A driving behavior recognition method, device and storage medium

Abstract

The present application discloses a driving behavior recognition method, device, and storage medium, which relate to the technical field of vehicle driving behavior recognition and are used to improve the accuracy of vehicle driving behavior recognition. The method includes: obtaining the historical acceleration of a target vehicle; determining a first moving average of the historical acceleration of the target vehicle; the first moving average is used to reflect the driving state of the target vehicle before the current moment; obtaining the current acceleration of the target vehicle; and determining that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first moving average is greater than a preset threshold.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle driving behavior recognition, and particularly relates to a driving behavior recognition method, device and storage medium. Background Art

[0002] As the use of automobiles becomes more and more common, the driving safety of automobiles has also received increasing attention. During the driving process of a vehicle, in case of an emergency, the driver often takes emergency braking, sharp turning and other risk avoidance actions, and actions such as emergency braking and sharp turning are extremely likely to cause traffic accidents. Therefore, devices such as driving recorders are expected by consumers to have the function of automatically identifying emergencies and automatically recording videos of emergencies.

[0003] In the prior art, multiple sensors (such as infrared sensors, direction sensors, etc.) are deployed in a vehicle to detect the driving data of the vehicle, and then the emergency braking and sharp turning behaviors are identified by determining whether the driving data is greater than a preset threshold. However, the method provided by the prior art is prone to false recognition (for example, false recognition caused by vehicle bumps due to factors such as speed bumps and potholed roads), and the accuracy of the recognition result is relatively low. Summary of the Invention

[0004] The present application provides a driving behavior recognition method, device and storage medium for improving the accuracy of vehicle driving behavior recognition.

[0005] In a first aspect, the present application provides a driving behavior recognition method, including: obtaining the historical acceleration of a target vehicle; determining a first moving average value of the historical acceleration of the target vehicle; the first moving average value is used to reflect the driving state of the target vehicle before the current moment; obtaining the current acceleration of the target vehicle; and determining that the target vehicle has a target driving behavior when the difference between the current acceleration and the first moving average value of the target vehicle is greater than a preset threshold.

[0006] It can be understood that by determining the first sliding average of the historical acceleration of the target vehicle and using the first sliding average to reflect the driving state of the target vehicle before the current moment; and then obtaining the current acceleration of the target vehicle, when the difference between the current acceleration of the target vehicle and the first sliding average is greater than a preset threshold, it is determined that the target vehicle has a target driving behavior. In this way, compared with the prior art method of directly identifying sudden braking and sharp turning behaviors based on whether the driving data is greater than a preset threshold, the embodiment of the present application considers the driving state of the target vehicle before the target driving behavior occurs, and introduces the first sliding average to characterize the driving state of the target vehicle (for example, the target vehicle is in a steady driving state, or the target vehicle is in a bumpy driving state, etc.). Then, based on the first sliding average and the current acceleration of the target vehicle, it is determined whether the target vehicle has a target driving behavior. In this way, it is possible to effectively filter out misidentifications caused by the normal bumps of the vehicle (for example, the vehicle bumps generated when the target vehicle is driving on a speed bump or a potholed road surface), and improve the accuracy of driving behavior recognition.

[0007] In a possible implementation manner, the historical acceleration of the target vehicle includes: the historical acceleration of the target vehicle within the first time period between the first time point before the current moment and the current moment; and the historical acceleration of the target vehicle within the second time period between the second time point before the current moment and the current moment; the first time point is earlier than the second time point; the above-mentioned determination of the first sliding average of the historical acceleration of the target vehicle includes: calculating the sliding average of the historical acceleration within the first time period and the sliding average of the historical acceleration within the second time period; when the difference between the sliding average of the historical acceleration within the first time period and the sliding average of the historical acceleration within the second time period is less than or equal to a preset change threshold, taking the sliding average of the historical acceleration within the first time period as the first sliding average; or, when the difference between the sliding average of the historical acceleration within the first time period and the sliding average of the historical acceleration within the second time period is greater than the preset change threshold, adjusting the historical acceleration within the first time period according to the sliding average of the historical acceleration within the second time period; taking the sliding average of the adjusted historical acceleration within the first time period as the first sliding average.

[0008] It can be understood that since the sliding average of the historical acceleration in the first time period has better anti-interference performance, and the sliding average of the historical acceleration in the second time period has better timeliness, therefore, in order to balance anti-interference performance and timeliness, the present application adopts the above technical solution to determine the first sliding average. When the difference between the sliding average of the historical acceleration in the first time period and the sliding average of the historical acceleration in the second time period is less than the preset change threshold, it indicates that the road condition of the target vehicle has not changed significantly. Therefore, the sliding average of the historical acceleration in the first time period is used as the first sliding average. When the difference between the sliding average of the historical acceleration in the first time period and the sliding average of the historical acceleration in the second time period is greater than or equal to the preset change threshold, it indicates that the road condition of the target vehicle may have changed significantly, and it is necessary to quickly adjust the first sliding average to the accurate range. Therefore, the historical acceleration in the first time period is adjusted according to the sliding average of the historical acceleration in the second time period, and the adjusted sliding average of the historical acceleration in the first time period is used as the first sliding average.

[0009] In another possible implementation manner, the above adjustment of the historical acceleration in the first time period according to the sliding average of the historical acceleration in the second time period includes: replacing the historical acceleration at each time point in the first time period with the sliding average of the historical acceleration in the second time period.

[0010] In another possible implementation manner, the acceleration includes: the acceleration in the front-rear direction of the target vehicle, and the front-rear direction of the target vehicle represents the direction of the body length of the target vehicle; the target driving behavior includes: an emergency braking behavior; the above determination that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first sliding average is greater than the preset threshold includes: determining that the target vehicle has an emergency braking behavior when the difference between the current acceleration in the front-rear direction of the target vehicle and the first sliding average of the historical acceleration in the front-rear direction of the target vehicle is greater than the first preset threshold.

[0011] It can be understood that when the target vehicle has an emergency braking behavior, the change characteristic of the acceleration in the front-rear direction of the target vehicle is the most significant. Therefore, the acceleration in the front-rear direction of the target vehicle is used as the basis for determining whether the target vehicle has an emergency braking behavior.

[0012] In another possible implementation, the acceleration includes the lateral acceleration of the target vehicle and the longitudinal acceleration of the target vehicle. The lateral direction of the target vehicle represents the direction of the vehicle body width of the target vehicle. The target driving behavior includes a sharp turning behavior. When the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, determining that the target vehicle has a target driving behavior includes: when the difference between the current lateral acceleration of the target vehicle and the first sliding average value of the historical lateral acceleration of the target vehicle is greater than a second preset threshold; and / or when the difference between the current longitudinal acceleration of the target vehicle and the first sliding average value of the historical longitudinal acceleration of the target vehicle is greater than a third preset threshold, determining that the target vehicle has a sharp turning behavior.

[0013] It can be understood that when the target vehicle has a sharp turning behavior, the change characteristics of the lateral acceleration of the target vehicle are the most significant, and at the same time, the longitudinal acceleration of the target vehicle also shows synchronous change characteristics. Therefore, the lateral acceleration of the target vehicle and the longitudinal acceleration of the target vehicle are used as the basis for judging whether the target vehicle has a sharp turning behavior.

[0014] In another possible implementation, when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, determining that the target vehicle has a target driving behavior includes: when the differences between a continuous plurality of current accelerations of the target vehicle and the first sliding average value are all greater than the preset threshold, determining that the target vehicle has a target driving behavior; the continuous plurality of current accelerations means the accelerations at a continuous plurality of moments starting from the current moment.

[0015] It can be understood that the data output by the acceleration sensor does not all accurately reflect the current acceleration of the target vehicle, and there may be some noise point data. If the target vehicle's sharp braking behavior is judged only based on one current longitudinal acceleration of the target vehicle, it may lead to misidentification. Therefore, in this application, the target vehicle's sharp braking behavior is judged based on a continuous plurality of current longitudinal accelerations of the target vehicle, which can improve the recognition accuracy.

[0016] In another possible implementation, the target vehicle includes a triaxial acceleration sensor; the triaxial acceleration sensor is used to detect the accelerations in the triaxial directions of the triaxial acceleration sensor; the triaxial directions of the triaxial acceleration sensor are the directions indicated by the three coordinate axes of the triaxial acceleration sensor; before determining that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, the method further includes: when the triaxial directions of the triaxial acceleration sensor are not the same as the triaxial directions corresponding to the target vehicle, respectively perform transformation processing on the current accelerations in the triaxial directions of the triaxial acceleration sensor and the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor; the second sliding average value is the sliding average value of the historical accelerations in the triaxial directions detected by the triaxial acceleration sensor before the current moment; determine the difference between the current acceleration of the target vehicle and the first sliding average value according to the current accelerations in the triaxial directions of the triaxial acceleration sensor after the transformation processing and the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor after the transformation processing.

[0017] It can be understood that the shapes of different acceleration sensors are various, and the installation positions of the acceleration sensors on different vehicles are also various. For example, when the acceleration sensor is installed on the windshield of the vehicle terminal, a large pitch angle often occurs in the front-rear direction. At this time, the acceleration in the front-rear direction read by the acceleration sensor cannot accurately reflect the acceleration in the front-rear direction of the vehicle terminal. To solve the above problems, in this application, through transformation processing, the accelerations detected in the triaxial directions of the acceleration sensor are converted to the triaxial directions of the vehicle terminal to unify the measurement standard and achieve the purpose of adapting to the installation attitude.

[0018] In another possible implementation, the above-mentioned step of respectively performing transformation processing on the current accelerations in the triaxial directions of the triaxial acceleration sensor and the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor includes: determining the included angle between the triaxial directions of the triaxial acceleration sensor and the triaxial directions corresponding to the target vehicle according to the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor; obtaining the current accelerations in the triaxial directions of the triaxial acceleration sensor after the transformation processing according to the current accelerations in the triaxial directions and the included angle; obtaining the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor after the transformation processing according to the second sliding average value of the historical accelerations in the triaxial directions of the triaxial acceleration sensor and the included angle.

[0019] It can be understood that, due to the included angle between the three axes of the triaxial acceleration sensor and the three axes corresponding to the target vehicle, which can satisfy the cosine theorem in trigonometric functions, therefore, according to this included angle, through transformation processing (such as trigonometric function transformation processing), the acceleration data in the three axes output by the triaxial acceleration sensor can be transformed to the three axes of the target vehicle to obtain the acceleration data in the three axes of the target vehicle.

[0020] In another possible implementation, the three axes of the above triaxial acceleration sensor are not the same as the three axes corresponding to the target vehicle, including: the front-back direction of the triaxial acceleration sensor is not the same as the front-back direction of the target vehicle; determining the difference between the current acceleration of the target vehicle and the first sliding average value according to the current acceleration in the three axes of the triaxial acceleration sensor after transformation processing and the second sliding average value of the historical acceleration in the three axes of the triaxial acceleration sensor after transformation processing includes: determining the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction according to the current acceleration in the front-back direction of the triaxial acceleration sensor after transformation processing and the second sliding average value of the historical acceleration in the front-back direction of the triaxial acceleration sensor after transformation processing;

[0021] Among them, the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction satisfies the following relationship:

[0022]

[0023] Among them, Δy represents the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction; y represents the current acceleration in the front-back direction of the triaxial acceleration sensor after transformation processing, and y satisfies: P Y represents the second sliding average value of the historical acceleration in the front-back direction of the triaxial acceleration sensor after transformation processing, and P Y satisfies: y0 represents the current acceleration in the front-back direction of the triaxial acceleration sensor, and P y represents the second sliding average value of the historical acceleration in the front-back direction of the triaxial acceleration sensor; θ represents the included angle between the front-back direction of the triaxial acceleration sensor and the corresponding front-back direction of the target vehicle, and θ satisfies: P x represents the second sliding average value of the triaxial acceleration sensor in the left-right direction; P y represents the second sliding average value of the triaxial acceleration sensor in the front-back direction; P z represents the second sliding average value of the triaxial acceleration sensor in the vertical direction.

[0024] In another possible implementation, the above method is applied to an identification device; the identification device is connected to the driving recorder of the target vehicle; the above method further includes: when the target vehicle exhibits a target driving behavior, triggering the automatic emergency video recording function of the driving recorder so that the driving recorder automatically records the video of the target driving behavior; the target driving behavior includes at least one of the following: hard braking behavior, sharp turning behavior.

[0025] It can be understood that hard braking behavior is extremely likely to cause rear-end collisions, and sharp turning behavior is extremely likely to cause collision accidents. Therefore, in this application, the target driving behavior is set as an emergency situation of the target vehicle, so that when the target vehicle exhibits a target driving behavior, the driving recorder of the target vehicle can be triggered to automatically record the video of the target driving behavior. In this way, the video of the emergency situation recorded by the driving recorder provides evidence for the accident caused by the target driving behavior and can be used as a basis for liability analysis and dispute resolution.

[0026] In a second aspect, this application provides an identification device, including: an acquisition module for acquiring the historical acceleration of the target vehicle; a determination module for determining the first moving average of the historical acceleration of the target vehicle; the first moving average is used to reflect the driving state of the target vehicle before the current moment; the above acquisition module is further used to acquire the current acceleration of the target vehicle; the above determination module is further used to determine that the target vehicle exhibits a target driving behavior when the difference between the current acceleration of the target vehicle and the first moving average is greater than a preset threshold.

[0027] In a possible implementation, the historical acceleration of the target vehicle includes: the historical acceleration of the target vehicle within the first time period between the first time point before the current moment and the current moment; and the historical acceleration of the target vehicle within the second time period between the second time point before the current moment and the current moment; the first time point is earlier than the second time point; the determination module is specifically configured to calculate the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period; when the difference between the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period is less than or equal to a preset change threshold, use the moving average of the historical acceleration within the first time period as the first moving average; or, when the difference between the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period is greater than the preset change threshold, adjust the historical acceleration within the first time period according to the moving average of the historical acceleration within the second time period; use the adjusted moving average of the historical acceleration within the first time period as the first moving average.

[0028] In another possible implementation manner, the above-mentioned determining module is specifically configured to replace the historical accelerations at each time point within the first time period with the moving average of the historical accelerations within the second time period.

[0029] In another possible implementation manner, the acceleration includes the acceleration in the front-rear direction of the target vehicle, and the front-rear direction of the target vehicle represents the direction of the body length of the target vehicle; the target driving behavior includes a hard braking behavior; the determining module is specifically configured to determine that the target vehicle has a hard braking behavior when the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average of the historical accelerations in the front-rear direction of the target vehicle is greater than a first preset threshold.

[0030] In another possible implementation manner, the acceleration includes the acceleration in the left-right direction of the target vehicle and the acceleration in the front-rear direction of the target vehicle, and the left-right direction of the target vehicle represents the direction of the body width of the target vehicle; the target driving behavior includes a sharp turning behavior; the determining module is specifically configured to determine that the target vehicle has a sharp turning behavior when the difference between the current acceleration in the left-right direction of the target vehicle and the first moving average of the historical accelerations in the left-right direction of the target vehicle is greater than a second preset threshold; and / or when the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average of the historical accelerations in the front-rear direction of the target vehicle is greater than a third preset threshold.

[0031] In another possible implementation manner, the above-mentioned determining module is specifically configured to determine that the target vehicle has a target driving behavior when the differences between the consecutive multiple current accelerations of the target vehicle and the first moving average are all greater than a preset threshold; the consecutive multiple current accelerations represent the accelerations at consecutive multiple time points starting from the current moment.

[0032] In another possible implementation manner, the target vehicle includes a triaxial acceleration sensor; the triaxial acceleration sensor is used to detect the accelerations in the three-axis directions of the triaxial acceleration sensor; the three-axis directions of the triaxial acceleration sensor are the directions indicated by the three coordinate axes of the triaxial acceleration sensor; the recognition device further includes a transformation module, which is configured to perform transformation processing on the current accelerations in the three-axis directions of the triaxial acceleration sensor and the second moving average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor respectively when the three-axis directions of the triaxial acceleration sensor are not the same as the corresponding three-axis directions of the target vehicle; the second moving average is the moving average of the historical accelerations in the three-axis directions detected by the triaxial acceleration sensor before the current moment; the determining module is further configured to determine the difference between the current acceleration of the target vehicle and the first moving average according to the current acceleration in the three-axis directions of the triaxial acceleration sensor after the transformation processing and the second moving average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor after the transformation processing.

[0033] In another possible implementation, the transformation module is specifically configured to determine the angle between the three-axis directions of the three-axis acceleration sensor and the corresponding three-axis directions of the target vehicle according to the second sliding average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor; obtain the current acceleration in the three-axis directions of the three-axis acceleration sensor after transformation processing according to the current acceleration in the three-axis directions and the angle; and obtain the second sliding average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after transformation processing according to the second sliding average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor and the angle.

[0034] In another possible implementation, the three-axis directions of the three-axis acceleration sensor are not the same as the corresponding three-axis directions of the target vehicle, including: the front-back direction of the three-axis acceleration sensor is not the same as the front-back direction of the target vehicle; the determination module is specifically configured to determine the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average in the front-back direction according to the current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing and the second sliding average of the historical accelerations in the front-back direction of the three-axis acceleration sensor after transformation processing.

[0035] Wherein, the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average in the front-back direction satisfies the following relationship:

[0036]

[0037] Wherein, Δy represents the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average in the front-back direction; y represents the current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing, and y satisfies: P Y represents the second sliding average of the historical accelerations in the front-back direction of the three-axis acceleration sensor after transformation processing, P Y satisfies: y0 represents the current acceleration in the front-back direction of the three-axis acceleration sensor, P y represents the second sliding average of the historical accelerations in the front-back direction of the three-axis acceleration sensor; θ represents the angle between the front-back direction of the three-axis acceleration sensor and the corresponding front-back direction of the target vehicle, and θ satisfies: P x represents the second sliding average of the three-axis acceleration sensor in the left-right direction; P y represents the second sliding average of the three-axis acceleration sensor in the front-back direction; P z represents the second sliding average of the three-axis acceleration sensor in the vertical direction.

[0038] In another possible implementation, the recognition device is connected to the driving recorder of the target vehicle; the recognition device further includes: a triggering module, configured to trigger the automatic emergency video recording function of the driving recorder when a target driving behavior occurs in the target vehicle, so that the driving recorder automatically records a video of the target driving behavior; the target driving behavior includes at least one of the following: a hard braking behavior, a sharp turning behavior.

[0039] In a third aspect, the present application provides a recognition device, including: one or more processors; one or more memories; wherein, the one or more memories are used to store computer program code, and the computer program code includes computer instructions. When the one or more processors execute the computer instructions, the recognition device executes any one of the driving behavior recognition methods provided in the first aspect above.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium, which stores computer execution instructions. When the computer execution instructions run on a computer, the computer executes any one of the driving behavior recognition methods provided in the first aspect above.

[0041] For the descriptions of the second aspect to the fourth aspect in the present application, reference may be made to the detailed description of the first aspect; and, for the beneficial effects of the descriptions of the second aspect to the fourth aspect, reference may be made to the beneficial effect analysis of the first aspect, which will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of the implementation environment involved in a driving behavior recognition method provided by an embodiment of the present application Figure 1 ;

[0043] Figure 2 Schematic diagram of the implementation environment involved in a driving behavior recognition method provided by an embodiment of the present application Figure 2 ;

[0044] Figure 3 Flow chart of a driving behavior recognition method provided by an embodiment of the present application Figure 1 ;

[0045] Figure 4 Schematic diagram of a time period provided by an embodiment of the present application Figure 1 ;

[0046] Figure 5 Flow chart of a driving behavior recognition method provided by an embodiment of the present application Figure 2 ;

[0047] Figure 6 Schematic diagram of a time period provided by an embodiment of the present application Figure 2 ;

[0048] Figure 7 A schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 1 ;

[0049] Figure 8 A schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 2 ;

[0050] Figure 9 A schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 3 ;

[0051] Figure 10 A schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 4 ;

[0052] Figure 11 A schematic diagram of three-axis directions provided by an embodiment of the present application;

[0053] Figure 12 A schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 5 ;

[0054] Figure 13(a) is a schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 6 ;

[0055] Figure 13(b) is a schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 7 ;

[0056] Figure 13(c) is a schematic diagram of an acceleration curve provided by an embodiment of the present application Figure 7 ;

[0057] Figure 14 A schematic diagram of the structure of an identification device provided by an embodiment of the present application Figure 1 ;

[0058] Figure 15 A schematic diagram of the structure of an identification device provided by an embodiment of the present application Figure 2 . Detailed implementation manners

[0059] The term "and / or" in this article is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. These three situations

[0060] The terms "first" and "second" in the specification and drawings of the present application are used to distinguish different objects or different processes for the same object, rather than to describe the specific order of the objects

[0061] In addition, the terms "comprising" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include other unlisted steps or units, or may optionally further include other steps or units inherent to these processes, methods, products, or devices.

[0062] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0063] In the description of the present application, unless otherwise specified, the meaning of "a plurality of" refers to two or more.

[0064] As described in the background art, the driving behavior recognition method provided by the prior art detects the driving data of a vehicle by deploying a plurality of sensors (such as infrared sensors, direction sensors, etc.) in the vehicle, and then identifies sudden braking and sharp turning behaviors based on the driving data. The problems existing in the prior art are: on the one hand, there are many dependencies (such as the need for a plurality of sensors); on the other hand, the initial state of the vehicle driving is not considered, resulting in misidentifications (for example, misidentifications caused by factors such as speed bumps and potholed roads that cause the vehicle to jolt), and the recognition results are inaccurate.

[0065] Based on the above technical problems, the embodiments of the present application provide a driving behavior recognition method, the idea of which is: by determining the first sliding average value of the historical acceleration of the target vehicle, using the first sliding average value to reflect the driving state of the target vehicle before the current moment; and then obtaining the current acceleration of the target vehicle, and when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, it is determined that the target vehicle has a target driving behavior. In this way, compared with the method of directly identifying sudden braking and sharp turning behaviors based on whether the driving data is greater than a preset threshold in the prior art, the embodiments of the present application consider the driving state of the target vehicle before the target driving behavior occurs, and introduce the first sliding average value to characterize the driving state of the target vehicle (for example, the target vehicle is in a steady driving state, or the target vehicle is in a bumpy driving state, etc.), and then based on the first sliding average value and the current acceleration of the target vehicle, it is determined whether the target vehicle has a target driving behavior. In this way, misidentifications caused by normal jolts of the vehicle (for example, vehicle jolts generated when the target vehicle is driving on a speed bump or a potholed road) can be effectively filtered, and the accuracy of driving behavior recognition is improved.

[0066] The implementation manners of the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

[0067] Please refer to Figure 1 , which shows a schematic diagram of an implementation environment involved in a driving behavior recognition method provided by an embodiment of the present application. As Figure 1 shown, the implementation environment may include: an acceleration sensor 100 and an identification device 200.

[0068] In some embodiments, the acceleration sensor 100 and the identification device 200 may communicate with each other. Optionally, the acceleration sensor 100 and the identification device 200 may be integrated together; or, the acceleration sensor 100 and the identification device 200 may be two independent devices.

[0069] In some embodiments, the acceleration sensor 100 and the identification device 200 may be disposed in a vehicle terminal.

[0070] The acceleration sensor 100 is configured to detect the acceleration of the vehicle terminal.

[0071] In some embodiments, the acceleration sensor 100 may be a G-sensor (Gravity sensor). The G-sensor is a three-axis acceleration sensor, including three different coordinate axes (where any two coordinate axes are perpendicular to each other), and can obtain the accelerations in the directions indicated by the three different coordinate axes (three-axis directions). It can be understood that since acceleration is a spatial vector, by measuring the accelerations in the three-axis directions of an object, the motion state of the object can be accurately understood.

[0072] In some embodiments, the three-axis directions of the acceleration sensor 100 need to be determined according to the object detected by the acceleration sensor 100 and the position where the acceleration sensor is installed. Exemplarily, if the three different coordinate axes of the acceleration sensor 100 are respectively represented by the X-axis, the Y-axis, and the Z-axis, and if the acceleration sensor 100 is installed on the vehicle terminal, where the included angle between the X-axis of the acceleration sensor 100 and the left-right direction of the vehicle terminal is less than a preset threshold (close or coincident), the included angle between the Y-axis and the front-back direction of the vehicle terminal is less than a preset threshold (close or coincident), and the included angle between the Z-axis and the vertical direction of the vehicle terminal is less than a preset threshold (close or coincident), then it indicates that the direction of the X-axis of the acceleration sensor 100 is the left-right direction, the Y-axis direction is the front-back direction, and the Z-axis direction is the vertical direction. That is, the acceleration sensor 100 can detect the acceleration in the left-right direction of the vehicle terminal, the acceleration in the front-back direction of the vehicle terminal, and the acceleration in the vertical direction of the vehicle terminal.

[0073] Among them, the front-rear direction of the vehicle terminal represents the direction of the vehicle terminal's body length; the left-right direction of the vehicle terminal represents the direction of the vehicle terminal's body width; the vertical direction of the vehicle terminal represents the direction of the vehicle terminal's body height.

[0074] It can be understood that the directions indicated by the three different coordinate axes of the acceleration sensor 100 change with the change of the position of the acceleration sensor 100, and there is no clear indication direction. Therefore, when the object detected by the acceleration sensor 100 is determined and the installation position of the acceleration sensor 100 on the object is determined, the three different coordinate axes of the acceleration sensor 100 can determine the specific indication directions, and at this time, the acceleration sensor 100 has a three-axis direction.

[0075] Exemplarily, if the acceleration sensor 100 is a G-sensor sensor, when the acceleration sensor 100 detects the acceleration of the vehicle terminal, if the three-axis direction of the acceleration sensor 100 is the same as the corresponding three-axis direction of the vehicle terminal, then the acceleration in the three-axis direction output by the acceleration sensor 100 is the acceleration in the three-axis direction of the vehicle terminal.

[0076] Another exemplarily, if the three-axis direction of the acceleration sensor 100 is not the same as the corresponding three-axis direction of the vehicle terminal, then it is necessary to perform a transformation process on the acceleration in the three-axis direction output by the acceleration sensor 100, and use the acceleration in the three-axis direction after the transformation process as the acceleration in the three-axis direction of the vehicle terminal.

[0077] It can be understood that when the three-axis direction of the triaxial acceleration sensor is not the same as the corresponding three-axis direction of the vehicle terminal, it means that there is an angle between the three-axis direction (or the direction of any one axis in the three-axis direction) of the triaxial acceleration sensor and the corresponding three-axis direction (or the direction of any one axis in the three-axis direction) of the vehicle terminal. Then, the above transformation process can be to perform a trigonometric function transformation on the acceleration in the three-axis direction of the triaxial acceleration sensor to obtain the acceleration in the three-axis direction of the vehicle terminal.

[0078] The recognition device 200 is used to obtain the acceleration of the vehicle terminal detected by the acceleration sensor 100, and determine the driving state of the vehicle terminal and whether the vehicle terminal has a target driving behavior according to the acceleration of the vehicle terminal.

[0079] Among them, the target driving behavior includes at least one of the following: sudden braking behavior, sharp turning behavior.

[0080] Exemplarily, the recognition device 200 may be a server. For example, it may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks, and big data servers.

[0081] In another example, the recognition device 200 may be an electronic device. For example, it may be a mobile phone, a tablet computer, a desktop computer, a laptop computer, a handheld computer, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a cellular phone, a personal digital assistant (PDA), an augmented reality (AR) / virtual reality (VR) device, etc. The present disclosure does not impose any special restrictions on the specific form of the recognition device 200.

[0082] Specifically, as Figure 2 shown, the acceleration sensor 100 detects the acceleration of the vehicle terminal and sends the acceleration of the vehicle terminal to the recognition device 200. After receiving the acceleration of the vehicle terminal (including historical acceleration and current acceleration), the recognition device 200 first determines the driving state of the vehicle terminal based on the historical acceleration of the vehicle terminal, and then determines whether the vehicle terminal exhibits a target driving behavior based on the current acceleration of the vehicle terminal and the driving state of the vehicle terminal.

[0083] The following specifically introduces the driving behavior recognition method provided by the present application.

[0084] An embodiment of the present application provides a driving behavior recognition method, which is applied to the recognition device as Figure 1 shown. As Figure 3 shown, the method includes the following steps:

[0085] S101. Obtain the historical acceleration of the target vehicle.

[0086] In some embodiments, the target vehicle includes an acceleration sensor, and the acceleration sensor is used to detect the acceleration of the target vehicle. Therefore, the recognition device obtains the historical acceleration of the target vehicle through the acceleration sensor.

[0087] Optionally, the acceleration sensor may be a three-axis acceleration sensor (such as a G-sensor). The three-axis acceleration sensor can detect the acceleration of the target vehicle in three axial directions. Among them, the three axial directions of the target vehicle include: the left-right direction of the target vehicle, the front-back direction of the target vehicle, and the vertical direction of the target vehicle.

[0088] It can be understood that the three axial directions of the target vehicle are the three axial directions based on the vehicle body. Among them, the left-right direction of the target vehicle represents the direction of the vehicle body width of the target vehicle; the front-back direction of the target vehicle represents the direction of the vehicle body length of the target vehicle; the vertical direction of the target vehicle represents the direction of the vehicle body height of the target vehicle.

[0089] Thus, the historical acceleration of the above target vehicle can be the historical acceleration in the three axial directions of the target vehicle, for example: the historical acceleration in the left-right direction of the target vehicle, the historical acceleration in the front-back direction of the target vehicle, and the historical acceleration in the vertical direction of the target vehicle.

[0090] In some embodiments, when the three axial directions of the three-axis acceleration sensor are the same as the three axial directions corresponding to the target vehicle, the above step S101 can be implemented as: using the historical acceleration in the three axial directions output by the three-axis acceleration sensor as the historical acceleration in the three axial directions of the target vehicle.

[0091] Among them, the three axial directions of the three-axis acceleration sensor are the directions indicated by the three coordinate axes of the three-axis acceleration sensor when the three-axis acceleration sensor is installed on the target vehicle.

[0092] In some embodiments, the three axial directions of the above three-axis acceleration sensor and the three axial directions corresponding to the target vehicle refer to the correspondence between the directions indicated by the three coordinate axes of the three-axis acceleration sensor and the three axial directions of the target vehicle based on the vehicle body when the three-axis acceleration sensor is installed on the target vehicle.

[0093] It can be understood that affected by the installation position, the three coordinate axes of the three-axis acceleration sensor may or may not coincide with the three axial directions of the target vehicle. If the three coordinate axes of the three-axis acceleration sensor coincide with the three axial directions of the target vehicle, then the three axial directions of the three-axis acceleration sensor are the same as the three axial directions corresponding to the target vehicle, and the acceleration in the three axial directions output by the three-axis acceleration sensor can represent the acceleration in the three axial directions of the target vehicle; if the three coordinate axes of the three-axis acceleration sensor do not coincide with the three axial directions of the target vehicle (i.e., there is an angle), then the three axial directions of the three-axis acceleration sensor are not the same as the three axial directions corresponding to the target vehicle, and the acceleration in the three axial directions of the three-axis acceleration sensor cannot represent the acceleration in the three axial directions of the target vehicle.

[0094] S102. Determine the first moving average of the historical acceleration of the target vehicle.

[0095] Among them, the first moving average is used to reflect the driving state of the target vehicle before the current moment (that is, when the target vehicle does not exhibit the target driving behavior). For example, the target vehicle is in a steady driving state, or the target vehicle is in a bumpy driving state, etc. Among them, the driving state of the target vehicle is reflected according to the magnitude of the first moving average. For example, when the first moving average is within the first preset range, it is determined that the target vehicle is in a steady driving state; when the first moving average is within the second preset range, it is determined that the target vehicle is in a bumpy driving state; among them, the upper limit value of the first preset range is less than the lower limit value of the second preset range.

[0096] In some embodiments, the first moving average can also be used to reflect the initial output state of the acceleration sensor before the current moment (that is, when the target vehicle does not exhibit the target driving behavior).

[0097] It can be understood that the moving average can reflect the overall change trend of data within a period of time. Therefore, the first moving average can reflect the driving state of the target vehicle before the current moment (a certain historical period).

[0098] Furthermore, the driving state of the target vehicle before the current moment can be used as a reference, and according to the current acceleration of the target vehicle, it can be determined whether the target vehicle exhibits the target driving behavior. Specifically, refer to the following steps S103 - S104.

[0099] In some embodiments, the historical acceleration of the target vehicle includes: the historical acceleration within the first time period between the first time point before the current moment and the current moment; and the historical acceleration within the second time period between the second time point before the current moment and the current moment.

[0100] Among them, the first time point is earlier than the second time point, and the first time period is greater than the second time period.

[0101] Exemplarily, as Figure 4 shown, assuming the current moment is 10:00 and the first time point is 9:40, then the first time period is the time period between 9:40 and 10:00; assuming the second time point is 9:50, then the second time period is the time period between 9:50 and 10:00.

[0102] In some embodiments, as Figure 5 shown, the above step S102 can be implemented as:

[0103] S1021. Calculate the moving average of the historical acceleration in the first time period and the moving average of the historical acceleration in the second time period.

[0104] In some embodiments, when the historical acceleration is the historical acceleration in three axial directions, the moving average is the moving average in three axial directions.

[0105] In some embodiments, the moving average method is adopted to calculate the first moving average of the historical acceleration in the first time period (or the first time window) and the second moving average of the historical acceleration in the second time period (or the second time window).

[0106] Among them, the moving average method, also known as the moving average method, the average method, the moving average filtering method, etc., is a signal smoothing method in the time domain. The idea of the moving average method is: taking the arithmetic average of the sampling points near a certain point as the smoothed value of this point. Usually, the number of seconds of the time period selected for a moving average is represented by a time window.

[0107] It can be understood that the present application uses the moving average method to determine the moving average. One moving average is obtained for the data within one time window (one time period). By sliding the time window, multiple moving averages can be obtained. Therefore, the moving average corresponding to each current moment is different. Exemplarily, assuming that the first time period is 30s, the moving average at 10:01 is the moving average of the historical acceleration within 30s before 10:01; the moving average at 10:02 is the moving average of the historical acceleration within 30s before 10:02.

[0108] Exemplarily, as Figure 6 shown, assuming that the output frequency of the acceleration sensor is n (that is, n data are output per second), if the first time period is T0 seconds and the second time period is T1 seconds, where T0 > T1, then the number of data for calculating the moving average in the first time period is T0 * n; the number of data for calculating the moving average in the second time period is T1 * n. Assuming that the sum of the historical accelerations of the target vehicle in the X-axis in the first time period is M x , the sum of the historical accelerations of the target vehicle in the Y-axis is M y , and the sum of the historical accelerations of the target vehicle in the Z-axis is M z , then the moving average of the historical acceleration of the target vehicle in the X-axis in the first time period is P x1 = M x / (T0 * n), the moving average of the historical acceleration of the target vehicle in the Y-axis is P y1 = M y / (T0 * n), and the moving average of the historical acceleration of the target vehicle in the Z-axis is P z1 = Mz / (T0*n). Assume that the sum of the historical accelerations of the target vehicle along the X-axis during the second time period is N x , and the sum of the historical accelerations along the Y-axis is N y , and the sum of the historical accelerations along the Z-axis is N z , then the moving average of the historical accelerations of the target vehicle along the X-axis during the second time period is P x2 = N x / (T1*n), and the moving average of the historical accelerations along the Y-axis is P y2 = N y / (T1*n), and the moving average of the historical accelerations along the Z-axis is P z2 = N z / (T1*n).

[0109] Exemplarily, as Figure 7 shown, the moving average of the historical accelerations during the first time period before the current moment is different from that during the second time period before the current moment, and compared with the moving average of the historical accelerations during the second time period, the fluctuation degree of the moving average of the historical accelerations during the first time period is smaller and more stable. Therefore, the calculated moving average values corresponding to different time periods are different. The longer the selected time period, the more stable the calculated moving average, and the stronger the anti-interference ability, but the timeliness is poor; the shorter the selected time period, the more timely the calculated moving average, but the data fluctuation is more obvious and the anti-interference ability is weak.

[0110] It can be understood that, since the moving average of the historical acceleration in the first time period has better anti-interference performance, and the moving average of the historical acceleration in the second time period has better timeliness, generally, the moving average of the historical acceleration in the first time period is selected as the first moving average. However, since the timeliness of the moving average of the historical acceleration in the first time period is poor, when the road condition of the target vehicle changes significantly (for example, when driving from a smooth road surface to a potholed road surface); or when the installation attitude of the acceleration sensor is adjusted significantly, the moving average of the historical acceleration in the first time period cannot be quickly updated to the accurate range. In this case, the moving average of the historical acceleration in the second time period can be selected to adjust the historical acceleration in the first time period, and the adjusted moving average of the historical acceleration in the first time period is used as the first moving average, so that the first moving average can reflect the driving state of the target vehicle in real time. Regarding the implementation of the adjustment of the historical acceleration in the first time period, for example, according to the moving average of the historical acceleration in the second time period, a preset adjustment parameter (the value can be adaptively set) can be added to or subtracted from the historical acceleration in the first time period, so that the difference between the adjusted moving average of the historical acceleration in the first time period and the moving average of the historical acceleration in the second time period is reduced. Or, for example, all or part of the historical acceleration in the first time period is replaced with all or part of the historical acceleration in the second time period according to the time point correspondence relationship. It should be understood that other adjustment methods that can achieve the foregoing adjustment purpose can be flexibly adopted, and the embodiments of the present application do not make specific limitations. In one implementation manner, please refer to the following steps S1022-S1023.

[0111] S1022. When the difference between the moving average of the historical acceleration in the first time period and the moving average of the historical acceleration in the second time period is less than or equal to a preset change threshold, the moving average of the historical acceleration in the first time period is used as the first moving average.

[0112] Among them, the above preset change threshold can be determined by means of experimental tests, simulation, expert experience, etc. Exemplarily, the preset change threshold can be 0.25g.

[0113] Exemplarily, if the moving average is the moving average of the historical acceleration in three-axis directions, the moving average of the historical acceleration in the first time period includes: P x1 、P y1 、P z1 ; the moving average of the historical acceleration in the second time period includes: P x2 、P y2 、P z2 ; if the preset change threshold is ΔV; then when |P x1-P x2 |Either <ΔV and / or|P y1 -P y2 |Either <ΔV and / or|P z1 -P z2 |When <ΔV, use the moving average of the historical accelerations in the first time period as the first moving average.

[0114] It can be understood that when the difference between the moving average of the historical accelerations in the first time period and the moving average of the historical accelerations in the second time period is less than the preset change threshold, it indicates that the road conditions of the target vehicle have not changed significantly. Therefore, use the moving average of the historical accelerations in the first time period as the first moving average.

[0115] S1023. When the difference between the moving average of the historical accelerations in the first time period and the moving average of the historical accelerations in the second time period is greater than the preset change threshold, adjust the historical accelerations in the first time period according to the moving average of the historical accelerations in the second time period; use the adjusted moving average of the historical accelerations in the first time period as the first moving average.

[0116] In some embodiments, adjusting the historical accelerations in the first time period according to the moving average of the historical accelerations in the second time period includes: replacing the historical accelerations at each time point in the first time period with the moving average of the historical accelerations in the second time period.

[0117] Exemplarily, assume that the number of historical accelerations in the first time period is n, and the number of historical accelerations in the second time period is m, where n > m; if the moving average is the moving average of the historical accelerations in the three-axis directions, then the first moving average includes: P x1 、P y1 、P z1 ; The second moving average includes: P x2 、P y2 、P z2 ; If the preset change threshold is ΔV; then when |P x1 -P x2 |≥ΔV, and / or |P y1 -P y2 |≥ΔV, and / or |P z1 -P z2 |≥ΔV, replace the n historical accelerations in the left-right direction output by the acceleration sensor in the first time period with P x2 ; Replace the n historical accelerations in the front-back direction with P y2 ; Replace the n historical accelerations in the vertical direction with P z2; The sliding averages of the historical accelerations during the adjusted first time period are respectively: P x2 、P y2 、P z2 , that is, the first sliding average is P x2 、P y2 、P z2 .

[0118] It can be understood that when the difference between the first sliding average and the second sliding average is greater than or equal to the preset change threshold, it indicates that the road condition of the target vehicle may have changed significantly, and it is necessary to quickly adjust the first sliding average to the accurate range. Therefore, the sliding average of the historical accelerations during the second time period can be selected to adjust the historical accelerations during the first time period, and the sliding average of the historical accelerations during the adjusted first time period is used as the first sliding average, so that the first sliding average can reflect the driving state of the target vehicle in real time.

[0119] Based on the above technical solution, determining the first sliding average according to the difference between the sliding average of the historical accelerations during the first time period and the sliding average of the historical accelerations during the second time period can quickly adjust the first sliding average to the accurate range when the road condition of the target vehicle changes significantly (or when the installation attitude of the acceleration sensor is adjusted significantly), and can balance anti-interference and timeliness.

[0120] S103. Obtain the current acceleration of the target vehicle.

[0121] In some embodiments, the recognition device obtains the current acceleration of the target vehicle through an acceleration sensor. Optionally, when the acceleration sensor is a three-axis acceleration sensor (such as a G-sensor), the current acceleration of the target vehicle can be the current accelerations in the three-axis directions of the target vehicle, for example: the current acceleration in the left-right direction of the target vehicle, the current acceleration in the front-rear direction of the target vehicle, and the current acceleration in the vertical direction of the target vehicle.

[0122] In some embodiments, when the three-axis directions of the three-axis acceleration sensor are the same as the three-axis directions corresponding to the target vehicle, the above step S103 can be implemented as: using the current accelerations in the three-axis directions output by the three-axis acceleration sensor as the current accelerations in the three-axis directions of the target vehicle.

[0123] In some other embodiments, when the three-axis directions of the three-axis acceleration sensor are not the same as the three-axis directions corresponding to the target vehicle, step S103 above may be implemented as: performing a transformation process on the current acceleration in the three-axis directions output by the three-axis acceleration sensor, and using the current acceleration in the three-axis directions output by the three-axis acceleration sensor after the transformation process as the current acceleration in the three-axis directions of the target vehicle. Exemplarily, the above transformation process may be a trigonometric function transformation process.

[0124] It can be understood that if the three-axis directions of the three-axis acceleration sensor are not the same as the three-axis directions corresponding to the target vehicle, it means that there is an angle between the three-axis directions of the three-axis acceleration sensor (or the direction of any one axis in the three-axis directions) and the three-axis directions corresponding to the target vehicle (or the direction of any one axis in the three-axis directions), and it satisfies the cosine theorem in trigonometric functions. Therefore, through a trigonometric function transformation process (cosine transformation), the current acceleration in the three-axis directions output by the three-axis acceleration sensor can be transformed to the three-axis directions of the target vehicle to obtain the current acceleration in the three-axis directions of the target vehicle.

[0125] In some embodiments, the acceleration sensor may output multiple data within one second, so the current acceleration of the target vehicle obtained may be multiple. Exemplarily, if the acceleration sensor can output 10 acceleration data within one second, then 10 current accelerations of the target vehicle are obtained. It can be understood that the higher the output frequency of the acceleration sensor, the more data the acceleration sensor outputs per second. Then, when identifying the driving behavior of the target vehicle, the more data is available for reference, which helps to improve the accuracy of the identification result.

[0126] S104. When the difference between the current acceleration of the target vehicle and the first moving average value is greater than a preset threshold, it is determined that the target vehicle has a target driving behavior.

[0127] Among them, the above preset threshold can be determined by means such as experimental tests, simulation, and expert experience; or the preset threshold can be adjusted according to actual usage requirements. Exemplarily, the value range of the preset threshold can be 0.5g - 1g.

[0128] In some embodiments, the target driving behavior includes at least one of the following: an emergency braking behavior, a sharp turning behavior.

[0129] It can be understood that when the target vehicle has an emergency braking behavior or a sharp turning behavior, the acceleration of the target vehicle will change greatly. Therefore, in the embodiments of the present application, whether the target vehicle has a target driving behavior is determined according to the change of the acceleration. Since judging the change of the data requires a reference value, obviously, for the present application, the reference value is the acceleration when the target vehicle is driving normally.

[0130] However, since the vehicle will also experience jolts and vibrations during normal driving (for example, passing through potholed roads or speed bumps, etc.), the data output by the acceleration sensor is always in a fluctuating state, and sometimes the fluctuation amplitude is relatively large. Then, at this time, the reference value cannot be accurately determined.

[0131] Exemplarily, if the target vehicle is driving on a smooth road including a speed bump, the accelerations in the three-axis directions output by the acceleration sensor can be as Figure 8 shown. Among them, when the target vehicle is driving on a smooth road, the fluctuation amplitude of the acceleration output by the acceleration sensor is relatively small, and the data is relatively stable as a whole. When the target vehicle is driving on the speed bump, the fluctuation amplitude of the acceleration output by the acceleration sensor is relatively large. Then, at this time, whether the acceleration with a relatively small fluctuation amplitude is selected as the reference value or the acceleration with a relatively large fluctuation amplitude is selected as the reference value, it is impossible to accurately determine whether the target vehicle has the target driving behavior.

[0132] Therefore, to solve the above problems, the embodiment of the present application uses the first sliding average value as the reference value. Since the sliding average value reflects the overall change trend of the acceleration of the target vehicle within a period of time, using the first sliding average value as the reference value can effectively exclude the influence of data fluctuations on the reference value and can relatively accurately reflect the driving state of the target vehicle before the current moment.

[0133] In some embodiments, in order to improve the recognition accuracy and reduce errors, the embodiment of the present application conducts on-vehicle tests on emergency braking behavior and sharp turning behavior respectively. The specific test process is as follows:

[0134] A1. Emergency braking behavior test

[0135] Let the target vehicle suddenly step on the brakes during smooth driving. After multiple experiments, the acceleration curve of emergency braking as shown in Figure 9 is obtained.

[0136] From Figure 9 it can be seen that when the target vehicle is driving smoothly, the fluctuation amplitude of the acceleration curves in the three-axis directions output by the acceleration sensor is relatively small; when the target vehicle has an emergency braking behavior, the fluctuation amplitude of the acceleration curve in the front-rear direction output by the acceleration sensor is relatively large, the fluctuation amplitude of the acceleration curve in the left-right direction is relatively small, and the fluctuation amplitude in the vertical direction is also relatively small.

[0137] A2. Sharp turning behavior test

[0138] Let the target vehicle make a sharp turn during smooth driving. After multiple experiments, the acceleration curve of sharp turning as shown in Figure 10 is obtained.

[0139] FromFigure 10 It can be seen that when the target vehicle is driving smoothly, the fluctuation amplitude of the acceleration curves in the three-axis directions of the target vehicle output by the acceleration sensor is small; when the target vehicle makes a sharp turn, the fluctuation amplitude of the acceleration curve in the left-right direction of the target vehicle output by the acceleration sensor is the largest. At the same time, the acceleration curves in the front-rear direction of the target vehicle also show synchronous change characteristics, but the fluctuation amplitude of the acceleration curve is smaller than that in the left-right direction of the target vehicle.

[0140] It can be seen that during the process of the target vehicle driving smoothly, during the process of sudden braking, and during the process of making a sharp turn, the acceleration in the vertical direction of the target vehicle output by the acceleration sensor does not show obvious rules. Therefore, it cannot be used as a basis for judging whether the target vehicle has a target driving behavior.

[0141] When the target vehicle makes a sudden braking, the change characteristics of the acceleration curve in the front-rear direction of the target vehicle output by the acceleration sensor are the most significant. Therefore, the acceleration in the front-rear direction of the target vehicle is used as the basis for judging whether the target vehicle makes a sudden braking.

[0142] When the target vehicle makes a sharp turn, the change characteristics of the acceleration curve in the left-right direction of the target vehicle output by the acceleration sensor are the most significant. At the same time, the acceleration curves in the front-rear direction of the target vehicle also show synchronous change characteristics. Therefore, the acceleration in the left-right direction of the target vehicle and the acceleration in the front-rear direction of the target vehicle are used as the basis for judging whether the target vehicle makes a sharp turn.

[0143] Therefore, the process of determining whether the target vehicle has a target driving behavior is as follows:

[0144] B1. Sudden braking behavior recognition

[0145] In some embodiments, when the acceleration includes the acceleration in the front-rear direction of the target vehicle and the target driving behavior includes sudden braking; the above step S104 can be implemented as: when the difference between the current acceleration in the front-rear direction of the target vehicle and the first sliding average value of the historical acceleration in the front-rear direction of the target vehicle is greater than the first preset threshold, it is determined that the target vehicle has a sudden braking behavior.

[0146] Furthermore, to avoid the influence of the noise point data output by the acceleration sensor on the recognition accuracy, the embodiments of the present application can determine whether the target vehicle has a target driving behavior according to the difference between the consecutive multiple current accelerations of the target vehicle and the first sliding average value.

[0147] That is, the above step S104 can also be implemented as follows: when the differences between the current accelerations in the front-rear direction of the target vehicle and the first sliding average of the historical accelerations in the front-rear direction of the target vehicle in a continuous plurality of cases are all greater than the first preset threshold, it is determined that the target vehicle has a hard braking behavior.

[0148] It can be understood that not all the data output by the acceleration sensor can accurately reflect the current acceleration of the target vehicle, and there may still be some noisy data points. If the determination is made only based on the current acceleration in one front-rear direction of the target vehicle, it may lead to misidentification. Therefore, in the embodiments of the present application, the determination of whether the target vehicle has a hard braking behavior is based on the continuous plurality of current accelerations in the front-rear direction of the target vehicle, which can improve the recognition accuracy.

[0149] Exemplarily, assume that the first preset threshold is ΔY0, the first sliding average of the historical front-rear direction of the target vehicle is P y1 , and the current acceleration in the front-rear direction of the target vehicle is y. Then, when in a continuous i (i is an integer greater than 0) cases of |y - P y1 | > ΔY0, it is determined that the target vehicle has a hard braking behavior.

[0150] B2. Recognition of sharp turning behavior

[0151] In some embodiments, when the acceleration includes the acceleration in the left-right direction of the target vehicle and the acceleration in the front-rear direction of the target vehicle, and the target driving behavior includes sharp turning behavior; the above step S104 can be implemented as follows: when the difference between the current acceleration in the left-right direction of the target vehicle and the first sliding average of the historical acceleration in the left-right direction of the target vehicle is greater than the second preset threshold; and / or when the difference between the current acceleration in the front-rear direction of the target vehicle and the first sliding average of the historical acceleration in the front-rear direction of the target vehicle is greater than the third preset threshold, it is determined that the target vehicle has a sharp turning behavior.

[0152] Furthermore, the above step S104 can also be implemented as follows: when the differences between the continuous plurality of current accelerations in the left-right direction of the target vehicle and the first sliding average of the historical accelerations in the left-right direction of the target vehicle are greater than the second preset threshold; and / or when the differences between the continuous plurality of current accelerations in the front-rear direction of the target vehicle and the first sliding average of the historical accelerations in the front-rear direction of the target vehicle are greater than the third preset threshold, it is determined that the target vehicle has a sharp turning behavior.

[0153] Among them, the above-mentioned consecutive current accelerations in the left and right directions represent the accelerations in the left and right directions at consecutive moments starting from the current moment; the above-mentioned consecutive current accelerations in the front and back directions represent the accelerations in the front and back directions at consecutive moments starting from the current moment. For example, if there are three consecutive accelerations, and the current moment is 10:10:10:10:10:10; assuming that from the current moment, the left and right accelerations at 10:10:10:10:10:10:10:20:20 is a1, the left and right accelerations at 10:10:10:20:20 is a2, and the left and right accelerations at 10:10:10:30:30 is a3, then the consecutive current accelerations in the left and right directions are: a1, a2, a3.

[0154] It can be understood that using the target vehicle's current accelerations in multiple consecutive left and right directions and / or multiple consecutive front and rear directions to determine whether the target vehicle has made a sharp turn can reduce the possible misidentification caused by noise point data (data that does not conform to the actual situation and is high or low) and improve recognition accuracy.

[0155] For example, assuming that the second preset threshold is ΔX0, the third preset threshold is ΔY1, and the first sliding average value of the historical acceleration in the left and right directions of the target vehicle is P x1 , the first sliding average of the historical acceleration of the target vehicle in the front and rear directions is P y1 , the current acceleration of the target vehicle in the left and right direction is x, and the current acceleration of the target vehicle in the front and rear direction is y, then when there are j consecutive (j is an integer greater than 0) |xP x1 |>ΔX0; and / or, j consecutive |yP y1 |>ΔY1, it is determined that the target vehicle makes a sharp turn.

[0156] It is understandable that, in the process of determining whether the target vehicle exhibits the target driving behavior, the more the number of the target vehicle's multiple consecutive current accelerations is used, the more accurate the recognition result is, but the sensitivity is lower; similarly, the larger the preset threshold is, the more accurate the recognition result is, but the sensitivity is lower. Therefore, the sensitivity of the recognition device in identifying the target driving behavior can be adjusted by changing the number of the target vehicle's multiple consecutive current accelerations and the size of the preset threshold, so that the method provided in the embodiment of the present application can be applied to different scenarios.

[0157] For example, for sudden braking behavior, due to different driving habits of different drivers, for example, different drivers brake with different strengths. Therefore, for drivers who brake with greater strength, the sensitivity of the recognition device should be lowered (for example, increasing the preset threshold), and for drivers who brake with less strength, the sensitivity of the recognition device should be higher (for example, lowering the preset threshold).

[0158] In some embodiments, when the three-axis directions of the three-axis acceleration sensor are not the same as the three-axis directions corresponding to the target vehicle, transformation processing is respectively performed on the current acceleration in the three-axis directions of the three-axis acceleration sensor and the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor. Furthermore, based on the current acceleration in the three-axis directions of the three-axis acceleration sensor after the transformation processing and the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after the transformation processing, the difference between the current acceleration of the target vehicle and the first moving average is determined.

[0159] As a possible implementation manner, the above-mentioned transformation processing of the current acceleration in the three-axis directions of the three-axis acceleration sensor and the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor respectively can be implemented as the following steps:

[0160] Step 1: Determine the included angle between the three-axis directions of the three-axis acceleration sensor and the three-axis directions corresponding to the target vehicle according to the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor.

[0161] Exemplarily, the included angle between the three-axis directions of the three-axis acceleration sensor and the three-axis directions corresponding to the target vehicle is determined according to the arctangent value of the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor. For example, this included angle can satisfy the following formula (1).

[0162] Step 2: Obtain the current acceleration in the three-axis directions of the three-axis acceleration sensor after the transformation processing according to the current acceleration in the three-axis directions and the included angle.

[0163] Exemplarily, the current acceleration in the three-axis directions of the three-axis acceleration sensor after the transformation processing is obtained according to the quotient of the current acceleration in the three-axis directions and the cosine value of the included angle. For example, the current acceleration in the three-axis directions of the three-axis acceleration sensor after the transformation processing can satisfy the following formula (2).

[0164] Step 3: Obtain the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after the transformation processing according to the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor and the included angle.

[0165] Exemplarily, the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after the transformation processing is obtained according to the quotient of the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor and the cosine value of the included angle. For example, the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after the transformation processing can satisfy the following formula (3).

[0166] Further, based on the absolute value of the difference between the current accelerations in the three-axis directions of the three-axis acceleration sensor after transformation processing and the second moving average of the historical accelerations in the three-axis directions of the three-axis acceleration sensor after variation processing, the difference between the current acceleration of the target vehicle and the first moving average is determined. Exemplarily, the difference between the current acceleration of the target vehicle and the first moving average satisfies the following formulas (4) and (5).

[0167] Exemplarily, as Figure 11 shown, it is assumed that the left-right direction of the vehicle terminal is represented by the X-axis, the front-back direction is represented by the Y-axis, and the vertical direction is represented by the Z-axis; the left-right direction of the three-axis acceleration sensor is represented by the X0-axis, the front-back direction is represented by the Y0-axis, and the vertical direction is represented by the Z0-axis; wherein, there is an included angle θ between the Y-axis direction of the vehicle terminal and the Y0-axis direction of the three-axis acceleration sensor, and the included angle θ can satisfy the following formula (1):

[0168]

[0169] wherein, P x represents the second moving average of the three-axis acceleration sensor in the left-right direction; P y represents the second moving average of the three-axis acceleration sensor in the front-back direction; P z represents the second moving average of the three-axis acceleration sensor in the vertical direction.

[0170] The current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing can satisfy the following formula (2):

[0171]

[0172] wherein, y represents the current acceleration in the front-back direction of the three-axis acceleration sensor after the transformation processing, and y0 represents the current acceleration in the front-back direction of the three-axis acceleration sensor before the transformation processing. The front-back direction of the three-axis acceleration sensor refers to the first direction whose direction trend corresponds to the vehicle body length direction based on the installation state of the three-axis acceleration sensor; the left-right direction of the three-axis acceleration sensor refers to the second direction whose direction trend corresponds to the vehicle body width direction based on the installation state of the three-axis acceleration sensor; the vertical direction of the three-axis acceleration sensor refers to the third direction that is perpendicular to the front-back direction and the left-right direction respectively based on the installation state of the three-axis acceleration sensor.

[0173] The second moving average of the historical acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing can satisfy the following formula (3):

[0174]

[0175] Among them, P Y represents the second moving average of the historical acceleration in the front-rear direction of the triaxial acceleration sensor after the transformation process, and P y represents the second moving average of the historical acceleration in the front-rear direction of the triaxial acceleration sensor before the transformation process is performed.

[0176] The difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average in the front-rear direction satisfies the cosine formula of the trigonometric function, that is, it can satisfy the following formula (4):

[0177]

[0178] Among them, Δy represents the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average in the front-rear direction, that is, Δy = |y - P Y |, and Δy0 represents the difference between the current acceleration in the front-rear direction of the triaxial acceleration sensor and the second moving average of the historical acceleration in the front-rear direction of the triaxial acceleration sensor, that is, Δy0 = |y0 - P y |.

[0179] Therefore, according to the above formula (1), formula (2), formula (3) and formula (4), the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average in the front-rear direction can be obtained. Exemplarily, it can satisfy the following formula (5):

[0180]

[0181] It can be understood that the shapes of different triaxial acceleration sensors are different, and the installation positions of the triaxial acceleration sensors on different vehicles are also different. For example, when the triaxial acceleration sensor is installed on the windshield of the target vehicle, the coordinate axis indicating the front-rear direction of the target vehicle in the triaxial acceleration sensor often generates a large pitch angle. At this time, the acceleration in the front-rear direction output by the triaxial acceleration sensor cannot accurately reflect the acceleration in the front-rear direction of the target vehicle. To solve the above problems, the embodiments of the present application respectively perform transformation processing on the accelerations in the three-axis directions output by the triaxial acceleration sensor and the second moving average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor, obtain the difference between the accelerations in the three-axis directions of the target vehicle and the first moving average, unify the measurement standard, and achieve the purpose of adapting to the installation attitude.

[0182] In some embodiments, the recognition device is connected to the driving recorder of the target vehicle; then the above method further includes: when the target driving behavior occurs in the target vehicle, triggering the automatic emergency video recording function of the driving recorder so that the driving recorder automatically records the video of the target driving behavior.

[0183] Among them, the automatic emergency recording function of the driving recorder means that when an emergency occurs to the target vehicle (such as a sudden braking behavior or a sharp turning behavior), the driving recorder will enter the emergency recording mode, record the video of the emergency, and lock the video of the recorded emergency and save it to the emergency recording folder. The content in this emergency recording folder will not be overwritten by other video materials and can only be deleted manually by the user.

[0184] It can be understood that sudden braking behavior is extremely likely to cause rear-end collisions, and sharp turning behavior is extremely likely to cause collision accidents. Therefore, in the embodiments of the present application, the target driving behavior is set as an emergency of the target vehicle, so that when the target vehicle has a target driving behavior, the driving recorder of the target vehicle can be triggered to automatically record the video of the target driving behavior. In this way, the video of the emergency recorded by the driving recorder provides evidence for the accident caused by the target driving behavior and can be used as a basis for liability analysis and dispute handling.

[0185] Based on the technical solution provided by the embodiments of the present application, at least the following beneficial effects can be achieved: by determining the first sliding average value of the historical acceleration of the target vehicle and using the first sliding average value to reflect the driving state of the target vehicle before the current moment; and then obtaining the current acceleration of the target vehicle, when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, it is determined that the target vehicle has a target driving behavior. In this way, compared with the prior art method of directly identifying sudden braking and sharp turning behaviors based on whether the driving data is greater than a preset threshold, the embodiments of the present application consider the driving state of the target vehicle before the target driving behavior occurs, and introduce the first sliding average value to characterize the driving state of the target vehicle (for example, the target vehicle is in a steady driving state, or the target vehicle is in a bumpy driving state, etc.), and then based on the first sliding average value and the current acceleration of the target vehicle, it is determined whether the target vehicle has a target driving behavior. In this way, misidentifications caused by normal bumps of the vehicle (for example, when the target vehicle is driving on a speed bump or a potholed road surface, the vehicle bumps) can be effectively filtered, and the accuracy of driving behavior identification is improved.

[0186] To test the recognition effect, the embodiments of the present application provide recognition data from on-vehicle tests. Exemplarily, such as Figure 12As shown, from top to bottom are the acceleration data of the target vehicle collected by three different acceleration sensors. Among them, from left to right, they are the acceleration curves output by the acceleration sensor when the target vehicle is driving on the platform road surface, the acceleration curves output by the acceleration sensor when the target vehicle passes over the speed bump, the acceleration curves output by the acceleration sensor when the target vehicle is driving on the potholed road surface, the acceleration curves output by the acceleration sensor when the target vehicle performs an emergency braking behavior, and the acceleration curves output by the acceleration sensor when the target vehicle performs a sharp turning behavior.

[0187] Furthermore, the recognition device is respectively set to high sensitivity, medium sensitivity, and low sensitivity, and the acceleration data of the above target vehicle is processed to obtain the recognition results shown in Figures 13(a), 13(b), and 13(c). Among them, the vertical line marked part is the acceleration data corresponding to the emergency braking behavior and the sharp turning behavior recognized by the recognition device (in actual use, the emergency braking behavior and the sharp turning behavior can be marked with different colors respectively, for example, the emergency braking behavior is marked with a black vertical line, and the sharp turning behavior is marked with a blue vertical line).

[0188] It can be seen that the driving behavior recognition method provided by the embodiment of the present application can accurately recognize the emergency braking behavior and the sharp turning behavior. At the same time, it can effectively filter the influence caused by the acceleration data fluctuations such as speed bumps and potholed road surfaces, reduce the probability of misrecognition, and improve the accuracy of driving behavior recognition.

[0189] Among them, Figure 13(a) is the acceleration data corresponding to the emergency braking behavior and the sharp turning behavior recognized when the recognition device is at high sensitivity; Figure 13(b) is the acceleration data corresponding to the emergency braking behavior and the sharp turning behavior recognized when the recognition device is at medium sensitivity; Figure 13(c) is the acceleration data corresponding to the emergency braking behavior and the sharp turning behavior recognized when the recognition device is at low sensitivity.

[0190] It can be seen that at high sensitivity, the results recognized by the recognition device are relatively dense; at low sensitivity, the results recognized by the recognition device are relatively sparse. Therefore, the sensitivity of the recognition device is different, and the recognized results are also different. Therefore, the embodiment of the present application can adjust the sensitivity of the recognition device so that the above driving behavior recognition method can be applicable to different scenarios.

[0191] As Figure 14 shown, the embodiment of the present application provides a recognition device for performing the driving behavior recognition method as Figure 4 shown. The recognition device 300 includes: an acquisition module 301 and a determination module 302. In some other embodiments, the recognition device 300 further includes: a transformation module 303 and a trigger module 304.

[0192] An acquisition module 301, configured to acquire the historical acceleration of a target vehicle.

[0193] A determination module 302, configured to determine a first moving average value of the historical acceleration of the target vehicle; the first moving average value is used to reflect the driving state of the target vehicle before the current moment.

[0194] The acquisition module 301 is further configured to acquire the current acceleration of the target vehicle.

[0195] The determination module 302 is further configured to determine that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first moving average value is greater than a preset threshold.

[0196] In a possible implementation manner, the historical acceleration of the target vehicle includes: the historical acceleration of the target vehicle within a first time period between a first time point before the current moment and the current moment; and the historical acceleration of the target vehicle within a second time period between a second time point before the current moment and the current moment; the first time point is earlier than the second time point; the determination module 302 is specifically configured to calculate the moving average value of the historical acceleration within the first time period and the moving average value of the historical acceleration within the second time period; when the difference between the moving average value of the historical acceleration within the first time period and the moving average value of the historical acceleration within the second time period is less than or equal to a preset change threshold, use the moving average value of the historical acceleration within the first time period as the first moving average value; or, when the difference between the moving average value of the historical acceleration within the first time period and the moving average value of the historical acceleration within the second time period is greater than the preset change threshold, adjust the historical acceleration within the first time period according to the moving average value of the historical acceleration within the second time period; use the adjusted moving average value of the historical acceleration within the first time period as the first moving average value.

[0197] In another possible implementation manner, the above determination module 302 is specifically configured to replace the historical acceleration of each time point within the first time period with the moving average value of the historical acceleration within the second time period.

[0198] In another possible implementation manner, the acceleration includes: the acceleration in the front-rear direction of the target vehicle, and the front-rear direction of the target vehicle represents the direction of the body length of the target vehicle; the target driving behavior includes: an emergency braking behavior; the determination module 302 is specifically configured to determine that the target vehicle has an emergency braking behavior when the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average value of the historical acceleration in the front-rear direction of the target vehicle is greater than a first preset threshold.

[0199] In another possible implementation, the acceleration includes the acceleration of the target vehicle in the left-right direction and the acceleration of the target vehicle in the front-rear direction. The left-right direction of the target vehicle represents the direction of the body width of the target vehicle. The target driving behavior includes a sharp-turning behavior. The determining module 302 is specifically configured to determine that the target vehicle has a sharp-turning behavior when the difference between the current acceleration of the target vehicle in the left-right direction and the first sliding average of the historical acceleration of the target vehicle in the left-right direction is greater than a second preset threshold; and / or when the difference between the current acceleration of the target vehicle in the front-rear direction and the first sliding average of the historical acceleration of the target vehicle in the front-rear direction is greater than a third preset threshold.

[0200] In another possible implementation, the determining module 302 is specifically configured to determine that the target vehicle has a target driving behavior when the differences between a continuous plurality of current accelerations of the target vehicle and the first sliding average are all greater than a preset threshold. The continuous plurality of current accelerations represent the accelerations at a continuous plurality of moments starting from the current moment.

[0201] In another possible implementation, the target vehicle includes a triaxial acceleration sensor. The triaxial acceleration sensor is used to detect the accelerations in the three-axis directions of the triaxial acceleration sensor. The three-axis directions of the triaxial acceleration sensor are the directions indicated by the three coordinate axes of the triaxial acceleration sensor. The transformation module 303 performs transformation processing on the current acceleration in the three-axis directions of the triaxial acceleration sensor and the second sliding average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor respectively when the three-axis directions of the triaxial acceleration sensor are not the same as the corresponding three-axis directions of the target vehicle. The second sliding average is the sliding average of the historical accelerations in the three-axis directions detected by the triaxial acceleration sensor before the current moment. The determining module 302 is further configured to determine the difference between the current acceleration of the target vehicle and the first sliding average according to the current acceleration in the three-axis directions of the triaxial acceleration sensor after the transformation processing and the second sliding average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor after the transformation processing.

[0202] In another possible implementation, the transformation module 303 is specifically configured to determine the included angle between the three-axis directions of the triaxial acceleration sensor and the corresponding three-axis directions of the target vehicle according to the second sliding average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor; obtain the current acceleration in the three-axis directions of the triaxial acceleration sensor after the transformation processing according to the current acceleration in the three-axis directions and the included angle; and obtain the second sliding average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor after the transformation processing according to the second sliding average of the historical accelerations in the three-axis directions of the triaxial acceleration sensor and the included angle.

[0203] In another possible implementation, the three-axis directions of the above three-axis acceleration sensor are not the same as the three-axis directions corresponding to the target vehicle, including: the front-back direction of the three-axis acceleration sensor is not the same as the front-back direction of the target vehicle; the determination module 302 is specifically configured to determine the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction according to the current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing and the second sliding average value of the historical acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing.

[0204] Among them, the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction satisfies the following relationship:

[0205]

[0206] Among them, Δy represents the difference between the current acceleration in the front-back direction of the target vehicle and the first sliding average value in the front-back direction; y represents the current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing, and y satisfies: P Y represents the second sliding average value of the historical acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing, and P Y satisfies: y0 represents the current acceleration in the front-back direction of the three-axis acceleration sensor, and P y represents the second sliding average value of the historical acceleration in the front-back direction of the three-axis acceleration sensor; θ represents the included angle between the front-back direction of the three-axis acceleration sensor and the corresponding front-back direction of the target vehicle, and θ satisfies: P x represents the second sliding average value of the three-axis acceleration sensor in the left-right direction; P y represents the second sliding average value of the three-axis acceleration sensor in the front-back direction; P z represents the second sliding average value of the three-axis acceleration sensor in the vertical direction.

[0207] In another possible implementation, the recognition device is connected to the driving recorder of the target vehicle; the trigger module 304 is configured to trigger the automatic emergency video recording function of the driving recorder in the case where the target vehicle has a target driving behavior, so that the driving recorder automatically records the video of the target driving behavior; the target driving behavior includes at least one of the following: a hard braking behavior, a sharp turning behavior.

[0208] In the case of implementing the functions of the above integrated modules in the form of hardware, the embodiments of the present application provide another possible structural schematic diagram of the recognition device involved in the above embodiments. As Figure 15As shown, the recognition device 400 includes: a processor 402, a communication interface 403, and a bus 404. Optionally, the recognition device may further include a memory 401.

[0209] The processor 402 may be a device that implements or executes various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor 402 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor 402 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0210] The communication interface 403 is used to connect to other devices through a communication network. The communication network may be an Ethernet, a radio access network, a wireless local area network (WLAN), etc.

[0211] The memory 401 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM), or other type of dynamic storage device that can store information and instructions. It may also be an electrically erasable programmable read-only memory (EEPROM), a magnetic disk storage medium, or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0212] As a possible implementation, the memory 401 may exist independently of the processor 402. The memory 401 may be connected to the processor 402 through the bus 404 and is used to store instructions or program code. When the processor 402 calls and executes the instructions or program code stored in the memory 401, the driving behavior recognition method provided in the embodiments of this application can be implemented.

[0213] In another possible implementation, the memory 401 may also be integrated with the processor 402.

[0214] The bus 404 can be an extended industry standard architecture (EISA) bus or the like. The bus 404 can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 15 only a thick line is used to represent it in Figure 15 , but it does not mean that there is only one bus or one type of bus.

[0215] Through the description of the above embodiments, those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above division of each functional module is used as an example. In practical applications, the above functions can be allocated to different functional modules as needed, that is, the internal structure of the recognition device is divided into different functional modules to complete all or part of the functions described above.

[0216] The embodiment of the present application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be instructed by computer instructions to complete the relevant hardware. This program can be stored in the above computer-readable storage medium. When this program is executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be the memory in any of the foregoing embodiments. The above computer-readable storage medium can also be an external storage device of the above recognition device, such as a plug-in hard disk equipped on the above recognition device, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the above computer-readable storage medium can also include both the internal storage unit of the above recognition device and the external storage device. The above computer-readable storage medium is used to store the above computer program and other programs and data required by the above recognition device. The above computer-readable storage medium can also be used to temporarily store the data that has been output or will be output.

[0217] The embodiment of the present application also provides a computer program product. The computer product includes a computer program. When the computer program product runs on a computer, the computer is enabled to execute any one of the driving behavior recognition methods provided in the above embodiments.

[0218] Although the present application has been described in connection with various embodiments, it will be understood by those skilled in the art that other variations of the disclosed embodiments can be understood and effected while practicing the claimed application, by reference to the drawings, the disclosure, and the appended claims. In the claims, the term "comprising" does not exclude other elements or steps, the word "a" or "an" does not exclude a plurality of instances, and a single processor or other unit may fulfill the functions recited in several claims. Certain measures are recited in mutually different dependent claims, but this does not indicate that these measures cannot be combined to advantage.

[0219] Although the present application has been described in connection with specific features and their embodiments, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and drawings are merely exemplary illustrations of the application defined by the appended claims and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the application. It is obvious that those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

[0220] The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A driving behavior recognition method, characterized in that, The method includes: Obtaining the historical acceleration of the target vehicle; the historical acceleration of the target vehicle includes: the historical acceleration of the target vehicle within a first time period between a first time point before the current moment and the current moment; and the historical acceleration of the target vehicle within a second time period between a second time point before the current moment and the current moment; the first time point is earlier than the second time point; Determining a first moving average of the historical acceleration of the target vehicle; the first moving average is used to reflect the driving state of the target vehicle before the current moment; The determining the first moving average of the historical acceleration of the target vehicle includes: Calculating the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period; When the difference between the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period is less than or equal to a preset change threshold, using the moving average of the historical acceleration within the first time period as the first moving average; When the difference between the moving average of the historical acceleration within the first time period and the moving average of the historical acceleration within the second time period is greater than the preset change threshold, adjusting the historical acceleration within the first time period according to the moving average of the historical acceleration within the second time period; using the adjusted moving average of the historical acceleration within the first time period as the first moving average; the adjusting the historical acceleration within the first time period according to the moving average of the historical acceleration within the second time period includes: replacing the historical acceleration at each time point within the first time period with the moving average of the historical acceleration within the second time period; Obtaining the current acceleration of the target vehicle; When the difference between the current acceleration of the target vehicle and the first moving average is greater than a preset threshold, determining that the target vehicle has a target driving behavior.

2. The method according to claim 1, wherein Acceleration includes: The acceleration in the front-rear direction of the target vehicle, and the front-rear direction of the target vehicle represents the direction of the vehicle body length of the target vehicle; The target driving behavior includes: an emergency braking behavior; the determining that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first moving average is greater than a preset threshold includes: When the difference between the current acceleration in the front-rear direction of the target vehicle and the first moving average of the historical acceleration in the front-rear direction of the target vehicle is greater than a first preset threshold, determining that the target vehicle has the emergency braking behavior.

3. The method according to claim 1, characterized in that Acceleration includes: The acceleration in the left - right direction of the target vehicle and the acceleration in the front - rear direction of the target vehicle, where the left - right direction of the target vehicle represents the direction of the vehicle body width of the target vehicle; the target driving behavior includes: a sharp - turn behavior; when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, determining that the target vehicle has a target driving behavior, including: When the difference between the current acceleration in the left - right direction of the target vehicle and the first sliding average value of the historical acceleration in the left - right direction of the target vehicle is greater than a second preset threshold; and / or, when the difference between the current acceleration in the front - rear direction of the target vehicle and the first sliding average value of the historical acceleration in the front - rear direction of the target vehicle is greater than a third preset threshold, determining that the target vehicle has the sharp - turn behavior.

4. The method according to claim 1, characterized in that When the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, determining that the target vehicle has a target driving behavior, including: When the differences between a continuous plurality of the current accelerations of the target vehicle and the first sliding average value are all greater than the preset threshold, determining that the target vehicle has the target driving behavior; the continuous plurality of current accelerations represent the accelerations at a continuous plurality of moments starting from the current moment.

5. The method according to claim 1, characterized in that, The target vehicle includes a tri - axis acceleration sensor; the tri - axis acceleration sensor is used to detect the accelerations in the three - axis directions of the tri - axis acceleration sensor; the three - axis directions of the tri - axis acceleration sensor are the directions indicated by the three coordinate axes of the tri - axis acceleration sensor; Before determining that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first sliding average value is greater than a preset threshold, the method further includes: When the three - axis directions of the tri - axis acceleration sensor are not the same as the corresponding three - axis directions of the target vehicle, respectively performing transformation processing on the current accelerations in the three - axis directions of the tri - axis acceleration sensor and the second sliding average value of the historical accelerations in the three - axis directions of the tri - axis acceleration sensor; the second sliding average value is the sliding average value of the historical accelerations in the three - axis directions detected by the tri - axis acceleration sensor before the current moment; Determining the difference between the current acceleration of the target vehicle and the first sliding average value according to the current acceleration in the three - axis directions of the tri - axis acceleration sensor after the transformation processing and the second sliding average value of the historical accelerations in the three - axis directions of the tri - axis acceleration sensor after the transformation processing.

6. The method according to claim 5, wherein The respectively performing transformation processing on the current accelerations in the three - axis directions of the tri - axis acceleration sensor and the second sliding average value of the historical accelerations in the three - axis directions of the tri - axis acceleration sensor includes: Determining the included angle between the three - axis directions of the tri - axis acceleration sensor and the corresponding three - axis directions of the target vehicle according to the second sliding average value of the historical accelerations in the three - axis directions of the tri - axis acceleration sensor. Obtain the current acceleration in the three-axis directions of the three-axis acceleration sensor after transformation processing based on the current acceleration in the three-axis directions and the included angle; Obtain the second moving average of the historical acceleration in the three-axis directions of the three-axis acceleration sensor after transformation processing based on the second moving average of the historical acceleration in the three-axis directions of the three-axis acceleration sensor and the included angle; 7. The method according to claim 6, characterized in that, The three-axis directions of the three-axis acceleration sensor are not the same as the three-axis directions corresponding to the target vehicle, including: the front-back direction of the three-axis acceleration sensor is not the same as the front-back direction of the target vehicle; Determine the difference between the current acceleration of the target vehicle and the first moving average based on the current acceleration in the three-axis directions of the three-axis acceleration sensor after transformation processing and the second moving average of the historical acceleration in the three-axis directions of the three-axis acceleration sensor after transformation processing, including: Determine the difference between the current acceleration in the front-back direction of the target vehicle and the first moving average in the front-back direction based on the current acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing and the second moving average of the historical acceleration in the front-back direction of the three-axis acceleration sensor after transformation processing; Among them, the difference between the current acceleration in the front-back direction of the target vehicle and the first moving average in the front-back direction satisfies the following relationship: where, Δy represents the difference between the current acceleration in the front-rear direction of the target vehicle and the first sliding average value in the front-rear direction; y represents the current acceleration in the front-rear direction of the triaxial acceleration sensor after the transformation process, and y satisfies: P Y represents the second sliding average value of the historical acceleration in the front-rear direction of the triaxial acceleration sensor after the transformation process, and P Y satisfies: y0 represents the current acceleration in the front-rear direction of the triaxial acceleration sensor, and P y represents the second sliding average value of the historical acceleration in the front-rear direction of the triaxial acceleration sensor; θ represents the angle between the front-rear direction of the triaxial acceleration sensor and the corresponding front-rear direction of the target vehicle, and θ satisfies: P x represents the second sliding average value of the triaxial acceleration sensor in the left-right direction; P y represents the second sliding average value of the triaxial acceleration sensor in the front-rear direction; P z represents the second sliding average value of the triaxial acceleration sensor in the vertical direction.

8. The method according to claim 1, characterized in that, The method is applied to an identification device; The identification device is connected to the driving recorder of the target vehicle; the method further includes: In the case where the target vehicle exhibits a target driving behavior, trigger the automatic emergency recording function of the driving recorder so that the driving recorder automatically records the video of the target driving behavior; the target driving behavior includes at least one of the following: sudden braking behavior, sharp turning behavior.

9. An identification device, characterized in that, The identification device includes: An acquisition module for acquiring the historical acceleration of the target vehicle; the historical acceleration of the target vehicle includes: the historical acceleration of the target vehicle within the first time period between the first time point before the current moment and the current moment; and the historical acceleration of the target vehicle within the second time period between the second time point before the current moment and the current moment; the first time point is earlier than the second time point; A determination module for determining the first moving average of the historical acceleration of the target vehicle; the first moving average is used to reflect the driving state of the target vehicle before the current moment; The determining module is specifically configured to calculate the sliding average of the historical acceleration in the first time period and the sliding average of the historical acceleration in the second time period; when the difference between the sliding average of the historical acceleration in the first time period and the sliding average of the historical acceleration in the second time period is less than or equal to a preset change threshold, use the sliding average of the historical acceleration in the first time period as the first sliding average; when the difference between the sliding average of the historical acceleration in the first time period and the sliding average of the historical acceleration in the second time period is greater than the preset change threshold, adjust the historical acceleration in the first time period according to the sliding average of the historical acceleration in the second time period; use the adjusted sliding average of the historical acceleration in the first time period as the first sliding average; the determining module is specifically configured to replace the historical acceleration at each time point in the first time period with the sliding average of the historical acceleration in the second time period; The obtaining module is further configured to obtain the current acceleration of the target vehicle; The determining module is further configured to determine that the target vehicle has a target driving behavior when the difference between the current acceleration of the target vehicle and the first sliding average is greater than a preset threshold.

10. An identification device, characterized in that, Comprising: One or more processors; One or more memories; Wherein, the one or more memories are used to store computer program code, the computer program code includes computer instructions, and when the one or more processors execute the computer instructions, the recognition device executes the driving behavior recognition method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions run on a computer, the computer executes the driving behavior recognition method according to any one of claims 1 to 8.

Images