initial perception data does not meet the evaluation requirement, classifying the initial perception data based on the initial identification information;

[0009] if the initial perception data meets the evaluation requirement, determining target perception data based on the initial identification information, each initial attribute label and each initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels;

[0010] The target perception data is classified based on the target attribute labels, the target information, and the reference information corresponding to the target attribute labels.

[0011] Optionally, the initial perception data is obtained by determining to-be-detected perception data, and performing frame extraction on the to-be-detected perception data based on an output frame rate of the to-be-detected perception data to obtain the initial perception data.

[0012] Optionally, the initial perception data is obtained by determining a target frame extraction interval of the to-be-detected perception data based on the output frame rate of the to-be-detected perception data and a corresponding relationship between the output frame rate and a frame extraction interval, and performing frame extraction on the to-be-detected perception data based on the target frame extraction interval to obtain the initial perception data.

[0013] Optionally, the determination of whether the initial perception data meets the evaluation requirement based on the initial identification information and the preset identification information set comprises: determining whether the initial identification information exists in the preset identification information set; if the initial identification information exists in the preset identification information set, it is determined that the initial perception data meets the evaluation requirement; and if the initial identification information does not exist in the preset identification information set, it is determined that the initial perception data does not meet the evaluation requirement.

[0014] Optionally, the target perception data is determined based on the initial identification information, the initial attribute labels, and the initial information, which comprises: determining a reference attribute label set based on the initial identification information and a corresponding relationship between the identification information and the reference attribute label, wherein the reference attribute label set comprises at least one reference attribute label; determining at least one target attribute label based on a matching degree between each initial attribute label and each reference attribute label, wherein the number of the target attribute labels is the same as the number of the reference attribute labels; determining target information corresponding to each target attribute label based on the initial information corresponding to each target attribute label; and determining the target perception data based on each target attribute label and the target information corresponding to each target attribute label.

[0015] Optionally, for any one target attribute label, the target perception data is classified based on the target attribute labels, the target information, and the reference information corresponding to the target attribute labels, which comprises: determining a data difference between the target information and the reference information; determining whether the data difference is less than or equal to an evaluation threshold of the target attribute label; if the data difference is less than or equal to the evaluation threshold of the target attribute label, it is determined that a category of the target perception data in the target attribute label is normal; and if the data difference is greater than the evaluation threshold of the target attribute label, it is determined that the category of the target perception data in the target attribute label is abnormal.

[0016] Optionally, after it is determined that the category of the target perception data in the target attribute label is abnormal, the method further comprises: storing the target attribute label and the data difference between the target information and the reference information.

[0017] Optionally, the method further comprises: sorting the anomaly classification result of the perception data according to a preset rule; and determining an evaluation report of the perception data based on the sorting result.

[0018] According to another aspect of the present application, a perception data processing apparatus is provided, which comprises:

[0019] an information obtaining module configured to obtain initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label;

[0020] an information judging module configured to determine whether the initial perception data meets an evaluation requirement based on the initial identification information and a preset identification information set;

[0021] a first executing module configured to, if the initial perception data does not meet the evaluation requirement, classify the initial perception data based on the initial identification information;

[0022] an information determining module configured to, if the initial perception data meets the evaluation requirement, determine target perception data based on the initial identification information, the initial attribute labels and the initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels;

[0023] a second executing module configured to classify the target perception data based on the target attribute labels, the target information and reference information corresponding to the target attribute labels.

[0024] According to another aspect of the present application, an electronic device is provided, which comprises:

[0025] at least one processor; and a memory connected with the at least one processor in communication;

[0026] wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the perception data processing method according to any one of the embodiments of the present application.

[0027] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to execute the perception data processing method according to any one of the embodiments of the present application.

[0028] The technical scheme of the embodiment of the present application comprises the following steps: obtaining initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label; determining whether the initial perception data meets the evaluation requirement based on the initial identification information and a preset identification information set; if the initial perception data does not meet the evaluation requirement, classifying the initial perception data based on the initial identification information; if the initial perception data meets the evaluation requirement, determining target perception data based on the initial identification information, each initial attribute label and each initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels; and classifying the target perception data based on each target attribute label, each target information and reference information corresponding to each target attribute label. The present application can determine whether the perception data meets the evaluation requirement based on the identification information of the perception data, classify the perception data based on the identification information when the perception data meets the evaluation requirement, determine the target perception data when the perception data meets the evaluation requirement, and classify the target perception data based on the target attribute label of the target perception data, the target information corresponding to the target attribute label and the reference information corresponding to the target attribute label, thereby improving the test efficiency and classification accuracy of the perception data, reducing the probability of missed test and misclassification, improving the processing quality of the perception data, saving the classification resources of the perception data without manually classifying each perception data one by one, and solving the problems of time-consuming and laborious manual analysis and classification of each perception data, limited processing efficiency and evaluation accuracy of the perception data, and large data volume of the perception data, which may lead to data omission and misclassification during manual classification and affect the processing quality of the perception data.

[0029] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0031] Figure 1 is a flow diagram of a perception data processing method provided by an embodiment of the present application;

[0032] Figure 2 is a flow diagram of a perception data processing method provided by an embodiment of the present application;

[0033] Figure 3 is a structural schematic diagram of a perception data processing device provided by embodiment three of the present application;

[0034] Figure 4 is a structural schematic diagram of an electronic device provided by embodiment four of the present application. DETAILED DESCRIPTION

[0035] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor should fall within the protection scope of the present application.

[0036] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0037] Embodiment one

[0038] Figure 1 is a flowchart of a perception data processing method provided by embodiment one of the present application. The embodiment can be applicable to the classification of perception data, the determination of attribute errors of perception data, etc. The method can be executed by a perception data processing device provided by the present application. The device can be realized in the form of software and / or hardware. In a specific embodiment, the device can be integrated in an electronic device. The following embodiments will be described by taking the device integrated in an electronic device as an example. Referring to Figure 1 , the method specifically includes the following steps:

[0039] S101, obtaining initial perception data.

[0040] The initial perception data can be understood as data that needs to be evaluated, analyzed and classified, including initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label.

[0041] The perception data processing process in the present application is an offline data processing process. The initial perception data can be measurement data of a certain stage provided by a sensor or output data of a certain stage provided by a data processing model. Obtaining the initial perception data can be understood as obtaining the perception data of a certain stage output by the sensor or the data processing model.

[0042] Specifically, the initial identification information can be understood as a target category type represented by the perception data, including a vehicle, a pedestrian, a cyclist, a lane line, etc. The initial attribute label can be understood as a characteristic data name of the target category, including a lateral position, a longitudinal position, a size, a length-width information, a lateral speed, a longitudinal speed, a moving direction, etc. of the target category. The initial information can be understood as specific characteristic data of the target category, including specific data of the initial attribute label, such as the lateral position, the longitudinal position, the size, the length-width information, the lateral speed, the longitudinal speed, the moving direction, etc. of the target category. The present embodiment does not limit this.

[0043] For example, assuming that the initial identification information of the initial perception data is a pedestrian, the initial attribute label includes the lateral position, the longitudinal position, the longitudinal distance, the lateral distance, the lateral speed, the longitudinal speed, the moving direction, etc. of the pedestrian, and the initial information is the specific data of each initial attribute label.

[0044] The advantage of such a setting is that accurate perception data can be quickly obtained to efficiently analyze and classify the perception data.

[0045] S102, determining whether the initial perception data meets the evaluation requirement based on the initial identification information and the preset identification information set.

[0046] The preset identification information set can be understood as a set composed of at least one identification information in the image to be processed. According to the principle of the shortest center distance, the identification information corresponding to the initial identification information (equivalent to the true value of the initial identification information) in the preset identification information set can be determined. When the initial identification information corresponds to the true value of the initial identification information, it is considered that the vehicle processor can analyze and classify the initial perception data, and the initial perception data meets the evaluation requirement. When the initial identification information does not correspond to the true value of the initial identification information, it is considered that the vehicle processor cannot analyze and classify the initial perception data, and the initial perception data does not meet the evaluation requirement.

[0047] Specifically, if the initial perception data does not meet the evaluation requirement, step S103 is performed, and if the initial perception data meets the evaluation requirement, step S104 is performed.

[0048] For example, assuming that the true value of the initial identification information is a vehicle, if the initial identification information is a vehicle, it is determined that the initial perception data meets the evaluation requirement, and if the initial identification information is a pedestrian, it is determined that the initial perception data does not meet the evaluation requirement.

[0049] The advantage of such an arrangement is that the initial perception data can be screened, and the perception data that does not meet the evaluation requirement can be directly classified, without the need to classify the initial attribute label and the initial information corresponding to the initial attribute label, thereby reducing the work task of the vehicle processor.

[0050] S103, classifying the initial perception data based on the initial identification information.

[0051] Specifically, classifying the initial perception data based on the initial identification information can be understood as dividing the perception data whose initial identification information is not in the preset identification information set into a type error category. The advantage of such an arrangement is that it can intuitively indicate that the perception data in this category is not in the processing scope of the current perception data (the perception data is type error data), so that the technical personnel can understand the classification standard of the perception data and the perception data that is not subjected to attribute label analysis.

[0052] For example, when the initial perception data does not meet the evaluation requirement, it is considered that the type of the perception data is incorrect, and the name, number, detection value and true value (ideal value), current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is divided into a type error category.

[0053] S104, determining target perception data based on the initial identification information, the initial attribute labels and the initial information.

[0054] The target perception data can be understood as the attribute label and the information corresponding to the attribute label in the initial perception data that need to be analyzed and classified. The target perception data includes at least one target attribute label and at least one target information corresponding to the target attribute label, and the number of target attribute labels is less than or equal to the number of initial attribute labels.

[0055] Specifically, an initial perception data includes initial identification information and multiple initial attribute labels, but not all initial attribute labels need to be analyzed and classified. The target attribute label can be understood as the initial attribute label that needs to be analyzed and classified, and the target information can be understood as the information corresponding to the target attribute label.

[0056] For example, assuming that the initial identification information is a pedestrian, the initial attribute label includes the horizontal position, vertical position, spatial position, vertical distance, horizontal distance, horizontal speed, vertical speed, and moving direction of the pedestrian, the attributes that need to be analyzed and classified include the horizontal position, vertical position, horizontal speed, vertical speed, and moving direction of the pedestrian, and the target attribute label includes the horizontal position, vertical position, horizontal speed, vertical speed, and moving direction of the pedestrian, the target information is the specific data of each target attribute label.

[0057] The advantage of this arrangement is to reduce the processing amount of data and improve the classification efficiency of perception data.

[0058] S105, based on each target attribute label, each target information, and the reference information corresponding to each target attribute label, classifying the target perception data.

[0059] The reference information corresponding to the target attribute label can be understood as the true value corresponding to the target attribute label.

[0060] Classifying the target perception data based on each target attribute label, each target information, and the reference information corresponding to each target attribute label can be understood as classifying the target perception data according to the error between the target information of each target attribute label and the reference information of each target attribute label. Specifically, if the error between the target information of the target attribute label and the reference information of the target attribute label is greater than the pre-set evaluation threshold, it is determined that the target perception data is unqualified on the target attribute label, and if the error between the target information of the target attribute label and the reference information of the target attribute label is less than or equal to the pre-set evaluation threshold, it is determined that the target perception data is qualified on the target attribute label. The advantage of this arrangement is that the classification standard of the target perception data can be quantified, and the target perception data can be quickly and accurately classified.

[0061] The present application only counts the attribute label classification and perception data whose error between the target information and the reference information is greater than the evaluation threshold. It is worth noting that the reference information (true value) of each target attribute label is different, and the specific size of the true value is related to the classification standard and classification logic. When the accuracy of the perception data processing model is high or the number of iterations is large, a smaller evaluation threshold can be set, and when the accuracy of the perception data processing model is low or the number of iterations is small, a larger evaluation threshold can be set. This embodiment does not limit this.

[0062] Further, the method further includes: sorting the abnormal classification results of the perception data according to a pre-set rule; and determining an evaluation report of the perception data based on the sorting result.

[0063] The abnormal classification result can be understood as the perception data whose error of the target information and the reference information is greater than the evaluation threshold, and the evaluation report can be understood as the classification display result of the perception data. Specifically, the sorting of the abnormal classification result of the perception data according to the preset rule can be understood as sorting the classification results under each target attribute label in descending order of error, and displaying the perception data in descending order of error, so that the technical personnel can see the perception data with greater error first, so as to evaluate and understand the specific information of the perception data.

[0064] Optionally, the classification result of the perception data can also be used to evaluate the quality of each perception data processing model version. When the same batch of perception data is processed using the same frame interval and the same threshold, the quality of different versions can be compared according to the number of problems of each attribute type, so as to analyze the differences between the versions. Assuming that the target attribute label includes type, longitudinal distance, lateral distance, target length, target width, target lateral speed, target longitudinal speed and target orientation, when the type error statistical quantity, the longitudinal distance error statistical quantity, the lateral distance error statistical quantity, the target length error statistical quantity, the target width error statistical quantity, the target lateral speed error statistical quantity, the target longitudinal speed error statistical quantity or the target orientation statistical quantity of the A version is more than that of the B version, it indicates that the algorithm model detection accuracy of the B version is higher than that of the A, and the quality of the B version is better than that of the A. When multiple versions are iterated, graphical comparison between different versions can be performed for each attribute category, taking the version number as the horizontal coordinate and the number of problems as the vertical coordinate to obtain a line comparison graph of the number of problems of different versions, such as a target type error quantity comparison graph of different versions, a lateral distance error quantity comparison graph of different versions, a longitudinal distance error quantity comparison graph of different versions, a target length error quantity comparison graph of different versions, a target width error quantity comparison graph of different versions, a target lateral speed error quantity comparison graph of different versions, a target longitudinal speed error quantity comparison graph of different versions, a target orientation error quantity comparison graph of different versions, etc., which can directly reflect the differences and change trends of different versions.

[0065] The technical scheme of the embodiment is characterized in that initial perception data is acquired, wherein the initial perception data comprises initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label; whether the initial perception data meets evaluation requirements is determined based on the initial identification information and a preset identification information set; if the initial perception data does not meet the evaluation requirements, the initial perception data is classified based on the initial identification information; if the initial perception data meets the evaluation requirements, target perception data is determined based on the initial identification information, each initial attribute label and each initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels; and the target perception data is classified based on each target attribute label, each target information and reference information corresponding to each target attribute label. The application can determine whether the perception data meets the evaluation requirements based on the identification information of the perception data, classify the perception data based on the identification information when the perception data meets the evaluation requirements, determine the target perception data when the perception data meets the evaluation requirements, and automatically classify the target perception data based on the target attribute labels of the target perception data, the target information corresponding to the target attribute labels and the reference information corresponding to the target attribute labels, thereby improving the test efficiency and classification accuracy of the perception data, reducing the probability of missed tests and misclassification, improving the processing quality of the perception data, saving the classification resources of the perception data, and eliminating the problems of time-consuming and laborious manual analysis and classification of each perception data, limited processing efficiency and evaluation accuracy of the perception data, and large data volume of the perception data, data omission and data misclassification caused by manual classification, and affected processing quality of the perception data.

[0066] Embodiment two

[0067] Figure 2 is a flowchart of a perception data processing method provided by the embodiment two of the application. The embodiment can be applied to the classification of perception data, the determination of attribute errors of perception data and the like. The method can be executed by a perception data processing device provided by the application. The device can be realized in the form of software and / or hardware. In a specific embodiment, the device can be integrated in an electronic device. The following embodiment will be described by taking the device integrated in the electronic device as an example. For reference Figure 2 , the method specifically comprises the following steps:

[0068] S201, determine to-be-detected perception data.

[0069] The to-be-detected perception data can be understood as perception data that needs to be analyzed and classified. Specifically, the to-be-detected perception data comprises identification information, at least one attribute label and data information corresponding to the at least one attribute label.

[0070] For example, determining the to-be-detected perception data can be understood as directly obtaining the measurement data of a certain stage detected by the sensor, obtaining the perception data of a certain stage output by the data processing model, or obtaining the perception output result corresponding to the perception data after backfilling.

[0071] The advantage of this arrangement is that the complete perception data of the vehicle can be determined, and the accuracy of the perception data is ensured.

[0072] S202, frame the to-be-detected perception data based on the output frame rate of the to-be-detected perception data to obtain initial perception data.

[0073] The data amount of the to-be-detected perception data is large. If all the to-be-detected perception data is processed, on the one hand, the data processing efficiency will be affected, and on the other hand, the similarity of continuous data is high. Analyzing and classifying continuous data will cause the problem of repeated data in the analysis result.

[0074] Frame extraction can be understood as indirectly selecting data from continuous data in a certain way. In this embodiment, the to-be-detected perception data is frame extracted based on the output frame rate of the to-be-detected perception data. This can effectively control the data amount of the to-be-detected perception data, and can also reduce the problem of repeated classification of similar perception data caused by continuous perception data.

[0075] The advantage of this arrangement is that it can reduce the processing amount of perception data, improve the data processing efficiency, avoid the problem of repeated classification of similar perception data caused by continuous perception data, and ensure the stability of the analysis and processing result of the perception data.

[0076] In an embodiment, S202 can specifically include: determining a target frame extraction interval of the to-be-detected perception data based on the output frame rate of the to-be-detected perception data and the corresponding relationship between the output frame rate and the frame extraction interval; frame extracting the to-be-detected perception data based on the target frame extraction interval to obtain the initial perception data.

[0077] The target frame extraction interval can be understood as the frame extraction interval of the to-be-detected perception data.

[0078] Specifically, the corresponding relationship between the output frame rate and the frame extraction interval can be set and adjusted according to the processing requirements of the perception data. For example, a larger frame extraction interval can be set for perception data with a high output frame rate, and a smaller frame extraction interval can be set for perception data with a low output frame rate. This can control the data amount of the perception data and ensure the processing quality of the perception data. This embodiment does not limit this.

[0079] For example, suppose the frame extraction interval corresponding to the output frame rate of A is interval 1, the frame extraction interval corresponding to the output frame rate of B is interval 3, and the frame extraction interval corresponding to the output frame rate of C is interval 3. If the output frame rate of the sensing data to be detected is B, then the target frame extraction interval of the sensing data to be detected is interval 2. Sensing data will be extracted from the sensing data to be detected based on interval 2, and the extracted sensing data will be determined as the initial sensing data.

[0080] S203. Determine whether the initial sensing data meets the evaluation requirements based on the initial identification information and the preset identification information set.

[0081] On the one hand, the preset identification information can be understood as the type of perceived data that the vehicle processor can recognize and process. The preset identification information set can be understood as a collection composed of various preset identification information. When the initial identification information is an identification information in the preset identification information set, it is considered that the vehicle processor can analyze and classify the initial perceived data, and the initial perceived data meets the evaluation requirements. When the initial identification information is not an identification information in the preset identification information set, it is considered that the vehicle processor cannot analyze and classify the initial perceived data, and the initial perceived data does not meet the evaluation requirements. On the other hand, the preset identification information set can be understood as a collection composed of at least one identification information in the image to be processed. Based on the principle of shortest center distance, the identification information in the preset identification information set corresponding to the initial identification information (equivalent to the truth value of the initial identification information) can be determined. When the initial identification information corresponds to the truth value of the initial identification information, it is considered that the vehicle processor can analyze and classify the initial perceived data, and the initial perceived data meets the evaluation requirements. When the initial identification information does not correspond to the truth value of the initial identification information, it is considered that the vehicle processor cannot analyze and classify the initial perceived data, and the initial perceived data does not meet the evaluation requirements. This embodiment does not limit this aspect.

[0082] Specifically, if the initial perception data does not meet the evaluation requirements, then step S204 is executed; if the initial perception data meets the evaluation requirements, then step S205 is executed.

[0083] In one embodiment, S203 may specifically include: determining whether initial identification information exists in the preset identification information set; if initial identification information exists in the preset identification information set, then determining that the initial perception data meets the evaluation requirements; if initial identification information does not exist in the preset identification information set, then determining that the initial perception data does not meet the evaluation requirements.

[0084] Specifically, determining whether the initial identification information exists in the preset identification information set can be understood as determining whether the initial identification information is consistent with the truth value of the initial identification information (the identification information in the preset identification information set that has the shortest center distance from the initial identification information).

[0085] For example, assuming that the true value of the initial identification information is a vehicle, if the initial identification information is a vehicle, it is determined that the initial perception data meets the evaluation requirement, and if the initial identification information is a pedestrian or a cyclist, it is determined that the initial perception data does not meet the evaluation requirement.

[0086] S204, classifying the initial perception data based on the initial identification information.

[0087] Specifically, classifying the initial perception data based on the initial identification information can be understood as dividing the perception data whose initial identification information is not in the preset identification information set into the category of identification information error. The advantage of this setting is that it can intuitively indicate that this category of perception data is not in the processing scope of the current perception data, so that the technical personnel can understand the classification standard of the perception data and the perception data that has not been analyzed for attribute labels.

[0088] For example, when the initial perception data does not meet the evaluation requirement, it is considered that the type of the perception data is incorrect, and the name, number, detection value and true value (ideal value), current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is divided into the category of type error.

[0089] S205, determining target perception data based on the initial identification information, the initial attribute labels and the initial information.

[0090] The target perception data can be understood as the attribute labels and the information corresponding to the attribute labels in the initial perception data that need to be analyzed and classified. The target perception data includes at least one target attribute label and at least one target information corresponding to the target attribute label. The number of target attribute labels is less than or equal to the number of initial attribute labels.

[0091] Specifically, an initial perception data includes initial identification information and multiple initial attribute labels, but not all initial attribute labels need to be analyzed and classified. The target attribute label can be understood as the initial attribute label that needs to be analyzed and classified, and the target information can be understood as the information corresponding to the target attribute label.

[0092] In an embodiment, S205 can specifically include: determining a reference attribute label set based on the initial identification information and the correspondence between the identification information and the reference attribute label, wherein the reference attribute label set includes at least one reference attribute label; determining at least one target attribute label based on the matching degree between each initial attribute label and each reference attribute label, wherein the number of target attribute labels is the same as the number of reference attribute labels; determining target information corresponding to each target attribute label based on the initial information corresponding to each target attribute label; and determining target perception data based on each target attribute label and the target information corresponding to each target attribute label.

[0093] The reference attribute label can be understood as an attribute label corresponding to the initial label information that needs to be analyzed and classified, and the reference attribute label set can be understood as a set composed of the reference attribute labels.

[0094] For example, assuming that the initial identification information is a pedestrian, the initial attribute label includes the horizontal position, vertical position, spatial position, horizontal distance, vertical distance, horizontal speed, vertical speed, and moving direction of the pedestrian, and the attributes that need to be analyzed and classified include the horizontal position, vertical position, horizontal speed, vertical speed, and moving direction of the pedestrian, the reference attribute label set includes the horizontal position, vertical position, horizontal speed, vertical speed, and moving direction of the pedestrian, the target attribute label is the horizontal position, vertical position, horizontal speed, vertical speed, and moving direction of the pedestrian, and the target information can be understood as the initial information corresponding to each target attribute label.

[0095] S206, based on each target attribute label, each target information, and the reference information corresponding to each target attribute label, classifying the target perception data.

[0096] The reference information corresponding to the target attribute label can be understood as the true value corresponding to the target attribute label. Classifying the target perception data based on each target attribute label, each target information, and the reference information corresponding to each target attribute label can be understood as classifying the target perception data according to the error between the target information of each target attribute label and the reference information of each target attribute label. Specifically, if the error between the target information of the target attribute label and the reference information of the target attribute label is greater than a pre-set evaluation threshold, it is determined that the target perception data is unqualified in the target attribute label, and if the error between the target information of the target attribute label and the reference information of the target attribute label is less than or equal to the pre-set evaluation threshold, it is determined that the target perception data is qualified in the target attribute label. The advantage of such a setting is that it can quantify the classification standard of the target perception data, and quickly and accurately classify the target perception data.

[0097] Optionally, for any one target attribute label, based on each target attribute label, each target information, and the reference information corresponding to each target attribute label, classifying the target perception data includes: determining the data difference between the target information and the reference information; determining whether the data difference is less than or equal to the evaluation threshold of the target attribute label; if the data difference is less than or equal to the evaluation threshold of the target attribute label, determining that the category of the target perception data in the target attribute label is normal; and if the data difference is greater than the evaluation threshold of the target attribute label, determining that the category of the target perception data in the target attribute label is abnormal.

[0098] Specifically, the plurality of target attribute labels can be judged simultaneously or one by one, and the judgment method and order are not limited by the present application.

[0099] In an embodiment, the classification manner of the target attribute label is described in a one-by-one judgment manner. Assuming that the target attribute labels to be classified are longitudinal position, transverse position, length, width, transverse speed, longitudinal speed and orientation (heading angle) in turn, the classification manner of the perception data includes: 1) judging whether the longitudinal position of the perception data and the longitudinal position true value error are less than or equal to the set longitudinal position threshold value. If the error is greater than the longitudinal position threshold value, the name, target number, target longitudinal position detection value, longitudinal distance true value, error value, current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is classified into the category of large longitudinal position deviation. If the error is less than or equal to the longitudinal position threshold value, it is considered that the perception data does not have longitudinal position error, and does not need to be classified according to the longitudinal position error; 2) judging whether the transverse position of the perception data and the transverse position true value error are less than or equal to the set transverse position threshold value. If the error is greater than the transverse position threshold value, the name, target number, target transverse position detection value, transverse distance true value, error value, current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is classified into the category of large transverse position deviation. If the error is less than or equal to the transverse position threshold value, it is considered that the perception data does not have transverse position error, and does not need to be classified according to the transverse position error; 3) judging whether the length of the perception data and the length true value error are less than or equal to the set length threshold value. If the error is greater than the length threshold value, the name, target number, target length detection value, length true value, error value, current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is classified into the category of large length deviation. If the error is less than or equal to the length threshold value, it is considered that the perception data does not have length error, and does not need to be classified according to the length error; 4) judging whether the width of the perception data and the width true value error are less than or equal to the set width threshold value. If the error is greater than the width threshold value, the name, target number, target width detection value, width true value, error value, current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is classified into the category of large width deviation. If the error is less than or equal to the width threshold value, it is considered that the perception data does not have width error, and does not need to be classified according to the width error; 5) judging whether the transverse speed of the perception data and the transverse speed true value error are less than or equal to the set transverse speed threshold value. If the error is greater than the transverse speed threshold value, the name, target number, target transverse speed detection value, transverse speed true value, error value, current frame number, time stamp and other related logs of the perception data are recorded, and the perception data is classified into the category of large transverse speed deviation. If the error is less than or equal to the transverse speed threshold value, it is considered that the perception data does not have transverse speed error, and does not need to be classified according to the transverse speed error.6) judging whether the longitudinal speed of the perception data and the longitudinal speed true value error are less than or equal to a set longitudinal speed threshold value, if the error is greater than the longitudinal speed threshold value, recording the name of the perception data, the target number, the target longitudinal speed detection value, the longitudinal speed true value, the error value, the current frame number, the time stamp and other related logs, and classifying the perception data into a large longitudinal speed deviation category, if the error is less than or equal to the longitudinal speed threshold value, considering that the perception data does not have longitudinal speed error, and longitudinal speed error classification is not needed; 7) judging whether the orientation of the perception data and the orientation true value error are less than or equal to a set orientation threshold value, if the error is greater than the orientation threshold value, recording the name of the perception data, the target number, the target orientation detection value, the orientation true value, the error value, the current frame number, the time stamp and other related logs, and classifying the perception data into a large orientation deviation category, if the error is less than or equal to the orientation threshold value, considering that the perception data does not have orientation error, and orientation error classification is not needed. Specifically, each threshold value is related to the classification standard of the perception data, and can be set and adjusted according to the classification requirement, which is not limited in the embodiment.

[0100] Further, after determining that the category of the target perception data in the target attribute label is abnormal, the method further includes: storing the data difference between the target attribute label and the target information and the reference information. The advantage of such setting is to facilitate the operation personnel to use the abnormal data, for example, evaluating the model version based on the abnormal data, generating a test report based on the abnormal data, and the like, which is not limited in the embodiment.

[0101] The technical scheme of the embodiment is characterized in that: the to-be-detected perception data is determined; the to-be-detected perception data is frame-extracted based on an output frame rate of the to-be-detected perception data, to obtain initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label, and initial information corresponding to the at least one initial attribute label; whether the initial perception data meets an evaluation requirement is determined based on the initial identification information and a preset identification information set; if the initial perception data does not meet the evaluation requirement, the initial perception data is classified based on the initial identification information; if the initial perception data meets the evaluation requirement, target perception data is determined based on the initial identification information, each initial attribute label, and each initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels; and the target perception data is classified based on each target attribute label, each target information, and reference information corresponding to each target attribute label. The application can frame-extract the to-be-detected perception data, reduce the data processing amount, save the operation resources of a processor, determine whether the perception data meets the evaluation requirement based on the identification information of the perception data, classify the perception data based on the identification information when the perception data meets the evaluation requirement, determine the target perception data when the perception data meets the evaluation requirement, and classify the target perception data based on the target attribute labels of the target perception data, the target information corresponding to the target attribute labels, and the reference information corresponding to the target attribute labels, so as to improve the test efficiency and classification accuracy of the perception data, reduce the probability of missed detection and misclassification, improve the processing quality of the perception data, save the classification resources of the perception data, and avoid the problems of time-consuming and laborious manual analysis and classification of each perception data, limited processing efficiency and evaluation accuracy of the perception data, and large data volume of the perception data, which may lead to data omission and misclassification during manual classification, and affect the processing quality of the perception data.

[0102] Embodiment three

[0103] Figure 3 is a structural schematic diagram of a perception data processing device provided by the embodiment three of the application. As shown in the figure, the device comprises an information acquisition module 301, an information judgment module 302, a first execution module 303, an information determination module 304, and a second execution module 305. Figure 3

[0104] The information acquisition module 301 is configured to acquire initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label, and initial information corresponding to the at least one initial attribute label.

[0105] The information judgment module 302 is configured to determine whether the initial perception data meets an evaluation requirement based on the initial identification information and a preset identification information set.​

[0106] The first execution module 303 is configured to classify the initial perception data based on the initial identification information if the initial perception data does not meet the evaluation requirement.

[0107] The information determination module 304 is configured to determine target perception data based on the initial identification information, the initial attribute labels and the initial information if the initial perception data meets the evaluation requirement, wherein the target perception data includes at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels.

[0108] The second execution module 305 is configured to classify the target perception data based on the target attribute labels, the target information and reference information corresponding to the target attribute labels.

[0109] Optionally, the information acquisition module 301 is specifically configured to determine the to-be-detected perception data; and frame the to-be-detected perception data based on an output frame rate of the to-be-detected perception data to obtain the initial perception data.

[0110] Optionally, the information acquisition module 301 is specifically configured to determine a target frame interval of the to-be-detected perception data based on the output frame rate of the to-be-detected perception data and a corresponding relationship between the output frame rate and the frame interval; and frame the to-be-detected perception data based on the target frame interval to obtain the initial perception data.

[0111] Optionally, the information judgment module 302 is specifically configured to determine whether the initial identification information exists in the preset identification information set; if the initial identification information exists in the preset identification information set, it is determined that the initial perception data meets the evaluation requirement; and if the initial identification information does not exist in the preset identification information set, it is determined that the initial perception data does not meet the evaluation requirement.

[0112] Optionally, the information determination module 304 is specifically configured to determine a reference attribute label set based on the initial identification information and a corresponding relationship between the identification information and the reference attribute labels, wherein the reference attribute label set includes at least one reference attribute label; determine at least one target attribute label based on matching degrees between the initial attribute labels and the reference attribute labels, wherein the number of the target attribute labels is the same as the number of the reference attribute labels; determine target information corresponding to the target attribute labels based on the initial information corresponding to the target attribute labels; and determine the target perception data based on the target attribute labels and the target information corresponding to the target attribute labels.

[0113] Optionally, for any one target attribute label, the second execution module 305 is specifically configured to determine a data difference between the target information and the reference information; determine whether the data difference is less than or equal to an evaluation threshold of the target attribute label; if the data difference is less than or equal to the evaluation threshold of the target attribute label, determine that the category of the target perception data in the target attribute label is normal; and if the data difference is greater than the evaluation threshold of the target attribute label, determine that the category of the target perception data in the target attribute label is abnormal.

[0114] Optionally, the apparatus further comprises a storage module configured to store the target attribute label and the data difference between the target information and the reference information after determining that the category of the target perception data in the target attribute label is abnormal.

[0115] Optionally, the apparatus further comprises a statistical module configured to sort the abnormal category results of the perception data according to a preset rule; and determine an evaluation report of the perception data based on the sorting result.

[0116] The perception data processing apparatus provided by the embodiments of the present application can execute the perception data processing method provided by any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0117] Embodiment four

[0118] Figure 4 is a structural schematic diagram of an electronic device provided by Embodiment Four of the present application. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application described and / or claimed in this document.

[0119] As Figure 4As shown, the electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected to the at least one processor 11 in communication. The memory stores a computer program executable by the at least one processor 11, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0120] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, a speaker, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0121] The processor 11 can be various general and / or special-purpose processing components having processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the processing method of perception data.

[0122] In some embodiments, the processing method of perception data can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the processing method of perception data described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the processing method of perception data by any other appropriate means, such as by means of firmware.

[0123] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system on a chip systems (SOCs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0124] Computer programs used to implement the processes of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer program

[0125] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. Computer-readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0126] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0127] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0128] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0129] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in a different order, as long as the desired results of the present disclosure are achieved, and the present disclosure is not limited herein.

[0130] The specific embodiments described above are not intended to be limiting, and persons skilled in the art will appreciate that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the disclosure. Accordingly, the disclosure is not limited to the specific embodiments described above, but only by the scope of the appended claims.

Claims

1. A method of processing perception data, the method comprising: The method comprises the following steps: acquiring initial perception data, wherein the initial perception data comprises initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label; determining whether the initial perception data meets evaluation requirements based on the initial identification information and a preset identification information set; if the initial perception data does not meet the evaluation requirements, classifying the initial perception data based on the initial identification information; if the initial perception data meets the evaluation requirements, determining target perception data based on the initial identification information, each initial attribute label and each initial information, wherein the target perception data comprises at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of the target attribute labels is less than or equal to the number of the initial attribute labels; classifying the target perception data based on each target attribute label, each target information and reference information corresponding to each target attribute label; the step of determining the target perception data based on the initial identification information, each initial attribute label and each initial information comprises the following steps: determining a reference attribute label set based on the initial identification information and a correspondence between identification information and reference attribute labels, wherein the reference attribute label set comprises at least one reference attribute label; determining the at least one target attribute label based on a matching degree between each initial attribute label and each reference attribute label, wherein the number of the target attribute labels is the same as the number of the reference attribute labels; determining target information corresponding to each target attribute label based on the initial information corresponding to each target attribute label; determining the target perception data based on each target attribute label and the target information corresponding to each target attribute label; if the initial perception data does not meet the evaluation requirements, classifying the initial perception data based on the initial identification information, specifically, classifying perception data whose initial identification information is not in the preset identification information set into a category of identification information errors; the reference information corresponding to the target attribute label is a true value corresponding to the target attribute label; wherein the perception data comprises vehicles, pedestrians, cyclists and lane lines.

2. The method of claim 1, wherein, The step of acquiring the initial perception data comprises the following steps: determining to-be-detected perception data; decimating the to-be-detected perception data based on an output frame rate of the to-be-detected perception data to obtain the initial perception data.

3. The method of claim 2, wherein, The step of decimating the to-be-detected perception data based on the output frame rate of the to-be-detected perception data to obtain the initial perception data comprises the following steps: determining a target decimation interval of the to-be-detected perception data based on the output frame rate of the to-be-detected perception data and a correspondence between an output frame rate and a decimation interval; decimating the to-be-detected perception data based on the target decimation interval to obtain the initial perception data.

4. The method of claim 1, wherein, The step of determining whether the initial perception data meets the evaluation requirements based on the initial identification information and the preset identification information set comprises the following steps: determining whether the initial identification information exists in the preset identification information set; If the initial identification information exists in the preset identification information set, it is determined that the initial perception data meets the evaluation requirement. If the initial identification information does not exist in the preset identification information set, it is determined that the initial perception data does not meet the evaluation requirement.

5. The method of claim 1, wherein, For any target attribute label, the target perception data is classified based on each target attribute label, each target information and reference information corresponding to each target attribute label, including: determining the data difference between the target information and the reference information; determining whether the data difference is less than or equal to the evaluation threshold of the target attribute label; if the data difference is less than or equal to the evaluation threshold of the target attribute label, it is determined that the category of the target perception data in the target attribute label is normal; if the data difference is greater than the evaluation threshold of the target attribute label, it is determined that the category of the target perception data in the target attribute label is abnormal.

6. The method of claim 5, wherein, Further comprising: sorting the abnormal classification results of the perception data according to a preset rule; determining the evaluation report of the perception data based on the sorting result.

7. A processing device of perception data, characterized in that, including: an information acquisition module for acquiring initial perception data, wherein the initial perception data includes initial identification information, at least one initial attribute label and initial information corresponding to the at least one initial attribute label; an information judgment module for determining whether the initial perception data meets the evaluation requirement based on the initial identification information and a preset identification information set; a first execution module for classifying the initial perception data based on the initial identification information if the initial perception data does not meet the evaluation requirement; an information determination module for determining target perception data based on the initial identification information, each initial attribute label and each initial information if the initial perception data meets the evaluation requirement, wherein the target perception data includes at least one target attribute label and target information corresponding to the at least one target attribute label, and the number of target attribute labels is less than or equal to the number of initial attribute labels; a second execution module for classifying the target perception data based on each target attribute label, each target information and reference information corresponding to each target attribute label; the information determination module is specifically configured to determine a reference attribute label set based on the initial identification information and the correspondence between the identification information and the reference attribute label, wherein the reference attribute label set includes at least one reference attribute label; determine at least one target attribute label based on the matching degree between each initial attribute label and each reference attribute label, wherein the number of target attribute labels is the same as the number of reference attribute labels; determine target information corresponding to each target attribute label based on initial information corresponding to each target attribute label; determine target perception data based on each target attribute label and target information corresponding to each target attribute label; if the initial perception data does not meet the evaluation requirement, the initial perception data is classified based on the initial identification information, specifically: the perception data whose initial identification information is not in the preset identification information set is divided into a category of identification information error. The reference information corresponding to the target attribute label is a true value corresponding to the target attribute label. The perception data includes a vehicle, a pedestrian, a cyclist, and a lane line.

8. An electronic device, comprising: The electronic device includes: at least one processor; and a memory connected to the at least one processor in communication; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the perception data processing method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the processor to execute the perception data processing method of any one of claims 1 to 6.

Images