A license plate number matching method and related device

Abstract

The application discloses a license plate number matching method and related device. The method comprises the following steps: obtaining the recognition result of a to-be-matched license plate, wherein the recognition result of the to-be-matched license plate comprises characters in the to-be-matched license plate and confidence of the characters; determining valid characters and invalid characters in the recognition result of the to-be-matched license plate according to the confidence of the characters; encoding the recognition result of the to-be-matched license plate according to the valid characters and the invalid characters, and obtaining encoded to-be-matched license plate data; and matching the encoded to-be-matched license plate data with pre-stored historical license plate data, and obtaining a historical license plate number matched with the recognition result of the to-be-matched license plate. The license plate number can be accurately and efficiently recognized in the case that the license plate picture is blurred, stained or blocked.

Description

Technical Field

[0001] This invention relates to a method and related apparatus for matching license plate numbers. Background Technology

[0002] In some real-world scenarios, such as Autonomous Parking Assist (AVP) systems, parking lot parking, traffic violations, fugitive pursuit, and autonomous driving, license plate recognition is required. However, in practice, such as when vehicles are entering or exiting parking lots, there are situations where the vehicle is far from the camera installed at the parking lot, or the angle at which the license plate is photographed is large. This can result in the license plate number being unclear in the photographed image and unable to be accurately recognized. In order to more accurately and efficiently identify and match license plate numbers, there is an urgent need to provide a license plate number matching method. Summary of the Invention

[0003] The inventors of this invention discovered that by converting the license plate numbers of vehicles that have entered and exited a parking lot into integers, sorting the license plate numbers, identifying the confidence level of each character in the license plate number to be matched based on an AI algorithm, converting the license plate number to be matched into an integer based on the confidence level, and then using a binary search algorithm to perform fuzzy matching on the license plate number to be matched, the matching speed can be greatly improved, and the matching result of the license plate number to be matched can be obtained accurately, quickly and efficiently.

[0004] In view of the above problems and findings, the present invention is proposed to provide a method and related apparatus for matching license plate numbers that overcomes or at least partially solves the above problems.

[0005] In a first aspect, embodiments of the present invention provide a method for matching license plate numbers, comprising:

[0006] Obtain the recognition result of the license plate to be matched, wherein the recognition result of the license plate to be matched includes the characters in the license plate to be matched and the confidence level of the characters;

[0007] Based on the confidence level of the characters, the valid and invalid characters in the recognition result of the license plate to be matched are determined; based on the valid and invalid characters, the recognition result of the license plate to be matched is encoded to obtain the encoded license plate data to be matched;

[0008] Obtain pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and obtain the historical license plate number that matches the recognition result of the license plate to be matched.

[0009] In one embodiment, determining the valid and invalid characters in the recognition result of the license plate to be matched based on the confidence level of the characters includes:

[0010] Characters with a confidence level higher than a preset confidence threshold are identified as valid characters; and / or,

[0011] Characters with a confidence level lower than a preset confidence level threshold are identified as invalid characters.

[0012] In one embodiment, encoding the recognition result of the license plate to be matched based on the valid characters and the invalid characters to obtain the encoded license plate data to be matched includes:

[0013] According to the pre-established encoding rules, the valid characters in the recognition result of the license plate to be matched are encoded, and the invalid characters in the recognition result of the license plate to be matched are ignored, so as to obtain the initial license plate code corresponding to the recognition result of the license plate to be matched.

[0014] The initial license plate code is shifted to obtain several corresponding transformed license plate codes.

[0015] From the license plate code group, valid license plate codes are selected; the license plate code group includes the initial license plate code and several corresponding transformed license plate codes; the valid license plate code is the license plate code in the license plate code group with the largest number of consecutive valid character codes starting from the highest data bit, and the valid character code is the code corresponding to the valid character in the recognition result of the license plate to be matched;

[0016] The valid license plate code corresponding to the recognition result of the license plate to be matched is converted to obtain the license plate data to be matched.

[0017] In one embodiment, the historical license plate data is implemented in the following manner:

[0018] Retrieve the original license plate numbers from the parking lot database;

[0019] According to the encoding rules, each character in the original license plate number is encoded to obtain the original license plate code. The original license plate code is then converted to obtain the original license plate data.

[0020] Move the original license plate code; obtain the redundant license plate code corresponding to the original license plate code; convert the redundant license plate code to obtain the redundant license plate data corresponding to the original license plate data;

[0021] The historical license plate data corresponding to the original license plate number includes the original license plate data and the corresponding redundant license plate data, wherein the number of redundant license plate data corresponding to the original license plate data is at least one.

[0022] In one embodiment, there are multiple records of historical license plate data;

[0023] After obtaining the historical license plate data, the method further includes:

[0024] Multiple historical license plate data are sorted according to their numerical values, and the sorted historical license plate data are stored in a preset sequence table.

[0025] In one embodiment, matching the encoded license plate data to be matched with the historical license plate data, and determining the historical license plate number that matches the recognition result of the license plate to be matched based on the matching result, includes:

[0026] Based on the license plate data to be matched, obtain the code corresponding to the license plate data to be matched. Based on the valid codes in the codes corresponding to the license plate data to be matched, use a binary search algorithm to determine the start position and end position of the historical license plate data from the sequence table.

[0027] The license plate data to be matched is ANDed with the historical license plate data between the start position and the end position of the historical license plate data, respectively. Based on the result of the operation, the historical license plate data that matches the license plate data to be matched is determined.

[0028] Based on the historical license plate data that matches the license plate data to be matched, the historical license plate number corresponding to the recognition result of the license plate to be matched is restored.

[0029] In one embodiment, if no historical license plate data matching the license plate data to be matched is found from the historical data, the method further includes:

[0030] Based on the recognition result of the license plate to be matched, obtain the license plate number to be matched;

[0031] The license plate data corresponding to the license plate number to be matched is stored in the sequence table. The license plate data corresponding to the license plate number to be matched includes the original license plate data corresponding to the license plate number to be matched and the redundant license plate data corresponding to the license plate number to be matched.

[0032] Secondly, embodiments of the present invention provide a license plate number matching device, comprising:

[0033] The acquisition module is used to acquire the recognition result of the license plate to be matched, wherein the recognition result of the license plate to be matched includes the characters in the license plate to be matched and the confidence level of the characters;

[0034] The encoding module is used to determine the valid and invalid characters in the recognition result of the license plate to be matched based on the confidence level of the characters; and to encode the recognition result of the license plate to be matched based on the valid and invalid characters to obtain the encoded license plate data to be matched.

[0035] The matching module is used to obtain pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and obtain the historical license plate number that matches the recognition result of the license plate to be matched.

[0036] Thirdly, embodiments of the present invention provide a computing device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the aforementioned license plate number matching method.

[0037] Fourthly, embodiments of the present invention provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the aforementioned license plate number matching method.

[0038] Fifthly, embodiments of the present invention provide a computer program product, the computer program product including a computer program, which, when executed by a processor, implements the aforementioned license plate number matching method.

[0039] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following:

[0040] This invention provides a license plate number matching method. It obtains the recognition result of the license plate to be matched, acquires each character in the recognition result and the corresponding confidence level of each character, determines the valid and invalid characters in the recognition result based on the confidence level of each character, encodes the recognition result based on the valid and invalid characters, obtains the encoded license plate data, acquires pre-stored historical license plate data, matches the encoded license plate data with the pre-stored historical license plate data, and obtains the historical license plate number that matches the recognition result of the license plate to be matched based on the matching result. Based on the confidence level of the characters in the recognition result of the license plate to be matched, it distinguishes between valid and invalid characters, and encodes the recognition result based on the valid and invalid characters, obtaining the license plate data to be matched after encoding. This improves the accuracy of the obtained license plate data to be matched, thereby improving the accuracy of the matching result between the license plate data to be matched and the pre-stored historical license plate data. Based on the matching result, it accurately and efficiently matches the historical license plate number corresponding to the license plate data to be matched.

[0041] Based on this, embodiments of the present invention provide a method and related apparatus for matching license plate numbers. The implementation method of the above-mentioned fuzzy matching of license plate numbers will be described in detail below with reference to the accompanying drawings. Attached Figure Description

[0042] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0043] Figure 1 This is a flowchart of the license plate number matching method in an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the encoding structure of the license plate number in an embodiment of the present invention;

[0045] Figure 3 This is a flowchart of the method for obtaining license plate data to be matched in an embodiment of the present invention;

[0046] Figure 4 A schematic diagram illustrating the process of encoding the recognition result of the license plate to be matched in this embodiment of the invention.

[0047] Figure 5 This is a schematic diagram illustrating the encoding conversion process corresponding to the recognition result of the license plate to be matched in an embodiment of the present invention;

[0048] Figure 6 This is a flowchart of the method for obtaining license plate data to be matched in an embodiment of the present invention;

[0049] Figure 7 This is a flowchart of the method for obtaining historical license plate data in an embodiment of the present invention;

[0050] Figure 8 This is a schematic diagram illustrating the process of obtaining original license plate data in an embodiment of the present invention;

[0051] Figure 9 This is a schematic diagram illustrating the process of obtaining redundant license plate data in an embodiment of the present invention;

[0052] Figure 10 This is a schematic diagram of the hash ring structure in an embodiment of the present invention;

[0053] Figure 11 This is a schematic diagram of the data corresponding to the hash ring in an embodiment of the present invention;

[0054] Figure 12 This is a schematic diagram of hash ring splitting in an embodiment of the present invention;

[0055] Figure 13 This is a schematic diagram illustrating the principle of using a binary search algorithm to find data in an embodiment of the present invention.

[0056] Figure 14 This is a flowchart illustrating the method for determining historical license plate data that matches the license plate data to be matched in this embodiment of the invention.

[0057] Figure 15This is a schematic diagram illustrating the binary search algorithm for searching historical license plate data in an embodiment of the present invention;

[0058] Figure 16 This is a schematic diagram of the matching interval in an embodiment of the present invention;

[0059] Figure 17 This is a schematic diagram of the vehicle license plate matching device in an embodiment of the present invention. Detailed Implementation

[0060] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0061] When the license plate image of a vehicle is unclear, existing technologies can identify the license plate number in the image through fuzzy matching. However, there is a lack of an accurate, efficient solution that is applicable to various scenarios related to license plate number search.

[0062] The inventors of this invention discovered that by converting the license plate numbers of vehicles that have entered and exited a parking lot into numerical values, and then sorting the license plate numbers of the parking lot vehicles according to the converted numerical values, and by retaining characters with confidence scores higher than the confidence score threshold and ignoring characters with confidence scores lower than the confidence score threshold using a binary search method, the historical license plate numbers corresponding to the recognition results of the license plate to be matched can be accurately, quickly and efficiently matched.

[0063] Based on this, embodiments of the present invention provide a method for implementing fuzzy matching of license plate numbers. The following describes the method for implementing fuzzy matching of license plate numbers in detail with reference to the accompanying drawings.

[0064] Example:

[0065] The license plate matching method provided in this embodiment refers to... Figure 1 As shown, it includes:

[0066] Step S1: Obtain the recognition result of the license plate to be matched. The recognition result of the license plate to be matched includes the characters in the license plate and the confidence level of the characters.

[0067] Step S2: Based on the confidence level of the characters, determine the valid and invalid characters in the recognition result of the license plate to be matched; based on the valid and invalid characters, encode the recognition result of the license plate to be matched to obtain the encoded license plate data to be matched;

[0068] Step S3: Obtain the pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and obtain the historical license plate number that matches the recognition result of the license plate to be matched.

[0069] This solution can be applied to various fields or scenarios related to license plate search. For example, parking lot parking scenarios, automated valet parking assistance systems (AVP), traffic violation scenarios, traffic scenarios, pursuit scenarios, autonomous driving systems, etc. In the embodiment of the present invention, the scenario of identifying the license plate number of a vehicle driving into a parking lot is used as an example to illustrate this solution.

[0070] In some optional embodiments, when a vehicle drives into a parking lot, the front-field camera of the parking lot captures the license plate of the vehicle, and through a license plate recognition algorithm, the captured picture of the vehicle license plate is recognized to obtain the recognition result of the license plate to be matched. The license plate recognition algorithm can recognize each character of the license plate number in the picture and the confidence level of each character. The confidence level represents the probability of accurately recognizing the character in the license plate number. The license plate recognition algorithm is a prior art, and the embodiment of the present invention will not elaborate on this, and in the prior art, there are various license plate recognition algorithms, and the embodiment of the present invention does not limit this. The following is an example for illustration. For example, by using the license plate recognition algorithm to recognize a blue license plate with the number "Jing ABCDEF", its recognition result is shown in Table 1 below:

[0071] Table 1:

[0072] license plate number Beijing A B C D E F Confidence 35 87 42 75 35 97 95

[0073] There are two rows of data in Table 1. The data in the first row are each character in the recognized license plate number, and the data in the second row are the confidence levels of each character. For example, the data in the second column of Table 1 indicates that the recognized character is "Jing", and the confidence level of the character "Jing" is 35.

[0074] In some optional embodiments, in the above step S2, according to the confidence level of the character, the valid characters and invalid characters in the recognition result of the license plate to be matched are determined. For example, it can be implemented in the following way:

[0075] Determine the characters with a confidence level higher than the preset confidence level threshold as valid characters; and / or

[0076] Determine the characters with a confidence level lower than the preset confidence level threshold as invalid characters.

[0077] In fact, if the captured license plate picture is relatively clear and the accuracy of the characters recognized by the license plate recognition algorithm is relatively high, there may be a situation where the confidence levels of all characters are higher than the confidence threshold, that is, all the obtained characters are valid characters. If the captured license plate picture is relatively blurred or the shooting angle deviates significantly, the accuracy of the characters recognized by the license plate recognition algorithm is relatively low, and there may be a situation where the confidence levels of all characters are lower than the confidence threshold, that is, all the obtained characters are invalid characters. Of course, there may also be a situation where among the characters of the recognized license plate number, there are both valid characters and invalid characters.

[0078] Set the confidence threshold according to the actual situation, and the embodiments of the present invention do not limit this. For example, the confidence threshold can be set to 50. Still taking the data in Table 1 as an example, in Table 1, the confidence levels of "Jing", "B", and "D" are all less than 50, which are invalid characters, and the confidence levels of "A", "C", "E", and "F" are all greater than 50, which are valid characters.

[0079] Generally, a license plate number consists of 7 - 8 characters. For example, "Jing A88888", "Wan A D11111", "Jing D0011 Police", "Civil Aviation B0150", "Gui X5995 Emergency", etc. Among the first two characters and the last two characters at the beginning of the license plate number, there may be Chinese characters, and the other characters are capital letters or numbers. Based on this, a coding rule is established in advance. Specifically, the coding rule can be adopted in the following manner:

[0080] (1) Determine the coding structure of the license plate.

[0081] Construct corresponding data bits for each character of the license plate number. The coding results corresponding to multiple characters form a 64 - bit integer. For example, the coding structure of the license plate number is as Figure 2 shown:

[0082] (a) P0 - P7: Store the encoding of each character of the license plate number respectively.

[0083] It should be noted that Figure 2 the data at the P7 position is determined according to the length of the license plate number. If the license plate number has 7 characters, the P7 position is empty.

[0084] (b) color: Store the color of the license plate.

[0085] (c) off: Store the offset of the characters in the license plate number.

[0086] (2) Determine the coding rule:

[0087] (a) Encoding rules for positions P0 - P7: Obtain all characters that make up the license plate number, including a finite number of Chinese characters, a finite number of capital letters, and a finite number of digits. According to the custom character order, perform Arabic numeral encoding on all characters that make up the license plate number. For example, the custom character order is A - Z, Beijing, Tianjin, Shanghai, Chongqing, Hebei, Henan, Yunnan, Liaoning, Heilongjiang, Hunan, Anhui, Fujian, Shandong, Xinjiang, Jiangsu, Zhejiang, Jiangxi, Hubei, Guangxi, Gansu, Shanxi, Inner Mongolia, Shaanxi, Jilin, Guizhou, Guangdong, Qinghai, Tibet, Sichuan, Ningxia, Hainan, Embassy, Military, Civilian, Emergency, Aviation, Urgent, Police, Inspection, School, Hanging... 0 - 9 (not enumerated one by one in this embodiment of the present invention). According to the above custom character order, encode all characters that make up the license plate number. For example, the encoding corresponding to A is 1, the encoding corresponding to B is 2, the encoding corresponding to C is 3, the encoding corresponding to D is 4, the encoding corresponding to Z is 26, the encoding corresponding to Beijing is 27, and the encoding corresponding to Tianjin is 28.

[0088] (b) Encoding rules for the color position: For example: Define the color order as blue, white, yellow, black, green, yellow - green. According to this color order, encode the color of the license plate to be matched. For example, the encoding corresponding to blue is 1, the encoding corresponding to white is 2, the encoding corresponding to yellow is 3, the encoding corresponding to black is 4, and the encoding corresponding to yellow - green is 6.

[0089] (c) Encoding rules for the off position: The initial value is 0. When the characters in the license plate number move, it is updated to the total number of digits of the offset license plate number.

[0090] In some optional embodiments, in step S2 above, according to the valid characters and invalid characters, encode the recognition result of the license plate to be matched to obtain the encoded data of the license plate to be matched. Specifically, as Figure 3 shown, it can be achieved through the following method:

[0091] Step S21: According to the pre - established encoding rules, encode the valid characters in the recognition result of the license plate to be matched, and ignore the invalid characters in the recognition result of the license plate to be matched, to obtain the initial license plate encoding corresponding to the recognition result of the license plate to be matched;

[0092] Step S22: Move the initial license plate encoding to obtain a corresponding number of transformed license plate encodings;

[0093] Step S23: From the license plate encoding group, screen out the valid license plate encodings; the license plate encoding group includes the initial license plate encoding and the corresponding number of transformed license plate encodings; the valid license plate encoding is the license plate encoding in the license plate encoding group with the largest number of consecutive valid character encodings starting from the highest data bit, and the valid character encoding is the encoding corresponding to the valid characters in the recognition result of the license plate to be matched;

[0094] Step S24: Convert the valid license plate code corresponding to the recognition result of the license plate to be matched to obtain the license plate data to be matched.

[0095] In some optional embodiments, during the execution of the above step S21, invalid characters in the recognition result of the license plate to be matched are ignored. In other words, based on the confidence level of the characters, characters with a confidence level lower than the confidence threshold (for example, the confidence threshold is 50) are considered to have too low credibility. Based on this, when encoding the recognition result of the license plate to be matched, the value of the data bit corresponding to the invalid character is set to 0. This will be explained with a specific example: Taking the data in Table 1 above as an example, each character in Table 1 is encoded to obtain the encoding corresponding to each character in the recognition result of the license plate to be matched. Refer to Figure 4 as shown Figure 4 In the figure, the data above the arrow represents the recognition result of the license plate to be matched before encoding, and the data below the arrow represents the initial license plate code obtained after encoding the recognition result of the license plate to be matched. The encoding corresponding to each character obtained, where "Jing" is an invalid character, so the value of the data bit at position P0 is set to 0, and the corresponding encoding is 0. "A" is a valid character, and according to the aforementioned encoding rule, "A" is encoded, so the encoding corresponding to "A" is 1. "B" is an invalid character, and the value of the data bit at position P2 is set to 0, so the encoding corresponding to "B" is 0. The encoding corresponding to "Blue" is 1.

[0096] After obtaining the initial license plate code, execute the above step S22 to move each encoding in the initial license plate code one character to the left to obtain a corresponding transformed license plate code. For example, taking the data in Table 1 as an example, the original license plate code corresponding to the license plate number "JingABCDEF" is "0103056". Move each encoding in this original license plate code one character to the left, that is: the encoding at the original P7 position is moved one bit to the left to the P6 position, the encoding at the original P6 position is moved one bit to the left to the P5 position, and so on. The encoding at the original P1 position is moved one bit to the left to the P0 position, and the encoding at the original P0 position, because it is already at the highest data bit and there is no position to move to the left, so it is moved to the vacated original P7 position to obtain the transformed license plate code, that is: "1030560", corresponding to the license plate number "ABCDEFJing". Execute the above process of moving the license plate code several times to obtain several corresponding transformed license plate codes until the encoding at the P7 position in the initial license plate code is moved to the P0 position. Combine the initial license plate code and the corresponding several transformed license plate codes to obtain a license plate code group;

[0097] From the license plate code group, select the license plate code with the longest consecutive valid character code starting from the highest data position (i.e., from the leftmost position P0). For example, as shown in Table 1, the license plate code with the longest consecutive valid character code is EF. Therefore, the license plate code with 'E' at position P0 is a valid license plate code. (See table 1 for reference.) Figure 5 As shown.

[0098] After obtaining the valid license plate code, a binary algorithm can be used to convert the valid license plate code to obtain the license plate data to be matched. For example, refer to... Figure 6 As shown, Figure 6 There are two sets of data, each with two rows. The first row of both sets represents the characters in the recognition result of the license plate to be matched after shifting. The second row of the first set represents the code corresponding to each character in the recognition result of the license plate to be matched after shifting. The second row of the second set represents the license plate data to be matched obtained by converting the code in the first row into binary. In this case, the binary value corresponding to the code "0" is 0000000, and the binary value corresponding to "1" is 00000001. In other words, taking the data in Table 1 as an example, after steps S21-S23, the license plate data to be matched is:

[0099] 010100001010000110000000000000000000001000000000001100000000001; For ease of calculation, the above binary license plate data to be matched can be represented in hexadecimal, that is, the license plate data to be matched is 0X50A1800002001801.

[0100] In some alternative embodiments, such as Figure 7 As shown, historical license plate data can be obtained in the following ways:

[0101] Step S61: Obtain the original license plate number from the parking lot database;

[0102] Step S62: Encode each character in the original license plate number according to the encoding rules to obtain the original license plate code. Convert the original license plate code to obtain the original license plate data.

[0103] Step S63: Move the original license plate code; obtain the redundant license plate code corresponding to the original license plate code, convert the redundant license plate code to obtain the redundant license plate data corresponding to the original license plate data;

[0104] The historical license plate data corresponding to the original license plate number includes the original license plate data and the corresponding redundant license plate data. Among them, there is at least one redundant license plate data corresponding to the original license plate data.

[0105] In the above step S61, the data stored in the parking lot database is the license plate numbers of all vehicles that have entered and exited the parking lot, including the license plate numbers of the vehicles currently in the parking lot at the current moment, and also including the license plate numbers of the vehicles that have entered and exited the parking lot but are not currently in the parking lot at the current moment.

[0106] In the above step S62, according to the encoding rules defined in the previous text, each character in the original license plate number is encoded respectively. For example, if the obtained original license plate number is "JingABCDEF" and the license plate is blue, according to the encoding rules defined in the previous text: the encoding corresponding to "Jing" is 27, the encoding corresponding to "A" is 1, the encoding corresponding to "B" is 2, the encoding corresponding to "C" is 3, the encoding corresponding to "D" is 4, the encoding corresponding to "E" is 5, the encoding corresponding to "F" is 6, and the encoding corresponding to "blue" is 1, the original license plate encoding is obtained. Then continue to execute the step of "converting the original license plate encoding", that is, using the binary algorithm to convert the encoding corresponding to each character in the original license plate number. For example, the binary value corresponding to "27" is 0011011, refer to Figure 8 as shown Figure 8 represents the process of obtaining the original license plate data according to the original license plate number. Figure 8 There are three rows of data in it. Among them, the first row of data represents the information of the original license plate number (including each character in the original license plate number and the license plate color), the second row of data represents the original license plate encoding obtained after encoding each character, and the third row of data represents the obtained original license plate data as:

[0107] 00000011011000000010000010000001100001000000101000011000000000001. For the convenience of calculation, the above historical license plate data in binary can be represented in hexadecimal, that is, the original license plate number 0X0360410308143001.

[0108] In some optional embodiments, in order to improve the matching speed of the license plate data to be matched, the original license plate encoding is shifted to obtain at least one redundant license plate encoding. In the above step S63, each encoding in the original license plate encoding is shifted one data bit to the left (in the direction of higher data bits), and the encoding located at the highest data bit is moved to the vacated data bit. Refer to Figure 9 as shown Figure 9The data in the first row represents the value of the original license plate number and its corresponding code in binary representation. For example, the code corresponding to "Jing" is 27, which is represented as 0011011 in binary. At this time, the data on the "off" bit is 0, indicating that the code corresponding to the character in this license plate number has not been shifted; after the entire original license plate code is shifted one bit to the left, the character "A" is located at the highest data bit, and the character "Jing" is shifted to the vacated position of P7. At the same time, the data on the "off" bit is incremented by 1, as Figure 9 shown in the second row of data in Figure 9 ; the codes in the original license plate code are shifted one bit to the left in sequence until the character E is located at the highest data bit. In this process, a total of 5 redundant license plate codes, namely II, III, IV, V, and VI, are obtained. It should be noted that the number of shifts of the codes in the original license plate code depends on the shortest continuous symbol number. The number of shifts is the same as the number of redundant license plate codes obtained, which is the difference between the number of characters in the original license plate number and the shortest continuous symbol number. Among them, the shortest continuous symbol number can be preset as 2 or 3 according to empirical values, or other values can be preset. The embodiments of the present invention do not limit this. For example, Figure 9 the number of characters in the original license plate number corresponding to the original data code in Figure 9 is 7, and the shortest continuous symbol number is set to 2. Therefore, the number of shifts, that is, the number of redundant license plate codes obtained, is 5.

[0109] In some optional embodiments, after obtaining multiple historical license plate data through the above steps S61 - S63, the multiple historical license plate data are sorted according to numerical values, and the sorted multiple historical license plate data are stored.

[0110] When matching with the license plate data to be matched, the pre - saved historical license plate data are directly obtained, which greatly saves the calculation cost when matching the recognition result of the license plate to be matched and improves the matching efficiency.

[0111] In some optional embodiments, if no historical license plate data matching the license plate data to be matched is found in the historical data, it may be that the vehicle enters the parking lot for the first time. Then, according to the recognition result of the license plate to be matched, the license plate number to be matched is obtained; the license plate data corresponding to the license plate number to be matched is stored in the sequence table. The license plate data corresponding to the license plate number to be matched includes the original license plate data corresponding to the license plate number to be matched and the redundant license plate data corresponding to the license plate number to be matched.

[0112] In a real-world parking lot scenario, new vehicles enter the parking lot at any time. In this case, the parking lot's database does not store the license plate numbers of the new vehicles. Therefore, it is necessary to store the license plate number of the new vehicle in the parking lot's database. Then, through steps S62-S63, the historical license plate data corresponding to the new vehicle's license plate number (including the original license plate data and redundant license plate data) is obtained and stored. This means inserting new data (the historical license plate data corresponding to the new vehicle's license plate number) into the existing multiple historical license plate data entries. Since the existing historical license plate data entries are ordered, it is necessary to ensure that the combination of the original historical license plate data and the new data remains ordered after insertion. Due to the large volume of historical license plate data, to improve the speed of data insertion and deletion, a distributed hash table can be used to store the historical license plate data. Specifically, this can be achieved in the following way:

[0113] (1), reference Figure 10 As shown, construct a hash ring, defining the data capacity of each segment (the capacity of each segment can be defined as 1024). The constructed data is as follows. Figure 11 As shown, Figure 11 The left side of the image shows the start and end positions of the data corresponding to the hash ring. Figure 11 The right side represents the ordered data within the hash ring;

[0114] (2), reference Figure 12 As shown, when the number of data exceeds the segment capacity, the queue is split into two from the middle (at position 512), and the splitting step is executed in a loop (ensuring that the numbers in each Ring are ordered and the segments on the left are ordered and continuous) until all the original license plate data and redundant license plate data are inserted.

[0115] In some optional embodiments, if a new vehicle enters the parking lot, after obtaining the license plate recognition result of the new vehicle, the license plate number corresponding to the new vehicle's license plate recognition result can be obtained in the following way:

[0116] (1) If the confidence level of each character in the recognition result of the new vehicle's license plate is higher than the confidence level threshold, then the recognition result of the new vehicle's license plate is considered to be accurate. Then, based on the characters in the recognition result of the new vehicle's license plate, the license plate number of the new vehicle is obtained and stored in the parking lot's database.

[0117] (2) If there are characters in the recognition result of the license plate of the new vehicle whose confidence level is lower than the confidence level threshold, the recognition result of the license plate of the new vehicle is considered to be inaccurate. In this case, the staff can manually record the license plate number corresponding to the recognition result of the new vehicle into the parking lot's database.

[0118] In some optional embodiments, in step S3 above, since the historical license plate data of the parking lot is huge, numbering in the tens of thousands, hundreds of thousands, or even millions, if all the historical license plate data is traversed, although only one operation is needed to obtain the result for each historical license plate data, the search efficiency is very low due to the huge volume of historical license plate data. Therefore, a highly efficient algorithm is needed to search for historical license plate data that matches the license plate data to be matched from a large amount of historical license plate data.

[0119] Among the many search algorithms available, binary search is highly efficient. To illustrate its principle, let's consider a search for the number 22 in a sorted array. Figure 13 As shown, the data to be queried (22) is compared with the middle data in the array. Figure 13 The middle data is 16. Since 22 is larger than 16, the first half of the array is discarded, and the second half is retained. The process of comparing the data to be queried (22) with the middle data is repeated. The middle data of the remaining array is 30. Since 22 is smaller than 30, the second half of the array is discarded, and the first half is retained. This process is repeated until the target data is found after four queries. In fact, the speed of the binary search algorithm is log2N. For 1 billion data points, the number of queries for the target data using the binary search algorithm is ≤30 (log2N).

[0120] 21000000000=30), with very high performance. Based on this, a binary search algorithm is adopted to quickly filter out a portion of historical license plate data from a large amount of historical license plate data as the matching interval for the license plate data to be matched. Then, the license plate data to be matched is compared with the historical license plate data in the matching interval, and finally the historical license plate data that matches the license plate data to be matched is determined, which can greatly narrow the comparison range of the license plate data to be matched.

[0121] Specifically, step S3 above, such as Figure 14 As shown, this can be achieved in the following way:

[0122] Step S81: Based on the license plate data to be matched, obtain the code corresponding to the license plate data to be matched. Based on the valid codes in the codes corresponding to the license plate data to be matched, use the binary search algorithm to determine the start position and end position of the historical license plate data from the sequential list.

[0123] Step S82: Perform an AND operation between the license plate data to be matched and the historical license plate data between the start position and the end position of the historical license plate data, respectively. Based on the operation result, determine the historical license plate data that matches the license plate data to be matched.

[0124] Step S83: According to the historical license plate data that matches the license plate data to be matched, restore the historical license plate number corresponding to the recognition result of the license plate to be matched.

[0125] Among them, in the sequential list, the data between the starting position and the ending position of the historical license plate data is determined as the matching interval. During the process of matching the historical license plate data, the characters with low confidence (i.e., invalid characters) in the recognition result of the license plate to be matched are ignored, and only the characters with high confidence (i.e., valid characters) are compared. In other words, if the character corresponding to the highest data bit of the license plate data to be matched is an invalid character, it will be ignored. Then, when determining the matching interval, multiple matching intervals will be obtained from the sequential list. To avoid this problem and make full use of the sorting queue to improve the query performance, when performing Step S3, the consecutive valid characters starting from the highest data bit in the valid license plate encoding are used to match the historical license plate data that matches the license plate data to be matched from multiple historical license plate data; and during the process of converting the initial license plate encoding into a numerical value, a binary algorithm is adopted to convert the encoding corresponding to each character in the recognition result of the license plate to be matched into a numerical value respectively. In other words, the encoding corresponding to each character also corresponds to a binary numerical value; based on this, a binary search algorithm is used to find the starting historical license plate data and the ending historical license plate data corresponding from multiple historical data, and a unique matching interval is obtained. The following is illustrated with an example;

[0126] Reference Figure 15 As shown in Figure 15 There are a total of 5 lines of data. Among them, the first line of data is the recognition result of the license plate to be matched (taking the data in Table 1 as an example). The character corresponding to the P0 position is the invalid character "Jing", so the encoding of the P0 position is 0. Also, because the invalid character cannot be used as the highest data bit of the license plate encoding, the characters in this license plate number are shifted to the left, and the character "A" is used as the highest bit. The result is as shown in Figure 15 the second line of data in Figure 15 The binary algorithm is used to convert the shifted license plate encoding, as shown in Figure 15 the third line of data in Figure 15 to obtain the license plate data to be matched; starting from the highest data bit, all characters from the position corresponding to the first invalid character (i.e., the character at the P1 position in the second line of data) to the P7 position are set as invalid characters, and the starting historical license plate data is obtained, as shown in Figure 15All characters from the position corresponding to the character at the P1 position of the second-line data to the P7 position are made valid, obtaining the terminated historical license plate data, such as Figure 15 shown in the fifth-line data in Figure 15 , that is, the termination position of the matching interval is: 0001000000111111111111111111111111111111111111111111111111111001; for ease of calculation, this matching interval is represented in hexadecimal as 0X1020000000000001 to 0X103FFFFFFFFFFFF9.

[0127] However, in Figure 15 the recognition result of the license plate to be matched shown in Figure 15 , for the license plate code after one move (the valid character 'A' is at the highest data bit), although the character at the highest data bit P0 position is a valid character, the character code with the largest number of consecutive valid characters is 'EF', and E is not at the highest data bit. Also, based on the above process of obtaining the license plate data to be matched, Figure 15 the initial license plate code of the license plate number 'JINGABCDEF' represented by the first-line data in Figure 15 is '0A0C0EF', and the valid license plate code after moving is as shown in Figure 16 the second-line data in Figure 16 , the valid license plate code is 'EF0A0C0'. Starting from the highest data bit, all characters from the position corresponding to the first invalid character (that is, Figure 16 the character at the P2 position of the second-line data) to the P7 position are made invalid, obtaining the starting historical license plate data, such as Figure 16 shown in the fourth-line data in Figure 16 , that is, the starting position of the matching interval: 0101000010100001100000000000000000000000000000000000000000000001. Starting from the highest data bit, all characters from the position corresponding to the first invalid character (that is, Figure 16 the character at the P2 position of the second-line data) to the P7 position are made valid, obtaining the terminated historical license plate data, such as Figure 16 shown in the fifth-line data in Figure 16 , that is, the termination position of the matching interval is: 0101000010100001101111111111111111111111111111111111111111111001; for ease of calculation, this matching interval is represented in hexadecimal as 0X50A1800000000001 to 0X50A1BFFFFFFFFFF9.

[0128] Compared to the matching range obtained by using a single "A" character, the matching range obtained by combining "EF" characters is smaller. In other words, the more consecutive valid characters there are, the more data is filtered out by binary search, and the less historical license plate data needs to be matched. Therefore, using the matching range obtained by valid license plate codes can improve the matching efficiency.

[0129] If the number of possible characters in a license plate is 80, and each character has an equal probability, then the number of valid consecutive characters and the proportion of historical license plate data that need to be matched in a fuzzy manner are as follows:

[0130] One consecutive valid character: 1 / 80 = 1.25%; two consecutive valid characters: 1 / 80 / 80 = 0.015625%; three consecutive valid characters: 1 / 80 / 80 / 80 = 0.00019531255%, and so on. This embodiment of the invention will not list them all. As can be seen from the aforementioned data, the more consecutive valid characters that appear, the better. Based on this, during the process of obtaining the license plate data to be matched, the valid code group is shifted to the highest data position, so that the number of consecutive valid characters starting from the highest data position is the largest, which can narrow the matching range of historical license plate data and improve the search performance of historical license plate data.

[0131] It is important to note that when using the binary search algorithm to query historical license plate data, the algorithm first uses the start and end positions of the hash ring to find the appropriate segment, then uses binary search to find the start and end positions within the segment, and finally performs fuzzy matching between the start and end positions.

[0132] In some optional embodiments, step S82 above involves performing an AND operation between the license plate data to be matched and historical license plate data. If the result of the AND operation between the license plate data to be matched and the historical license plate data matches the license plate data to be matched, then the historical license plate data is determined to be the historical license plate data that matches the recognition result of the license plate to be matched. For example, if the license plate data to be matched is 0X50A1800002001801, and there is historical license plate data 0X50A1801B02081841 in the matching interval, the following formula is used to perform an AND operation between the two:

[0133] 0X50A1801B02081841&0X50A1800002001801; (1)

[0134] In the above formula, '&' is the bitwise AND operator in computer programming. The calculation rules are: 0&0=0, 0&1=0, 1&0=0, 1&1=1. For example, 1011&1101=1001.

[0135] According to the above calculation rules, the result of the calculation obtained by formula (1) is 0X50A1800002001801, which is consistent with the license plate data to be matched. Therefore, 0X50A1801B02081841 is determined to be the historical license plate data that matches the license plate data to be matched.

[0136] Since the historical license plate data obtained that matches the license plate data to be matched may be the original license plate data or redundant license plate data, the obtained historical license plate data is restored. The resulting historical license plate number may be the license plate number corresponding to the original license plate data or the license plate number corresponding to the redundant license plate data. If it is the license plate number corresponding to the original license plate data, then the license plate number is directly used as the historical license plate number corresponding to the recognition result of the license plate to be matched. If it is the license plate number corresponding to the redundant license plate data, then the license plate number corresponding to the redundant license plate data needs to be restored to the actual license plate number according to the number of moves recorded by the "off" data bit. The restored license plate number is then used as the historical license plate number corresponding to the recognition result of the license plate to be matched.

[0137] The inventors of this invention randomly generated ten million license plate numbers and used both the aforementioned license plate number matching method (this algorithm) and a traditional algorithm to perform calculations. The calculation results were compared, as shown in Table 2 below. Table 2 compares the algorithm and the traditional algorithm from four dimensions:

[0138] (1) Memory usage dimension: The traditional algorithm uses 1.12GB of memory, while this algorithm uses 2.19GB;

[0139] (2) Startup time dimension: The startup time of the traditional algorithm is 31 seconds, while that of this algorithm is 111 seconds;

[0140] (3) Insertion speed dimension: The insertion speed of the traditional algorithm is 1 / 10000 times / second, while that of this algorithm is 10000 times / second;

[0141] (4) Search speed dimension: The search speed of traditional algorithms is 1 / 1000 times / second, while that of this algorithm is greater than 2,380,000 times / second;

[0142] Compared to traditional algorithms, this algorithm's insertion speed is 30 billion times faster and its search speed is 100 million times faster.

[0143] Table 2

[0144] content Traditional Algorithm This algorithm Performance improvement Memory usage (GB) 1.12 2.19 1 / 2 Startup time (seconds) 31 111 1 / 3 Insertion rate (times / second) 1 / 10000 10000 100 million times Search speed (times / second) 1 / 1000 ＞2380000 30 billion times

[0145] Based on the same inventive concept, embodiments of the present invention also provide a license plate number matching device, the structure of which is as follows: Figure 17 As shown, it includes:

[0146] The acquisition module 101 is used to acquire the recognition result of the license plate to be matched, which includes the characters in the license plate and the confidence level of the characters.

[0147] The encoding module 102 is used to determine the valid and invalid characters in the recognition result of the license plate to be matched based on the confidence level of the characters; and to encode the recognition result of the license plate to be matched based on the valid and invalid characters to obtain the encoded license plate data to be matched.

[0148] The matching module 103 is used to obtain pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and obtain the historical license plate number that matches the recognition result of the license plate to be matched.

[0149] In some optional embodiments, the encoding module 102 is further configured to determine characters with a confidence level higher than a preset confidence level threshold as valid characters; and / or determine characters with a confidence level lower than a preset confidence level threshold as invalid characters.

[0150] In some optional embodiments, the encoding module 102 is further configured to, according to a pre-established encoding rule, encode the valid characters in the recognition result of the license plate to be matched, ignore invalid characters in the recognition result of the license plate to be matched, and obtain the initial license plate code corresponding to the recognition result of the license plate to be matched; move the initial license plate code to obtain a number of corresponding transformed license plate codes; and select the valid license plate code from the license plate code group; the license plate code group includes the initial license plate code and a number of corresponding transformed license plate codes; the valid license plate code is the license plate code with the most consecutive valid character codes starting from the highest data bit in the license plate code group, and the valid character code is the code corresponding to the valid character in the recognition result of the license plate to be matched; and convert the valid license plate code corresponding to the recognition result of the license plate to be matched to obtain the license plate data to be matched.

[0151] In some optional embodiments, the encoding module 102 is further configured to obtain the original license plate number in the parking lot database; encode each character in the original license plate number according to the encoding rules to obtain the original license plate code; convert the original license plate code to obtain the original license plate data; move the original license plate code; obtain the redundant license plate code corresponding to the original license plate code; convert the redundant license plate code to obtain the redundant license plate data corresponding to the original license plate data; the historical license plate data corresponding to the original license plate number includes the original license plate data and the corresponding redundant license plate data, wherein the number of redundant license plate data corresponding to the original license plate data is at least one.

[0152] In some optional embodiments, after obtaining the pre-stored historical license plate data, the encoding module 102 is further used to sort the multiple historical license plate data according to their numerical values ​​and store the sorted multiple historical license plate data into a preset sequence table.

[0153] In some optional embodiments, the matching module 103 is further configured to: obtain the code corresponding to the license plate data to be matched based on the license plate data to be matched; use a binary search algorithm to determine the start position and end position of the historical license plate data from the sequence list based on the valid codes in the codes corresponding to the license plate data to be matched; perform an AND operation between the license plate data to be matched and the historical license plate data between the start position and the end position of the historical license plate data respectively; determine the historical license plate data that matches the license plate data to be matched based on the operation result; and reconstruct the historical license plate number corresponding to the recognition result of the license plate to be matched based on the historical license plate data that matches the license plate data to be matched.

[0154] In some optional embodiments, if no historical license plate data matching the license plate data to be matched is found from the historical data, the matching module 103 is further used to obtain the license plate number to be matched based on the recognition result of the license plate to be matched; and store the license plate data corresponding to the license plate number to be matched into a sequence table, wherein the license plate data corresponding to the license plate number to be matched includes the original license plate data corresponding to the license plate number to be matched and the redundant license plate data corresponding to the license plate number to be matched.

[0155] Regarding the license plate matching device in the above embodiments, the specific way in which each module performs its operation has been described in detail in the embodiments of the method, and will not be elaborated here.

[0156] Based on the same inventive concept, embodiments of the present invention also provide a computing device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the aforementioned vehicle license plate number matching method.

[0157] Based on the same inventive concept, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned license plate number matching method.

[0158] Based on the same inventive concept, a computer program product, comprising a computer program, which, when executed by a processor, implements the aforementioned license plate number matching method.

[0159] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0160] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0161] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0162] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0164] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for matching license plate numbers, characterized in that, include: Obtain the recognition result of the license plate to be matched, wherein the recognition result of the license plate to be matched includes the characters in the license plate to be matched and the confidence level of the characters; Based on the confidence level of the characters, the valid and invalid characters in the recognition result of the license plate to be matched are determined; according to the pre-established encoding rules, the valid characters in the recognition result of the license plate to be matched are encoded, and the invalid characters in the recognition result of the license plate to be matched are ignored, so as to obtain the initial license plate code corresponding to the recognition result of the license plate to be matched. The initial license plate code is shifted to obtain several corresponding transformed license plate codes; valid license plate codes are selected from the license plate code group; the license plate code group includes the initial license plate code and several corresponding transformed license plate codes; the valid license plate code is the license plate code in the license plate code group with the largest number of consecutive valid character codes starting from the highest data bit, and the valid character code is the code corresponding to the valid character in the recognition result of the license plate to be matched; The valid license plate code corresponding to the recognition result of the license plate to be matched is converted to obtain the license plate data to be matched; Obtain pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and obtain the historical license plate number that matches the recognition result of the license plate to be matched.

2. The method as described in claim 1, characterized in that, The step of determining the valid and invalid characters in the recognition result of the license plate to be matched based on the confidence level of the characters includes: Characters with a confidence level higher than a preset confidence threshold are identified as valid characters; and / or, Characters with a confidence level lower than a preset confidence level threshold are identified as invalid characters.

3. The method as described in claim 2, characterized in that, The historical license plate data is obtained through the following methods: Retrieve the original license plate numbers from the parking lot database; According to the encoding rules, each character in the original license plate number is encoded to obtain the original license plate code. The original license plate code is then converted to obtain the original license plate data. Move the original license plate code; Obtain the redundant license plate code corresponding to the original license plate code, convert the redundant license plate code, and obtain the redundant license plate data corresponding to the original license plate data; The historical license plate data corresponding to the original license plate number includes the original license plate data and the corresponding redundant license plate data, wherein the number of redundant license plate data corresponding to the original license plate data is at least one.

4. The method as described in claim 3, characterized in that, There are multiple records of the historical license plate data; After obtaining the historical license plate data, the method further includes: Multiple historical license plate data are sorted according to their numerical values, and the sorted historical license plate data are stored in a preset sequence table.

5. The method as described in claim 4, characterized in that, The step of matching the encoded license plate data to be matched with the historical license plate data, and determining the historical license plate number that matches the recognition result of the license plate to be matched based on the matching result, includes: Based on the license plate data to be matched, obtain the code corresponding to the license plate data to be matched. Based on the valid codes in the codes corresponding to the license plate data to be matched, use a binary search algorithm to determine the start position and end position of the historical license plate data from the sequence table. The license plate data to be matched is ANDed with the historical license plate data between the start position and the end position of the historical license plate data, respectively. Based on the result of the operation, the historical license plate data that matches the license plate data to be matched is determined. Based on the historical license plate data that matches the license plate data to be matched, the historical license plate number corresponding to the recognition result of the license plate to be matched is restored.

6. The method as described in claim 5, characterized in that, If no matching historical license plate data is found in the historical license plate data, the method further includes: Based on the recognition result of the license plate to be matched, obtain the license plate number to be matched; The license plate data corresponding to the license plate number to be matched is stored in the sequence table. The license plate data corresponding to the license plate number to be matched includes the original license plate data corresponding to the license plate number to be matched and the redundant license plate data corresponding to the license plate number to be matched.

7. A license plate number matching device, characterized in that, include: The acquisition module is used to acquire the recognition result of the license plate to be matched, wherein the recognition result of the license plate to be matched includes the characters in the license plate to be matched and the confidence level of the characters; The encoding module is used to determine the encoding of valid characters in the recognition result of the license plate to be matched according to the confidence level of the characters and to encode the characters according to the pre-established encoding rules, while ignoring invalid characters in the recognition result of the license plate to be matched, so as to obtain the initial license plate code corresponding to the recognition result of the license plate to be matched. The initial license plate code is shifted to obtain several corresponding transformed license plate codes; valid license plate codes are selected from the license plate code group; the license plate code group includes the initial license plate code and several corresponding transformed license plate codes; the valid license plate code is the license plate code in the license plate code group with the largest number of consecutive valid character codes starting from the highest data bit, and the valid character code is the code corresponding to the valid character in the recognition result of the license plate to be matched; The valid license plate code corresponding to the recognition result of the license plate to be matched is converted to obtain the license plate data to be matched; The matching module is used to acquire pre-stored historical license plate data, match the encoded license plate data to be matched with the historical license plate data, and determine the historical license plate number that matches the recognition result of the license plate to be matched based on the matching result.

8. A computing device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the license plate number matching method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the license plate number matching method according to any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the license plate number matching method according to any one of claims 1-6.

Images