Parking space stop lever identification method, vehicle, storage medium and computer program product

Abstract

The invention provides a parking space stop lever identification method, a vehicle, a storage medium and a computer program product. The method comprises the following steps: acquiring image acquisition data and ultrasonic detection data of a target vehicle; determining visual position data of the parking space stop lever based on the image acquisition data, and determining ultrasonic position data of the parking space stop lever based on the ultrasonic detection data; the visual position data and the ultrasonic position data of the parking space stop lever are subjected to correlation processing, a position correlation result is obtained, and the position correlation result is used for determining whether the visual position data are matched with the ultrasonic position data or not; and determining that the visual position data is matched with the ultrasonic position data based on the position association result, and performing fusion processing on the visual position data and the ultrasonic position data to obtain a target recognition result. According to the invention, the technical problems of low accuracy and poor anti-interference performance of a parking space stop lever identification method provided in the prior art are solved.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a parking lever recognition method, a vehicle, a storage medium, and a computer program product. Background Art

[0002] In the automatic parking system of automated driving vehicles, accurate identification of parking levers is crucial for the safe driving of vehicles. The parking lever identification methods in the related art mainly rely on a single sensor technology, such as visual recognition technology and ultrasonic detection technology. Among them, the visual recognition technology captures images through a camera and uses image processing algorithms to analyze and identify parking levers, and the ultrasonic detection technology determines the distance of obstacles in front by emitting ultrasonic waves and receiving reflected waves. However, both parking lever identification methods in the related art have the problems of low accuracy and poor anti-interference, which makes it difficult to accurately identify parking levers.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present disclosure provide a parking lever recognition method, a vehicle, a storage medium and a computer program product to at least solve the technical problems of low accuracy and poor anti-interference performance of the parking lever recognition method provided in the related art.

[0005] According to one aspect of an embodiment of the present disclosure, a method for identifying a parking lever is provided, comprising: acquiring image acquisition data and ultrasonic detection data of a target vehicle; determining visual position data of the parking lever based on the image acquisition data, and determining ultrasonic position data of the parking lever based on the ultrasonic detection data; correlating the visual position data and ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data; in response to determining that the visual position data and the ultrasonic position data match each other based on the position correlation result, fusing the visual position data and the ultrasonic position data to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0006] Optionally, obtaining image acquisition data includes: obtaining real-time video stream data, wherein the real-time video stream data is acquired by an image acquisition device deployed at a preset position by the target vehicle; preprocessing the real-time video stream data to obtain image acquisition data, wherein the preprocessing includes at least one of the following: image denoising processing, contrast enhancement processing.

[0007] Optionally, obtaining ultrasonic detection data includes: obtaining a pulse reflection time corresponding to an ultrasonic sensor array in a target vehicle; determining target distance data based on the pulse reflection time; and performing denoising processing on the target distance data to obtain ultrasonic detection data.

[0008] Optionally, determining the visual position data of the parking barrier based on the image acquisition data includes: segmenting the image acquisition data to obtain a target recognition area; and determining the visual position data of the parking barrier in the target recognition area using an inverse perspective transformation algorithm.

[0009] Optionally, determining the ultrasonic position data of the parking barrier based on the ultrasonic detection data includes: clustering the ultrasonic detection data to obtain multiple candidate point clusters; screening the candidate point clusters using preset object shapes to obtain target point clusters; and determining the ultrasonic position data of the parking barrier based on the target point cluster.

[0010] Optionally, the method for identifying a parking lever in the embodiment of the present disclosure further includes: in response to determining based on the position association result that the visual position data and the ultrasonic position data do not match, deleting the visual position data.

[0011] Optionally, the parking lever identification method in the embodiment of the present disclosure further includes: performing feature extraction based on the target identification area to obtain a feature extraction result, wherein the feature extraction result is used to represent feature attribute information of the parking lever; and displaying the feature extraction result on a graphical user interface of the target vehicle.

[0012] According to another aspect of the embodiment of the present disclosure, a parking lever identification device is also provided, including: an acquisition module, used to acquire image acquisition data and ultrasonic detection data of a target vehicle; a determination module, used to determine visual position data of the parking lever based on the image acquisition data, and to determine ultrasonic position data of the parking lever based on the ultrasonic detection data; an association module, used to associate the visual position data and ultrasonic position data of the parking lever to obtain a position association result, wherein the position association result is used to determine whether the visual position data matches the ultrasonic position data; a processing module, used to, in response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, fuse the visual position data and the ultrasonic position data to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0013] Optionally, the acquisition module is further used to: pre-process the real-time video stream data to obtain image acquisition data, wherein the pre-processing includes at least one of the following: image denoising processing and contrast enhancement processing.

[0014] Optionally, the acquisition module is also used to: acquire the pulse reflection time corresponding to the ultrasonic sensor array in the target vehicle; determine the target distance data based on the pulse reflection time; and perform denoising on the target distance data to obtain ultrasonic detection data.

[0015] Optionally, the determination module is further used to: segment the image acquisition data to obtain a target recognition area; and determine the visual position data of the parking lever in the target recognition area using an inverse perspective transformation algorithm.

[0016] Optionally, the determination module is further used to: cluster the ultrasonic detection data to obtain multiple candidate point clusters; filter the candidate point clusters using preset object shapes to obtain target point clusters; and determine the ultrasonic position data of the parking lever based on the target point clusters.

[0017] According to one embodiment of the present disclosure, a vehicle is further provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the method for identifying a parking lever in the embodiment of the present disclosure.

[0018] According to one embodiment of the present disclosure, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored executable program, wherein when the executable program is running, the device where the storage medium is located is controlled to execute the parking lever identification method in the embodiment of the present disclosure.

[0019] According to one embodiment of the present disclosure, a computer program product is further provided. The computer program product includes computer instructions. When the computer instructions are executed by a processor, the method for identifying a parking lever in the embodiment of the present disclosure is implemented.

[0020] In the embodiment of the present disclosure, by acquiring image acquisition data and ultrasonic detection data of the target vehicle, and determining the visual position data of the parking barrier based on the image acquisition data, and determining the ultrasonic position data of the parking barrier based on the ultrasonic detection data, the visual position data and the ultrasonic position data of the parking barrier are then correlated to obtain a position correlation result, and finally, in response to determining that the visual position data and the ultrasonic position data match each other based on the position correlation result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, thereby achieving the purpose of accurately identifying and locating the parking barrier, thereby realizing the technical effect of improving the recognition accuracy and anti-interference ability of the parking barrier, and thereby solving the technical problems of low accuracy and poor anti-interference ability of the parking barrier recognition method provided in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings described herein are used to provide a further understanding of the present disclosure and constitute a part of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure. In the drawings:

[0022] Figure 1 is a flow chart of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0023] Figure 2 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0024] Figure 3 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0025] Figure 4 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0026] Figure 5 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0027] Figure 6 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure;

[0028] Figure 7 It is a structural block diagram of a parking lever recognition device according to one embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present disclosure, the technical scheme in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present disclosure.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0031] The parking lever recognition method in the related art mainly relies on a single sensor technology, such as visual recognition technology and ultrasonic detection technology. The visual recognition technology captures images through a camera and uses image processing algorithms to analyze and recognize parking levers, while the ultrasonic detection technology determines the distance of obstacles ahead by emitting ultrasonic waves and receiving reflected waves. However, both parking lever recognition methods in the related art have the problems of low accuracy and poor anti-interference, making it difficult to accurately recognize parking levers.

[0032] Specifically, visual recognition technology is easily affected by environmental factors such as light, shadows, and water stains, and the recognition accuracy will be reduced at night or in low light conditions; ultrasonic detection technology is easily affected by environmental noise, which leads to the inability to distinguish obstacles when there is a lot of noise in the vehicle's environment, and it is difficult to accurately identify the shape of the parking lever.

[0033] According to an embodiment of the present disclosure, a method embodiment of a method for identifying a parking lever is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] The method embodiment can be executed in an electronic device or a similar computing device including a memory and a processor. Taking running on a computer terminal as an example, the computer terminal may include one or more processors (processors may include but are not limited to central processing units (CPU), graphics processing units (GPU), digital signal processing (DSP) chips, microprocessors (MCU), programmable logic devices (Field Programmable Gate Array, FPGA), neural network processors (Neural-network Processor Unit, NPU), tensor processors (TensorProcessing Unit, TPU), artificial intelligence (Artificial Intelligence, AI) type processors and other processing devices) and a memory for storing data. Optionally, the above-mentioned computer terminal may also include a transmission device, an input and output device, and a display device for communication functions. It can be understood by those of ordinary skill in the art that the above-mentioned structural description is only for illustration, and it does not limit the structure of the above-mentioned computer terminal. For example, the computer terminal may also include more or fewer components than the above-mentioned structural description, or have a configuration different from the above-mentioned structural description.

[0035] The memory can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the method for identifying the parking lever in the embodiment of the present disclosure. The processor executes various functional applications and data processing by running the computer program stored in the memory, that is, the above-mentioned method for identifying the parking lever is realized. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory may further include a memory remotely arranged relative to the processor, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0036] The transmission device is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0037] The display device may be, for example, a touch screen type liquid crystal display (LCD) and a touch display (also referred to as a "touch screen" or "touch display screen"). The liquid crystal display may enable a user to interact with a user interface of a mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), and a user may interact with the GUI by finger contacts and / or gestures on a touch-sensitive surface, wherein the human-computer interaction functions here may optionally include the following interactions: creating web pages, drawing, word processing, making electronic documents, games, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital videos, playing digital music and / or web browsing, etc. The executable instructions for executing the above human-computer interaction functions are configured / stored in a computer program product or readable storage medium executable by one or more processors.

[0038] Figure 1 is a flow chart of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 1 As shown, the method comprises the following steps:

[0039] Step S11, acquiring image acquisition data and ultrasonic detection data of the target vehicle;

[0040] Step S12, determining visual position data of the parking barrier based on the image acquisition data, and determining ultrasonic position data of the parking barrier based on the ultrasonic detection data;

[0041] Step S13, correlating the visual position data and the ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data;

[0042] Step S14, in response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0043] The above-mentioned image acquisition data refers to the image data obtained after preprocessing the real-time video stream of the environment continuously captured by the camera module on the autonomous driving vehicle, including but not limited to roads, other vehicles, pedestrians, traffic signs and various fixed obstacles, such as parking levers, curbs, flower bed boundaries, etc. Among them, the above-mentioned camera module can be located in the center of the front and rear bumpers of the vehicle, and at the left and right rearview mirrors of the vehicle.

[0044] The above-mentioned ultrasonic detection data refers to the environmental obstacle distance information collected by the ultrasonic sensor array installed on the autonomous driving vehicle during operation, which is used to provide detailed distance information of the vehicle's surrounding environment. Especially in low visibility, at night or when visual sensors are limited, ultrasonic detection data can help the vehicle perceive the existence of obstacles and the relative position of the obstacles and the vehicle itself.

[0045] For example, during the vehicle parking process, the environmental obstacle information is a parking lever. It is assumed that four wide-angle lenses are installed in the center of the front and rear bumpers of the autonomous driving vehicle, as well as the left and right rearview mirrors of the vehicle, and an ultrasonic sensor is configured around each wide-angle lens to form multi-point detection. Each ultrasonic sensor is responsible for distance measurement at a specific angle, thereby enhancing the accurate capture of the parking lever position.

[0046] The visual position data of the parking lever mentioned above refers to the position of the obstacle in the image identified by the image acquisition data captured by the camera module on the autonomous vehicle. By using image processing and visual positioning algorithms, the two-dimensional position of the obstacle in the image is converted into relative position information in three-dimensional space, and the three-dimensional position of the obstacle relative to the vehicle coordinate system is output, that is, the position data of the potential visual obstacle.

[0047] The ultrasonic position data mentioned above refers to the straight-line distance of the obstacle detected by the ultrasonic sensor relative to the ultrasonic sensor. The ultrasonic sensor array is distributed around the vehicle, and each sensor is responsible for detecting the distance of the obstacle in a specific direction. By analyzing the round-trip time of the ultrasonic pulse signal, the distance between the ultrasonic sensor and the obstacle can be calculated. Subsequently, combined with the position information of the ultrasonic sensor on the vehicle and the vehicle's coordinate system, the distance information between the ultrasonic sensor and the obstacle can be converted into the potential position coordinates of the obstacle in three-dimensional space, that is, the position data of the potential ultrasonic gear lever.

[0048] The above-mentioned association processing is data association, which uses a position matching method to correspond and associate the potential visual obstacles identified by the camera module with the potential ultrasonic obstacles detected by the ultrasonic sensor array, ensuring that the same obstacle detected by the two sensors can be correctly identified and processed at the data level.

[0049] The above-mentioned fusion processing refers to the process of fusing the ultrasonic position data and visual position data of obstacles using methods such as weighted averaging and deep learning models.

[0050] Specifically, during the parking process, when the environmental obstacle information is a parking barrier, the parking barrier needs to be identified and located. First, the visual position data of the parking barrier is obtained based on the camera module configured by the vehicle, and the ultrasonic position data of the parking barrier is obtained based on the ultrasonic sensor; secondly, the visual position data of the parking barrier and the ultrasonic position data of the parking barrier are correlated to determine whether the visual position data and the ultrasonic position data of the parking barrier match; finally, when the position data match each other, the visual position data and the ultrasonic position data of the parking barrier are fused to obtain the actual position information of the parking barrier, and then the driving decision of the vehicle is made based on the actual position information of the parking barrier.

[0051] Based on the above steps S11 to S14, by acquiring the image acquisition data and ultrasonic detection data of the target vehicle, and determining the visual position data of the parking lever based on the image acquisition data, and determining the ultrasonic position data of the parking lever based on the ultrasonic detection data, the visual position data and the ultrasonic position data of the parking lever are then correlated to obtain a position correlation result, and finally, in response to determining that the visual position data and the ultrasonic position data match each other based on the position correlation result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, thereby achieving the purpose of accurately identifying and locating the parking lever, thereby achieving the technical effect of improving the recognition accuracy and anti-interference performance of the parking lever, and thereby solving the technical problems of low accuracy and poor anti-interference performance of the parking lever recognition method provided in the related art.

[0052] The following is a further introduction to the method for identifying the parking lever in the embodiment of the present disclosure.

[0053] Figure 2 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 2 As shown, the method comprises the following steps:

[0054] Step S21, acquiring real-time video stream data, wherein the real-time video stream data is acquired by an image acquisition device deployed at a preset position of the target vehicle;

[0055] Step S22, preprocessing the real-time video stream data to obtain image acquisition data, wherein the preprocessing includes at least one of the following: image denoising processing and contrast enhancement processing.

[0056] The above-mentioned image acquisition equipment includes but is not limited to wide-angle lenses, fisheye lenses, panoramic lenses, etc.

[0057] The above-mentioned preprocessing refers to a series of image processing operations performed on real-time video stream data to improve image quality and enhance the accuracy and efficiency of subsequent image analysis. Preprocessing methods include but are not limited to image denoising, contrast enhancement, image correction, color conversion, etc. Among them, common image denoising algorithms include but are not limited to mean filtering, Gaussian filtering, bilateral filtering, etc.; contrast enhancement processing methods include but are not limited to histogram equalization, adaptive contrast enhancement, retinal model algorithm, etc.

[0058] Exemplarily, a camera module consisting of a wide-angle lens is set. Since the viewing angle of a wide-angle lens is usually greater than 120 degrees, a total of 4 wide-angle lenses are installed in the center of the front and rear bumpers of the vehicle, as well as the left and right rearview mirrors of the vehicle, to ensure that the field of view covers the parking area around the vehicle. The above-mentioned camera module is used to obtain real-time video stream data, and the acquired real-time video stream data is subjected to image denoising using a bilateral filtering algorithm to reduce image clutter caused by light changes, electronic device noise, etc., so as to make the image clearer and smoother. Subsequently, a histogram equalization algorithm is used to perform contrast enhancement processing to enhance the edges and details of obstacles in the image, so that the obstacles are more prominent in the image, thereby obtaining image acquisition data.

[0059] For example, when a vehicle drives out of a backlit shadow area, the parking lever in the real-time video stream data acquired by the camera module may be blurred in the image due to insufficient light. After preprocessing, the parking lever feature is displayed more clearly in the image and can be accurately identified even under backlit conditions.

[0060] Based on the above optional embodiments, by acquiring real-time video stream data and preprocessing the real-time video stream data to obtain image acquisition data, the image quality in poor lighting conditions or complex environments can be effectively improved, thereby improving the accuracy of the recognition algorithm.

[0061] Figure 3 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 3 As shown, the method comprises the following steps:

[0062] Step S31, obtaining the pulse reflection time corresponding to the ultrasonic sensor array in the target vehicle;

[0063] Step S32, determining target distance data based on the pulse reflection time;

[0064] Step S33, denoising the target distance data to obtain ultrasonic detection data.

[0065] The above-mentioned pulse reflection time refers to the time it takes for the ultrasonic wave to be emitted from the sensor, reach the target object, and then be reflected back to the sensor. Specifically, when the ultrasonic pulse emitted by the ultrasonic sensor encounters a parking lever or other obstacle, it will return along the original path and be received by the ultrasonic sensor. The time it takes for the ultrasonic pulse to travel back and forth twice, which is recorded by the ultrasonic sensor, is the pulse reflection time.

[0066] The above target distance data refers to the actual distance from the ultrasonic sensor to the parking lever or other obstacles. The target distance data can be calculated by the following formula, where c is the ultrasonic sound speed, usually 340 m / s, t is the pulse reflection time, and D is the target distance data.

[0067]

[0068] For example, a total of 4 ultrasonic sensors are installed in the center of the front and rear bumpers of the vehicle, as well as the left and right rearview mirrors of the vehicle. When the vehicle enters the underground parking lot for parking, the ultrasonic sensor emits ultrasonic pulses in all directions. Assuming that at a certain moment, the ultrasonic sensor detects that there is a parking lever in front of the vehicle, and the recorded ultrasonic pulse reflection time is 0.0068 seconds, it can be calculated that the distance from the ultrasonic sensor to the parking lever at this time is 1.156 meters, that is, the target distance data is 1.156 meters. To ensure the accuracy of the data, the acquired target distance data is denoised to obtain ultrasonic detection data to eliminate the influence of environmental noise on the ultrasonic detection data.

[0069] Based on the above optional embodiment, the pulse reflection time corresponding to the ultrasonic sensor array in the target vehicle is obtained, and the target distance data is determined based on the pulse reflection time, and then the target distance data is denoised to obtain ultrasonic detection data, which can improve the accuracy of distance measurement and at the same time improve the stability and reliability of ultrasonic detection data in complex environments.

[0070] Figure 4 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 4 As shown, the method comprises the following steps:

[0071] Step S41, segmenting the image acquisition data to obtain a target recognition area;

[0072] Step S42: In the target recognition area, the visual position data of the parking lever is determined using an inverse perspective transformation algorithm.

[0073] The above-mentioned segmentation process refers to the use of image processing technology to separate a specific area in the image, that is, an area that may contain a parking barrier, from the background. Segmentation processing generally includes but is not limited to threshold segmentation, edge detection, region growing, machine learning and other methods. Through segmentation processing, the vehicle system can focus on analyzing and identifying the part of the image related to the parking barrier, reduce the amount of calculation, and improve recognition efficiency and accuracy.

[0074] The above-mentioned Inverse Perspective Mapping (IPM) algorithm refers to an image processing technology that converts the perspective image captured by the camera into a top-down view, which is used to convert the perspective projection in the image into an orthogonal projection to eliminate perspective distortion. The IPM algorithm is often used in image correction, robot navigation and autonomous vehicles to analyze and process the actual physical position and size in the image.

[0075] Exemplarily, during a parking process of a target vehicle, a threshold segmentation algorithm is used to segment the image acquisition data obtained by the camera module to separate the image area containing the parking lever, and then in the area containing the parking lever, the IPM algorithm is used to convert the perspective image of the parking lever captured by the camera into a top view, thereby converting the position information of the parking lever in the image coordinate system to the vehicle coordinate system, and then determining the visual position data of the parking lever, including its specific position and size relative to the vehicle.

[0076] Based on the above optional embodiment, the image acquisition data is segmented to obtain a target recognition area, and then in the target recognition area, the visual position data of the parking lever is determined using an inverse perspective transformation algorithm. This can effectively distinguish the parking lever from the background, reduce the interference of non-target areas on the recognition process, and facilitate the positioning of the parking lever in the vehicle coordinate system, thereby improving the accuracy and reliability of recognition.

[0077] Figure 5 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 5 As shown, the method comprises the following steps:

[0078] Step S51, clustering the ultrasonic detection data to obtain multiple candidate point clusters;

[0079] Step S52, screening the candidate point clusters using the preset object shape to obtain the target point cluster;

[0080] Step S53: determining the ultrasonic position data of the parking barrier based on the target point cluster.

[0081] The above clustering process refers to an unsupervised learning method in the data analysis process, which is used to automatically group the elements in the data set into different point clusters according to the principle of similarity without giving specific category information in advance. Common clustering algorithms include hierarchical clustering algorithm, K-means clustering algorithm and density-based clustering algorithm.

[0082] For example, during the parking process of the target vehicle, the ultrasonic sensor array continuously transmits and receives echoes, generating a series of ultrasonic detection data. The ultrasonic detection data includes not only parking levers, but also other irrelevant environmental features, such as walls, other vehicles or pedestrians. In order to accurately identify the parking lever, the coordinate system of the vehicle is combined, the positions of the surrounding ultrasonic points are calculated by direct measurement and triangulation, and clustering methods are used to form point clusters. Based on the comparison of the shape of the point clusters with the possible shapes of the levers, the point clusters that are not levers are excluded, and the actual positions of multiple potential parking levers are obtained.

[0083] Based on the above optional embodiment, the ultrasonic detection data is clustered to obtain multiple candidate point clusters, and the candidate point clusters are screened using preset object shapes to obtain target point clusters, and then the ultrasonic position data of the parking lever is determined based on the target point cluster, which can reduce interference from non-target obstacles and thus reduce the possibility of misidentification.

[0084] Optionally, the method for identifying a parking lever in the embodiment of the present disclosure further includes: in response to determining based on the position association result that the visual position data and the ultrasonic position data do not match, deleting the visual position data.

[0085] For example, in the barrier detection method based on vision and ultrasonic fusion, the data processing unit will continuously compare the visual position data with the ultrasonic position data to check whether there is consistency and correlation between the two. When the difference in the position information of the two exceeds the preset threshold or standard, that is, the visual position data does not match the ultrasonic position data, it means that there may be insufficient light, shadows, light reflections, etc., which causes the visual position data to be unable to reflect the accurate position of the parking barrier, and thus the visual position data needs to be deleted.

[0086] Based on the above optional embodiments, in response to determining based on the position association result that the visual position data does not match the ultrasonic position data, deleting the visual position data can avoid erroneous judgments caused by errors in single sensor data, thereby improving recognition accuracy and anti-interference.

[0087] Optionally, the parking lever identification method in the embodiment of the present disclosure further includes: performing feature extraction based on the target identification area to obtain a feature extraction result, wherein the feature extraction result is used to represent feature attribute information of the parking lever; and displaying the feature extraction result on a graphical user interface of the target vehicle.

[0088] The above-mentioned characteristic attribute information refers to various quantitative data or descriptive information extracted from the target recognition area that can characterize the characteristics of the parking lever, including but not limited to shape, color, texture, position coordinates, size information, number or identification information, etc.

[0089] For example, during the parking process, the camera module of the target vehicle captures the potential parking lever area, and uses edge detection, color separation, text recognition, and shape recognition algorithms to obtain the shape, color, texture, and any specific logo or number of the parking lever. By comparing with the built-in parking lever feature data of the vehicle, the feature information is further verified and refined to generate feature extraction results. Subsequently, the obtained feature extraction results are transmitted to the vehicle's graphical user interface, the identification position and shape of the parking lever are highlighted, and the relative distance between the parking lever and the vehicle is intuitively displayed in the form of a dynamic arrow or distance reading, ensuring that the driver can quickly understand the current parking situation.

[0090] Based on the above optional embodiment, feature extraction is performed based on the target recognition area to obtain feature extraction results, and then the feature extraction results are displayed on the graphical user interface of the target vehicle, which can enhance the driver's environmental perception ability and provide environmental decision support for the driver.

[0091] Figure 6 is a schematic diagram of a method for identifying a parking lever according to one embodiment of the present disclosure, such as Figure 6 As shown, the method includes the following steps: first, using a camera module and an ultrasonic sensor to collect data to obtain real-time video stream data and pulse reflection time; second, for the real-time video stream data, firstly perform data preprocessing to obtain image acquisition data, and then perform image segmentation on the image acquisition data to determine the target recognition area, and then obtain visual position information according to the target recognition area; for the pulse reflection time, calculate the target distance data based on the pulse reflection time, and then perform denoising on the target distance data to obtain ultrasonic detection data, and perform clustering based on the ultrasonic detection data to obtain target point clusters, thereby obtaining ultrasonic position information based on the target point clusters; finally, perform position association on the visual position data and the ultrasonic position data to obtain a position association result, and then perform fusion processing on the position association result to obtain a target recognition result, thereby providing a driving decision for the driver based on the target recognition result.

[0092] For example, a total of 4 wide-angle lenses are installed in the center of the front and rear bumpers of the vehicle, as well as the positions of the left and right rearview mirrors of the vehicle, and 1 ultrasonic sensor is installed around the wide-angle lenses respectively, so as to form multi-point detection. During the parking process, firstly, the camera module and ultrasonic sensor are used to collect data to obtain real-time video stream data and pulse reflection time; secondly, Gaussian filtering and contrast enhancement processing are performed on the real-time video stream data to reduce noise interference and highlight the parking lever to obtain image acquisition data, and then the edge segmentation algorithm is used to segment the image acquisition data to separate the image area containing the parking lever, and the IPM algorithm is used to convert the perspective image of the parking lever captured by the camera into a top view, and the position information of the parking lever in the image coordinate system is converted to the vehicle coordinate system, so that the visual position data of the parking lever can be determined; for the pulse reflection time, the target distance data is calculated based on the product of the ultrasonic sound velocity and the pulse reflection time, and then the target distance data is denoised using Gaussian filtering to obtain ultrasonic detection data. At the same time, combined with the coordinate system where the vehicle is located, the positions of the surrounding ultrasonic points are calculated by direct measurement and triangulation, and a clustering method is used to form a point cluster. According to the comparison between the shape of the point cluster and the possible shape of the gear lever, the point cluster that is not a gear lever is excluded, thereby obtaining the ultrasonic position data of the parking lever; finally, based on the visual position data and ultrasonic position data obtained above, position association is performed, and when the difference in the position information of the two exceeds a preset threshold or standard, the visual position data is deleted. The parking lever recognition method disclosed in the present invention combines visual recognition and ultrasonic detection, overcomes the limitations of a single technology, not only improves the recognition accuracy, but also enhances the anti-interference ability of the parking lever recognition method under different light and noise conditions.

[0093] Figure 7 is a structural block diagram of a parking lever recognition device according to one embodiment of the present disclosure, such as Figure 7 As shown, the device comprises:

[0094] An acquisition module 701 is used to acquire image acquisition data and ultrasonic detection data of a target vehicle;

[0095] A determination module 702, configured to determine visual position data of the parking barrier based on the image acquisition data, and to determine ultrasonic position data of the parking barrier based on the ultrasonic detection data;

[0096] The association module 703 is used to associate the visual position data and the ultrasonic position data of the parking lever to obtain a position association result, wherein the position association result is used to determine whether the visual position data matches the ultrasonic position data;

[0097] The processing module 704 is used to respond to the determination based on the position association result that the visual position data and the ultrasonic position data match each other, fuse the visual position data and the ultrasonic position data, and obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0098] Optionally, the acquisition module 701 is further used to: pre-process the real-time video stream data to obtain image acquisition data, wherein the pre-processing includes at least one of the following: image denoising processing and contrast enhancement processing.

[0099] Optionally, the acquisition module 701 is further used to: acquire the pulse reflection time corresponding to the ultrasonic sensor array in the target vehicle; determine the target distance data based on the pulse reflection time; and perform denoising on the target distance data to obtain ultrasonic detection data.

[0100] Optionally, the determination module 702 is further used to: segment the image acquisition data to obtain a target recognition area; and determine the visual position data of the parking lever in the target recognition area using an inverse perspective transformation algorithm.

[0101] Optionally, the determination module 702 is further used to: cluster the ultrasonic detection data to obtain multiple candidate point clusters; filter the candidate point clusters using preset object shapes to obtain target point clusters; and determine the ultrasonic position data of the parking lever based on the target point clusters.

[0102] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0103] According to another aspect of an embodiment of the present disclosure, a vehicle is further provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the method for identifying a parking lever in an embodiment of the present disclosure.

[0104] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0105] S1, obtaining image acquisition data and ultrasonic detection data of the target vehicle;

[0106] S2, determining visual position data of the parking barrier based on the image acquisition data, and determining ultrasonic position data of the parking barrier based on the ultrasonic detection data;

[0107] S3, correlating the visual position data and the ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data;

[0108] S4, in response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0109] According to another aspect of an embodiment of the present disclosure, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored executable program, wherein when the executable program runs, the device where the storage medium is located is controlled to execute the parking lever identification method in the embodiment of the present disclosure.

[0110] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0111] S1, obtaining image acquisition data and ultrasonic detection data of the target vehicle;

[0112] S2, determining visual position data of the parking barrier based on the image acquisition data, and determining ultrasonic position data of the parking barrier based on the ultrasonic detection data;

[0113] S3, correlating the visual position data and the ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data;

[0114] S4, in response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0115] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.

[0116] According to another aspect of the embodiments of the present disclosure, a computer program product is further provided. The computer program product includes computer instructions. When the computer instructions are executed by a processor, the method for identifying a parking lever in the embodiments of the present disclosure is implemented.

[0117] Optionally, in this embodiment, the computer program product may be configured as a computer program for executing the following steps:

[0118] S1, obtaining image acquisition data and ultrasonic detection data of the target vehicle;

[0119] S2, determining visual position data of the parking barrier based on the image acquisition data, and determining ultrasonic position data of the parking barrier based on the ultrasonic detection data;

[0120] S3, correlating the visual position data and the ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data;

[0121] S4, in response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

[0122] The serial numbers of the above-mentioned embodiments of the present disclosure are only for description and do not represent the advantages or disadvantages of the embodiments.

[0123] In the above embodiments of the present disclosure, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0124] In the several embodiments provided in the present disclosure, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0125] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0126] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0127] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0128] The above is only a preferred embodiment of the present disclosure. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present disclosure. These improvements and modifications should also be regarded as the scope of protection of the present disclosure.

Claims

1. A method for identifying a parking lever, characterized in that: include: Acquire image acquisition data and ultrasonic detection data of the target vehicle; Determining visual position data of the parking barrier based on the image acquisition data, and determining ultrasonic position data of the parking barrier based on the ultrasonic detection data; Correlating the visual position data and the ultrasonic position data of the parking lever to obtain a position correlation result, wherein the position correlation result is used to determine whether the visual position data matches the ultrasonic position data; In response to determining that the visual position data and the ultrasonic position data match each other based on the position association result, the visual position data and the ultrasonic position data are fused to obtain a target recognition result, wherein the target recognition result is used to make driving decisions for the target vehicle.

2. The method for identifying a parking lever according to claim 1, characterized in that: Acquiring the image acquisition data includes: Acquiring real-time video stream data, wherein the real-time video stream data is acquired by an image acquisition device deployed at a preset position of the target vehicle; The real-time video stream data is preprocessed to obtain the image acquisition data, wherein the preprocessing includes at least one of the following: image denoising processing and contrast enhancement processing.

3. The method for identifying a parking lever according to claim 1, characterized in that: Acquiring the ultrasonic detection data includes: Obtaining a pulse reflection time corresponding to an ultrasonic sensor array in the target vehicle; Determining target distance data based on the pulse reflection time; The target distance data is subjected to denoising processing to obtain the ultrasonic detection data.

4. The method for identifying a parking lever according to claim 1, characterized in that: Determining the visual position data of the parking lever based on the image acquisition data includes: Segmenting the image acquisition data to obtain a target recognition area; In the target recognition area, the visual position data of the parking lever is determined using an inverse perspective transformation algorithm.

5. The method for identifying a parking lever according to claim 1, characterized in that: Determining the ultrasonic position data of the parking lever based on the ultrasonic detection data includes: Performing clustering processing on the ultrasonic detection data to obtain multiple candidate point clusters; The candidate point clusters are screened using a preset object shape to obtain a target point cluster; The ultrasonic position data of the parking lever is determined based on the target point cluster.

6. The method for identifying a parking lever according to claim 1, characterized in that: The method further comprises: In response to determining that the visual position data does not match the ultrasonic position data based on the position association result, the visual position data is deleted.

7. The method for identifying a parking lever according to claim 4, characterized in that: The method further comprises: Performing feature extraction based on the target recognition area to obtain a feature extraction result, wherein the feature extraction result is used to represent feature attribute information of the parking lever; The feature extraction results are displayed on a graphical user interface of the target vehicle.

8. A vehicle, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for identifying a parking lever according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is run, the device where the storage medium is located is controlled to execute the parking lever recognition method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The computer program product comprises computer instructions, which, when executed by a processor, implement the method for identifying a parking lever as claimed in any one of claims 1 to 7.

Images