Vehicle type identification method, device, equipment and medium based on radar chart

Abstract

The present invention discloses a vehicle type identification method, apparatus, device, and medium based on radar images. The method comprises: determining a current radar image including a target vehicle; separating the background and foreground in the current radar image to obtain a background-removed target radar image; and identifying the vehicle type of the target vehicle in the current radar image based on the target radar image. The technical solution of this application achieves faster and more accurate vehicle type identification, thereby improving vehicle classification accuracy, by separating the background and foreground in the current radar image to obtain a background-removed radar image, and then identifying the vehicle type of the target vehicle in the current radar image based on the background-removed radar image.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle type classification and identification, and in particular to a vehicle type identification method, device, equipment and medium based on radar charts. Background Art

[0002] As the transportation industry continues to develop towards intelligence, digitization and informatization, vehicle type classification and identification has become an important part of traffic management. Through vehicle type classification, traffic can be unblocked and managed by using the results of different vehicle types.

[0003] Related solutions typically employ methods such as video image detection, ultrasonic detection, and induction coil recognition for vehicle type identification. However, video image detection has high requirements for image quality and environmental conditions, making it difficult to accurately detect vehicle types in conditions with high interference and poor lighting, such as rain or snow. Induction coil recognition offers low investment costs, strong stability, and minimal environmental impact, but is complex and expensive to maintain, and can cause significant damage to the ground. While ultrasonic detection can extract vehicle outline data for simple vehicle type classification, its accuracy is low and the equipment installation is complex. Therefore, simply and accurately implementing vehicle type classification is crucial. Summary of the Invention

[0004] The present invention provides a vehicle type identification method, device, equipment and medium based on radar charts, so as to realize rapid and accurate detection of vehicle types.

[0005] According to one aspect of the present invention, a method for vehicle type identification based on radar chart is provided, the method comprising:

[0006] determining a current radar image including a target vehicle;

[0007] Separate the background and foreground in the current radar image to obtain a target radar image after background removal;

[0008] The target vehicle in the current radar image is identified according to the target radar image.

[0009] According to another aspect of the present invention, a vehicle type identification device based on a radar chart is provided, characterized in that the device comprises:

[0010] a radar image determination module, configured to determine a current radar image including a target vehicle;

[0011] The radar image separation module is used to separate the background and foreground in the current radar image to obtain the target radar image after the background is removed;

[0012] The vehicle type recognition module is used to identify the vehicle type of the target vehicle in the current radar image based on the target radar image.

[0013] According to another aspect of the present invention, an electronic device is provided, comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the vehicle model identification method based on radar charts described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the vehicle model identification method based on radar charts according to any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided, comprising a computer program. When executed by a processor, the computer program implements the method for vehicle type identification based on radar charts according to any embodiment of the present invention.

[0019] The technical solution of an embodiment of the present invention determines a current radar image including a target vehicle, separates the background and foreground in the current radar image to obtain a background-removed target radar image, and then identifies the vehicle type of the target vehicle in the current radar image based on the target radar image. The technical solution of this application achieves faster and more accurate vehicle type identification, thereby improving the accuracy of vehicle classification, by separating the background and foreground in the current radar image to obtain a background-removed radar image, and then identifying the vehicle type of the target vehicle in the current radar image based on the background-removed radar image.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1is a flow chart of a vehicle type identification method based on a radar chart according to the first embodiment of the present invention;

[0023] Figure 2 is a flow chart of another vehicle type identification method based on radar charts according to the second embodiment of the present invention;

[0024] Figure 3 1 is a schematic structural diagram of a vehicle type identification device based on a radar chart according to a third embodiment of the present invention;

[0025] Figure 4 The present invention is a schematic structural diagram of an electronic device for implementing a vehicle type identification method based on a radar chart according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only 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 making creative efforts should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "current", "target", etc. in the description and claims of the present invention 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 numbers used in this way can be interchanged where appropriate so that the embodiments of the present invention 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 clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] Example 1

[0029] Figure 1 A flowchart of a vehicle type identification method based on a radar image is provided for the first embodiment of the present invention. This embodiment is applicable to situations where vehicle type identification is performed on radar images. The method can be performed by a vehicle type identification device based on a radar image. The device can be implemented in the form of hardware and / or software. The vehicle type identification device based on a radar image can be configured in any electronic device with network communication capabilities.

[0030] like Figure 1 As shown, the method includes:

[0031] S110: Determine a current radar image including a target vehicle.

[0032] A radar image is an image formed when a radar transmitter transmits radio waves toward a target and a receiver receives the scattered echoes. The radar image is divided into several small squares, each of which is called a pixel. By analyzing and processing the information at each pixel, the characteristics of each detection point within the radar detection area corresponding to that pixel can be obtained. The current radar image is represented by indicating the position, color, and brightness of each pixel, thereby obtaining information about each detection point, such as the signal strength at each detection point. The current radar image can be the radar image obtained by scanning the radar detection area at the current moment.

[0033] As an optional but non-limiting implementation, determining the current radar image including the target vehicle may include the following process:

[0034] The microwave radar is used to scan the target vehicle entering the radar detection area to obtain the current radar image at the current moment.

[0035] The value of each pixel in the current radar image is used to describe the signal strength of the radar reflection wave when the radar scans the detection position point, and the current radar image is a grayscale image.

[0036] Among them, a grayscale image can be an image with only one sampled color for each pixel, usually displayed as a grayscale from the darkest black to the brightest white. Pure white in a grayscale image represents that the color light here is at the highest brightness, and the brightness level is 255. The brightness level of the darkest black in a grayscale image is 0, that is, the grayscale value of a grayscale image is 0-255.

[0037] Specifically, if the microwave radar's detection area is located on a highway or low-speed road, the millimeter-wave radar scans the detection area at the current moment and then uses a grayscale value lookup table to populate each pixel to obtain the current radar image represented by a grayscale image. The grayscale value represents the signal strength of the radar reflected wave at the detection location corresponding to each pixel during the scan.

[0038] This technical solution uses microwave radar to scan the radar detection area at the current moment to obtain the current radar image at the current moment. Based on the grayscale value of each pixel point on the current radar image, the signal strength of the radar reflection wave at the detection position point corresponding to each pixel point can be accurately reflected, so as to facilitate the subsequent processing of the current radar image.

[0039] As an optional but non-limiting implementation method, in this embodiment, the target vehicle can be a vehicle that needs to be analyzed to obtain a vehicle type classification. Optionally, the target vehicle can be any one or more vehicles in the radar image. The target vehicle types include bicycles, motorcycles or electric vehicles, cars, vans, pickup trucks, large trucks and buses.

[0040] This technical solution uses microwave radar to scan the radar detection area at the current moment to obtain the current radar image at the current moment. Based on the grayscale value of each pixel point on the current radar image, the signal strength of the radar reflection wave at the detection position point corresponding to each pixel point can be accurately reflected, so as to facilitate the subsequent processing of the current radar image.

[0041] S120 , separating the background and foreground in the current radar image to obtain a target radar image after the background is removed.

[0042] The background may be an image of an inherent object in the corresponding radar detection area in the radar image, for example, an image of a road in the radar detection area may be the background. The foreground may be a vehicle area graphic in the corresponding radar detection area in the radar image, specifically a pedestrian, bicycle, motorcycle or electric bicycle, car, van, pickup truck, truck, bus, etc. The present application mainly separates the foreground and background to obtain a separated radar image without the foreground, so as to facilitate comparative analysis of the radar image obtained by scanning the radar detection area based on the separated image.

[0043] The current radar image includes a foreground and a background. The foreground may be a vehicle area graphic, specifically a pedestrian, bicycle, motorcycle or battery vehicle, car, van, pickup truck, truck or bus.

[0044] S130 : Identify the vehicle type of the target vehicle in the current radar image based on the target radar image.

[0045] The technical solution of the embodiment of the present invention separates the background and foreground in the current radar image to obtain a radar image after background removal, and then identifies the model of the target vehicle in the current radar image based on the radar image after background removal, thereby achieving faster and more accurate model identification and improving the accuracy of vehicle classification.

[0046] Example 2

[0047] Figure 2 This is a flow chart of a vehicle type recognition method based on radar images provided in the second embodiment of the present invention. This embodiment is based on the previous embodiment and further optimizes the process of "separating the background and foreground in the current radar image to obtain the target radar image after the background is removed" in the previous embodiment. Figure 2 As shown, the method includes:

[0048] S210: Determine a current radar image including a target vehicle.

[0049] S220: Determine a preset number of previous radar images collected before the current radar image.

[0050] The previous radar image includes a radar image acquired in a nearby time before the current radar image is acquired or a radar image acquired when a radar detection area is closed.

[0051] The preset number may be the number of radar images required to be acquired by scanning the radar detection area before acquiring the current radar image, determined based on actual needs. The previous radar image may be all radar images acquired by scanning the radar detection area using a microwave radar (such as a millimeter-wave radar) before acquiring the current radar image, where the previous radar image only contains the background. Alternatively, if the radar detection area is closed and free of other interfering objects, the previous radar image can be acquired by scanning the radar detection area using a millimeter-wave radar.

[0052] Optionally, when there are no vehicles or pedestrians in the radar detection area, N frames of radar images of the radar detection area are continuously accumulated. That is, after accumulating N frames of the previous radar image, an average of the accumulated images is calculated. The accumulated average image remains stable for a short period of time after being determined. A new accumulated average image is then required unless the radar detection area changes, such as when new equipment (such as a fence) is installed in a tunnel.

[0053] S230 , performing image cumulative averaging on a preset number of previous radar images to obtain a cumulative average image corresponding to the previous radar image.

[0054] Specifically, a millimeter-wave radar is used to scan the radar detection area to obtain a first preset number of previous radar images, and each image can be recorded as F, which has P rows and Q columns, that is, a grayscale image composed of P*Q pixels. The matrix representation is as follows:

[0055]

[0056] Then perform image cumulative averaging on the first preset number of previous radar images to obtain the cumulative average image of the previous radar image Expressed as formula:

[0057]

[0058] Among them, F i is the grayscale image of each image in the previous radar image.

[0059] S240 , separating the background and foreground in the current radar image according to the accumulated average image to obtain a target radar image after background removal.

[0060] The microwave radar is used to scan the radar detection area to obtain a preset number of previous radar images, and each image can be recorded as F. Then, the preset number of previous radar images are cumulatively averaged to obtain the cumulative average image F of the previous radar image. Next, the current radar image obtained by scanning the radar detection area by the microwave radar at the current moment is obtained. Since the target vehicle in the current radar image does not appear in the previous radar image, the cumulative average image of the previous radar image is It can be used as the background image of the radar image. Then, the background and foreground in the current radar image can be separated based on the cumulative average image of the previous radar image to accurately obtain the target radar image after background removal.

[0061] This technical solution performs image cumulative averaging on a preset number of previous radar images, making the obtained cumulative average image of the previous radar images more accurate. This allows the background and foreground in the current radar image to be separated based on the cumulative average image of the previous radar images, thereby obtaining an accurate target radar image after background removal. This facilitates subsequent more accurate acquisition of the target vehicle's vehicle size and location information.

[0062] As an optional but non-limiting implementation, separating the background and foreground in the current radar image based on the cumulative average image to obtain a target radar image after background removal may include but is not limited to the following steps A1-A2:

[0063] Step A1: Perform image difference processing on the current radar image and the cumulative average image of the previous radar image to obtain an image after image difference processing.

[0064] Step A2: Binarize the image after image difference processing to separate the background and foreground in the current radar image to obtain a target radar image after background removal.

[0065] Image difference processing can be performed by taking the difference between two similar images. Binarization processing can be performed by making each pixel in the image have only two possible values ​​or grayscale states, that is, the grayscale value of any pixel in the image is either 0 or 255, representing black and white respectively. Binarization can be performed using the following formula:

[0066]

[0067] Among them, f ij ' is the grayscale value of the corresponding pixel after the radar image is binarized, is the grayscale value of the corresponding pixel in the radar image, T is the preset grayscale value, and the preset grayscale value can be the critical value for converting the grayscale value of the corresponding pixel in the radar image to 0 or 255. When the grayscale value of the corresponding pixel in the radar image is greater than or equal to the preset grayscale value, the grayscale value of the corresponding pixel is converted to 255, otherwise it is converted to 0.

[0068] Specifically, after performing image cumulative averaging on a first preset number of previous radar images to obtain a cumulative average image of the previous radar images, the current radar image is obtained, and image difference processing is performed on the current radar image and the cumulative average image of the previous radar image to obtain an image F after image difference processing. Δ , which can be expressed as:

[0069]

[0070] Where F is the grayscale image of the current radar image.

[0071] Next, the image F after image difference processing Δ Perform binarization processing, that is, F Δ The pixel value of each pixel in the image is converted to 0 or 255, and the pixel value of 0 is the background, and the pixel value of 255 is the foreground. Therefore, the background and foreground in the current radar image can be separated according to the image after image difference processing after binarization, thereby obtaining the separated radar image.

[0072] This technical solution performs image difference processing on the cumulative average image of the current radar image and the previous radar image to obtain an image after image difference processing. Then, by binarizing the image after image difference processing, the background and foreground in the current radar image are separated to obtain a target radar image after background removal. This achieves accurate separation of the background and foreground of the current radar image, making the target radar image after background removal more accurate, and can achieve more accurate acquisition of vehicle size information and vehicle position information of the target vehicle.

[0073] S250: Identify the vehicle type of the target vehicle in the current radar image based on the target radar image.

[0074] When binarizing the radar image, some pixels in the foreground with grayscale values ​​lower than the preset grayscale value are converted to 0, resulting in internal holes between different foreground sub-regions and / or gaps in neighboring regions in the target radar image. Therefore, morphological processing is required to eliminate the internal holes and / or gaps in neighboring regions between different foreground sub-regions.

[0075] As an optional but non-limiting implementation, performing vehicle type recognition on the target vehicle in the current radar image based on the target radar image may include but is not limited to the process of steps B1-B2:

[0076] Step B1: performing morphological processing on the target radar image to obtain a processed radar image; the target vehicle in the target radar image is segmented into different foreground sub-regions due to foreground and background separation.

[0077] Step B2: input the processed radar image into a preset vehicle type recognition model, and determine the vehicle type of the target vehicle in the current radar image through the preset vehicle type recognition model.

[0078] Specifically, the target radar image is obtained. During this process, the target vehicle area may be divided into different sub-areas. Therefore, it is necessary to perform morphological processing on the target radar image to obtain a processed radar image, and then perform Gaussian smoothing on the processed radar image to eliminate noise caused by radar detection. Finally, edge detection is performed on the processed radar image after Gaussian smoothing to accurately distinguish the edges of the target vehicle in the target radar image, thereby accurately obtaining an edge detection map of the target radar image.

[0079] This technical solution performs morphological processing on the target radar image to obtain a processed radar image, so that the processed radar image can more accurately represent the target vehicle area. In addition, Gaussian smoothing is performed on the processed radar image to further eliminate some small noise points caused by radar detection, further enhancing the accuracy of the image. Finally, edge detection is performed on the Gaussian smoothed processed radar image to obtain an edge detection map of the target radar image, which is conducive to subsequently obtaining accurate vehicle size information and vehicle position information of the target vehicle through the edge detection map.

[0080] As an optional but non-limiting implementation, performing morphological processing on the target radar image to obtain a processed radar image may include but is not limited to steps C1-C2:

[0081] Step C1: performing a morphological dilation operation on the target radar image to obtain an expanded radar image; the morphological dilation operation is used to eliminate internal holes between different foreground sub-regions and / or gaps between neighboring regions.

[0082] Since the target vehicle area may be divided into different sub-areas during the target radar image formation process, a morphological dilation operation is performed on the target radar image to obtain an expanded radar image in order to eliminate internal holes and / or gaps between adjacent sub-areas corresponding to the target vehicle area. Since the area becomes larger after dilation, a morphological erosion operation is performed on the expanded radar image to obtain an eroded radar image. This restores the area to its pre-dilation state, allowing the processed radar image to more accurately represent the target vehicle area.

[0083] Step C2: performing a morphological erosion operation on the expanded radar image to obtain an eroded radar image, which serves as the processed radar image.

[0084] Specifically, when performing morphological dilation operations on different foreground sub-regions, each different foreground sub-region will become larger, resulting in the size of the foreground sub-region after the morphological dilation operation being unable to correspond to the size of the current radar image. Therefore, it is necessary to perform an erosion operation on the foreground sub-region after the dilation operation so that its area can be restored to the size before the dilation, that is, corresponding to the area size in the current target image.

[0085] As an optional but non-limiting implementation, the preset vehicle type recognition model is obtained by training based on a target detection network using preset training samples.

[0086] Specifically, the vehicle area graphics are labeled, and the categories include pedestrians, bicycles, motorcycles or electric vehicles, cars, vans, small trucks, large trucks and buses to obtain training samples.

[0087] The training samples are randomly divided into a training set and a validation set according to a preset ratio, and the training set and the validation set are used to train the preset vehicle model recognition model. In this embodiment, the preset vehicle model recognition model is preferably a YOLO network model, and the preset ratio is 4:1.

[0088] Specifically, the YOLO network model divides the input image into S*S grids. If the center of a target vehicle falls within a cell, the corresponding cell is responsible for detecting the target vehicle. For each target vehicle, the probability of belonging to each of the eight object categories is output. The validation set is labeled data and is not used during training. The validation algorithm compares the predicted targets with the labeled targets to determine the accuracy of the predictions, which is used to evaluate the model's predictive ability for unknown samples.

[0089] As an optional but non-limiting implementation, the target vehicle in the current radar image is identified based on the target radar image, which may include but is not limited to the process of steps D1-D3:

[0090] Step D1: Determine the radar position and detection angle of the radar detection area.

[0091] Step D2: Construct an outline of the pixel points describing the target vehicle based on the values ​​of each pixel point in the target radar image and the radar position and detection angle of the radar detection area.

[0092] Step D3: Determine the model of the target vehicle based on the constructed pixel point outline.

[0093] Specifically, the radar position and detection angle of the radar detection area corresponding to the target vehicle area are obtained. This is because the pixel points with different radar reflection wave signal intensities in the target radar image can reflect the distance between the radar position and the target vehicle. Then, by calculating the radar position and detection angle of the radar detection area corresponding to the detection vehicle area and the values ​​of each pixel point in the target radar image, the plane outline and / or three-dimensional outline of the pixel point can be constructed, and finally the vehicle size information of the target vehicle corresponding to the target vehicle area can be accurately obtained.

[0094] Constructing the outline of the pixels of the vehicle area to be detected may include constructing a plane outline and / or a three-dimensional outline of the pixels of the vehicle area to be detected. Because the radar cannot scan the entire vehicle, but only a portion of the vehicle, the constructed plane outline is a complete outline, while the constructed three-dimensional outline is a partial three-dimensional outline of the vehicle, rather than a complete outline.

[0095] The pixel points with different radar reflection wave signal intensities in the target radar image can reflect the distance between the radar position and the vehicle to be detected. After obtaining the radar position and detection angle of the radar detection area corresponding to the vehicle to be detected area, the detection angle of the radar toward the detection position corresponding to each pixel point can be determined according to the radar detection angle. According to the detection angle toward each pixel point corresponding to the vehicle to be detected in the target radar image and the relative distance from each pixel point corresponding to the vehicle to be detected in the target radar image to the radar (the relative distance from each pixel point corresponding to the vehicle to be detected to the radar can be estimated based on the grayscale value of each pixel point. The grayscale value of each pixel point is used to describe the signal strength of the radar reflection wave when the radar scans the detection position point. The signal strength is inversely proportional to the relative distance), the relative position between each pixel point corresponding to the vehicle to be detected can be analyzed. At the same time, the road surface undulation in the radar detection area can be determined according to the radar position. Combined with the road surface undulation, the three-dimensional outline of the vehicle to be detected in the radar detection area can be constructed.

[0096] This technical solution accurately constructs the plane outline and / or three-dimensional outline of the pixel points in the target vehicle area by taking the values ​​of each pixel point in the target radar image and the radar position and detection angle of the radar detection area. Based on the constructed plane outline and / or three-dimensional outline of the pixel points, vehicle model recognition can be achieved according to the plane outline and / or three-dimensional outline.

[0097] The technical solution of the embodiment of the present invention determines a current radar image of a radar detection area and separates the foreground and background of the current radar image to accurately obtain a target radar image with the background removed. A morphological dilation operation is then performed on the target radar image to eliminate internal holes and / or gaps between different sub-regions corresponding to the foreground, so that the image can more accurately represent the target vehicle area. A morphological erosion operation is performed on the dilated radar image to obtain an eroded radar image as a processed radar image. The processed radar image is input into a preset vehicle type recognition model so that the processed image has the same size as the current radar image. The vehicle type of the target vehicle in the current radar image is determined by the preset vehicle type recognition model, thereby achieving faster and more accurate vehicle type recognition and improving the accuracy of vehicle classification.

[0098] Example 3

[0099] Figure 3 This is a schematic diagram of the structure of a vehicle type identification device based on radar charts provided in the third embodiment of the present invention. Figure 3 As shown, the device includes:

[0100] a radar image determination module 310 for determining a current radar image including a target vehicle;

[0101] The radar image separation module 320 is used to separate the background and foreground in the current radar image to obtain a target radar image after the background is removed;

[0102] The vehicle type recognition module 330 is configured to recognize the vehicle type of the target vehicle in the current radar image based on the target radar image.

[0103] Optionally, the radar image determination module 310 is specifically configured to:

[0104] The microwave radar is used to scan the target vehicle entering the radar detection area at the current moment to obtain the current radar image;

[0105] The value of each pixel in the current radar image is used to describe the signal strength of the radar reflection wave when the radar scans the detection position point, and the current radar image is a grayscale image.

[0106] Based on the above embodiment, optionally, the radar image separation module 320 includes:

[0107] An image determining unit, configured to: determine a preset number of previous radar images acquired before a current radar image;

[0108] The previous radar image includes a radar image acquired in a nearby time before the current radar image is acquired or a radar image acquired by closing a radar detection area;

[0109] An image accumulation and averaging unit is used to: perform image accumulation and averaging on a preset number of previous radar images to obtain an accumulated average image corresponding to the previous radar image;

[0110] The background removal unit is used to separate the background and foreground in the current radar image according to the accumulated average image to obtain a target radar image after background removal;

[0111] Based on the above embodiment, optionally, the background removal unit is specifically configured to: perform image difference processing on the current radar image and the cumulative average image of the previous radar image to obtain an image after image difference processing;

[0112] By performing binarization on the image after image difference processing, the background and foreground in the current radar image are separated to obtain the target radar image after background removal.

[0113] Based on the above embodiment, optionally, the vehicle type identification module 330 is specifically configured to:

[0114] Performing morphological processing on the target radar image to obtain a processed radar image; the target vehicle in the target radar image is segmented into different foreground sub-regions due to foreground and background separation;

[0115] Inputting the processed radar image into a preset vehicle type recognition model, and determining the vehicle type of the target vehicle in the current radar image by the preset vehicle type recognition model;

[0116] The preset vehicle type recognition model is obtained by training based on the target detection network using preset training samples;

[0117] Based on the above embodiment, optionally, the vehicle type identification module 330 is further configured to:

[0118] Performing a morphological dilation operation on the target radar image to obtain an expanded radar image; the morphological dilation operation is used to eliminate internal holes between different foreground sub-regions and / or gaps between adjacent regions;

[0119] Performing a morphological erosion operation on the expanded radar image to obtain an eroded radar image, and using the eroded radar image as the processed radar image;

[0120] Based on the above embodiment, optionally, the vehicle type identification module 330 is further configured to:

[0121] Determine the radar position and detection angle in the radar detection area;

[0122] Constructing an outline of the pixel points describing the target vehicle based on the values ​​of each pixel point in the target radar image and the radar position and detection angle of the radar detection area;

[0123] Based on the constructed pixel outline, the model of the target vehicle is determined.

[0124] The vehicle type identification device based on radar charts provided in the embodiments of the present invention can execute the vehicle type identification method based on radar charts provided in any of the above-mentioned embodiments of the present invention, and has the corresponding functions and beneficial effects of executing the vehicle type identification method based on radar charts. For detailed processes, please refer to the relevant operations of the vehicle type identification method based on radar charts in the above-mentioned embodiments.

[0125] Example 4

[0126] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0127] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0128] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0129] Processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. Processor 11 executes the various methods and processes described above, such as the radar image-based vehicle model recognition method.

[0130] In some embodiments, the vehicle type identification method based on a radar image can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the vehicle type identification method based on a radar image described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the vehicle type identification method based on a radar image by any other appropriate means (e.g., by means of firmware).

[0131] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0132] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0133] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0134] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0135] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0136] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0137] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0138] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A vehicle type recognition method based on radar chart, characterized in that: The method comprises: Scanning a target vehicle entering a radar detection area by a microwave radar to obtain a current radar image at the current moment. The value of each pixel in the current radar image is used to describe the signal strength of the radar reflection wave when the microwave radar scans the detection position point. The current radar image is a grayscale image. Separate the background and foreground in the current radar image to obtain a target radar image after background removal; Performing vehicle type recognition on a target vehicle in a current radar image based on the target radar image; The performing vehicle type recognition on the target vehicle in the current radar image based on the target radar image includes: Determining a radar position and a first detection angle in a radar detection area; determining, according to the first detection angle, a second detection angle of the microwave radar toward a detection position point corresponding to each pixel point in the target radar image; Determining relative positions between the pixels corresponding to the target vehicle based on the second detection angle and the relative distances between the pixels corresponding to the target vehicle in the target radar image and the microwave radar, and constructing a pixel profile of the target vehicle based on the relative positions; Based on the constructed pixel outline, the model of the target vehicle is determined.

2. The method according to claim 1, characterized in that Separate the background and foreground of the current radar image to obtain the target radar image after background removal, including: Determining a preset number of previous radar images acquired before the current radar image; The previous radar image includes a radar image acquired in a nearby time before the current radar image is acquired or a radar image acquired by closing a radar detection area; Performing image cumulative averaging on a preset number of previous radar images to obtain a cumulative average image corresponding to the previous radar image; The background and foreground in the current radar image are separated based on the cumulative average image to obtain the target radar image after background removal.

3. The method according to claim 2, characterized in that The background and foreground of the current radar image are separated based on the cumulative average image to obtain the target radar image after background removal, including: Performing image difference processing on the current radar image and the cumulative average image of the previous radar image to obtain an image after image difference processing; By performing binarization on the image after image difference processing, the background and foreground in the current radar image are separated to obtain the target radar image after background removal.

4. The method according to claim 1, wherein Performing vehicle type recognition on a target vehicle in a current radar image based on the target radar image includes: Performing morphological processing on the target radar image to obtain a processed radar image; the target vehicle in the target radar image is segmented into different foreground sub-regions due to foreground and background separation; The processed radar image is input into a preset vehicle type recognition model, and the vehicle type recognition of the target vehicle in the current radar image is determined by the preset vehicle type recognition model.

5. The method according to claim 4, characterized in that Performing morphological processing on the target radar image to obtain a processed radar image includes: Performing a morphological dilation operation on the target radar image to obtain an expanded radar image; the morphological dilation operation is used to eliminate internal holes between different foreground sub-regions and / or gaps between adjacent regions; A morphological erosion operation is performed on the expanded radar image to obtain an eroded radar image, which is used as the processed radar image.

6. The method according to claim 4, characterized in that The preset vehicle type recognition model is obtained by training based on a target detection network using preset training samples.

7. A vehicle type identification device based on radar chart, characterized in that: The device comprises: a radar image determination module, configured to scan a target vehicle entering a radar detection area using a microwave radar to obtain a current radar image at the current moment, wherein the value of each pixel in the current radar image is used to describe the signal strength of the radar reflection wave when the microwave radar scans the detection position point, and the current radar image is a grayscale image; The radar image separation module is used to separate the background and foreground in the current radar image to obtain the target radar image after the background is removed; A vehicle type recognition module is used to identify the vehicle type of the target vehicle in the current radar image based on the target radar image; The vehicle type recognition module is further configured to determine a radar position and a first detection angle in a radar detection area; Determining a second detection angle of the microwave radar toward a detection position corresponding to each pixel in the target radar image according to the first detection angle; Determining relative positions between the pixels corresponding to the target vehicle based on the second detection angle and the relative distances between the pixels corresponding to the target vehicle in the target radar image and the microwave radar, and constructing a pixel profile of the target vehicle based on the relative positions; Based on the constructed pixel outline, the model of the target vehicle is determined.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the vehicle type identification method based on radar charts according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the vehicle type identification method based on radar charts according to any one of claims 1 to 6 when executed.

Images