Ultrasonic data processing method and device, equipment and storage medium

By converting ultrasonic data into visual images and labeling target areas, the data set is generated, and the problem of poor performance of traditional ultrasonic data processing methods in complex environments is solved, and the accuracy and robustness of parking space detection are improved.

CN119064906APending Publication Date: 2024-12-03ZHEJIANG GEELY HLDG GRP CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411184436.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

Traditional ultrasonic data processing methods perform poorly in complex environments, have high computational complexity, rely on manual design features, are difficult to deal with complex scenarios and corner cases, and have poor real-time performance.

Method used

The ultrasonic data is converted into intuitive visual images, and the target area is marked in the image to generate a data set for training deep learning models for parking space detection.

Benefits of technology

The computational complexity of deep learning model training is reduced, the algorithm generalization ability is improved, and the accuracy and robustness of parking space detection are enhanced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119064906A_ABST
    Figure CN119064906A_ABST
Patent Text Reader

Abstract

The invention discloses an ultrasonic data processing method and device, equipment and a storage medium. The method comprises the following steps: acquiring ultrasonic data; the obtained ultrasonic data are preprocessed; converting the preprocessed ultrasonic data into a visual image; marking a target area in the visual image; and arranging the visual image after the target area is marked, and generating a data set. According to the method, the ultrasonic data is converted into the visual image and marked, so that the original non-visual ultrasonic data becomes visual and easy to understand, the problem that the ultrasonic data is difficult to be directly used for deep learning model training is solved, the calculation complexity of subsequent deep learning model training is reduced, and the algorithm generalization ability is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an ultrasonic data processing method, device, equipment and storage medium. Background Art

[0002] With the continuous development of autonomous driving technology, automatic parking systems have become one of the important functions of modern smart cars. In the process of automatic parking, accurate location detection and obstacle recognition are key technologies. The core of these technologies lies in the acquisition and processing of sensor data. Among them, ultrasonic sensors are widely used in automatic parking systems due to their low cost, simple structure, and strong resistance to environmental interference.

[0003] Traditional ultrasonic sensor data processing methods mainly rely on manually designed feature extraction algorithms. These algorithms calculate the location and distance of obstacles by processing ultrasonic echo data, such as echo timestamp, echo height, echo confidence, and echo width. These feature extraction methods perform well in certain specific scenarios, but their performance decreases significantly in complex environments, such as those with a large number of obstacles or corner cases, and are difficult to meet the needs of practical applications.

[0004] In addition, there are also methods that convert ultrasonic data into two-dimensional or three-dimensional point cloud images, and then use point cloud processing algorithms or traditional machine learning methods to detect storage locations. These methods include: 1. Point cloud conversion and processing: The echo data generated by the ultrasonic sensor is processed to form a point cloud image, which can be converted into a two-dimensional image or directly processed in three dimensions through projection. The processing methods of point cloud data include filtering, clustering, segmentation and other steps, aiming to extract useful features from the point cloud for storage location detection, and the processing process is relatively complicated. 2. Rule-based algorithms: These algorithms analyze point cloud data based on predefined rules, such as judging obstacles by distance thresholds and identifying storage locations by shape features. These methods rely on manually designed features and rules and are difficult to cope with complex and changeable actual environments. 3. Traditional machine learning methods: By extracting features from point cloud data, traditional machine learning algorithms such as support vector machines (SVM) and decision trees are used for classification and regression. These methods have improved the accuracy of detection to a certain extent, but still face the problems of insufficient feature design and model generalization capabilities.

[0005] Although these current methods have solved the problem of ultrasonic data processing to a certain extent, there are still some shortcomings. First, the processing of point cloud data is complex, computationally intensive, and has poor real-time performance. Second, rule-based and traditional machine learning methods rely on manually designed features and are difficult to cope with complex scenarios and corner cases in practical applications. In addition, these methods usually require a lot of testing and debugging to ensure stability and reliability in different environments. Summary of the invention

[0006] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an ultrasonic data processing method, device, equipment and storage medium, aiming to convert ultrasonic data into intuitive visual images and generate high-quality data sets to solve the problems of poor generalization ability and high computational complexity of traditional algorithms.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In a first aspect, the present invention provides an ultrasonic data processing method, comprising:

[0009] Acquire ultrasonic data;

[0010] Preprocessing the acquired ultrasonic data;

[0011] Convert the preprocessed ultrasonic data into visual images;

[0012] Mark the target area in the visualization image;

[0013] The visualized images after marking the target area are organized and a data set is generated.

[0014] Furthermore, the ultrasonic data includes a transmission timestamp, an echo timestamp, an echo height, an echo confidence level, and an echo width.

[0015] Furthermore, the converting the pre-processed ultrasonic data into a visual image includes:

[0016] Arrange the wave timestamps in order as the horizontal coordinates of the pixels;

[0017] Calculate the distance of the obstacle based on the sending timestamp and the echo timestamp;

[0018] The distance to the obstacle is normalized and used as the vertical coordinate of the pixel point;

[0019] The echo height, echo confidence and echo width are numerically quantized respectively to be used as the RGB values ​​of the pixel points respectively, and the numerical quantization range is 0-255.

[0020] Furthermore, the target area is a parking space area.

[0021] Furthermore, marking the target area in the visualized image includes:

[0022] Identify the upper edge and the lower edge of the parking space in the visualization image;

[0023] Add markings at the top and bottom edges of the parking space.

[0024] Furthermore, after arranging the annotated visual images and generating a data set, the following steps are included:

[0025] Use the generated dataset to train a deep learning model;

[0026] Use the trained deep learning model to perform parking space detection on the ultrasonic data to be detected to output the parking space detection results.

[0027] Furthermore, before marking the target area in the visualized image, the method includes:

[0028] The pixel position in the horizontal direction of the visualization image corresponds to the horizontal position of the vehicle's world coordinates.

[0029] In a second aspect, the present invention further provides an ultrasonic data processing device, comprising:

[0030] An acquisition unit, used for acquiring ultrasonic data;

[0031] A preprocessing unit, used for preprocessing the acquired ultrasonic data;

[0032] A visualization unit, used for converting the preprocessed ultrasonic data into a visualization image;

[0033] A labeling unit, used to label the target area in the visualization image;

[0034] The data set generation unit is used to organize the visualized images after the target area is marked and generate a data set.

[0035] In a third aspect, the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the ultrasonic data processing method as described above when executing the computer program.

[0036] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein the storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the ultrasonic data processing method as described above.

[0037] Compared with the prior art, the present invention has the following beneficial effects: the ultrasonic data processing method includes: acquiring ultrasonic data; preprocessing the acquired ultrasonic data; converting the preprocessed ultrasonic data into a visual image; marking the target area in the visual image; organizing the visual image after marking the target area and generating a data set. The present invention converts ultrasonic data into intuitive visual images and marks them, making the originally non-intuitive ultrasonic data intuitive and easy to understand, solving the problem that ultrasonic data is difficult to directly use for deep learning model training, reducing the computational complexity of subsequent deep learning model training, and improving the algorithm generalization ability.

[0038] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the following preferred embodiments are specifically cited and described in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0040] Figure 1 The main flow chart of the ultrasonic data processing method provided by the specific embodiment of the present invention;

[0041] Figure 2 A sub-flow chart of an ultrasonic data processing method provided by a specific embodiment of the present invention;

[0042] Figure 3 A sub-flow chart of an ultrasonic data processing method provided by a specific embodiment of the present invention;

[0043] Figure 4 A flowchart of another implementation of the ultrasonic data processing method provided by a specific embodiment of the present invention;

[0044] Figure 5 A flowchart of another implementation of the ultrasonic data processing method provided by a specific embodiment of the present invention;

[0045] Figure 6 A schematic block diagram of an ultrasonic data processing device provided in a specific embodiment of the present invention;

[0046] Figure 7 A schematic block diagram of a computer device provided for a specific embodiment of the present invention;

[0047] Figure 8 A schematic diagram of visualized ultrasonic data annotation provided by a specific embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0050] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0051] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0052] like Figure 1 As shown, an embodiment of the present invention provides an ultrasonic data processing method, which can be applied to an automatic parking system of a vehicle. The method includes the following steps: S10-S50.

[0053] S10: Acquire ultrasonic data.

[0054] Ultrasonic data is obtained from the ultrasonic sensor installed on the vehicle. The ultrasonic data includes the wave transmission timestamp, echo timestamp, echo height, echo confidence, echo width, etc.

[0055] Emission timestamp: It is the exact time when the ultrasonic sensor emits the sound wave. This timestamp is used to record the start time of the sound wave emission and is the starting point of the distance measurement calculation.

[0056] Echo timestamp: It is the time when the ultrasonic sensor receives the echo when the sound wave encounters an obstacle and reflects back. This timestamp records the end time of the sound wave returning.

[0057] Echo Height: This is the strength or height of the received echo signal. The height of the echo can indicate the properties of the reflecting surface, such as hardness and material type. Generally, harder or smoother surfaces will produce higher echoes.

[0058] Echo confidence: This is a measure of the quality of the echo signal, usually based on the clarity, strength and other noise data of the echo. A high confidence level indicates good signal quality and more reliable measurement data.

[0059] Echo Width: This is the width of the echo signal, which can be understood as the duration of the echo signal in time. Echo width helps to judge the size or shape of an object, because large or complex objects may scatter more energy, resulting in a wider echo.

[0060] S20: pre-processing the acquired ultrasonic data.

[0061] Preprocessing mainly includes filtering out noise and outliers to ensure the accuracy and reliability of data. Specifically, this step includes filtering out the small tail problem caused by radar hardware and some co-frequency interference data. These small tail signals may be caused by the sidelobe effect of the radar antenna, pulse delay, hardware noise, etc. On the other hand, when the radar works in the same frequency radio band, the transmission signals of other electronic devices may interfere with the radar signal. Therefore, this link is an important link to ensure data quality and improve data accuracy and reliability.

[0062] S30, converting the preprocessed ultrasonic data into a visual image.

[0063] Filling the ultrasonic data into the image as pixel coordinates and RGB values ​​makes the ultrasonic data intuitive and easy to understand. This step improves the intuitiveness and interpretability of the data, making it more suitable for training and analysis of deep learning models. In the automatic parking system, intuitive images can more clearly represent the location, shape and other relevant features of obstacles, which are difficult to directly provide with ultrasonic raw data.

[0064] like Figure 2 As shown, in one embodiment, step S30 specifically includes the following steps: S301-S304.

[0065] S301, arranging the wave transmission timestamps in order as the horizontal coordinates of the pixel points.

[0066] Each time an ultrasonic wave is emitted, the emission time is recorded, which is the emission timestamp. Arranging the emission timestamps in chronological order can form the horizontal coordinate of the pixel point.

[0067] By using the wave emission timestamp as the horizontal coordinate of the pixel point, the change of the ultrasonic signal over time can be intuitively observed.

[0068] S302: Calculate the distance of the obstacle according to the transmission timestamp and the echo timestamp.

[0069] The difference between the emission timestamp and the echo timestamp can be used to calculate the time interval from the ultrasonic wave being emitted by the sensor to the echo being received. This time interval is then multiplied by the speed of sound and divided by two to obtain the distance between the obstacle and the ultrasonic sensor.

[0070] S303: Normalize the distance to the obstacle and use it as the vertical coordinate of the pixel point.

[0071] After the ultrasonic wave is emitted, it will encounter an obstacle and reflect back. The distance of the obstacle can be calculated based on the time it takes to reflect back. This distance is normalized. The purpose of normalization is to unify the distance of all obstacles within a specific range. The normalized distance is used as the vertical coordinate of the pixel point.

[0072] S304 , numerically quantize the echo height, echo confidence, and echo width respectively to use them as RGB values ​​of the pixel points respectively, and the numerical quantization range is 0-255.

[0073] Echo height, echo confidence and echo width are the characteristics of ultrasonic wave reflection, which can be understood as the strength, reliability and width of the reflected ultrasonic wave. These characteristics are quantized to the range of 0-255, because in the image, the value of each channel of RGB color is between 0-255, so these characteristics are converted into the color of the image.

[0074] Echo height, echo confidence, and echo width can be quantified to different channels of RGB color. For example, echo height can be quantified and mapped to the red channel. Echo height may represent the amplitude or energy of the reflected echo. Higher amplitudes can be interpreted as brighter red. Echo confidence is quantified and mapped to the green channel, which usually indicates the degree of confidence or accuracy in the measurement results. Higher confidence is represented by brighter green. Echo width is quantified and mapped to the blue channel. Echo width may represent the duration or pulse width of the reflected signal. A larger width may be mapped to a brighter blue.

[0075] For steps S301-S304, it can be understood that since the horizontal coordinate of each pixel represents the wave timestamp, these timestamps are arranged in sequence, so the horizontal coordinates of all pixels are actually on the same horizontal line, forming a time axis. This time axis shows the time sequence from the first wave to the last wave. However, the vertical coordinate of each pixel represents the distance between the ultrasonic sensor and the obstacle, and the vertical coordinate value will change according to the calculated distance of the obstacle, so the vertical coordinate of each pixel may be different. In addition, the color of each pixel is determined by the echo height, echo confidence and echo width, so the color of each pixel may be different.

[0076] S40: Mark the target area in the visualized image.

[0077] Since the originally non-intuitive ultrasonic data has been converted into a visual image, it is more conducive to marking the target area. The target area can be the area of ​​interest to the user. The marking method can be implemented with the help of marking software. The marking information can then be extracted and used for further analysis or training of deep learning models.

[0078] In this embodiment, the target area is a parking area, and the marked parking area will be used for subsequent deep learning model training.

[0079] like Figure 3 In one embodiment, step S40 specifically includes the following steps: S401-S402.

[0080] S401: Identify the upper edge and the lower edge of the parking space in the visualized image.

[0081] S402: Add marks at the upper and lower edges of the parking space.

[0082] For steps S401-S402, the upper edge of the parking space is the edge close to the entrance of the parking space, and the lower edge of the parking space is the edge close to the end of the parking space.

[0083] By marking the upper and lower edges of the parking spaces, these marking information will be input into the subsequent deep learning model as training data. Through the marking process, a large number of pictures can be generated, and the consistency and accuracy of the marking can be guaranteed, which not only improves the efficiency of model training, but also enhances the performance of the model in the detection of parking space targets. Figure 8 The upper edge and the lower edge of the parking space marked on the visualization image are indicated, and the positions of the upper edge and the lower edge of the parking space can be clearly seen.

[0084] like Figure 4 As shown, in one embodiment, before step S40, step S35 is included.

[0085] S35. Correspond the pixel position in the horizontal direction of the visualized image to the horizontal position of the world coordinates of the vehicle.

[0086] The pixel positions in the horizontal direction of the visualization image can be mapped to the world coordinates of the vehicle. This processing here facilitates the verification of the detection results, making the detection results more accurate. It can directly verify whether the detected parking space is an available parking space, thereby improving reliability and practicality.

[0087] S50, organizing the visualized images after marking the target area and generating a data set.

[0088] After collecting a large number of annotated visualization images, they can be cleaned to remove duplicate images, correct incorrect annotations, fill in missing data, etc. The purpose of data cleaning is to ensure the quality of the data set and reduce errors and deviations in subsequent analysis.

[0089] like Figure 5 As shown, in one embodiment, after step S50, the following steps are included: S60-S70.

[0090] S60. Use the generated dataset to train a deep learning model.

[0091] S70: Use the trained deep learning model to perform parking space detection on the ultrasonic data to be detected, so as to output a parking space detection result.

[0092] For steps S60-S70, the visual images with the upper and lower edges of the parking spaces marked are organized into a data set, and the generated data set is used to train the deep learning model to learn the features of the ultrasonic data and improve the accuracy and robustness of parking space target detection. When the deep learning model training is completed, the trained deep learning model is used to perform parking space detection on the ultrasonic data to be detected to obtain the parking space detection result, and the parking space detection result can be used to correspond to the world coordinates of the vehicle to verify whether the detected parking space is an available parking space, which can ensure the practicality and accuracy of the detection result, which can be completed by testing and verification in actual scenarios.

[0093] The present invention converts ultrasonic data into intuitive visual images and annotates them, making the originally non-intuitive ultrasonic data intuitive and easy to understand, solving the problem that ultrasonic data is difficult to use directly for deep learning model training, reducing the computational complexity of subsequent deep learning model training, and improving the generalization ability of the algorithm.

[0094] The embodiment of the present invention further provides an ultrasonic data processing device, which is used to execute the steps in any embodiment of the above-mentioned ultrasonic data processing method. Figure 6 , Figure 6FIG. 1 is a schematic block diagram of an ultrasonic data processing device 100 provided in an embodiment of the present application. The ultrasonic data processing device 100 specifically includes:

[0095] The acquisition unit 110 is used to acquire ultrasonic data. The preprocessing unit 120 is used to preprocess the acquired ultrasonic data. The visualization unit 130 is used to convert the preprocessed ultrasonic data into a visualized image. The annotation unit 140 is used to annotate the target area in the visualized image. The data set generation unit 150 is used to organize the visualized image after the target area is annotated and generate a data set.

[0096] In one embodiment, the visualization unit 130 includes:

[0097] The arrangement module is used to arrange the transmission timestamps in order as the horizontal coordinates of the pixels. The calculation module is used to calculate the distance of the obstacle according to the transmission timestamp and the echo timestamp. The normalization processing module is used to perform normalization processing on the distance of the obstacle and use it as the vertical coordinate of the pixel. The numerical quantization processing module is used to perform numerical quantization processing on the echo height, echo confidence and echo width respectively to use them as the RGB values ​​of the pixel, and the numerical quantization range is 0-255.

[0098] In one embodiment, the labeling unit 140 includes:

[0099] The recognition module is used to recognize the upper edge and the lower edge of the parking space in the visual image. The marking module is used to add marks at the upper edge and the lower edge of the parking space.

[0100] In one embodiment, the ultrasonic data processing device 100 further includes:

[0101] The mapping module is used to correspond the pixel position in the horizontal direction of the visualization image to the horizontal position of the vehicle's world coordinates.

[0102] In one embodiment, the ultrasonic data processing device 100 further includes:

[0103] The training module is used to train the deep learning model using the generated data set. The result output module is used to use the trained deep learning model to perform parking space detection on the ultrasonic data to be detected, so as to output the parking space detection result.

[0104] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the ultrasonic data processing device 100 and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0105] The ultrasonic data processing device can be implemented in the form of a computer program. Figure 7 Runs on the computer device shown.

[0106] See also Figure 7 , Figure 7 700 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 700 may be a server, wherein the server may be an independent server or a server cluster composed of multiple servers.

[0107] like Figure 7 As shown, the computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the ultrasonic data processing method as described above are implemented.

[0108] The computer device 700 may be a terminal or a server. The computer device 700 includes a processor 720 , a memory, and a network interface 750 connected via a system bus 710 , wherein the memory may include a non-volatile storage medium 730 and an internal memory 740 .

[0109] The non-volatile storage medium 730 may store an operating system 731 and a computer program 732. When the computer program 732 is executed, the processor 720 may execute an ultrasonic data processing method.

[0110] The processor 720 is used to provide computing and control capabilities and support the operation of the entire computer device 700.

[0111] The internal memory 740 provides an environment for the operation of the computer program 732 in the non-volatile storage medium 730. When the computer program 732 is executed by the processor 720, the processor 720 can execute the ultrasonic data processing method.

[0112] The network interface 750 is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will appreciate that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the computer device 700 to which the present application scheme is applied. The specific computer device 700 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. The processor 720 is used to run the program code stored in the memory to implement the steps of the ultrasonic data processing method.

[0113] Those skilled in the art will understand that Figure 7The embodiments of the computer device shown in the figure do not constitute a limitation on the specific composition of the computer device. In other embodiments, the computer device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. For example, in some embodiments, the computer device may only include a memory and a processor. In such embodiments, the structure and function of the memory and the processor are the same as those of the embodiment of the present invention. Figure 7 The embodiments shown are consistent and will not be described again here.

[0114] It should be understood that in the embodiment of the present application, the processor 720 may be a central processing unit (CPU), and the processor 720 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0115] In another embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium may be a non-volatile computer-readable storage medium. The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the ultrasonic data processing method disclosed in the embodiment of the present invention is implemented.

[0116] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the above-described equipment, devices and units can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. Those of ordinary skill in the art can appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0117] In the several embodiments provided by the present invention, it should be understood that the disclosed equipment, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. Units with the same function may also be combined into one unit. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0118] 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 network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0119] In addition, each functional unit in each embodiment of the present invention 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.

[0120] 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 storage medium. Based on this understanding, the technical solution of the present invention 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. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), disk or optical disk and other media that can store program codes.

[0121] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. Ultrasonic data processing method, It is characterized in that include: Acquire ultrasonic data; Preprocessing the acquired ultrasonic data; Convert the preprocessed ultrasonic data into visual images; Mark the target area in the visualization image; The visualized images after marking the target area are organized and a data set is generated.

2. The ultrasonic data processing method according to claim 1, It is characterized in that The ultrasonic data includes a transmission timestamp, an echo timestamp, an echo height, an echo confidence level, and an echo width.

3. The ultrasonic data processing method according to claim 2, It is characterized in that The step of converting the pre-processed ultrasonic data into a visual image comprises: Arrange the wave timestamps in order as the horizontal coordinates of the pixels; Calculate the distance of the obstacle based on the sending timestamp and the echo timestamp; The distance to the obstacle is normalized and used as the vertical coordinate of the pixel point; The echo height, echo confidence and echo width are numerically quantized respectively to be used as the RGB values ​​of the pixel points respectively, and the numerical quantization range is 0-255.

4. The ultrasonic data processing method according to claim 1, It is characterized in that The target area is a parking space area.

5. The ultrasonic data processing method according to claim 4, It is characterized in that The step of marking the target area in the visualized image includes: Identify the upper edge and the lower edge of the parking space in the visualization image; Add markings at the top and bottom edges of the parking space.

6. The ultrasonic data processing method according to claim 4, It is characterized in that After the labeled visualization images are sorted and a data set is generated, the following steps are included: Use the generated dataset to train a deep learning model; Use the trained deep learning model to perform parking space detection on the ultrasonic data to be detected to output the parking space detection results.

7. The ultrasonic data processing method according to claim 1, It is characterized in that Before marking the target area in the visual image, the method includes: The pixel position in the horizontal direction of the visualization image corresponds to the horizontal position of the vehicle's world coordinates.

8. Ultrasonic data processing device, It is characterized in that include: An acquisition unit, used for acquiring ultrasonic data; A preprocessing unit, used for preprocessing the acquired ultrasonic data; A visualization unit, used for converting the preprocessed ultrasonic data into a visualization image; A labeling unit, used to label the target area in the visualization image; The data set generation unit is used to organize the visualized images after the target area is marked and generate a data set.

9. A computer device, It is characterized in that The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the ultrasonic data processing method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, It is characterized in that The storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the ultrasonic data processing method according to any one of claims 1 to 7.