Pedestrian crossing area identification method, device and electronic equipment

Abstract

The embodiment of the application provides a kind of identification method, device and electronic equipment of pedestrian crossing area, comprising: obtaining the video stream of road surface that traffic road side camera shoots;Pedestrian crossing area is extracted from the road surface image of the video stream, and the coordinate range of the pedestrian crossing area in the road surface image is marked;The pedestrian crossing area is identified from the road surface image based on Fourier transform according to the coordinate range.

Description

Technical Field

[0001] This application relates to the field of vehicle networking technology, and in particular to a method, device and electronic device for identifying pedestrian crossing areas. Background Technology

[0002] With the development of vehicle-road cooperative technology, autonomous driving and intelligent connected driving need to focus on high-risk areas such as lanes, pedestrian crossings, and turning areas. Therefore, accurate identification of lanes, pedestrian crossings, and turning areas can effectively improve the computational efficiency of vehicle safety warnings.

[0003] In scenarios involving pedestrian crossing area identification, vehicle-road cooperative holographic intersection systems often use manual drawing to delineate pedestrian crossing areas, and then use scene algorithms to assist vehicle-road identification, reminders, and warnings within these areas. However, this method requires manual drawing in the early stages, which is time-consuming and labor-intensive, and it cannot respond promptly to scenarios where pedestrian crossings are redefined, resulting in poor accuracy in pedestrian crossing area detection. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, and electronic device for identifying pedestrian crossing areas, so as to solve the problem of poor accuracy in pedestrian crossing area detection.

[0005] To solve the above-mentioned technical problems, the embodiments of this application are implemented as follows:

[0006] In a first aspect, embodiments of this application provide a method for identifying pedestrian crossing areas, comprising: acquiring a video stream of the road surface captured by a roadside camera; extracting a road surface image containing a pedestrian crossing area from the video stream, and marking the coordinate range of the pedestrian crossing area in the road surface image; and identifying the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform.

[0007] Secondly, embodiments of this application provide a pedestrian crossing area identification device, comprising: an acquisition module for acquiring a video stream of the road surface captured by a roadside camera; an extraction module for extracting a road surface image containing a pedestrian crossing area from the video stream and marking the coordinate range of the pedestrian crossing area in the road surface image; and an identification module for identifying the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform.

[0008] Thirdly, embodiments of this application provide an electronic device, including a processor, a communication interface, a memory, and a communication bus; wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the pedestrian crossing area identification method steps as described in the first aspect.

[0009] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the pedestrian crossing area identification method steps as described in the first aspect.

[0010] As can be seen from the technical solutions provided in the embodiments of this application above, by acquiring the video stream of the road surface captured by the roadside camera, extracting the road surface image containing the pedestrian crossing area from the video stream, marking the coordinate range of the pedestrian crossing area in the road surface image, and identifying the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform, the pedestrian crossing area in the road surface can be automatically identified based on the video stream of the actual road surface. This eliminates the need for manual pre-drawing of the pedestrian crossing area, saving time and effort. When the pedestrian crossing is re-delineated, the pedestrian crossing area can also be re-identified based on the acquired video stream, improving the accuracy of pedestrian crossing area detection. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic flowchart of a first method for identifying pedestrian crossing areas provided in an embodiment of this application;

[0013] Figure 2 A second flowchart illustrating the method for identifying pedestrian crossing areas provided in an embodiment of this application;

[0014] Figure 3 A schematic diagram of a third method for identifying pedestrian crossing areas provided in an embodiment of this application;

[0015] Figure 4 A schematic diagram of the module composition of the pedestrian crossing area identification device provided in the embodiments of this application;

[0016] Figure 5This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0017] This application provides a method, apparatus, and electronic device for identifying pedestrian crossing areas, which solves the problem of poor accuracy in pedestrian crossing area detection.

[0018] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0019] For example, such as Figure 1 As shown in the figure, this application embodiment provides a method for identifying pedestrian crossing areas. The execution subject of this method can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers, and the server can be a server capable of identifying pedestrian crossing areas. The method for identifying pedestrian crossing areas specifically includes the following steps:

[0020] In step S101, the video stream of the road surface captured by the roadside camera is obtained.

[0021] Specifically, video streams of the road surface are collected by cameras located on the side of the traffic road near the pedestrian crossing. These video streams contain road surface information, pedestrians, and vehicles. When collecting road surface information, the cameras can be controlled to capture images of the road surface using the traffic light triggering mechanism. Specifically, for a given pedestrian crossing, the cameras are controlled to sample during the red and green light cycles, based on the traffic light signal control period. This allows for the acquisition of video streams with the maximum or minimum number of pedestrians and vehicles.

[0022] In step S103, the road surface image containing the pedestrian crossing area is extracted from the video stream, and the coordinate range of the pedestrian crossing area in the road surface image is marked.

[0023] Specifically, after acquiring the video stream of the road surface, the video stream is input into the DarkNet deep learning model. The DarkNet deep learning model performs contour recognition on moving objects in the video stream, obtaining multiple images output by the DarkNet deep learning model. Image filling is then performed on these multiple images. The image filling process involves filtering moving objects from the multiple images and color-filling the filtered areas to obtain a road surface image containing the pedestrian crossing area. Moving objects include, but are not limited to, pedestrians and vehicles. The resulting road surface image containing the pedestrian crossing area contains only the road surface and road markings. Based on the length, width, and color features of the pedestrian crossing lines, the road surface image is binarized and then traversed horizontally across the entire road surface image. The coordinate range of the binarized pedestrian crossing area within the road surface image is marked according to the pedestrian crossing features. This coordinate range includes, but is not limited to, the range from the upper left to the lower right coordinates of the pedestrian crossing area, or the range from the lower left to the upper right coordinates.

[0024] Furthermore, there is a strong color difference between the road surface and the road markings. In order to avoid the impact of lighting on the recognition accuracy of the pedestrian crossing area, the road surface and the road markings can be uniformly processed in grayscale across the entire image during or after image filling. This will eliminate the interference of lighting and improve the recognition accuracy of the pedestrian crossing area.

[0025] In step S105, pedestrian crossing areas are identified from the road surface image based on the coordinate range using Fourier transform.

[0026] Specifically, after determining the coordinate range of the pedestrian crossing area in the road image, the approximate range of the pedestrian crossing area is obtained. Then, based on the Fourier transform, the boundary image of the pedestrian crossing area is extracted from the road image according to the approximate range of the pedestrian crossing area. Specifically, the target image within the coordinate range of the road image is subjected to low-pass filtering through Fourier transform. The inverse Fourier transform is then performed on the low-pass filtered target image to identify the pedestrian crossing area from the target image, that is, to identify the sidewalk area in the road image, thus identifying the pedestrian crossing area from the road image.

[0027] The Fourier transform is as follows:

[0028] Suppose f(x,y) is a two-dimensional discrete signal of M times N obtained by equally spaced sampling from the road surface image in the above embodiment, and F(u,v) represents the Fourier transform of f(x,y). The coordinate system in which F(u,v) lies is called the frequency domain. Here, the two-dimensional discrete signal is the grayscale distribution value of the road surface image, M represents the number of columns in the road surface image, N represents the number of rows in the road surface image, x and y are discrete real variables, and u and v are discrete frequency variables. Then, the two-dimensional discrete Fourier transform is defined as follows:

[0029]

[0030]

[0031] In the processing of road surface images, if M equals N, the two-dimensional discrete Fourier transform is defined as:

[0032]

[0033]

[0034] in, It is a positive transform kernel. It is the inverse transform kernel, which defines the square of the spectrum of the two-dimensional discrete Fourier transform as the power spectrum P(u,v) of f(x,y), as follows:

[0035] P(u,v)=|F(u,v)| 2 =R 2 (u,v)+I 2 (u,v), where R 2 (u,v) represents the square of the real part of F(u,v), I 2 (u,v) represents the square of the imaginary part of F(u,v).

[0036] The power spectrum reflects the energy distribution of a two-dimensional discrete signal in the spatial frequency domain. Using the power spectrum P(u,v) as a threshold, when the identified pedestrian crossing area is unreasonable, the boundary of the pedestrian crossing area can be re-extracted by adjusting the power spectrum. In this embodiment, the power spectrum P(u,v) can range from [127, 255]. After low-pass filtering of the road surface image using Fourier transform, an inverse Fourier transform is performed to obtain the boundary feature image of the pedestrian crossing area in the road surface image, thus identifying the pedestrian crossing area in the road surface image. Using Fourier transform to process the road surface image reduces the computational complexity of traditional pedestrian crossing area identification, and using the entire road surface image for pedestrian crossing area boundary detection further improves the accuracy of pedestrian crossing area identification.

[0037] As can be seen from the technical solutions provided in the embodiments of this application above, by acquiring the video stream of the road surface captured by the roadside camera, extracting the road surface image containing the pedestrian crossing area from the video stream, marking the coordinate range of the pedestrian crossing area in the road surface image, and identifying the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform, the pedestrian crossing area in the road surface can be automatically identified based on the video stream of the actual road surface. This eliminates the need for manual pre-drawing of the pedestrian crossing area, saving time and effort. When the pedestrian crossing is re-delineated, the pedestrian crossing area can also be re-identified based on the acquired video stream, improving the accuracy of pedestrian crossing area detection and reducing the limitations of later maintenance.

[0038] For example, such as Figure 2 As shown in the figure, this application embodiment provides a method for identifying pedestrian crossing areas. The execution subject of this method can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers, and the server can be a server capable of identifying pedestrian crossing areas. The method for identifying pedestrian crossing areas specifically includes the following steps:

[0039] In step S201, the video stream of the road surface captured by the roadside camera is obtained.

[0040] In step S203, the road surface image containing the pedestrian crossing area is extracted from the video stream, and the coordinate range of the pedestrian crossing area in the road surface image is marked.

[0041] In step S205, pedestrian crossing areas are identified from the road surface image based on the coordinate range using Fourier transform.

[0042] In step S207, the Hough transform is used to verify whether the identified pedestrian crossing area is correctly delineated. If the verification fails, the pedestrian crossing area is re-identified. If the verification passes, the pedestrian crossing area is determined to be correctly delineated.

[0043] Specifically, verifying the correct delineation of the identified pedestrian crossing area using Hough transform includes: converting the image space pixel coordinates of the boundary line of the identified pedestrian crossing area to polar coordinates, where the angle of the polar coordinates is the tilt angle of the zebra crossing in the identified pedestrian crossing area; calculating the boundary coordinates of the markings of the identified pedestrian crossing area based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the tilt angle range of the zebra crossing; calculating the boundary coordinates of the zebra crossing in the identified pedestrian crossing area based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the attribute characteristics of the zebra crossing; redefining the target pedestrian crossing area based on the boundary coordinates of the markings, the boundary coordinates of the zebra crossing, and the tilt angle of the zebra crossing; and determining whether the difference between the pixel grayscale value within the target pedestrian crossing area and the pixel grayscale value of other areas outside the target pedestrian crossing area is less than a threshold. If it is less than the threshold, the verification passes; if it is not less than the threshold, the verification fails.

[0044] More specifically, after identifying the pedestrian crossing area from the road image using Fourier transform, the identified pedestrian crossing area can be further verified using Hough transform to check its rationality, i.e., to verify whether the pedestrian crossing area identified from the road image based on the coordinate range using Fourier transform is correctly delineated. Specifically, after distortion correction of the boundary image of the identified pedestrian crossing area, the entire image is traversed vertically to extract the pixel coordinates (x, y) of the boundary line of the pedestrian crossing area.

[0045] For pedestrian crossing areas, zebra crossings can be vertical or inclined. The slope of the zebra crossing is approximately in the range of [60, 90]. In this embodiment, the image space coordinates (x, y) of the pedestrian crossing area are converted to polar coordinates (ρ, θ). At this time, a point on the image space coordinates (linear coordinate system xy) of the pedestrian crossing area corresponds to a curve on the parameter space (polar coordinate system ρ-θ), namely a sine curve (ρ=xcosθ+ysinθ).

[0046] After converting the image spatial coordinates (x, y) of the pedestrian crossing area to polar coordinates (ρ, θ), θ is discretized, with values ​​of 60, 61, 62, ..., 90. ρ is calculated based on the image spatial coordinates (x, y) of the pedestrian crossing area and each angle θ. The frequency of approximately identical ρ values ​​within the same interval at each angle θ is counted. The image spatial coordinates (x, y) with the highest frequency of occurrence are identified as the boundary line of the pedestrian crossing area. Then, based on the image spatial coordinates (x, y) of the pedestrian crossing area and the width and length of the zebra crossing, the boundary coordinates of the zebra crossing (e.g., upper left and lower right coordinates) are calculated. Finally, based on the two side boundaries of the pedestrian crossing area, the boundary coordinates of the zebra crossing, and the angle of the zebra crossing, the four fixed-point pixel coordinates of the pedestrian crossing area are identified, and the pedestrian crossing area is redefined.

[0047] After redefining the pedestrian crossing area using Hough transform, it is determined whether the pixel grayscale values ​​within the redefined pedestrian crossing area are approximately or identical to the pixel grayscale values ​​of other areas outside the pedestrian crossing area. If they are approximately or identical, it indicates that the pedestrian crossing area identified based on Fourier transform has been correctly delineated. If the difference between the two grayscale values ​​exceeds a threshold, it indicates that the pedestrian crossing area identified based on Fourier transform has been unreasonably delineated. Steps S201 to S205 are then repeated to re-identify the pedestrian crossing area. Step S207 is then used to verify the pedestrian crossing area identified in step S205 again until the pedestrian crossing area identified based on Fourier transform passes the verification in step S207.

[0048] It is worth noting that steps S201 to S205 have the same or similar implementation methods, and the similarities can be referred to each other. The embodiments of this application will not be described again here.

[0049] As can be seen from the technical solutions provided in the embodiments of this application above, pedestrian crossing areas on the road surface can be automatically identified based on the video stream of the actual road surface, eliminating the need for manual pre-drawing of pedestrian crossing areas, saving time and effort. When pedestrian crossings are re-delineated, the pedestrian crossing areas can also be re-identified based on the acquired video stream, improving the accuracy of pedestrian crossing area detection. Furthermore, the accuracy and reasonableness of pedestrian crossing areas identified based on Fourier transform are verified using Hough transform. When the pedestrian crossing areas identified based on Fourier transform are unreasonable, the pedestrian crossing areas can be re-identified, further improving the accuracy and reliability of pedestrian crossing area detection.

[0050] For example, such as Figure 3As shown in the figure, this application embodiment provides a method for identifying pedestrian crossing areas. The execution subject of this method can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers, and the server can be a server capable of identifying pedestrian crossing areas. The method for identifying pedestrian crossing areas specifically includes the following steps:

[0051] In step S301, the video stream of the road surface captured by the roadside camera is obtained.

[0052] In step S303, the road surface image containing the pedestrian crossing area is extracted from the video stream, and the coordinate range of the pedestrian crossing area in the road surface image is marked.

[0053] In step S304, the road surface image is subjected to median filtering.

[0054] Specifically, in order to further eliminate impulse noise in the road surface image containing the pedestrian crossing area and thus improve the image quality of the road surface image containing the pedestrian crossing area, median filtering can be applied to the road surface image containing the pedestrian crossing area extracted from the video stream to further eliminate impulse noise in the image containing the pedestrian crossing area.

[0055] In step S305, pedestrian crossing areas are identified from the road surface image based on the coordinate range using Fourier transform.

[0056] Specifically, pedestrian crossing areas are identified from median-filtered road surface images based on the coordinate range using Fourier transform.

[0057] It is worth noting that steps S301 to S302 have the same or similar implementation as steps S101 to S102 in the above embodiments, and they can be referred to each other. The embodiments of this application will not be described again here.

[0058] As can be seen from the technical solutions provided in the embodiments of this application above, pedestrian crossing areas in the road surface can be automatically identified based on the video stream of the actual road surface, eliminating the need for manual pre-drawing of pedestrian crossing areas, saving time and effort. When pedestrian crossings are re-delineated, the pedestrian crossing areas can also be re-identified based on the acquired video stream, improving the accuracy of pedestrian crossing area detection. Furthermore, impulse noise in the road surface image of the pedestrian crossing area is further eliminated, thereby improving the image quality of the road surface image containing the pedestrian crossing area and further enhancing the accuracy of pedestrian crossing area detection.

[0059] Based on the same technical concept, and corresponding to the pedestrian crossing area identification method provided in the above embodiments, this application also provides a pedestrian crossing area identification device. Figure 4This application provides a schematic diagram of the module composition of a pedestrian crossing area identification device, which is used to perform... Figures 1 to 3 The method for identifying pedestrian crossing areas described, such as Figure 4 The pedestrian crossing area recognition device 400 includes: an acquisition module 401 for acquiring a video stream of the road surface captured by a roadside camera; an extraction module 402 for extracting a road surface image containing the pedestrian crossing area from the video stream and marking the coordinate range of the pedestrian crossing area in the road surface image; and a recognition module 403 for recognizing the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform.

[0060] As can be seen from the technical solutions provided by the embodiments of this application above, pedestrian crossing areas in the road surface can be automatically identified based on the video stream of the actual road surface collected, without the need for manual pre-drawing of pedestrian crossing areas, saving time and effort. When pedestrian crossings are re-delineated, pedestrian crossing areas can also be re-identified based on the collected video stream, improving the accuracy of pedestrian crossing area detection.

[0061] In one possible implementation, the identification module 403 is also used to verify whether the identified pedestrian crossing area is correctly delineated through Hough transform. If the verification fails, the pedestrian crossing area is re-identified. If the verification passes, the pedestrian crossing area is determined to be correctly delineated.

[0062] In one possible implementation, the recognition module 403 is further configured to convert the image space pixel coordinates of the boundary line of the recognized pedestrian crossing area into polar coordinates, wherein the angle of the polar coordinates is the tilt angle of the zebra crossing in the recognized pedestrian crossing area; calculate the boundary coordinates of the marking of the recognized pedestrian crossing area based on the image space pixel coordinates of the boundary line of the recognized pedestrian crossing area and the tilt angle range of the zebra crossing; calculate the boundary coordinates of the zebra crossing in the recognized pedestrian crossing area based on the image space pixel coordinates of the boundary line of the recognized pedestrian crossing area and the attribute features of the zebra crossing; redefine the target pedestrian crossing area based on the boundary coordinates of the marking, the boundary coordinates of the zebra crossing, and the tilt angle of the zebra crossing; determine whether the difference between the pixel grayscale value in the target pedestrian crossing area and the pixel grayscale value in other areas outside the target pedestrian crossing area is less than a threshold; if it is less than the threshold, the verification is passed; if it is not less than the threshold, the verification is failed.

[0063] In one possible implementation, the extraction module 402 is further configured to input the video stream into the DarkNet deep learning model, perform contour recognition on the moving objects in the video stream through the DarkNet deep learning model, and obtain multiple images output by the DarkNet deep learning model; perform image filling on the multiple images to obtain a road surface image, wherein the image filling indicates filtering moving objects in the multiple images and color filling the areas after filtering moving objects.

[0064] In one possible implementation, the recognition module 403 is also used to perform grayscale processing on the road surface image.

[0065] In one possible implementation, the recognition module 403 is further configured to perform low-pass filtering on the target image within the coordinate range of the road surface image through Fourier transform, and perform inverse Fourier transform on the target image after low-pass filtering to identify the pedestrian crossing area from the target image.

[0066] In one possible implementation, the recognition module 403 is also used to perform median filtering on the road surface image.

[0067] The pedestrian crossing area identification device provided in this application embodiment can realize the various processes in the embodiments corresponding to the above-mentioned pedestrian crossing area identification method. To avoid repetition, it will not be described again here.

[0068] It should be noted that the pedestrian crossing area identification device provided in this application embodiment and the pedestrian crossing area identification method provided in this application embodiment are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned pedestrian crossing area identification method, and the repeated parts will not be described again.

[0069] Corresponding to the pedestrian crossing area identification method provided in the above embodiments, based on the same technical concept, this application also provides an electronic device for performing the above-described pedestrian crossing area identification method. Figure 5 To illustrate the structure of an electronic device according to various embodiments of this application, as shown in the following diagrams... Figure 5 As shown. Electronic devices can vary considerably due to differences in configuration or performance, and may include one or more processors 501 and memory 502. Memory 502 may store one or more application programs or data. Memory 502 may be temporary or persistent storage. The application programs stored in memory 502 may include one or more modules (not shown), and each module may include a series of computer-executable instructions for the electronic device.

[0070] Furthermore, the processor 501 may be configured to communicate with the memory 502 and execute a series of computer-executable instructions stored in the memory 502 on the electronic device. The electronic device may also include one or more power supplies 503, one or more wired or wireless network interfaces 504, one or more input / output interfaces 505, and one or more keyboards 506.

[0071] Specifically, in this embodiment, the electronic device includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, the communication interface, and the memory communicate with each other through the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the various steps in the above method embodiments and has the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of this application will not be described again here.

[0072] This embodiment also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the steps in the above method embodiments and has the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of this application will not be described again here.

[0073] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) in which computer-usable program code is carried.

[0074] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0075] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0076] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0077] In a typical configuration, an electronic device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0078] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0079] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0080] It should also be noted that the terms "comprising," "carrying," or any other variations thereof are intended to cover non-exclusive carrying, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0081] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) in which computer-usable program code is carried.

[0082] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for identifying pedestrian crossing areas, characterized in that, The method for identifying the pedestrian crossing area includes: Acquire video streams of the road surface captured by roadside cameras; Extract road surface images containing pedestrian crossing areas from the video stream, and mark the coordinate range of the pedestrian crossing areas in the road surface images; The pedestrian crossing area is identified from the road surface image based on the coordinate range using Fourier transform; The Hough transform is used to verify whether the identified pedestrian crossing area is correctly delineated. If the verification fails, the pedestrian crossing area is re-identified. If the verification passes, the pedestrian crossing area is determined to be correctly delineated. The verification of whether the pedestrian crossing area identified by Hough transform is correctly delineated includes: The image space pixel coordinates of the boundary line of the identified pedestrian crossing area are converted into polar coordinates, and the angle of the polar coordinates is the tilt angle of the zebra crossing in the identified pedestrian crossing area. The coordinates of the marked boundary of the identified pedestrian crossing area are calculated based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the tilt angle range of the zebra crossing. The zebra crossing boundary coordinates in the identified pedestrian crossing area are calculated based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the attribute features of the zebra crossing. The target pedestrian crossing area is redefined based on the coordinates of the marking boundary, the coordinates of the zebra crossing boundary, and the tilt angle of the zebra crossing. Determine whether the difference between the pixel grayscale value within the target pedestrian crossing area and the pixel grayscale value in other areas outside the target pedestrian crossing area is less than a threshold. If it is less than the threshold, the verification passes; if it is not less than the threshold, the verification fails.

2. The method for identifying pedestrian crossing areas according to claim 1, characterized in that, The step of extracting the road surface image containing the pedestrian crossing area from the video stream includes: The video stream is input into the DarkNet deep learning model, and the DarkNet deep learning model is used to perform contour recognition on the moving objects in the video stream to obtain multiple images output by the DarkNet deep learning model. The road surface image is obtained by performing image filling on the multiple images, wherein the image filling indicates filtering the moving objects in the multiple images and color filling the areas after filtering the moving objects.

3. The method for identifying pedestrian crossing areas according to claim 2, characterized in that, After performing image filling on the multiple images to obtain the road surface image, the method further includes: The road surface image is processed into grayscale.

4. The method for identifying pedestrian crossing areas according to claim 1, characterized in that, Identifying the pedestrian crossing region from the road surface image based on the coordinate range using Fourier transform includes: The target image within the coordinate range of the road surface image is subjected to low-pass filtering through Fourier transform, and the target image after low-pass filtering is subjected to inverse Fourier transform to identify the pedestrian crossing area from the target image.

5. The method for identifying pedestrian crossing areas according to claim 1, characterized in that, After extracting the road surface image containing the pedestrian crossing area from the video stream, the method further includes: The road surface image is then subjected to median filtering.

6. A pedestrian crossing area identification device, characterized in that, The identification device for the pedestrian crossing area includes: The acquisition module is used to acquire video streams of the road surface captured by roadside cameras on the traffic route; The extraction module is used to extract road surface images containing pedestrian crossing areas from the video stream and mark the coordinate range of the pedestrian crossing areas in the road surface images; The recognition module is used to identify the pedestrian crossing area from the road surface image based on the coordinate range using Fourier transform; The Hough transform is used to verify whether the identified pedestrian crossing area is correctly delineated. If the verification fails, the pedestrian crossing area is re-identified. If the verification passes, the pedestrian crossing area is determined to be correctly delineated. The recognition module is further configured to convert the image space pixel coordinates of the boundary line of the recognized pedestrian crossing area into polar coordinates, wherein the angle of the polar coordinates is the tilt angle of the zebra crossing in the recognized pedestrian crossing area. The coordinates of the marked boundary of the identified pedestrian crossing area are calculated based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the tilt angle range of the zebra crossing. The zebra crossing boundary coordinates in the identified pedestrian crossing area are calculated based on the image space pixel coordinates of the boundary line of the identified pedestrian crossing area and the attribute features of the zebra crossing. The target pedestrian crossing area is redefined based on the coordinates of the marking boundary, the coordinates of the zebra crossing boundary, and the tilt angle of the zebra crossing. Determine whether the difference between the pixel grayscale value within the target pedestrian crossing area and the pixel grayscale value in other areas outside the target pedestrian crossing area is less than a threshold. If it is less than the threshold, the verification passes; if it is not less than the threshold, the verification fails.

7. An electronic device, characterized in that, The system includes a processor, a communication interface, a memory, and a communication bus; wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to implement the steps of the pedestrian crossing area identification method as described in any one of claims 1-5.

8. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for identifying pedestrian crossing areas as described in any one of claims 1-5.

Images