Obstacle orientation optimization method and apparatus, storage medium, and electronic device

Abstract

An obstacle orientation optimization method and apparatus, a storage medium, and an electronic device. The method comprises: acquiring the orientations of a target obstacle within a preset time; performing statistical analysis on the orientation included angles of the target obstacle in two adjacent image frames to determine image filtering modes; and optimizing the orientation of the target obstacle on the basis of the image filtering modes.

Description

Obstacle orientation optimization method, optimization device, storage medium and electronic equipment

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This disclosure claims priority to Chinese patent application No. 2023111428252, filed on September 5, 2023, entitled “Optimization method, optimization device, storage medium and electronic device for obstacle orientation,” the entire contents of which are incorporated by reference into this disclosure. Technical Field

[0003] The present disclosure relates to the field of autonomous driving technology, and in particular to a method for optimizing obstacle orientation, a device for optimizing obstacle orientation, a computer-readable storage medium, and an electronic device. Background Art

[0004] In perception and positioning, tracking algorithms are a crucial component of autonomous driving systems. The purpose of target tracking is to use sensor measurements to estimate the state of a target of interest, such as the position, speed, and orientation of traffic participants (other vehicles, people, etc.), in real time. In recent years, with the advancement of deep learning, visual perception has become a key application in autonomous driving systems. Sensor data is used to identify physical objects, including their type, location, detection frame, orientation, and attributes. This information provides prior knowledge for downstream control systems, enabling vehicle control. The tracking module, in particular, requires real-time estimation of the target using information such as position, detection frame, and orientation, obtaining information about the target's position, speed, orientation, and other state information.

[0005] In target detection and tracking technology for autonomous driving, vehicle orientation is usually filtered and optimized for obstacles. However, in some special scenarios, such as when the orientation flips or rotates, it becomes impossible to correctly predict and track the orientation.

[0006] Public content

[0007] The present disclosure aims to address, at least to some extent, one of the technical problems in the related art. To this end, the first objective of the present disclosure is to propose a method for optimizing the orientation of an obstacle. This method statistically analyzes the angle between the orientations of a target obstacle in two adjacent image frames within a preset time interval, determines different image filtering modes, and optimizes the orientation of the target obstacle based on these image filtering modes, thereby improving the accuracy of determining the orientation of the target obstacle.

[0008] The second objective of the present disclosure is to provide a device for optimizing obstacle orientation.

[0009] A third object of the present disclosure is to provide a computer-readable storage medium.

[0010] A fourth objective of the present disclosure is to provide an electronic device.

[0011] To achieve the above-mentioned objectives, an embodiment of the first aspect of the present disclosure proposes a method for optimizing the orientation of an obstacle, the method comprising: obtaining the orientation of a target obstacle within a preset time; performing statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images to determine an image filtering mode; and optimizing the orientation of the target obstacle according to the image filtering mode.

[0012] According to the obstacle orientation optimization method of the disclosed embodiment, the orientation of the target obstacle within a preset time is obtained, and the angle between the orientations of the target obstacle in two adjacent image frames is statistically analyzed to determine an image filtering mode. The orientation of the target obstacle is then optimized based on the image filtering mode. Thus, by performing a statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames within a preset time, determining different image filtering modes, and optimizing the orientation of the target obstacle based on the image filtering modes, the method can improve the accuracy of determining the orientation of the target obstacle.

[0013] In addition, the obstacle orientation optimization method according to the above embodiment of the present disclosure may also have the following additional technical features:

[0014] According to some embodiments of the present disclosure, the statistical analysis of the orientation angle of the target obstacle in two adjacent frames of images includes: determining the statistical number of times the orientation angle is greater than a first preset angle within the preset time, recorded as a first statistical number; determining the statistical number of times the orientation angle is less than a second preset angle within the preset time, recorded as a second statistical number; determining the statistical number of times the orientation angle is greater than or equal to the second preset angle and less than or equal to the first preset angle within the preset time, recorded as a third statistical number.

[0015] According to some embodiments of the present disclosure, determining the image filtering mode includes: when the first statistical number is less than a first preset value and the third statistical number is zero, determining the image filtering mode to be the first image filtering mode; when the second statistical number is equal to the first preset value, determining the image filtering mode to be the second image filtering mode; wherein the first preset value is determined based on the number of image frames of the target obstacle within the preset time.

[0016] According to some embodiments of the present disclosure, optimizing the orientation of the target obstacle according to the image filtering mode includes: when the image filtering mode is the first image filtering mode, optimizing the orientation of the target obstacle according to the average of the orientation angles of the target obstacle.

[0017] According to some embodiments of the present disclosure, optimizing the orientation of the target obstacle according to the image filtering mode includes: when the image filtering mode is the second image filtering mode, optimizing the orientation of the target obstacle using a Kalman filtering algorithm.

[0018] According to some embodiments of the present disclosure, optimizing the orientation of the target obstacle according to the image filtering mode includes: outputting the orientation of the target obstacle when the third statistical number is greater than zero.

[0019] According to some embodiments of the present disclosure, obtaining the orientation of the target obstacle within a preset time includes: obtaining image data of the target obstacle; processing the image data to obtain a detection frame of the target obstacle; and determining the orientation of the target obstacle based on the orientation of the detection frame of the target obstacle.

[0020] To achieve the above-mentioned objectives, the second embodiment of the present disclosure proposes an obstacle orientation optimization device, which includes: an acquisition module for acquiring the orientation of a target obstacle within a preset time; a determination module for performing statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images to determine an image filtering mode; and an optimization module for optimizing the orientation of the target obstacle according to the image filtering mode.

[0021] According to the obstacle orientation optimization device of the disclosed embodiment, an acquisition module acquires the orientation of a target obstacle within a preset time period, a determination module performs statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames to determine an image filtering mode, and an optimization module optimizes the orientation of the target obstacle based on the image filtering mode. Thus, by performing statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames within a preset time period, determining different image filtering modes, and optimizing the orientation of the target obstacle based on the image filtering modes, the device can improve the accuracy of determining the orientation of the target obstacle.

[0022] To achieve the above-mentioned objectives, the third embodiment of the present disclosure proposes a computer-readable storage medium on which an obstacle orientation optimization program is stored. When the obstacle orientation optimization program is executed by a processor, the above-mentioned obstacle orientation optimization method is implemented.

[0023] The computer-readable storage medium of the embodiment of the present disclosure can improve the accuracy of determining the target obstacle orientation by executing the above-mentioned obstacle orientation optimization method.

[0024] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present disclosure proposes an electronic device, including a memory, a processor, and an obstacle orientation optimization program stored in the memory and runnable on the processor. When the processor executes the obstacle orientation optimization program, the above-mentioned obstacle orientation optimization method is implemented.

[0025] According to the electronic device of the embodiment of the present disclosure, by executing the above-mentioned obstacle orientation optimization method, the accuracy of determining the target obstacle orientation can be improved.

[0026] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] FIG1 is a flow chart of a method for optimizing obstacle orientation according to some embodiments of the present disclosure;

[0028] Figures 2a and 2b are schematic diagrams of the orientation of obstacles obtained in the related art;

[0029] FIG3 is a schematic diagram of an obstacle orientation optimized in the related art;

[0030] FIG4 is a schematic diagram of an obstacle orientation optimized according to some embodiments of the present disclosure;

[0031] FIG5 is a schematic diagram of obstacle orientation according to some embodiments of the present disclosure;

[0032] FIG6 is a schematic diagram of obstacle orientation according to some embodiments of the present disclosure;

[0033] FIG7 is a schematic diagram of an obstacle orientation optimized according to some embodiments of the present disclosure;

[0034] FIG8 is a block diagram of an apparatus for optimizing obstacle orientation according to some embodiments of the present disclosure;

[0035] FIG9 is a block diagram of an electronic device according to some embodiments of the present disclosure. DETAILED DESCRIPTION

[0036] The following describes in detail embodiments of the present disclosure, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0037] The following describes an obstacle orientation optimization method, an obstacle orientation optimization device, a computer-readable storage medium, and an electronic device proposed in embodiments of the present disclosure with reference to the accompanying drawings.

[0038] FIG1 is a flow chart of a method for optimizing obstacle orientation according to some embodiments of the present disclosure.

[0039] As shown in FIG1 , the method for optimizing the obstacle orientation according to an embodiment of the present disclosure may include the following steps:

[0040] S1, obtaining the direction of the target obstacle within a preset time, wherein the preset time can be calibrated according to actual conditions.

[0041] S2, performing statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images to determine the image filtering mode.

[0042] S3, optimizing the direction of the target obstacle according to the image filtering mode.

[0043] Specifically, in the target detection and tracking technology of autonomous driving, tracking requires optimization of the target information obtained by detection. Usually, the vehicle orientation is filtered and optimized for driving obstacles. However, there are some special scenarios that conventional Kalman filters cannot handle. For example, there are some errors in the orientation of the detection task input, which is reversed or rotated from the actual real data, as shown in Figure 2a and Figure 2b. i represents the current time, t i-1 , t i-2 ,…,t i-8 Represents the eight moments before the current moment, and the bold position in the figure is the direction of the target obstacle. If only the Kalman filter algorithm is used, it is impossible to correctly predict and track the direction, and the obtained effect is not good, as shown in Figure 3; through the control method of the present application, this type of error (the type shown in Figure 3) in the tracking process of the target obstacle is optimized, and the optimization result is shown in Figure 4, where T in Figures 3 and 4 represents the current moment, and T-1, T-2, T-3 and T-4 represent the four moments before the current moment. In this way, the obstacle direction can be optimized and the accuracy of determining the target obstacle direction can be improved.

[0044] Specifically, the target obstacle can be a vehicle. Image data of the target obstacle can be acquired over a period of time using a camera or radar sensor. This image data can include images or videos. Processing this image data can determine the target obstacle's orientation over that period of time. Assuming there are N frames of imagery within a preset timeframe, the target obstacle's orientation angle between two adjacent frames is calculated, with the number of orientation angles being N-1. Statistical analysis is then performed on the target obstacle's orientation angles between the two adjacent frames. An image filtering mode is determined based on this orientation angle between the two adjacent frames. The target obstacle's orientation is then optimized based on the image filtering mode, and the optimized target obstacle's orientation is output.

[0045] For example, assuming N is an odd number, such as 5, the target obstacle's orientation angle between two adjacent frames is calculated for five frames. The number of orientation angles between two adjacent frames is four, and the range of these angles is 0-180°. When the orientation angle is small, the orientation of the target obstacle in the two adjacent frames is assumed to be roughly the same; when the orientation angle is large, the target obstacle is considered to have significantly flipped between the two adjacent frames. The four orientation angles are statistically analyzed to determine the appropriate filtering method. For example, if all four orientation angles are relatively small, a conventional Kalman filter can be used. However, if three orientation angles are small and one is large, a probabilistic statistical filtering method is required.

[0046] Therefore, by performing statistical analysis on the orientation angle, different filtering modes can be selected, thereby improving the accuracy of determining the orientation of the obstacle.

[0047] The following describes in detail the method for optimizing the obstacle orientation according to an embodiment of the present disclosure.

[0048] In some embodiments of the present disclosure, a statistical analysis of the orientation angle of a target obstacle in two adjacent frames of imagery is performed, including: determining the number of times within a preset time period that the orientation angle is greater than a first preset angle, recorded as a first statistical number; determining the number of times within a preset time period that the orientation angle is less than a second preset angle, recorded as a second statistical number; and determining the number of times within a preset time period that the orientation angle is greater than or equal to the second preset angle and less than or equal to the first preset angle, recorded as a third statistical number. The first preset angle and the second preset angle can be calibrated based on actual conditions. For example, the first preset angle can be 150°, and the second preset angle can be 30°.

[0049] Specifically, still taking the preset time including N frames of images as an example, the number of orientation angles is N-1, and the orientation angles of the target obstacle in two adjacent frames of images are calculated in sequence, and the size of the orientation angles is judged. Among them, when the orientation angle is large, such as the orientation angle is greater than the first preset angle (such as 150°), the corresponding counter is increased by 1; when the orientation angle is small, such as the orientation angle is less than the second preset angle (such as 30°), the corresponding counter is increased by 1; when the orientation angle is at an intermediate value, such as the orientation angle is greater than or equal to the second preset angle (such as 30°) and less than or equal to the first preset angle (such as 150°), the corresponding counter is increased by 1, and finally the first statistical number, the second statistical number, and the third statistical number are obtained, so that the subsequent embodiments can determine the image filtering mode according to the values ​​of the first statistical number, the second statistical number, and the third statistical number, so that the corresponding filtering mode can be determined by statistical analysis of the orientation angle, thereby improving the accuracy of determining the orientation of the target obstacle. It should be noted that the sum of the first statistical number, the second statistical number, and the third statistical number is N-1.

[0050] In some embodiments of the present disclosure, determining the image filtering mode includes: when the first statistical number is less than a first preset value and the third statistical number is zero, determining the image filtering mode to be the first image filtering mode; when the second statistical number is equal to the first preset value, determining the image filtering mode to be the second image filtering mode; wherein the first preset value is determined based on the number of image frames of the target obstacle within a preset time. For example, when the number of image frames of the target obstacle is N, the first preset value may be N-1.

[0051] Specifically, after determining the first, second, and third statistical times through the above embodiment, the image filtering mode is determined based on the magnitude of the first, second, and third statistical times. When the first statistical time is less than the first preset value and the third statistical time is zero, that is, there is an orientation angle greater than 150° and less than or equal to 180°, and there is no orientation angle between 30° and 150°, indicating that the orientation of the target obstacle is basically consistent, but exhibits significant errors at certain times. In this case, the first image filtering mode can be adopted, such as using a more probable orientation for correction to optimize the orientation of the target obstacle. When the second statistical time is equal to the first preset value, that is, all orientation angles are less than 30°, indicating that the orientation of the target obstacle is relatively consistent, with smaller errors within a specific time. In this case, the second image filtering mode can be adopted, such as using a conventional Kalman filter to optimize the orientation of the target obstacle.

[0052] For example, as shown in Figure 5, the target obstacle's orientation angle between the two frames at time T and time T-1 exceeds 150°, the target obstacle's orientation angle between the two frames at time T-4 and time T-3 does not exceed 30°, the target obstacle's orientation angle between the two frames at time T-3 and time T-2 does not exceed 30°, and the target obstacle's orientation angle between the two frames at time T-2 and time T-1 does not exceed 30°. The orientation at time T is corrected using the orientations corresponding to time T-4, time T-3, time T-2, and time T-1. For another example, as shown in Figure 6, the target obstacle's orientation angle between the two frames at time T and time T-1 does not exceed 30°, the target obstacle's orientation angle between the two frames at time T-4 and time T-3 does not exceed 30°, the target obstacle's orientation angle between the two frames at time T-3 and time T-2 does not exceed 30°, and the target obstacle's orientation angle between the two frames at time T-2 and time T-1 does not exceed 30°. It should be noted that the bold positions in FIG5 and FIG6 are the directions of the target obstacles.

[0053] In some embodiments of the present disclosure, the orientation of the target obstacle is optimized according to the image filtering mode, including: when the image filtering mode is the first image filtering mode, optimizing the orientation of the target obstacle according to the average of the orientation angles of the target obstacle.

[0054] Specifically, by statistically analyzing the target obstacle's orientation angle in two adjacent image frames, the image filtering mode is determined to be the first image filtering mode. As shown in Figure 7, using the horizontal direction as a reference line, the angle between the target obstacle's orientation and the reference line is obtained, recorded as the target obstacle's orientation angle. The target obstacle's orientation angles corresponding to times T-4, T-3, T-2, T-1, and T are obtained, and then averaged. The target obstacle's orientation at time T is corrected based on the average value. The corrected target obstacle's orientation at time T is shown in Figure 7. This method can correct for significant deviations in the target obstacle's orientation in individual image frames, optimize the target obstacle's orientation, and output the optimized target obstacle's orientation, as shown in Figure 4. This improves the accuracy of determining the target obstacle's orientation during target obstacle tracking.

[0055] In some embodiments of the present disclosure, the orientation of the target obstacle is optimized according to the image filtering mode, including: when the image filtering mode is the second image filtering mode, using a Kalman filtering algorithm to optimize the orientation of the target obstacle.

[0056] Specifically, by statistically analyzing the included angle of the target obstacle's orientation in two adjacent image frames, the second image filtering mode is determined. As shown in Figure 6, the orientation of the target obstacle in the two adjacent image frames is relatively stable, with the orientation being essentially consistent. This allows the Kalman filter algorithm to optimize the target obstacle's orientation and output the optimized target obstacle's orientation, enabling accurate prediction and tracking of the target obstacle's position. It should be noted that the Kalman filter algorithm is a relatively mature technology and will not be described in detail here.

[0057] In some embodiments of the present disclosure, optimizing the orientation of the target obstacle according to the image filtering mode includes: outputting the orientation of the target obstacle when the third statistical number is greater than zero.

[0058] Specifically, by performing a statistical analysis on the included angle of the target obstacle's orientation in two adjacent frames of images, when the third statistical number is greater than zero, it indicates that the orientation of the target obstacle has no statistical characteristics and presents an inconsistent distribution, making it impossible to perform a probability analysis. In this case, the image filtering mode is the no-filter mode, and the acquired orientation of the target obstacle is directly output.

[0059] In some embodiments of the present disclosure, obtaining the orientation of a target obstacle within a preset time includes: obtaining image data of the target obstacle; processing the image data to obtain a detection frame of the target obstacle; and determining the orientation of the target obstacle based on the orientation of the detection frame of the target obstacle.

[0060] Specifically, during the target obstacle orientation tracking process, image data of the target obstacle is first acquired over a period of time using a camera or radar sensor. This image data is then processed to obtain a detection frame for the target obstacle. The target obstacle's orientation is then determined based on the orientation of the detection frame. Specifically, the bolded portion of the detection frame represents the target obstacle's orientation. After obtaining the target obstacle's orientation, an image filtering mode is determined based on the angle between the target obstacle's orientations in two adjacent image frames. The target obstacle's orientation is then optimized based on the image filtering mode, and the optimized target obstacle's orientation is output. This improves the accuracy of determining the target obstacle's orientation during the target obstacle orientation tracking process.

[0061] Therefore, this application divides the orientation angle of the target obstacle in two adjacent frames of images into three different scenarios by performing statistical analysis. Different scenarios use different filtering modes, including Kalman filtering algorithm mode (second image filtering mode), mean filtering algorithm mode (first image filtering mode) and no filtering mode, so as to optimize the filtering results and improve the accuracy of determining the orientation of the target obstacle.

[0062] In summary, according to the obstacle orientation optimization method of the disclosed embodiment, the orientation of the target obstacle within a preset time is obtained, the angle between the orientations of the target obstacle in two adjacent image frames is statistically analyzed to determine an image filtering mode, and the orientation of the target obstacle is optimized based on the image filtering mode. Thus, by performing a statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames within a preset time, determining different image filtering modes, and optimizing the orientation of the target obstacle based on the image filtering modes, this method can improve the accuracy of determining the orientation of the target obstacle.

[0063] Corresponding to the above embodiment, the present disclosure also proposes a device for optimizing the direction of an obstacle.

[0064] As shown in FIG. 8 , the obstacle orientation optimization device 100 according to an embodiment of the present disclosure may include: an acquisition module 110 , a determination module 120 , and an optimization module 130 .

[0065] The acquisition module 110 is used to obtain the orientation of the target obstacle within a preset time. The determination module 120 is used to perform statistical analysis on the angle between the orientations of the target obstacle in two adjacent frames of image to determine the image filtering mode. The optimization module 130 is used to optimize the orientation of the target obstacle based on the image filtering mode.

[0066] In some embodiments of the present disclosure, the determination module 120 performs a statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images, specifically for: determining the statistical number of times the orientation angle is greater than a first preset angle within a preset time, recorded as a first statistical number; determining the statistical number of times the orientation angle is less than a second preset angle within a preset time, recorded as a second statistical number; determining the statistical number of times the orientation angle is greater than or equal to the second preset angle and less than or equal to the first preset angle within a preset time, recorded as a third statistical number.

[0067] In some embodiments of the present disclosure, the determination module 120 determines the image filtering mode specifically for: when the first statistical number is less than the first preset value and the third statistical number is zero, determining the image filtering mode to be the first image filtering mode; when the second statistical number is equal to the first preset value, determining the image filtering mode to be the second image filtering mode; wherein the first preset value is determined based on the number of image frames of the target obstacle within a preset time.

[0068] In some embodiments of the present disclosure, the optimization module 130 optimizes the orientation of the target obstacle according to the image filtering mode, specifically for: when the image filtering mode is the first image filtering mode, optimizing the orientation of the target obstacle according to the mean of the orientation angles of the target obstacle.

[0069] In some embodiments of the present disclosure, the optimization module 130 optimizes the orientation of the target obstacle according to the image filtering mode, specifically, when the image filtering mode is the second image filtering mode, using the Kalman filtering algorithm to optimize the orientation of the target obstacle.

[0070] In some embodiments of the present disclosure, the optimization module 130 optimizes the orientation of the target obstacle according to the image filtering mode, and is specifically configured to output the orientation of the target obstacle when the third statistical number is greater than zero.

[0071] In some embodiments of the present disclosure, the acquisition module 110 acquires the orientation of the target obstacle within a preset time, and is specifically used to: acquire image data of the target obstacle; process the image data to obtain a detection frame of the target obstacle; and determine the orientation of the target obstacle based on the orientation of the detection frame of the target obstacle.

[0072] It should be noted that for details not disclosed in the obstacle orientation optimization device of the embodiment of the present disclosure, please refer to the details disclosed in the obstacle orientation optimization method of the embodiment of the present disclosure, and no further details will be given.

[0073] According to the obstacle orientation optimization device of the disclosed embodiment, an acquisition module acquires the orientation of a target obstacle within a preset time period, a determination module performs statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames to determine an image filtering mode, and an optimization module optimizes the orientation of the target obstacle based on the image filtering mode. Thus, by performing statistical analysis on the angle between the orientations of the target obstacle in two adjacent image frames within a preset time period, determining different image filtering modes, and optimizing the orientation of the target obstacle based on the image filtering modes, the device can improve the accuracy of determining the orientation of the target obstacle.

[0074] Corresponding to the above embodiments, the present disclosure also proposes a computer-readable storage medium.

[0075] The computer-readable storage medium of the present disclosure stores an obstacle orientation optimization program thereon, and when the obstacle orientation optimization program is executed by a processor, the above-mentioned obstacle orientation optimization method is implemented.

[0076] The computer-readable storage medium of the embodiment of the present disclosure can improve the accuracy of determining the target obstacle orientation by executing the above-mentioned obstacle orientation optimization method.

[0077] Corresponding to the above embodiment, the present disclosure also proposes an electronic device.

[0078] As shown in Figure 9, the electronic device 200 of an embodiment of the present disclosure includes a memory 210, a processor 220, and an obstacle orientation optimization program stored in the memory 210 and executable on the processor 220. When the processor 220 executes the obstacle orientation optimization program, the above-mentioned obstacle orientation optimization method is implemented.

[0079] According to the electronic device of the embodiment of the present disclosure, by executing the above-mentioned obstacle orientation optimization method, the accuracy of determining the target obstacle orientation can be improved.

[0080] It should be noted that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0081] It should be understood that various parts of the present disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0082] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present disclosure. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0083] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. Throughout the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0084] In this disclosure, unless otherwise expressly specified or limited, terms such as "installed," "connected," "connect," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integration; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components or interaction between two components, unless otherwise expressly limited. Those skilled in the art will understand the specific meanings of the above terms in this disclosure based on specific circumstances.

[0085] Although the embodiments of the present disclosure have been shown and described above, it is understood that the above embodiments are illustrative and are not to be construed as limitations on the present disclosure. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present disclosure.

Claims

1. A method for optimizing obstacle orientation, the method comprising: Get the direction of the target obstacle within the preset time; Performing statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images to determine an image filtering mode; The orientation of the target obstacle is optimized according to the image filtering mode.

2. The method for optimizing obstacle orientation according to claim 1, wherein: The statistical analysis of the orientation angle of the target obstacle in two adjacent frames of images includes: Determine the number of times the orientation angle is greater than a first preset angle within the preset time, and record it as a first number of times; Determine the number of times the orientation angle is less than a second preset angle within the preset time, and record it as a second number of times; Determine the statistical number of times that the orientation angle is greater than or equal to the second preset angle and less than or equal to the first preset angle within the preset time, and record it as a third statistical number.

3. The method for optimizing obstacle orientation according to claim 2, wherein: Determining the image filtering mode comprises: When the first statistical number is less than a first preset value and the third statistical number is zero, determining that the image filtering mode is the first image filtering mode; When the second statistical number is equal to the first preset value, determining that the image filtering mode is a second image filtering mode; Wherein, the first preset value is determined according to the number of image frames of the target obstacle within the preset time.

4. The method for optimizing obstacle orientation according to claim 3, wherein: The optimizing the orientation of the target obstacle according to the image filtering mode includes: When the image filtering mode is the first image filtering mode, the orientation of the target obstacle is optimized according to the average of the orientation angles of the target obstacle.

5. The method for optimizing obstacle orientation according to claim 3 or 4, wherein: The optimizing the orientation of the target obstacle according to the image filtering mode includes: When the image filtering mode is the second image filtering mode, a Kalman filtering algorithm is used to optimize the orientation of the target obstacle.

6. The method for optimizing obstacle orientation according to any one of claims 3 to 5, wherein: The optimizing the orientation of the target obstacle according to the image filtering mode includes: When the third statistical number is greater than zero, the direction of the target obstacle is output.

7. The method for optimizing obstacle orientation according to any one of claims 1 to 6, characterized in that: The obtaining of the direction of the target obstacle within the preset time includes: Acquire image data of target obstacles; Processing the image data to obtain a detection frame of the target obstacle; The orientation of the target obstacle is determined according to the orientation of the detection frame of the target obstacle.

8. An obstacle orientation optimization device, the device comprising: An acquisition module is used to obtain the direction of the target obstacle within a preset time; A determination module, used for performing statistical analysis on the orientation angle of the target obstacle in two adjacent frames of images to determine an image filtering mode; An optimization module is used to optimize the orientation of the target obstacle according to the image filtering mode.

9. A computer-readable storage medium having stored thereon an obstacle orientation optimization program, wherein when the obstacle orientation optimization program is executed by a processor, the obstacle orientation optimization method according to any one of claims 1 to 7 is implemented.

10. An electronic device comprising a memory, a processor, and an obstacle orientation optimization program stored in the memory and executable on the processor, wherein when the processor executes the obstacle orientation optimization program, the obstacle orientation optimization method according to any one of claims 1 to 7 is implemented.

Images