License plate recognition method, image processing device, and computer-readable storage medium

Abstract

The embodiments of the present application disclose a license plate recognition method, an image processing device, and a computer-readable storage medium for identifying counterfeit license plates and genuine license plates. The method of the embodiments of the present application includes: obtaining multiple frames of images to be recognized continuously captured by an image acquisition device at a gate site, extracting image regions of the same license plate to be recognized from each frame of the images to be recognized, obtaining multiple license plate regions to be recognized and the positions of the license plate regions to be recognized in the images to be recognized, obtaining a target image recognition region generated for the gate site, wherein the target image recognition region is used to represent the image region where the license plate of a vehicle moving through the gate site may appear, judging whether the number of license plate regions to be recognized located in the target image recognition region meets a preset condition based on the positions of the multiple license plate regions to be recognized, obtaining a judgment result, and determining whether the license plate to be recognized is a counterfeit license plate or a genuine license plate based on the judgment result.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of image processing technology, and more specifically, to a license plate recognition method, an image processing device, and a computer-readable storage medium. Background Art

[0002] In recent years, with the widespread adoption of vehicles, unmanned vehicle access management based on license plate recognition technology has gradually become the mainstream vehicle access management method. For example, in parking lots, this unmanned vehicle access management method uses license plate recognition technology to capture license plates of passing vehicles, recognize the captured license plate images, and automatically control vehicle access based on the recognition results, saving labor costs for property managers.

[0003] However, whether the license plate is from a normal vehicle or a forged one, such as a mobile phone display or a printed license plate image, the license plate recognition technology can identify it as a license plate and control the gate device to open and release the vehicle. Therefore, under the current vehicle access management method, if someone holds a mobile phone displaying a license plate image or a printed license plate image, the gate device can open and release the vehicle.

[0004] Therefore, a technical solution is needed to identify the situation where a fake license plate is used as a real license plate to control the gate device to release the vehicle, so as to improve the accuracy of the gate release. Summary of the Invention

[0005] The embodiments of the present application provide a license plate recognition method, an image processing device, and a computer-readable storage medium, which can identify fake license plates and real license plates.

[0006] In a first aspect, an embodiment of the present application provides a license plate recognition method, comprising:

[0007] Acquire multiple frames of images to be identified that are continuously captured by an image acquisition device at the gate site;

[0008] Extracting the image region of the same license plate to be recognized from each frame of the image to be recognized, and obtaining multiple license plate regions to be recognized and the positions of the license plate regions to be recognized in the image to be recognized;

[0009] Obtaining a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear;

[0010] According to the positions of the plurality of license plate areas to be identified, determining whether the number of license plate areas to be identified in the target image recognition area meets a preset condition, and obtaining a determination result;

[0011] Based on the judgment result, it is determined whether the license plate to be identified is a fake license plate or a real license plate.

[0012] Optionally, obtaining a target image recognition area generated on-site for the gate includes:

[0013] Acquire vehicle images continuously captured by an image acquisition device of a plurality of vehicles passing through a gate, wherein each vehicle corresponds to a plurality of frames of vehicle images;

[0014] For each vehicle, extract the license plate image region of each vehicle from the multiple vehicle image frames to obtain the license plate region of each vehicle and the position of the license plate region of each vehicle, wherein each vehicle may have multiple license plate regions.

[0015] A target image recognition area is obtained based on multiple license plate areas of multiple vehicles.

[0016] Optionally, obtaining a target image recognition area based on multiple license plate areas of multiple vehicles includes:

[0017] Determine a first coordinate point and a second coordinate point of each license plate region according to the position of each license plate region in the vehicle image, wherein the first coordinate point is located in the upper left region of the license plate region and the second coordinate point is located in the lower right region of the license plate region;

[0018] determining a target first vehicle among the plurality of vehicles based on first coordinate points of the plurality of license plate areas of the plurality of vehicles, and determining a target second vehicle among the plurality of vehicles based on second coordinate points of the plurality of license plate areas of the plurality of vehicles;

[0019] Fitting an upper boundary line according to first coordinate points of a plurality of license plate areas of a target first vehicle, and fitting a lower boundary line according to second coordinate points of a plurality of license plate areas of a target second vehicle;

[0020] The image area between the upper boundary line and the lower boundary line is determined as the target image recognition area.

[0021] Optionally, determining a target first vehicle among a plurality of vehicles based on first coordinate points of a plurality of license plate areas of a plurality of vehicles, and determining a target second vehicle among a plurality of vehicles based on second coordinate points of a plurality of license plate areas of a plurality of vehicles, includes:

[0022] For the same vehicle, determining a first average coordinate value of the vehicle based on first coordinate points of multiple license plate areas of the vehicle, and determining a target first vehicle among the multiple vehicles based on the first average coordinate values ​​of the multiple vehicles;

[0023] For the same vehicle, a second average coordinate value of the vehicle is determined according to the second coordinate points of multiple license plate areas of the vehicle, and a target second vehicle is determined among the multiple vehicles according to the second average coordinate values ​​of the multiple vehicles.

[0024] Optionally, determining a first average coordinate value of the vehicle based on first coordinate points of multiple license plate areas of the vehicle, and determining a target first vehicle among the multiple vehicles based on the first average coordinate values ​​of the multiple vehicles includes:

[0025] Determining a first average ordinate value of the vehicle based on ordinate values ​​of upper left corner points of multiple license plate areas of the vehicle;

[0026] According to the first average ordinate values ​​of the plurality of vehicles, determining a vehicle corresponding to the smallest first average ordinate value as a target first vehicle;

[0027] Determining a second average coordinate value of the vehicle according to second coordinate points of multiple license plate areas of the vehicle, and determining a target second vehicle among the multiple vehicles according to the second average coordinate values ​​of the multiple vehicles, including:

[0028] determining a second average ordinate value of the vehicle based on the ordinate values ​​of the lower right corner points of the plurality of license plate regions of the vehicle;

[0029] According to the second average ordinate values ​​of the plurality of vehicles, a vehicle corresponding to the largest second average ordinate value is determined as the target second vehicle.

[0030] Optionally, determining whether the license plate to be identified is a fake license plate or a genuine license plate based on the judgment result includes:

[0031] If the judgment result is that the number of the license plate areas to be identified in the target image recognition area is less than the preset value, then the license plate to be identified is determined to be a fake license plate;

[0032] If the judgment result is that the number of the license plate areas to be identified in the target image recognition area is greater than or equal to a preset value, it is determined that the license plate to be identified is a real license plate.

[0033] Optionally, obtaining a target image recognition area based on multiple license plate areas of multiple vehicles includes:

[0034] Extract image features of the license plate area of ​​each vehicle;

[0035] Dividing the plurality of license plate regions into a plurality of license plate region sets according to image features of the license plate region of each vehicle;

[0036] For each license plate region set, determining a bounding box according to the position of each license plate region in the license plate region set in the vehicle image;

[0037] The image area contained in the bounding box corresponding to each license plate area set is used as the target image recognition area.

[0038] Optionally, the image features are pixels;

[0039] According to the image features of the license plate area of ​​each vehicle, the multiple license plate areas are divided into multiple license plate area sets, including:

[0040] Determine the maximum pixel and the minimum pixel based on the pixels of the license plate area of ​​each vehicle;

[0041] Divide the pixel range consisting of the maximum pixel and the minimum pixel into equal intervals to obtain multiple pixel range intervals;

[0042] According to the pixels of the license plate area, the license plate area corresponding to each pixel range interval is obtained from the multiple license plate areas to obtain multiple license plate area sets.

[0043] Optionally, each license plate area includes a left boundary line, a right boundary line, an upper boundary line, and a lower boundary line.

[0044] Determine a bounding box according to the position of each license plate region in the license plate region set in the vehicle image, including:

[0045] Determine the first target license plate area with the leftmost left boundary line, the second target license plate area with the rightmost right boundary line, the third target license plate area with the uppermost upper boundary line, and the fourth target license plate area with the lowermost lower boundary line in the license plate area set;

[0046] A bounding box is determined based on the left boundary line of the target first license plate area, the right boundary line of the target second license plate area, the upper boundary line of the target third license plate area, and the lower boundary line of the target fourth license plate area.

[0047] Optionally, determining whether the license plate to be identified is a fake license plate or a genuine license plate based on the judgment result includes:

[0048] If the judgment result is that the total number of the license plate areas to be identified in all target image recognition areas is less than a preset value, the license plate to be identified is determined to be a fake license plate;

[0049] If the judgment result is that the total number of the license plate areas to be identified in all target image recognition areas is greater than or equal to a preset value, it is determined that the license plate to be identified is a real license plate.

[0050] In a second aspect, an embodiment of the present application provides an image processing device, including:

[0051] An acquisition unit, used for acquiring multiple frames of images to be identified that are continuously acquired by an image acquisition device at the gate site;

[0052] An extraction unit is used to extract the image area of ​​the same license plate to be recognized from each frame of the image to be recognized, and obtain multiple license plate areas to be recognized and the positions of the license plate areas to be recognized in the image to be recognized;

[0053] an obtaining unit, configured to obtain a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear;

[0054] A judgment unit, configured to judge whether the number of license plate regions to be recognized located in the target image recognition area meets a preset condition based on the positions of the plurality of license plate regions to be recognized, and obtain a judgment result;

[0055] The determination unit is used to determine whether the license plate to be identified is a fake license plate or a real license plate based on the judgment result.

[0056] In a third aspect, an embodiment of the present application provides an image processing device, including:

[0057] CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply;

[0058] The memory is either transient or persistent storage;

[0059] The central processing unit is configured to communicate with the memory and execute instructions in the memory to perform the aforementioned license plate recognition method.

[0060] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which includes instructions. When the instructions are executed on a computer, the computer executes the aforementioned license plate recognition method.

[0061] In a fifth aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when run on a computer, enables the computer to execute the aforementioned license plate recognition method.

[0062] In the sixth aspect, an embodiment of the present application provides a chip system, which includes at least one processor and a communication interface. The communication interface and at least one processor are interconnected through lines, and the at least one processor is used to run computer programs or instructions to execute the aforementioned license plate recognition method.

[0063] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0064] The embodiment of the present application can determine whether the license plate to be identified is a fake license plate or a real license plate based on the positions of multiple license plate areas to be identified, and whether the number of license plate areas to be identified located in the target image recognition area meets the preset conditions, thereby improving the accuracy of determining whether the identified license plate is a fake license plate or a real license plate, thereby reducing the number of artificially forged fake license plates. For example, license plate images displayed on mobile phones or printed license plate images can be identified as license plates by license plate recognition technology, and after being identified as license plates, the gate equipment is controlled to open for release, thereby reducing the probability of fare evasion and improving the robustness of the license plate anti-counterfeiting effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic diagram of the architecture of an image recognition system disclosed in an embodiment of the present application;

[0066] Figure 2 A flowchart of a license plate recognition method disclosed in an embodiment of the present application;

[0067] Figure 3 A flowchart of another license plate recognition method disclosed in an embodiment of the present application;

[0068] Figure 4 A schematic diagram of the license plate area disclosed in an embodiment of the present application;

[0069] Figure 5 A schematic diagram of the upper and lower boundary lines disclosed in the embodiments of this application;

[0070] Figure 6 A schematic diagram of a scenario disclosed in an embodiment of the present application for determining whether the license plate area to be identified is located in the target image recognition area;

[0071] Figure 7 A flowchart of another license plate recognition method disclosed in an embodiment of the present application;

[0072] Figure 8 A schematic diagram of a license plate area divided based on pixels disclosed in an embodiment of the present application;

[0073] Figure 9 This is a schematic diagram of another license plate area divided based on pixels disclosed in an embodiment of the present application;

[0074] Figure 10 This is a schematic diagram of another license plate area divided based on pixels disclosed in an embodiment of the present application;

[0075] Figure 11 A schematic diagram of another scenario disclosed in an embodiment of the present application for determining whether the license plate area to be identified is located in the target image recognition area;

[0076] Figure 12A schematic diagram of another scenario disclosed in an embodiment of the present application for determining whether the license plate area to be identified is located in the target image recognition area;

[0077] Figure 13 A schematic structural diagram of an image processing device disclosed in an embodiment of the present application;

[0078] Figure 14 A schematic structural diagram of another image processing device disclosed in an embodiment of the present application;

[0079] Figure 15 This is a schematic structural diagram of another image processing device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0080] The embodiments of the present application provide a license plate recognition method, an image processing device, and a computer-readable storage medium, which can identify fake license plates and real license plates.

[0081] See also Figure 1 In the embodiment of the present application, the image recognition architecture includes:

[0082] An image processing device 101 and an image acquisition device 102 set up at the gate site.

[0083] When performing image recognition, image processing device 101 connects to image acquisition device 102, which captures and sends images to image processing device 101. Image processing device 101 can be a standalone processor or a server cluster, which can perform a series of image processing operations. Image acquisition device 102 can be a standalone processor or a server cluster, which can also capture images.

[0084] based on Figure 1 The image recognition architecture shown in Figure 2 In one embodiment of the license plate recognition method of the present application, the method includes:

[0085] 201. Acquire multiple frames of images to be identified that are continuously captured by an image acquisition device at the gate site.

[0086] In this embodiment, the image processing device may obtain multiple frames of images to be recognized sent by the image acquisition device, and store the multiple frames of images to be recognized in a local database for subsequent use.

[0087] 202. Extract the image region of the same license plate to be recognized from each frame of the image to be recognized, and obtain multiple license plate regions to be recognized and the positions of the license plate regions to be recognized in the image to be recognized.

[0088] After obtaining multiple frames of images to be identified continuously captured by the image acquisition device at the gate site, the image processing device can extract the image area of ​​the same license plate to be identified from each frame of the image to be identified, thereby obtaining multiple license plate areas to be identified and the positions of the license plate areas to be identified in the images to be identified.

[0089] 203. Obtain a target image recognition area generated for the gate site, where the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear.

[0090] The image processing device may obtain in advance a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear.

[0091] In this embodiment, there is no time sequence relationship between step 203, step 201 and step 202.

[0092] 204. According to the positions of the plurality of license plate regions to be identified, determine whether the number of license plate regions to be identified in the target image recognition area meets a preset condition, and obtain a determination result.

[0093] After obtaining multiple license plate areas to be identified, the positions of the license plate areas to be identified in the image to be identified, and the target image recognition area, the image processing device can determine whether the number of license plate areas to be identified located in the target image recognition area meets preset conditions based on the positions of the multiple license plate areas to be identified, thereby obtaining a judgment result.

[0094] 205. According to the judgment result, determine whether the license plate to be identified is a fake license plate or a real license plate.

[0095] After obtaining the judgment result, the image processing device can determine whether the license plate to be identified is a fake license plate or a real license plate based on the judgment result.

[0096] In the embodiment of the present application, the image processing device can determine whether the license plate to be identified is a fake license plate or a real license plate based on the positions of multiple license plate areas to be identified by determining whether the number of license plate areas to be identified located in the target image recognition area meets preset conditions, thereby improving the accuracy of determining whether the identified license plate is a fake license plate or a real license plate, and reducing the number of artificially forged fake license plates, such as license plate images displayed on mobile phones or printed license plate images, which can be identified as license plates by license plate recognition technology, and after being identified as license plates, the gate device is controlled to open for release, thereby reducing the probability of fare evasion and improving the robustness of the license plate anti-counterfeiting effect.

[0097] In the embodiment of the present application, there are multiple ways to obtain the target image recognition area generated on-site for the gate, which are described below:

[0098] 1. The image processing device obtains the target image recognition area based on the upper left corner and lower right corner of the license plate area:

[0099] In the embodiment of the present application, the image processing device can obtain the target image recognition area based on the upper left corner point and the lower right corner point of the license plate area. For the specific process of obtaining the target image recognition area based on the upper left corner point and the lower right corner point of the license plate area, please refer to Figure 3 ,The process of obtaining the target image recognition area includes:

[0100] 301. An image processing device obtains vehicle images continuously captured by an image acquisition device from a plurality of vehicles passing through a gate.

[0101] Before performing license plate recognition, the image processing device may first obtain vehicle images continuously captured by the image acquisition device of multiple vehicles passing through the gate, wherein each vehicle corresponds to multiple frames of vehicle images.

[0102] Specifically, for example, the image processing device may first obtain vehicle images continuously captured by the image acquisition device for 100 vehicles passing through the gate, where each vehicle corresponds to multiple frames of vehicle images. Of course, the number of vehicles can be set according to actual conditions and is not limited to the above values.

[0103] 302. The image processing device extracts the license plate image area of ​​each vehicle from the multiple frames of vehicle images corresponding to each vehicle, and obtains the license plate area of ​​each vehicle and the position of the license plate area of ​​each license plate.

[0104] After acquiring vehicle images continuously captured by the image acquisition device for multiple vehicles passing through the gate, the image processing device can extract the vehicle license plate image area from the multiple frames of vehicle images corresponding to each vehicle, thereby obtaining the license plate area of ​​each vehicle and the location of the license plate area of ​​each license plate, wherein each vehicle may have multiple license plate areas. Figure 4 , Figure 4 This is an example diagram of the license plate area disclosed in the embodiments of this application. The area enclosed by the box in the diagram represents the license plate area, and the position of the area enclosed by the box represents the position of the license plate area.

[0105] 303. The image processing device determines the upper left corner point and the lower right corner point of each license plate area according to the position of each license plate area in the vehicle image.

[0106] After obtaining the license plate area of ​​each vehicle and the position of the license plate area of ​​each license plate, the image processing device can determine the upper left corner point and the lower right corner point of each license plate area according to the position of each license plate area in the vehicle image.

[0107] It is understandable that in addition to determining the upper left corner and lower right corner of each license plate area, it is also possible to determine the point one tenth of the distance from the upper left corner and the point one tenth of the distance from the lower right corner in each license plate area. It is also possible to determine the coordinate point of any point in the upper left area of ​​the license plate area and the coordinate point of any point in the lower right area of ​​the license plate area. Here, any coordinate point in the upper left area of ​​the license plate area is the first coordinate point, and any coordinate point in the lower right area of ​​the license plate area is the second coordinate point. The specific details are not limited here. Please continue to refer to Figure 4 , the upper left corner of the box represents the upper left corner of the license plate area, and the lower right corner of the box represents the lower right corner of the license plate area.

[0108] 304. The image processing device determines, for the same vehicle, a first average vertical coordinate of the vehicle based on the vertical coordinate values ​​of the upper left corner points of multiple license plate areas of the vehicle, and determines, based on the first average vertical coordinate values ​​of the multiple vehicles, the vehicle corresponding to the smallest first average vertical coordinate value as the target first vehicle. For the same vehicle, the image processing device determines, based on the vertical coordinate values ​​of the lower right corner points of multiple license plate areas of the vehicle, a second average vertical coordinate of the vehicle based on the vertical coordinate values ​​of the multiple vehicles, and determines, based on the second average vertical coordinate values ​​of the multiple vehicles, the vehicle corresponding to the largest second average vertical coordinate value as the target second vehicle.

[0109] Specifically, after determining the upper left corner point and the lower right corner point of each license plate area, the image processing device can determine the first average vertical coordinate of the vehicle for the same vehicle based on the vertical coordinate values ​​of the upper left corner points of multiple license plate areas of the vehicle, and determine the vehicle corresponding to the smallest first average vertical coordinate value among 100 vehicles as the target first vehicle based on the first average vertical coordinate values ​​of multiple vehicles, and can determine the second average vertical coordinate value of the vehicle for the same vehicle based on the vertical coordinate values ​​of the lower right corner points of multiple license plate areas of the vehicle, and determine the vehicle corresponding to the largest second average vertical coordinate value among 100 vehicles as the target second vehicle based on the second average vertical coordinate values ​​of multiple vehicles.

[0110] It is understandable that in addition to using the vertical coordinate value as the basis for determining the target vehicle, the horizontal coordinate value can also be used as the basis for determining the target vehicle, and other numerical values ​​representing the position can also be used as the basis for determining the target vehicle. The specific details are not limited here.

[0111] It can also be understood that in addition to determining the vehicle corresponding to the smallest first average vertical coordinate value as the target first vehicle and the vehicle corresponding to the largest second average vertical coordinate value as the target second vehicle, the vehicle corresponding to the largest first average vertical coordinate value can also be determined as the target first vehicle, and the vehicle corresponding to the lowest second average vertical coordinate value can be determined as the target second vehicle. It can also be that any first average vertical coordinate value determines the target first vehicle, and any second average vertical coordinate value determines the target second vehicle. The specific details are not limited here.

[0112] 305. The image processing device fits an upper boundary line according to first coordinate points of multiple license plate areas of the target first vehicle, and fits a lower boundary line according to second coordinate points of multiple license plate areas of the target second vehicle.

[0113] After determining the target first vehicle and the target second vehicle, the image processing device can fit the upper boundary line based on the first coordinate points of multiple license plate areas of the target first vehicle, and can fit the lower boundary line based on the second coordinate points of multiple license plate areas of the target second vehicle.

[0114] It is understandable that, in addition to using the least squares method to fit the upper boundary line of the first coordinate points of the multiple license plate areas of the target first vehicle, and using the least squares method to fit the lower boundary line of the second coordinate points of the multiple license plate areas of the target second vehicle, an approximation algorithm can also be used to fit the upper and lower boundary lines, and other fitting algorithms can also be used to fit the upper and lower boundary lines, which are not limited here. Figure 5 , Figure 5 This is an example diagram of the upper and lower boundary lines disclosed in the embodiments of this application. The upper straight line in the diagram is the upper boundary line, and the lower straight line is the lower boundary line.

[0115] 306. The image processing device determines the image area between the upper boundary line and the lower boundary line as a target image recognition area, where the target image recognition area is used to represent the image area where the license plate of the vehicle moving through the gate site may appear.

[0116] After fitting the upper and lower boundary lines, the image processing device can determine the image area between the upper and lower boundary lines as the target image recognition area, where the target image recognition area is used to represent the image area where the license plate of the vehicle moving through the gate may appear. Figure 5 ,The area between the two straight lines in the figure represents the target image recognition area.

[0117] The above is the process of the image processing device obtaining the target image recognition area based on the upper left corner point and the lower right corner point of the license plate area. The following describes how to use the target image recognition area to judge the authenticity of the license plate to be identified.

[0118] The image processing device can obtain multiple frames of images to be recognized sent by the image acquisition device, and store the multiple frames of images to be recognized in a local database for subsequent use.

[0119] After obtaining multiple frames of images to be identified sent by the image acquisition device, the image processing device can extract the image area of ​​the same license plate to be identified from each frame of the image to be identified, thereby obtaining multiple license plate areas to be identified and the positions of the license plate areas to be identified in the images to be identified.

[0120] After obtaining multiple license plate regions to be recognized, their locations within the image to be recognized, and determining them as target image recognition regions, the image processing device can determine, based on the locations of the multiple license plate regions to be recognized, whether the number of license plate regions to be recognized within the target image recognition region satisfies a preset condition, and thereby obtain a determination result. Specifically, as long as any point within the license plate region to be recognized is within the target image recognition region, the license plate region to be recognized is determined to be within the target image recognition region.

[0121] It can be understood that the image processing device initially obtains multiple frames of images to be identified captured by the image acquisition device in a continuous manner, so the basis for determining whether the license plate to be identified is a fake license plate or a real license plate can be determined by judging whether the number of license plate areas to be identified located in the target image recognition area meets the preset conditions.

[0122] After obtaining the judgment result, the image processing device can determine whether the license plate to be identified is a fake license plate or a real license plate based on the judgment result. If the judgment result is that the number of license plate areas to be identified located in the target image recognition area is less than a preset value, the license plate to be identified is determined to be a fake license plate. If the judgment result is that the number of license plate areas to be identified located in the target image recognition area is greater than or equal to the preset value, the license plate to be identified is determined to be a real license plate.

[0123] Generally speaking, the preset value is around 10. For example, if the preset value is set to 10, if the number of license plate areas corresponding to the license plate to be recognized that are located in the target image recognition area is 4, then the license plate to be recognized is determined to be a fake license plate. If the number of license plate areas corresponding to the license plate to be recognized that are located in the target image recognition area is 12, then the license plate to be recognized is determined to be a genuine license plate. Of course, the preset value can be determined according to actual needs and is not limited to the above value.

[0124] See also Figure 6 , Figure 6This is an example diagram for determining whether the area of ​​a license plate to be identified is located in the target image recognition area, as disclosed in an embodiment of the present application. An image in the upper left portion of the image displays a license plate picture displayed by a person holding a mobile phone or a printed license plate picture, and an image in the lower right portion of the image also displays a license plate picture displayed by a person holding a mobile phone or a printed license plate picture. These two images represent two frames of images to be identified that were continuously captured by an image acquisition device at the gate site and obtained by an image processing device. These two frames of images to be identified correspond to the same license plate to be identified. The area of ​​the license plate to be identified and its position in the image to be identified are extracted from these two frames of images to be identified, respectively. The area of ​​the license plate picture displayed by a person holding a mobile phone or the printed license plate picture in the diagram is the area of ​​the license plate to be identified. It can be seen that all points in the area of ​​the license plate picture displayed by a person holding a mobile phone or the printed license plate picture are not within the range between the two straight lines in the diagram, so it can be determined that the license plate to be identified is a fake license plate.

[0125] In this embodiment, the image processing device can obtain the target image recognition area based on the upper left corner point and the lower right corner point of the license plate area, determine the target first license plate by determining the minimum average vertical coordinate value, determine the target second vehicle by determining the maximum average vertical coordinate, and fit the upper boundary line and the lower boundary line according to the upper left corner point of the target first license plate and the lower right corner point of the target second license plate, and determine the image area between the upper boundary line and the lower boundary line as the target image recognition area, thereby reducing the range of the target image recognition area and improving the accuracy of identifying whether the license plate is a fake license plate or a real license plate.

[0126] 2. The image processing device obtains the target image recognition area based on the pixels, left boundary, right boundary, upper boundary and lower boundary of the license plate area of ​​each vehicle:

[0127] In the embodiment of the present application, the image processing device can obtain the target image recognition area based on the pixels, left boundary, right boundary, upper boundary and lower boundary of the license plate area of ​​each vehicle. For the specific process of obtaining the target image recognition area based on the pixels, left boundary, right boundary, upper boundary and lower boundary of the license plate area of ​​each vehicle, please refer to Figure 7 ,Another process of obtaining the target image recognition area includes:

[0128] 701. An image processing device obtains vehicle images continuously captured by an image capturing device from a plurality of vehicles passing through a gate.

[0129] 702. The image processing device extracts the license plate image area of ​​each vehicle from the multiple frames of vehicle images corresponding to each vehicle, and obtains the license plate area of ​​each vehicle and the position of the license plate area of ​​each license plate.

[0130] In this embodiment, steps 701 to 702 are the same as those described above. Figure 3 Steps 301 to 302 in the illustrated embodiment are similar and will not be described again here.

[0131] 703. The image processing device extracts pixels of the license plate area of ​​each vehicle.

[0132] After obtaining the license plate area of ​​each vehicle, the image processing device can extract pixels of the license plate area of ​​each vehicle, and the pixels are used to represent the size of the license plate area of ​​each vehicle.

[0133] A pixel is the smallest unit in an image represented by a sequence of numbers and can be used to indicate the size of the image. For images with the same resolution, larger pixels indicate larger image size, while smaller pixels indicate smaller image size. A typical image resolution is 72, meaning there are 72 pixels per inch. One inch equals 2.54 centimeters, so by conversion, we can derive 28 pixels per centimeter. For example, a 15 cm by 15 cm image equals 420 by 420 pixels. For example, when extracting the license plate image area from a vehicle image, the size of the license plate image area is 0.357 cm by 1.071 cm, or approximately 10 by 30 pixels.

[0134] Of course, pixel is only an image unit representing the size of the license plate area of ​​each vehicle. In addition, it can also be an area unit or any other parameter that can represent the size of the image area. The specifics are not limited here.

[0135] 704. The image processing device determines the maximum pixel and the minimum pixel based on the pixels of the license plate area of ​​each vehicle, divides the pixel range composed of the maximum pixel and the minimum pixel into equal intervals to obtain multiple pixel range intervals, and obtains the license plate area corresponding to each pixel range interval from the multiple license plate areas based on the pixels of the license plate area to obtain multiple license plate area sets.

[0136] After extracting the pixels of the license plate area of ​​each vehicle, the image processing device can determine the maximum pixel and the minimum pixel based on the pixels of the license plate area of ​​each vehicle, and divide the pixel range composed of the maximum pixel and the minimum pixel into equal intervals to obtain multiple pixel range intervals. Then, based on the pixels of the license plate area, the license plate area corresponding to each pixel range interval is obtained from the multiple license plate areas to obtain multiple license plate area sets.

[0137] For example, the number of pixels in the license plate area of ​​each vehicle extracted by the image processing device is 100, 120, 110, 200, 150, 400, 350, etc. The pixels of the largest license plate area and the smallest license plate area can be determined first, thereby determining that the pixels of the largest license plate area are 400 and the pixels of the smallest license plate area are 100. Then, the pixel range consisting of 100 and 400 is divided into 100-pixel intervals, resulting in pixel ranges of 100-200, 201-300, and 301-400. Then, based on the pixels of the license plate area, the license plate area corresponding to each pixel range is obtained from multiple license plate areas, thereby obtaining a set of license plate areas of 100-200 pixels, a set of license plate areas of 201-300 pixels, and a set of license plate areas of 301-400 pixels. Please refer to [ 100-200 pixels ] for more information. Figure 8 、 Figure 9 and Figure 10 , Figure 8 、 Figure 9 and Figure 10 This is an example diagram of the license plate area divided according to pixels disclosed in the embodiment of this application. Figure 8 The area enclosed by the small squares in the large square represents the set of license plate areas corresponding to the range of 100-200 pixels. Figure 9 The area enclosed by the small squares in the large square represents the set of license plate areas corresponding to the range of 200-300 pixels. Figure 10 The area enclosed by the small boxes in the medium and large boxes represents the set of license plate areas corresponding to the range of 300-400 pixels.

[0138] Of course, the pixels of the maximum license plate area and the pixels of the minimum license plate area can be determined based on the pixels of the license plate area actually extracted, and are not limited to the above values.

[0139] 705. The image processing device determines, in the license plate area set, the target first license plate area with the leftmost left boundary line, the target second license plate area with the rightmost right boundary line, the target third license plate area with the uppermost upper boundary line, and the target fourth license plate area with the lowermost lower boundary line, and determines a bounding box based on the left boundary line of the target first license plate area, the right boundary line of the target second license plate area, the upper boundary line of the target third license plate area, and the lower boundary line of the target fourth license plate area.

[0140] After obtaining multiple license plate area sets, the image processing device can determine the target first license plate area with the leftmost left boundary line, the target second license plate area with the rightmost right boundary line, the target third license plate area with the uppermost upper boundary line, and the target fourth license plate area with the lowermost lower boundary line in the license plate area set, and can determine the bounding box based on the left boundary line of the target first license plate area, the right boundary line of the target second license plate area, the upper boundary line of the target third license plate area, and the lower boundary line of the target fourth license plate area.

[0141] Specifically, taking the license plate area set of 100-200 pixels as an example, this license plate area set has many license plate areas. The image processing device can determine the target first license plate area with the leftmost left boundary line, the target second license plate area with the rightmost right boundary line, the target third license plate area with the uppermost upper boundary line, and the target fourth license plate area with the lowermost lower boundary line. The left boundary line of the target first license plate area is the left boundary line of the bounding box, the right boundary line of the target second license plate area is the right boundary line of the bounding box, the upper boundary line of the target third license plate area is the upper boundary line of the bounding box, and the lower boundary line of the target fourth license plate area is the lower boundary line of the bounding box. By analogy, the image processing device can determine the bounding box of the license plate area set of 201-300 pixels and the bounding box of the license plate area set of 301-400 pixels. Please continue to refer to Figure 8 、 Figure 9 and Figure 10 , Figure 8 、 Figure 9 and Figure 10 The largest box in represents the bounding box.

[0142] 706. The image processing device uses the image area contained in the bounding box corresponding to each license plate area set as the target image recognition area.

[0143] After determining the bounding box, the image processing device can use the image area contained in the bounding box corresponding to each license plate area set as the target image recognition area. Figure 8 、 Figure 9 and Figure 10 , Figure 8 、 Figure 9 and Figure 10 The area enclosed by the largest box in represents the target image recognition area.

[0144] The above is the process of the image processing device obtaining the target image recognition area based on the pixels, left boundary, right boundary, upper boundary and lower boundary of the license plate area of ​​each vehicle. The following describes how to use the target image recognition area to judge the authenticity of the license plate to be identified.

[0145] The image processing device can obtain multiple frames of images to be recognized sent by the image acquisition device, and store the multiple frames of images to be recognized in a local database for subsequent use.

[0146] After obtaining multiple frames of images to be identified sent by the image acquisition device, the image processing device can extract the image area of ​​the same license plate to be identified from each frame of the image to be identified, thereby obtaining multiple license plate areas to be identified and the positions of the license plate areas to be identified in the images to be identified.

[0147] After obtaining multiple license plate areas to be identified, the positions of the license plate areas to be identified in the image to be identified, and determining them as target image recognition areas, the image processing device can judge whether the number of license plate areas to be identified located in the target image recognition area meets preset conditions based on the positions of the multiple license plate areas to be identified, and obtain a judgment result.

[0148] Specifically, as long as any point in the license plate area to be recognized is located in the target image recognition area, it is determined that the license plate area to be recognized is located in the target image recognition area.

[0149] It can be understood that the image processing device initially obtains multiple frames of images to be identified captured by the image acquisition device in a continuous manner, so the number of license plate areas to be identified located in the target image recognition area can be judged to determine whether the license plate to be identified is a fake license plate or a real license plate.

[0150] After obtaining the judgment result, the image processing device can determine whether the license plate to be identified is a fake license plate or a real license plate based on the judgment result. If the judgment result is that the total number of license plate areas to be identified located in all target image recognition areas is less than a preset value, it can be determined that the license plate to be identified is a fake license plate. If the judgment result is that the total number of license plate areas to be identified located in all target image recognition areas is greater than or equal to the preset value, it can be determined that the license plate to be identified is a real license plate.

[0151] Please continue reading Figure 8 、 Figure 9 and Figure 10 , Figure 8 The area enclosed by the largest box in the image represents the target image recognition area corresponding to 100-200 pixels. Figure 9 The area enclosed by the largest box in the image represents the target image recognition area corresponding to 200-300 pixels. Figure 10 The area enclosed by the largest box in the image represents the target image recognition area corresponding to 300-400 pixels. Figure 8 、 Figure 9 and Figure 10 The pixels corresponding to the area enclosed by the largest box can be set according to actual conditions and are not limited to the above values.

[0152] See also Figure 11 , Figure 11 This is a schematic diagram of another method disclosed in an embodiment of the present application for determining whether the license plate area to be identified is located in the target image recognition area. Figure 11 The area enclosed by the largest box in the image represents the target image recognition area corresponding to 100-200 pixels. Figure 11The lower right part is an image to be identified. The content of the image to be identified is a license plate picture printed by a person holding a mobile phone or a printed license plate picture. The license plate picture in the figure is the license plate area to be identified. It can be seen that the license plate area to be identified is not located in the target image recognition area corresponding to 100-200 pixels.

[0153] See also Figure 12 , Figure 12 This is a schematic diagram of another method disclosed in an embodiment of the present application for determining whether the license plate area to be identified is located in the target image recognition area. Figure 12 The area enclosed by the largest box in the image represents the target image recognition area corresponding to 300-400 pixels. Figure 12 The lower right part is an image to be recognized. The content of the image to be recognized is a license plate picture printed by a person holding a mobile phone or a printed license plate picture. The license plate picture in the figure is the license plate area to be recognized. It can be seen that the license plate area to be recognized is not located in the target image recognition area corresponding to 300-400 pixels. Of course, Figure 11 and Figure 12 The pixels corresponding to the area enclosed by the largest box can be set according to actual conditions and are not limited to the above values.

[0154] Generally, the preset value is around 10. Specifically, for example, if the preset value is set to 10, if the license plate area to be identified corresponding to a license plate to be identified is located in the license plate area set of 100-200 pixels, the number of target image recognition areas determined by the bounding box is 1, the number of target image recognition areas determined by the bounding box is located in the license plate area set of 201-300 pixels, and the number of target image recognition areas determined by the bounding box is located in the license plate area set of 301-400 pixels, then the total number is 4, and the license plate to be identified is determined to be a fake license plate. If the license plate area to be identified corresponding to a license plate to be identified is located in the license plate area set of 100-200 pixels, the number of target image recognition areas determined by the bounding box of the license plate area set is 2, the number of target image recognition areas determined by the bounding box of the license plate area set of 201-300 pixels is 5, and the number of target image recognition areas determined by the bounding box of the license plate area set of 301-400 pixels is 5, then the total number is 12, and the license plate to be identified is determined to be a real license plate.

[0155] It can be understood that in addition to determining that the license plate areas to be identified corresponding to the vehicles to be identified are located in all target image recognition areas, the total number of license plate areas to be identified that are located in the target image recognition area can also be determined by matching the target image recognition area corresponding to the pixel range interval of the license plate areas to be identified, and then adding up the number of vehicles to be identified whose vehicle areas to be identified are located in the target image recognition area to determine the total number of vehicles to be identified whose license plate areas to be identified are located in the target image recognition area.

[0156] For example, if a license plate to be identified has 12 license plate areas to be identified, the pixels from small to large are 95, 120, 200, 210, 240, 260, 270, 280, 290, 310, 330, and 340 respectively. Among them, the license plate area of ​​pixel 120 is located in the target image recognition area corresponding to 100-200 pixels, the license plate area of ​​pixel 200 is not located in the target image recognition area corresponding to 100-200 pixels, the license plate areas of pixels 200, 280, and 290 are located in the target image recognition area corresponding to 200-300 pixels, the license plate areas of pixels 210, 240, 260, and 270 are not located in the target image recognition area corresponding to 200-300 pixels, the license plate area of ​​pixel 310 is located in the target image recognition area corresponding to 300-400 pixels, and the license plate areas of pixels 330 and 340 are not located in the target image recognition area corresponding to 300-400 pixels. Therefore, the total number of license plate areas of the license plate to be recognized located in the recognition image area is 5, which is less than the preset value 10, so it is determined that the license plate to be recognized is a fake license plate.

[0157] Determining the total number of license plate areas to be identified that are located in the target image recognition area may also be determining the total number of license plate areas to be identified corresponding to vehicles to be identified that are located in any one or more target image recognition areas, which is not specifically limited here.

[0158] In this embodiment, the image processing device can obtain multiple target image recognition areas based on the pixels, left boundary, right boundary, upper boundary and lower boundary of the license plate area of ​​each vehicle, while ensuring the accuracy of identifying whether the license plate is a fake license plate or a real license plate, and to a certain extent improving the fault tolerance of license plate recognition.

[0159] The above describes the license plate recognition method in the embodiment of the present application. The following describes the image processing device in the embodiment of the present application. Figure 13 , an embodiment of the image processing device in the embodiment of the present application includes:

[0160] An acquisition unit 1301 is configured to acquire multiple frames of to-be-recognized images 301 continuously acquired at the gate site by the image acquisition device 102;

[0161] The extraction unit 1302 is configured to extract the image region of the same license plate to be recognized from each frame of the image to be recognized 301 acquired by the acquisition unit 1301, and obtain multiple license plate regions to be recognized and the positions of the license plate regions to be recognized in the image to be recognized;

[0162] An obtaining unit 1303 is configured to obtain a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear;

[0163] The judging unit 1304 is configured to judge, based on the positions of the plurality of license plate regions to be recognized obtained by the extracting unit 1302, whether the number of license plate regions to be recognized obtained by the extracting unit 1302 located in the target image recognition area obtained by the obtaining unit 1303 satisfies a preset condition, and obtain a judgment result;

[0164] The determination unit 1305 is used to determine whether the license plate to be identified is a fake license plate or a real license plate according to the judgment result obtained by the judgment unit 1304.

[0165] In the embodiment of the present application, the image processing device can determine whether the number of license plate areas to be identified located in the target image recognition area meets the preset conditions based on the positions of multiple license plate areas to be identified, so as to determine whether the license plate to be identified is a fake license plate or a real license plate, thereby improving the accuracy of determining whether the identified license plate is a fake license plate or a real license plate, thereby reducing the number of artificially forged fake license plates. For example, license plate images displayed on mobile phones or printed license plate images can be identified as license plates by license plate recognition technology, and after being identified as license plates, the gate device is controlled to open for release, thereby reducing the probability of fare evasion and improving the robustness of the license plate anti-counterfeiting effect.

[0166] The image processing device in the embodiment of the present application is described in detail below. Figure 14 Another embodiment of the image processing device in the embodiment of the present application includes:

[0167] An acquisition unit 1401 is configured to acquire multiple frames of images to be identified that are continuously acquired by an image acquisition device at the gate site;

[0168] An extraction unit 1402 is configured to extract an image region of the same license plate to be recognized from each frame of the image to be recognized, thereby obtaining multiple license plate regions to be recognized and positions of the license plate regions to be recognized in the image to be recognized;

[0169] An obtaining unit 1403 is configured to obtain a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear;

[0170] The judgment unit 1404 is used to judge whether the number of license plate regions to be recognized located in the target image recognition area meets a preset condition based on the positions of the multiple license plate regions to be recognized, and obtain a judgment result;

[0171] The determination unit 1405 is used to determine whether the license plate to be identified is a fake license plate or a real license plate according to the judgment result.

[0172] The acquisition unit 1403 is specifically used to obtain vehicle images continuously captured by the image acquisition device for multiple vehicles passing through the gate, wherein each vehicle corresponds to multiple frames of vehicle images, and for the multiple frames of vehicle images corresponding to each vehicle, the vehicle's license plate image area is extracted from the multiple frames of vehicle images to obtain the license plate area of ​​each vehicle and the position of the license plate area of ​​each vehicle, wherein each vehicle has multiple license plate areas, and the target image recognition area is obtained based on the multiple license plate areas of the multiple vehicles.

[0173] The acquisition unit 1403 is specifically used to determine the first coordinate point and the second coordinate point of each license plate area according to the position of each license plate area in the vehicle image, wherein the first coordinate point is located in the upper left area of ​​the license plate area, and the second coordinate point is located in the lower right area of ​​the license plate area, determine the target first vehicle among the multiple vehicles according to the first coordinate points of multiple license plate areas of the multiple vehicles, and determine the target second vehicle among the multiple vehicles according to the second coordinate points of multiple license plate areas of the multiple vehicles, fit the upper boundary line according to the first coordinate points of the multiple license plate areas of the target first vehicle, and fit the lower boundary line according to the second coordinate points of the multiple license plate areas of the target second vehicle, and determine the image area between the upper boundary line and the lower boundary line as the target image recognition area.

[0174] The determination unit 1405 is specifically used to determine, for the same vehicle, a first average coordinate value of the vehicle based on the first coordinate points of multiple license plate areas of the vehicle, and determine a target first vehicle among multiple vehicles based on the first average coordinate values ​​of the multiple vehicles; and for the same vehicle, determine a second average coordinate value of the vehicle based on the second coordinate points of multiple license plate areas of the vehicle, and determine a target second vehicle among multiple vehicles based on the second average coordinate values ​​of the multiple vehicles.

[0175] The determination unit 1405 is specifically used to determine the first average vertical coordinate value of the vehicle based on the vertical coordinate values ​​of the upper left corner points of multiple license plate areas of the vehicle, and determine the vehicle corresponding to the smallest first average vertical coordinate value as the target first vehicle based on the first average vertical coordinate values ​​of multiple vehicles; determine the second average vertical coordinate value of the vehicle based on the vertical coordinate values ​​of the lower right corner points of multiple license plate areas of the vehicle, and determine the vehicle corresponding to the largest second average vertical coordinate value as the target second vehicle based on the second average vertical coordinate values ​​of multiple vehicles.

[0176] The determination unit 1405 is specifically used to determine that the license plate to be identified is a fake license plate if the judgment result is that the number of license plate areas to be identified located in the target image recognition area is less than a preset value; if the judgment result is that the number of license plate areas to be identified located in the target image recognition area is greater than or equal to the preset value, then determine that the license plate to be identified is a real license plate.

[0177] The acquisition unit 1403 is specifically used to extract the image features of the license plate area of ​​each vehicle, divide the multiple license plate areas into multiple license plate area sets based on the image features of the license plate area of ​​each vehicle, and for each license plate area set, determine a bounding box based on the position of each license plate area in the license plate area set in the vehicle image, and use the image area contained in the bounding box corresponding to each license plate area set as the target image recognition area.

[0178] The acquisition unit 1403 is specifically used to determine the maximum pixel and the minimum pixel based on the pixels of the license plate area of ​​each vehicle, divide the pixel range composed of the maximum pixel and the minimum pixel into equal intervals to obtain multiple pixel range intervals, and obtain the license plate area corresponding to each pixel range interval from the multiple license plate areas based on the pixels of the license plate area to obtain multiple license plate area sets.

[0179] The determination unit 1405 is specifically used to determine the target first license plate area with the leftmost left boundary line, the target second license plate area with the rightmost right boundary line, the target third license plate area with the uppermost upper boundary line, and the target fourth license plate area with the lowermost lower boundary line in the license plate area set, and determine the bounding box based on the left boundary line of the target first license plate area, the right boundary line of the target second license plate area, the upper boundary line of the target third license plate area, and the lower boundary line of the target fourth license plate area.

[0180] The determination unit 1405 is specifically used to determine that the license plate to be identified is a fake license plate if the judgment result is that the total number of license plate areas to be identified located in all target image recognition areas is less than a preset value; if the judgment result is that the total number of license plate areas to be identified located in all target image recognition areas is greater than or equal to the preset value, then determine that the license plate to be identified is a real license plate.

[0181] In this embodiment, each unit in the image processing device performs the above Figure 3 and Figure 7 The operation of the image processing device in the illustrated embodiment will not be described in detail here.

[0182] See below Figure 15 Another embodiment of the image processing device in the embodiment of the present application includes:

[0183] CPU 1501, memory 1505, input / output interface 1504, wired or wireless network interface 1503 and power supply 1502;

[0184] The memory 1505 is a temporary storage memory or a permanent storage memory;

[0185] The CPU 1501 is configured to communicate with the memory 1505 and execute the instructions in the memory 1505 to perform the aforementioned Figure 3 and Figure 7The method in the embodiment shown.

[0186] The embodiment of the present application also provides a computer-readable storage medium, which includes instructions. When the instructions are executed on a computer, the computer executes the aforementioned Figure 3 and Figure 7 The method in the embodiment shown.

[0187] The present application also provides a computer program product comprising instructions, which, when executed on a computer, causes the computer to execute the aforementioned Figure 3 and Figure 7 The method in the embodiment shown.

[0188] The embodiment of the present application also provides a chip system, which includes at least one processor and a communication interface, wherein the communication interface and the at least one processor are interconnected through a line, and the at least one processor is used to run a computer program or instruction to execute the aforementioned Figure 3 and Figure 7 The method in the embodiment shown.

[0189] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0190] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0191] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0192] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0193] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk, and other media that can store program code.

Claims

1. A license plate recognition method, characterized in that: Applied to an image processing device, the image processing device is communicatively connected to an image acquisition device provided at a gate site, the method comprising: Acquire multiple frames of images to be identified that are continuously captured by the image acquisition device on the gate site; Extracting the image region of the same license plate to be recognized from each frame of the image to be recognized, respectively, to obtain a plurality of license plate regions to be recognized and positions of the license plate regions to be recognized in the image to be recognized; Obtaining a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear; According to the positions of the plurality of license plate areas to be identified, determining whether the number of license plate areas to be identified located in the target image recognition area meets a preset condition, and obtaining a determination result; According to the judgment result, determining whether the license plate to be identified is a fake license plate or a real license plate; The obtaining of a target image recognition area generated on-site for the gate includes: Acquire vehicle images continuously captured by the image acquisition device on a plurality of vehicles passing through the gate, wherein each vehicle corresponds to multiple frames of vehicle images; For each vehicle, extracting the license plate image region of each vehicle from the multiple vehicle image frames to obtain the license plate region of each vehicle and the position of the license plate region of each vehicle, wherein each vehicle may have multiple license plate regions. A target image recognition area is obtained based on the multiple license plate areas of the multiple vehicles.

2. The license plate recognition method according to claim 1, characterized in that: The step of obtaining a target image recognition area based on the plurality of license plate areas of the plurality of vehicles includes: Determining a first coordinate point and a second coordinate point of each license plate region according to a position of each license plate region in the vehicle image, wherein the first coordinate point is located in an upper left region of the license plate region, and the second coordinate point is located in a lower right region of the license plate region; determining a target first vehicle among the plurality of vehicles based on first coordinate points of a plurality of license plate areas of the plurality of vehicles, and determining a target second vehicle among the plurality of vehicles based on second coordinate points of a plurality of license plate areas of the plurality of vehicles; Fitting an upper boundary line according to first coordinate points of a plurality of license plate areas of the target first vehicle, and fitting a lower boundary line according to second coordinate points of a plurality of license plate areas of the target second vehicle; An image area between the upper boundary line and the lower boundary line is determined as a target image recognition area.

3. The license plate recognition method according to claim 2, characterized in that: The method of determining a target first vehicle among the plurality of vehicles based on first coordinate points of a plurality of license plate areas of the plurality of vehicles, and determining a target second vehicle among the plurality of vehicles based on second coordinate points of a plurality of license plate areas of the plurality of vehicles, comprises: For the same vehicle, determining a first average coordinate value of the vehicle according to first coordinate points of multiple license plate areas of the vehicle, and determining a target first vehicle among the multiple vehicles according to the first average coordinate values ​​of the multiple vehicles; For the same vehicle, a second average coordinate value of the vehicle is determined according to the second coordinate points of multiple license plate areas of the vehicle, and a target second vehicle is determined among the multiple vehicles according to the second average coordinate values ​​of the multiple vehicles.

4. The license plate recognition method according to claim 3, characterized in that: The step of determining a first average coordinate value of the vehicle based on first coordinate points of multiple license plate areas of the vehicle, and determining a target first vehicle from the multiple vehicles based on the first average coordinate values ​​of the multiple vehicles, comprises: Determining a first average ordinate value of the vehicle according to ordinate values ​​of upper left corner points of multiple license plate areas of the vehicle; According to the first average ordinate values ​​of the plurality of vehicles, determining a vehicle corresponding to the smallest first average ordinate value as a target first vehicle; The step of determining a second average coordinate value of the vehicle based on the second coordinate points of the plurality of license plate areas of the vehicle, and determining a target second vehicle from the plurality of vehicles based on the second average coordinate values ​​of the plurality of vehicles, comprises: determining a second average ordinate value of the vehicle according to the ordinate values ​​of the lower right corner points of the plurality of license plate areas of the vehicle; According to the second average ordinate values ​​of the plurality of vehicles, a vehicle corresponding to the largest second average ordinate value is determined as the target second vehicle.

5. The license plate recognition method according to any one of claims 1 to 4, characterized in that: Determining whether the license plate to be identified is a fake license plate or a genuine license plate based on the judgment result includes: If the judgment result is that the number of the license plate areas to be identified in the target image recognition area is less than a preset value, then the license plate to be identified is determined to be a fake license plate; If the judgment result is that the number of the license plate areas to be identified located in the target image recognition area is greater than or equal to a preset value, it is determined that the license plate to be identified is a real license plate.

6. The license plate recognition method according to claim 1, characterized in that: The step of obtaining a target image recognition area based on the plurality of license plate areas of the plurality of vehicles includes: Extract image features of the license plate area of ​​each vehicle; Dividing the plurality of license plate areas into a plurality of license plate area sets according to image features of the license plate area of ​​each vehicle; For each license plate region set, determining a bounding box according to a position of each license plate region in the license plate region set in the vehicle image; The image area contained in the bounding box corresponding to each license plate area set is used as the target image recognition area.

7. The license plate recognition method according to claim 6, characterized in that: The image features are pixels; The step of dividing the plurality of license plate areas into a plurality of license plate area sets according to the image features of the license plate area of ​​each vehicle comprises: Determining a maximum pixel and a minimum pixel based on the pixels of the license plate area of ​​each vehicle; Dividing the pixel range consisting of the maximum pixel and the minimum pixel into equal intervals to obtain a plurality of pixel range intervals; According to the pixels of the license plate area, the license plate area corresponding to each pixel range interval is obtained from the multiple license plate areas to obtain multiple license plate area sets.

8. The license plate recognition method according to claim 6, characterized in that: Each of the license plate areas includes a left boundary line, a right boundary line, an upper boundary line and a lower boundary line. Determining a bounding box according to a position of each license plate area in the set of license plate areas in the vehicle image includes: Determine the first target license plate area with the leftmost left boundary line, the second target license plate area with the rightmost right boundary line, the third target license plate area with the uppermost upper boundary line, and the fourth target license plate area with the lowermost lower boundary line in the license plate area set; A bounding box is determined according to the left boundary line of the target first license plate area, the right boundary line of the target second license plate area, the upper boundary line of the target third license plate area, and the lower boundary line of the target fourth license plate area.

9. The license plate recognition method according to any one of claims 6 to 8, characterized in that: Determining whether the license plate to be identified is a fake license plate or a genuine license plate based on the judgment result includes: If the judgment result is that the total number of the license plate areas to be identified in all the target image recognition areas is less than a preset value, then the license plate to be identified is determined to be a fake license plate; If the judgment result is that the total number of the license plate areas to be identified located in all the target image recognition areas is greater than or equal to a preset value, it is determined that the license plate to be identified is a real license plate.

10. An image processing device, which is communicatively connected to an image acquisition device set up at a gate site, characterized in that: include: An acquisition unit, configured to acquire a plurality of frames of images to be identified that are continuously acquired by the image acquisition device at the gate site; an extraction unit, configured to extract an image region of the same license plate to be recognized from each frame of the image to be recognized, and obtain a plurality of license plate regions to be recognized and positions of the license plate regions to be recognized in the image to be recognized; an obtaining unit, configured to obtain a target image recognition area generated for the gate site, wherein the target image recognition area is used to represent an image area where a license plate of a vehicle moving through the gate site may appear; a judgment unit, configured to judge whether the number of the license plate regions to be recognized located in the target image recognition area meets a preset condition based on the positions of the plurality of license plate regions to be recognized, and obtain a judgment result; a determination unit, configured to determine, based on the judgment result, whether the license plate to be identified is a fake license plate or a genuine license plate; The acquisition unit is specifically used to obtain vehicle images continuously captured by the image acquisition device of multiple vehicles passing through the gate, wherein each vehicle corresponds to multiple frames of vehicle images, and for the multiple frames of vehicle images corresponding to each vehicle, the license plate image area of ​​the vehicle is extracted from the multiple frames of vehicle images to obtain the license plate area of ​​each vehicle and the position of the license plate area of ​​each vehicle, wherein each vehicle has multiple license plate areas, and the target image recognition area is obtained based on the multiple license plate areas of the multiple vehicles.

11. An image processing device, characterized in that: include: CPU, memory, input and output interfaces, wired or wireless network interfaces, and power supply; The memory is a transient storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instructions in the memory to perform the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium comprises instructions, and when the instructions are executed on a computer, the computer is caused to perform the method according to any one of claims 1 to 9.

Images