Methods, systems, devices, and storage media for mining high value traffic flow segments

Abstract

The application provides a method, system, device and storage medium for mining high-value traffic flow segments, the method comprising: analyzing acquired traffic data to determine a feature threshold, determining a basic sub-scene according to the feature threshold, and combining the basic sub-scene to generate a common traffic flow segment; marking a high-value traffic flow segment in the common traffic flow segment, and constructing a feature model according to the marked high-value traffic flow segment; determining a multi-feature joint threshold of the high-value traffic flow segment according to the feature model; and screening the high-value traffic flow segment from a to-be-mined traffic flow segment according to the multi-feature joint threshold. The application can screen high-value traffic flow from common traffic flow, and improves the utilization rate of data.

Description

Technical Field

[0001] This invention relates to the field of vehicle-road cooperation, and more specifically, to a method, system, device, and storage medium for mining high-value traffic flow segments. Background Technology

[0002] In the field of vehicle-road cooperation, traffic data processing is crucial. Among these, the extraction of high-value traffic segments is of great significance for realizing intelligent transportation systems. Traditional autonomous driving research also studies high-value traffic segments, but it typically uses limited data sources, such as single-vehicle road data or data generated through multi-vehicle simulations. This leads to discrepancies between the obtained traffic flow segments and the actual traffic flow segments. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a method, system, electronic device, and computer-readable storage medium for mining high-value traffic flow segments. This invention constructs a feature model based on the marked high-value traffic flow segments, determines a multi-feature joint threshold for high-value traffic flow segments based on the feature model, and filters high-value traffic flow segments based on the multi-feature joint threshold, thereby improving data utilization.

[0004] To achieve the above objectives, one aspect of the present invention provides a method for mining high-value traffic flow segments, comprising the following steps: analyzing acquired traffic data to determine feature thresholds; determining basic sub-scenes based on the feature thresholds; and combining the basic sub-scenes to generate ordinary traffic flow segments; marking high-value traffic flow segments in the ordinary traffic flow segments; and constructing a feature model based on the marked high-value traffic flow segments; determining a multi-feature joint threshold for high-value traffic flow segments based on the feature model; and filtering high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

[0005] In some implementations, the step of analyzing the acquired traffic data to determine the feature threshold includes: combining map data to create a basic feature wide table that includes motion state information, basic attribute information, map information and information of adjacent traffic participants, and determining the feature threshold based on the basic feature wide table.

[0006] In some implementations, the step of determining the feature threshold based on the basic feature wide table includes: plotting a curve based on the basic feature wide table, and using the intervals in the curve where feature values ​​change abruptly as the feature thresholds.

[0007] In some implementations, the step of constructing a feature model based on the marked high-value traffic flow segments includes: constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments respectively.

[0008] In some implementations, the steps of constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the labeled high-value traffic flow segments include: constructing a multi-feature joint distribution scatter matrix based on the features of the labeled high-value traffic flow segments and non-high-value traffic flow segments; and constructing a multi-feature Gaussian regression model based on the relationships between the features in the labeled high-value traffic flow segments.

[0009] In some implementations, the step of determining the multi-feature joint threshold of high-value traffic flow segments based on the feature model includes: constructing a covariance matrix of multiple features; constructing a probability density function of a Gaussian distribution based on the covariance matrix; and calculating the threshold of each feature in the probability density function using a preset confidence level to form a multi-feature joint threshold.

[0010] In some implementations, the method further includes: reviewing the selected high-value traffic flow segments and marking misjudged traffic flow segments; and inputting the selected correct high-value traffic flow segments and the misjudged traffic flow segments into a regression model to update the multi-feature joint threshold.

[0011] In another aspect, this invention provides a system for mining high-value traffic flow segments, comprising: an analysis module configured to analyze acquired traffic data to determine feature thresholds, determine basic sub-scenes based on the feature thresholds, and combine the basic sub-scenes to generate ordinary traffic flow segments; a construction module configured to mark high-value traffic flow segments in the ordinary traffic flow segments and construct feature models based on the marked high-value traffic flow segments; a threshold module configured to determine a multi-feature joint threshold for high-value traffic flow segments based on the feature models; and an execution module configured to filter high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

[0012] In another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory storing computer instructions executable on the processor, the instructions, when executed by the processor, implementing the steps of the method described above.

[0013] In another aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method steps.

[0014] The present invention has the following beneficial technical effects: it constructs a feature model based on the marked high-value traffic flow segments, determines the multi-feature joint threshold of the high-value traffic flow segments based on the feature model, and filters the high-value traffic flow segments based on the multi-feature joint threshold, thereby improving the utilization rate of data. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic diagram illustrating an embodiment of the method for mining high-value traffic flow segments provided by the present invention;

[0017] Figure 2 A schematic diagram of an embodiment of the present invention for determining thresholds for rapid acceleration and rapid deceleration;

[0018] Figure 3 This is a schematic diagram of an embodiment of constructing a multi-feature covariance matrix provided by the present invention;

[0019] Figure 4 This is a schematic diagram of an embodiment of the probability density function for constructing a multivariate Gaussian distribution provided by the present invention;

[0020] Figure 5 A schematic diagram illustrating an embodiment of the system for mining high-value traffic flow segments provided by the present invention;

[0021] Figure 6 A schematic diagram of the hardware structure of an embodiment of the electronic device for mining high-value traffic flow segments provided by the present invention;

[0022] Figure 7 This is a schematic diagram of an embodiment of a computer storage medium for mining high-value traffic flow segments provided by the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to specific examples and the accompanying drawings.

[0024] It should be noted that all uses of "first" and "second" in the embodiments of the present invention are for the purpose of distinguishing two entities or parameters with the same name but different names. It is clear that "first" and "second" are only for the convenience of expression and should not be construed as limiting the embodiments of the present invention. Subsequent embodiments will not explain this in detail.

[0025] In a first aspect, an embodiment of a method for mining high-value traffic flow segments is proposed. Figure 1 The diagram shown is an embodiment of the method for mining high-value traffic flow segments provided by the present invention. Figure 1As shown, the embodiments of the present invention include the following steps:

[0026] S1. Analyze the acquired traffic data to determine the feature threshold, determine the basic sub-scene based on the feature threshold, and combine the basic sub-scenes to generate ordinary traffic flow segments.

[0027] S2. High-value traffic flow segments in the ordinary traffic flow segments are marked, and feature models are constructed based on the marked high-value traffic flow segments;

[0028] S3. Determine the multi-feature joint threshold of high-value traffic flow segments based on the feature model; and

[0029] S4. Select high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

[0030] The acquired traffic data is analyzed to determine feature thresholds, and basic sub-scenes are determined based on these thresholds. These basic sub-scenes are then combined to generate ordinary traffic flow segments. The acquired traffic data is cleaned and its quality improved: A big data ETL (extract, transform, load) process is used to clean the perceived data to improve its quality; obvious abnormal data in features such as speed, acceleration, lateral offset, and direction of motion are removed; abnormal perception points such as ID (identifier) ​​jumps and type jumps are identified and removed, and data within two seconds before and after these jumps are also removed; position smoothing and speed smoothing are performed on the position, speed, and acceleration data in the perceived data.

[0031] In some implementations, the step of analyzing the acquired traffic data to determine feature thresholds includes: creating a broad table of basic features that includes motion state information, basic attribute information, map information, and information on adjacent traffic participants, based on map data; and determining feature thresholds based on the broad table of basic features. The broad table of basic features for traffic participants created by combining map data includes: motion state (speed, acceleration, orientation, etc.) and basic attribute information; map information such as the intersection, lane, and traffic lights; and basic information on adjacent traffic participants in the same lane and in the lanes before and after.

[0032] In some implementations, the step of determining the feature threshold based on the basic feature wide table includes: plotting a curve based on the basic feature wide table, and using the intervals in the curve where feature values ​​change abruptly as the feature thresholds. Figure 2 A schematic diagram illustrating the determination of thresholds for rapid acceleration and rapid deceleration provided by the present invention, as shown below. Figure 2As shown, the horizontal axis represents the magnitude of acceleration, and the vertical axis represents the number of objects corresponding to the acceleration. The rate of change is shown in the 0.1 intervals. Based on the rate of change of the number of objects corresponding to the acceleration (the acceleration decreases relatively evenly to the left of the first vertical line and to the right of the second vertical line in the figure, which does not indicate an acceleration anomaly), the threshold for rapid acceleration can be determined as: acceleration greater than 2.7 m / s². 2 The threshold for rapid deceleration is: acceleration less than -4.2 m / s². 2 .

[0033] Basic sub-scenes are calculated using feature thresholds, such as speeding, sudden deceleration, sudden acceleration, deviation from the road, starting, stopping, and slow movement. Combined with intersection information, composite sub-scenes are calculated, such as lane changing, driving against traffic, crossing the road, illegal parking, U-turns, and turns. All traffic segments to be mined are analyzed to identify the sub-scenes needed for each traffic flow segment. In terms of time and space, basic sub-scenes are combined to generate a large number of ordinary traffic flow segments. For example, for illegal parking: identify sub-scenes of vehicles stopping, vehicles slowing down, and vehicles changing lanes and overtaking in the same time and area, and combine these three labels to generate illegal parking traffic flow segments.

[0034] High-value traffic flow segments within the general traffic flow segments are marked, and feature models are constructed based on these marked segments. Environmental information is supplemented to the general traffic flow segments, including intersection shape, area type, lane type, traffic light status, weather conditions (meteorology, wind speed, temperature), and time period (morning / evening rush hour, day / night). High-value traffic flow segments are manually marked for the first time: based on video viewing and visual traffic flow analysis, high-value and non-high-value traffic flow segments are manually marked.

[0035] In some implementations, the step of constructing a feature model based on the marked high-value traffic flow segments includes: constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments respectively.

[0036] In some implementations, the steps of constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the labeled high-value traffic flow segments include: constructing a multi-feature joint distribution scatter matrix based on the features of the labeled high-value traffic flow segments and non-high-value traffic flow segments; and constructing a multi-feature Gaussian regression model based on the relationships between the features in the labeled high-value traffic flow segments. For key features of the labeled high-value and non-high-value traffic flow segments, such as speed, acceleration, distance, duration, and offset, a multi-feature joint distribution scatter matrix is ​​created using tools such as MATLAB; using the labeled high-value traffic flow data, a multi-feature Gaussian regression model is constructed to capture the complex relationships between the features.

[0037] In some implementations, the step of constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments includes: evaluating the fitting effect of the Gaussian regression model on the training data using mean squared error, root mean square error, and log loss; and reconstructing the Gaussian regression model in response to a failure to pass the evaluation.

[0038] The multi-feature joint threshold for high-value traffic flow segments is determined based on the feature model.

[0039] In some implementations, the step of determining the multi-feature joint threshold of high-value traffic flow segments based on the feature model includes: constructing a covariance matrix of multiple features; constructing a probability density function of a Gaussian distribution based on the covariance matrix; and calculating the threshold of each feature in the probability density function using a preset confidence level to form a multi-feature joint threshold.

[0040] Figure 3 This is a schematic diagram illustrating the construction of a multi-feature covariance matrix provided by the present invention; as shown. Figure 3 As shown, given a dataset X, assuming there are n samples and m features, the covariance matrix S of dataset X can be obtained.

[0041] Figure 4 This is a schematic diagram illustrating the construction of the probability density function of a multivariate Gaussian distribution provided by the present invention; extending the Gaussian distribution to a multidimensional space, it can be assumed that each space in the multidimensional space is completely independent, and then according to... Figure 4 Determine an independent multivariate Gaussian distribution.

[0042] Using a 95% confidence level, calculate the thresholds for each feature of the probability density function.

[0043] High-value traffic flow segments are screened from the traffic flow segments to be mined based on the multi-feature joint threshold. Using the labeled high-value traffic flows, a threshold range of 95% (2 standard deviations) is determined, and high-value traffic flows are mined from ordinary traffic flow segments using the joint threshold of each feature (within the threshold range).

[0044] In some implementations, the method further includes: reviewing the selected high-value traffic flow segments and marking misclassified traffic flow segments; and inputting the correctly selected high-value traffic flow segments and the misclassified traffic flow segments into a regression model to update the multi-feature joint threshold. High-value traffic flow segments are periodically provided to the simulation training team for autonomous driving scenario training, and misclassified high-value traffic flows are marked during the process. The model is iterated daily based on the high-value traffic flow segments mined the previous day and the misclassified traffic flow segments marked by the simulation team to iterate the multi-feature joint threshold.

[0045] It should be noted that the steps in each embodiment of the above-described method for mining high-value traffic flow segments can be interleaved, substituted, added, or deleted. Therefore, these reasonable permutations and combinations of the method for mining high-value traffic flow segments should also fall within the scope of protection of this invention, and the scope of protection of this invention should not be limited to the embodiments.

[0046] Based on the above objectives, a second aspect of this invention proposes a system for mining high-value traffic flow segments. For example... Figure 5 As shown, system 200 includes the following modules: an analysis module configured to analyze the acquired traffic data to determine feature thresholds, determine basic sub-scenes based on the feature thresholds, and combine the basic sub-scenes to generate ordinary traffic flow segments; a construction module configured to mark high-value traffic flow segments in the ordinary traffic flow segments and construct feature models based on the marked high-value traffic flow segments; a threshold module configured to determine a multi-feature joint threshold for high-value traffic flow segments based on the feature models; and an execution module configured to filter high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

[0047] In some implementations, the analysis module is further configured to: combine map data to create a basic feature wide table including motion state information, basic attribute information, map information and adjacent traffic participant information, and determine feature thresholds based on the basic feature wide table.

[0048] In some implementations, the analysis module is further configured to: draw a curve based on the basic feature wide table, and use the intervals in the curve where feature values ​​change abruptly as feature thresholds.

[0049] In some implementations, the building module is also configured to: construct a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments.

[0050] In some implementations, the building module is also configured to: construct a multi-feature joint distribution scatter matrix based on the features of labeled high-value traffic flow segments and non-high-value traffic flow segments; and construct a multi-feature Gaussian regression model based on the relationships between features in the labeled high-value traffic flow segments.

[0051] In some implementations, the threshold module is further configured to: construct a covariance matrix of multiple features, construct a probability density function of a Gaussian distribution based on the covariance matrix, and calculate the threshold of each feature in the probability density function using a preset confidence level to form a joint threshold of multiple features.

[0052] In some embodiments, the system further includes a verification module configured to: verify the selected high-value traffic flow segments and mark misjudged traffic flow segments; and input the correctly selected high-value traffic flow segments and the misjudged traffic flow segments into a regression model to update the multi-feature joint threshold.

[0053] Based on the above objectives, a third aspect of the present invention provides an electronic device, comprising: at least one processor; and a memory storing computer instructions executable by the processor to perform the following steps: S1, analyzing acquired traffic data to determine feature thresholds, determining basic sub-scenes based on the feature thresholds, and combining the basic sub-scenes to generate ordinary traffic flow segments; S2, marking high-value traffic flow segments in the ordinary traffic flow segments, and constructing a feature model based on the marked high-value traffic flow segments; S3, determining a multi-feature joint threshold for high-value traffic flow segments based on the feature model; and S4, filtering high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

[0054] In some implementations, the step of analyzing the acquired traffic data to determine the feature threshold includes: combining map data to create a basic feature wide table that includes motion state information, basic attribute information, map information and information of adjacent traffic participants, and determining the feature threshold based on the basic feature wide table.

[0055] In some implementations, the step of determining the feature threshold based on the basic feature wide table includes: plotting a curve based on the basic feature wide table, and using the intervals in the curve where feature values ​​change abruptly as the feature thresholds.

[0056] In some implementations, the step of constructing a feature model based on the marked high-value traffic flow segments includes: constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments respectively.

[0057] In some implementations, the steps of constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the labeled high-value traffic flow segments include: constructing a multi-feature joint distribution scatter matrix based on the features of the labeled high-value traffic flow segments and non-high-value traffic flow segments; and constructing a multi-feature Gaussian regression model based on the relationships between the features in the labeled high-value traffic flow segments.

[0058] In some implementations, the step of determining the multi-feature joint threshold of high-value traffic flow segments based on the feature model includes: constructing a covariance matrix of multiple features; constructing a probability density function of a Gaussian distribution based on the covariance matrix; and calculating the threshold of each feature in the probability density function using a preset confidence level to form a multi-feature joint threshold.

[0059] In some implementations, the steps further include: reviewing the selected high-value traffic flow segments and marking misjudged traffic flow segments; and inputting the selected correct high-value traffic flow segments and the misjudged traffic flow segments into a regression model to update the multi-feature joint threshold.

[0060] like Figure 6 The diagram shown is a hardware structure schematic of an embodiment of the electronic device for mining high-value traffic flow segments provided by the present invention.

[0061] For example Figure 6 Taking the device shown as an example, the device includes a processor 301 and a memory 302.

[0062] Processor 301 and memory 302 can be connected via a bus or other means. Figure 6 Taking the example of a connection between China and Israel via a bus.

[0063] The memory 302, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for mining high-value traffic flow segments in this embodiment. The processor 301 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions, and modules stored in the memory 302, thereby realizing the method for mining high-value traffic flow segments.

[0064] Memory 302 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of methods for mining high-value traffic flow segments, etc. Furthermore, memory 302 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 302 may optionally include memory remotely located relative to processor 301, and these remote memories may be connected to the local module via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0065] One or more methods for mining high-value traffic flow segments are stored in memory 302. When executed by processor 301, the method for mining high-value traffic flow segments in any of the above method embodiments is executed.

[0066] Any embodiment of the electronic device that performs the above-described method for mining high-value traffic flow segments can achieve the same or similar effects as any of the aforementioned method embodiments.

[0067] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs a method for mining high-value traffic flow segments.

[0068] like Figure 7 The diagram shown is a schematic representation of an embodiment of the computer storage medium for mining high-value traffic flow segments provided by the present invention. Figure 7 Taking the computer storage medium shown as an example, the computer-readable storage medium 401 stores a computer program 402 that, when executed by a processor, performs the above method.

[0069] Finally, it should be noted that those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program for mining high-value traffic flow segments can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The storage medium for the program can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. The above computer program embodiments can achieve the same or similar effects as any of the corresponding foregoing method embodiments.

[0070] The above are exemplary embodiments disclosed in this invention. However, it should be noted that various changes and modifications can be made without departing from the scope of the embodiments of this invention as defined by the claims. The functions, steps, and / or actions of the methods according to the disclosed embodiments described herein do not need to be performed in any particular order. Furthermore, although the elements disclosed in the embodiments of this invention may be described or claimed individually, they may be understood as multiple unless explicitly limited to a singular number.

[0071] It should be understood that, as used herein, the singular form “a” is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, “and / or” refers to any and all possible combinations of one or more of the associated listed items.

[0072] The embodiment numbers disclosed in the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0073] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0074] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples. Within the framework of the invention, technical features of the above embodiments or different embodiments can be combined, and many other variations of different aspects of the invention exist, which are not provided in the details for the sake of brevity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A method for mining high-value traffic flow segments, characterized in that, Includes the following steps: The acquired traffic data is analyzed to determine feature thresholds, basic sub-scenes are determined based on the feature thresholds, and the basic sub-scenes are combined to generate ordinary traffic flow segments. High-value traffic flow segments in the ordinary traffic flow segments are marked, and feature models are constructed based on the marked high-value traffic flow segments; Determine the multi-feature joint threshold for high-value traffic flow segments based on the aforementioned feature model; as well as High-value traffic flow segments are selected from the traffic flow segments to be mined based on the multi-feature joint threshold. The acquired traffic data will be analyzed to determine feature thresholds, including: Combine map data to create a basic feature wide table, and then draw a curve based on the basic feature wide table; The intervals where the feature values ​​in the curve change abruptly are used as feature thresholds.

2. The method for mining high-value traffic flow segments according to claim 1, characterized in that, Based on map data, create a basic feature wide table, including: By combining map data, a basic feature wide table is created that includes motion status information, basic attribute information, map information, and information on adjacent traffic participants.

3. The method for mining high-value traffic flow segments according to claim 1, characterized in that, The steps for constructing a feature model based on labeled high-value traffic flow segments include: Based on the marked high-value traffic flow segments, construct a multi-feature joint distribution scatter matrix and a Gaussian regression model respectively.

4. The method for mining high-value traffic flow segments according to claim 3, characterized in that, The steps of constructing a multi-feature joint distribution scatter matrix and a Gaussian regression model based on the marked high-value traffic flow segments include: Construct a multi-feature joint distribution scatter matrix based on the features of labeled high-value traffic flow segments and non-high-value traffic flow segments; and A multi-feature Gaussian regression model is constructed based on the relationships between features in labeled high-value traffic flow segments.

5. The method for mining high-value traffic flow segments according to claim 3, characterized in that, The step of determining the joint threshold of multiple features for high-value traffic flow segments based on the feature model includes: Construct a multi-feature covariance matrix, and construct a Gaussian probability density function based on the covariance matrix; and The threshold of each feature in the probability density function is calculated using a preset confidence level to form the multi-feature joint threshold.

6. The method for mining high-value traffic flow segments according to claim 1, characterized in that, The method further includes: The selected high-value traffic flow segments are reviewed, and misclassified traffic flow segments are marked; and The selected high-value traffic flow segments and the misjudged traffic flow segments are input into the regression model to update the multi-feature joint threshold.

7. A system for mining high-value traffic flow segments, characterized in that, include: The analysis module is configured to analyze the acquired traffic data to determine feature thresholds, determine basic sub-scenes based on the feature thresholds, and combine the basic sub-scenes to generate ordinary traffic flow segments. The module is configured to mark high-value traffic flow segments in the ordinary traffic flow segments and to build a feature model based on the marked high-value traffic flow segments. The acquired traffic data is analyzed to determine feature thresholds, including: combining map data to create a basic feature wide table, and plotting a curve based on the basic feature wide table; the intervals where feature values ​​in the curve change abruptly are used as feature thresholds. A threshold module, configured to determine a multi-feature joint threshold for high-value traffic flow segments based on the feature model; and The execution module is configured to filter high-value traffic flow segments from the traffic flow segments to be mined based on the multi-feature joint threshold.

8. An electronic device, characterized in that, include: At least one processor; as well as A memory storing computer instructions executable on the processor, which, when executed by the processor, implement the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.

Images