Vehicle access intelligent identification method based on behavior analysis
By adopting an intelligent recognition method based on behavior analysis in the management of transportation access and exit, first identifying the number of the license plate head and comparing it, the problem of large computing power occupies and long waiting time in traditional methods is solved, and efficient vehicle exit recognition is achieved and management.
Patent Information
- Application Number
- CN202411479871.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-23
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The traditional method of access and exit management of transportation tools relies on the complete identification of license plates, which leads to large computing power consumption and long waiting time for vehicles to exit gates, making it difficult to quickly and accurately identify vehicles to exit, and cannot effectively deal with changes in traffic flow.
Using an intelligent recognition method based on behavior analysis, the license plate image is collected through image acquisition hardware. The license plate recognition software first recognizes the number of the license plate head and compares it with the license plate number library in the gate, reduces the computing power consumption and comparative calculation amount, and improves the vehicle Output efficiency.
It realizes fast and accurate vehicle exit identification, reduces computing power occupation and comparison calculation, improves vehicle exit efficiency, and adapts to the actual situation of changes in traffic flow.
Smart Images

Figure CN119007182A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method for intelligently identifying the entry and exit of vehicles based on behavior analysis. Background Art
[0002] In the scenario of vehicle access management, the efficiency problem during peak hours is particularly prominent, and efficient vehicle access management has become a crucial link in solving the problem of smooth traffic. Traditional vehicle access management often relies on the complete recognition of license plates. It only recognizes all license plate numbers and compares them with the stored numbers at the gate. It lacks overall control over vehicle access and lacks in-depth analysis and utilization of vehicle behavior. It is difficult to quickly and accurately complete vehicle exit identification. There are unreasonable situations in resource allocation (computing power usage), resulting in a long waiting time for vehicles to exit the gate. The vehicle access management solution is relatively traditional and fixed, and cannot cope well with the actual situation of constantly changing traffic flow.
[0003] Among the current related technologies, the license plate recognition method has the technical problem of large computing power consumption and long waiting time for vehicles to exit the gate because it needs to recognize all license plate numbers and compare them with the stored numbers at the gate. Summary of the invention
[0004] The present application provides a method for intelligent identification of vehicle entry and exit based on behavior analysis. When a vehicle exits the gate, the image acquisition hardware collects the license plate image, and the license plate recognition software identifies the k-digit number at the head of the image (7>k≥2). The recognition result is compared with the entry license plate number library to obtain a first-level entry matching license plate. If the number of matching license plates is 1, the exit vehicle identification result is configured according to the license plate. By analyzing the vehicle exit behavior, the partial number on the head of the license plate is first identified, and then compared with the entry license plate number library, thereby reducing the computing power occupied and the amount of comparison calculation, and achieving the technical effect of improving the efficiency of vehicle exit.
[0005] The present application provides a method for intelligently identifying vehicle entry and exit based on behavior analysis, comprising: When a vehicle is leaving the gate, the first vehicle license plate image is collected through image acquisition hardware; the head k-digit number of the first vehicle license plate image is recognized by license plate recognition software to obtain the head k-digit number recognition result, 7>k≥2, k is an integer; the head k-digit number recognition result is compared with the entry license plate number library to obtain a first-level entry matching license plate; when the number of the first-level entry matching license plates is equal to 1, the exit vehicle recognition result is configured according to the first-level entry matching license plate.
[0006] The present application proposes an intelligent identification method for vehicle entry and exit based on behavioral analysis. First, when a vehicle exits the gate, the image acquisition hardware collects the license plate image, and the license plate recognition software identifies the k-digit number on the head of the image (7>k≥2). The identification result is compared with the entry license plate number library to obtain a first-level entry matching license plate. If the number of matching license plates is 1, the exit vehicle identification result is configured according to the license plate. By analyzing the vehicle exit behavior, the partial number on the head of the license plate is first identified, and then compared with the entry license plate number library, which reduces the computing power occupied and the amount of comparison calculation, and achieves the technical effect of improving the efficiency of vehicle exit. BRIEF DESCRIPTION OF THE DRAWINGS
[0007] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0008] Figure 1 A flow chart of a method for intelligently identifying vehicle entry and exit based on behavior analysis provided in an embodiment of the present application; Figure 2 A schematic diagram of the license plate number library storage process for a method for intelligent identification of vehicle entry and exit based on behavior analysis provided in an embodiment of the present application. DETAILED DESCRIPTION
[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0010] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0011] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.
[0012] The present application embodiment provides a method for intelligently identifying vehicle entry and exit based on behavior analysis, such as Figure 1 As shown, the method includes: Step S100, when the vehicle is exiting the gate, the image of the first vehicle license plate is collected through the image acquisition hardware. Specifically, when the vehicle is in the exiting state, the image acquisition hardware (usually a camera) installed at the gate will automatically sense the approach of the vehicle and trigger the shooting mechanism, and accurately capture the vehicle license plate area in the lens using the optical imaging principle, and record the image information of the license plate, including the characters, colors, patterns and other features on the license plate. The collected first vehicle license plate image will be immediately transmitted to the license plate recognition software to provide reliable data support for the subsequent exit process.
[0013] In one possible implementation, Figure 2 As shown, when the vehicle is leaving the gate, the image of the first vehicle license plate is collected by the image acquisition hardware, and step S100 further includes step S110, when the vehicle is entering the gate, the image of the second vehicle license plate is collected by the image acquisition hardware. Specifically, when the vehicle is entering the gate, the image of the second vehicle license plate is collected by the image acquisition hardware, and when the vehicle drives to the gate and prepares to enter, the image acquisition hardware (usually a camera) starts to work. The camera is installed in a position where the vehicle license plate can be clearly photographed. When the vehicle enters its shooting range, the camera automatically senses and triggers the shooting mechanism, and uses the optical imaging principle to accurately capture the vehicle's license plate area in the lens, and record the second vehicle license plate image, in preparation for subsequent license plate recognition and information storage.
[0014] Step S120, the license plate recognition software performs 7-digit number recognition on the second vehicle license plate image to obtain a vehicle license plate recognition result. Specifically, the license plate recognition software performs 7-digit number recognition on the second vehicle license plate image to obtain a vehicle license plate recognition result. The license plate recognition software uses a specific image recognition algorithm to analyze the collected second vehicle license plate image. Since it is when entering the gate, in order to accurately record the vehicle information, the software will recognize the complete 7-digit number on the license plate, and extract the complete license plate number information by recognizing the characters, colors and other features on the license plate, and use it as the vehicle license plate recognition result.
[0015] Step S130, the vehicle license plate recognition result and the gate entry timestamp are stored in the gate entry license plate number library. Specifically, the vehicle license plate recognition result and the gate entry timestamp are stored in the gate entry license plate number library. After obtaining the vehicle license plate recognition result, in order to facilitate the subsequent comparison and management of the vehicle when leaving the gate, this recognition result is stored in the gate license plate number library together with the timestamp of the vehicle entering the gate. The gate entry timestamp records the specific time when the vehicle enters the gate. Storing it together with the license plate recognition result can provide more comprehensive information for subsequent vehicle management and query. When the vehicle leaves the gate, the license plate information at the time of exiting the gate can be compared with the information in the gate entry license plate number library to achieve efficient intelligent recognition and management of vehicle entry and exit.
[0016] In a possible implementation, when a vehicle is exiting a gate, an image of a first vehicle license plate is captured by image acquisition hardware, and step S100 further includes step S140, in which a vehicle motion positioning sequence of a lane within a preset range centered on the gate is captured by a millimeter-wave radar. Specifically, a vehicle motion positioning sequence of a lane within a preset range centered on the gate is captured by a millimeter-wave radar. A millimeter-wave radar is a device that can accurately detect the position and motion state of a target object. It is arranged within a preset range centered on the gate, and can cover areas where vehicles may pass through the gate. The millimeter-wave radar continuously emits millimeter-wave signals and receives reflected signals. By processing and analyzing the signals, the position information of the vehicles within the range can be obtained, and a vehicle motion positioning sequence is formed over time. The sequence records the position changes of the vehicle at different times, and provides basic data for subsequent judgment of the vehicle's gate entry and exit status.
[0017] Step S150, obtaining the vehicle head orientation according to the vehicle motion positioning sequence. Specifically, obtaining the vehicle head orientation according to the vehicle motion positioning sequence, and further analyzing the vehicle motion positioning sequence, the vehicle's driving direction can be inferred, thereby determining the vehicle head orientation. For example, the vehicle's motion direction vector can be calculated according to the vehicle's position change trend at different times, thereby determining the vehicle head orientation. With the help of specific algorithms and data analysis techniques, accurate vehicle head orientation information can be obtained.
[0018] Step S160, when the vehicle head is facing the gate and the vehicle is in the gate, the vehicle is exiting the gate. Specifically, when the vehicle head is facing the gate and the vehicle is in the gate, the vehicle is exiting the gate. After obtaining the vehicle head direction information, a judgment is made in combination with the vehicle position information. If the vehicle head is facing the gate, it means that the vehicle may be preparing to pass through the gate. When the vehicle is in the gate, it can be determined that the vehicle is in the exit state, because the vehicle is in the gate and the vehicle head is facing the gate, which usually means that the vehicle is driving towards the gate and preparing to go out.
[0019] Step S170, when the vehicle head is facing the gate and the vehicle is outside the gate, the vehicle is entering the gate. Specifically, when the vehicle head is facing the gate and the vehicle is outside the gate, the vehicle is entering the gate. When the vehicle head is facing the gate, if the vehicle is outside the gate, then it can be determined that the vehicle is in the gate-entering state, because in this case, the vehicle is outside the gate and the vehicle head is facing the gate, which usually means that the vehicle is driving towards the gate and preparing to enter.
[0020] In a possible implementation, when the vehicle head is facing the gate and the vehicle is outside the gate, the vehicle is entering the gate, and step S170 further includes step S171, when the vehicle head is not facing the gate, the vehicle is not entering or exiting the gate. Specifically, the vehicle movement positioning sequence of the lanes in a preset range with the gate as the center is collected by millimeter wave radar, and the vehicle head direction is obtained according to the sequence analysis. If the vehicle head is not facing the gate, it is determined that the vehicle is not in the state of entering or exiting the gate, that is, in this case, the vehicle may just be driving on the nearby road and has nothing to do with the gate entry and exit behavior, which helps to accurately distinguish between vehicles entering or exiting the gate and passing vehicles, and improve the accuracy and efficiency of vehicle entry and exit management.
[0021] Step S200, the license plate recognition software performs head k-digit number recognition on the first vehicle license plate image to obtain the head k-digit number recognition result, 7>k≥2, k is an integer. Specifically, the license plate recognition software processes the first vehicle license plate image obtained by the image acquisition hardware when the vehicle is ready to exit the gate. In order to improve efficiency, the method of first recognizing the head part of the license plate number is adopted. According to the setting 7>k≥2 (k is an integer), the k characters or numbers at the head of the license plate are extracted from the license plate image from left to right, and the specific area of the license plate image is analyzed by image feature extraction, character recognition and other technologies to determine the k-digit number at the head of the license plate, and finally obtain the head k-digit number recognition result. Although the result is only part of the license plate number information, it has a certain degree of distinction and can be used for subsequent comparison with the entry license plate number library to determine the identity of the exiting vehicle.
[0022] In a possible implementation, the first vehicle license plate image is subjected to k-digit number recognition by license plate recognition software to obtain a k-digit number recognition result, 7>k≥2, k is an integer, and step S200 further includes step S210, wherein the license plate recognition software includes a background color recognition network and a symbol segmentation network. Specifically, the license plate recognition software includes a background color recognition network and a symbol segmentation network, and the license plate recognition software is composed of two important parts, and the background color recognition network and the symbol segmentation network work together to complete the recognition task of the k-digit number at the head of the license plate.
[0023] Step S220, the first vehicle license plate image is recognized by the background color recognition network to obtain the background color features of the license plate. Specifically, the first vehicle license plate image is recognized by the background color recognition network to obtain the background color features of the license plate. The background color recognition network uses a specific image recognition algorithm to analyze the first vehicle license plate image. Since the license plate usually has a specific color combination, by recognizing the background color of the license plate, it can provide important clues for subsequent symbol segmentation. The network will extract the color features in the license plate image and determine the background color features of the license plate after processing, such as white characters on a blue background or black characters on a yellow background.
[0024] Step S230, sending the background color feature of the license plate to the symbol segmentation network to remove the background color of the first vehicle license plate image from left to right, and obtaining a segmentation symbol sequence. Specifically, sending the background color feature of the license plate to the symbol segmentation network to remove the background color of the first vehicle license plate image from left to right, and obtaining a segmentation symbol sequence, once the background color feature of the license plate is determined, the information will be passed to the symbol segmentation network, the symbol segmentation network processes the license plate image from left to right according to the background color feature, and removes the background color of the license plate, because the characters and numbers on the license plate are obviously different from the background color, by removing the background color, it is easier to segment the symbols on the license plate, and through this process, a segmentation symbol sequence is obtained, which contains each character and number on the license plate.
[0025] Step S240, extract the head k+1 bit symbol according to the segmentation symbol sequence, and set it as the head k-bit number recognition result. Specifically, extract the head k+1 bit symbol according to the segmentation symbol sequence, and set it as the head k-bit number recognition result. After obtaining the segmentation symbol sequence, extract the head k+1 bit symbol according to the requirements of the scheme, wherein the k+1 bit symbol is for the purpose of being more accurate in the subsequent determination of the head k-bit number recognition result. In some cases, it is necessary to determine the specific content of the k bit according to the information of the k+1 bit, and set the extracted head k+1 bit symbol as the head k-bit number recognition result for subsequent operations such as comparison with the gate entry license plate number library.
[0026] In a possible implementation, the first vehicle license plate image is recognized by the background color recognition network to obtain the background color feature of the license plate. Step S220 further includes step S221, configuring a background degree discrimination formula, wherein the background degree discrimination formula is the ratio of the length of the color union from left to right to the length of the license plate from left to right. Specifically, a background degree discrimination formula is configured, wherein the background degree discrimination formula is the ratio of the length of the color union from left to right to the length of the license plate from left to right. In order to accurately identify the background color feature of the license plate, a background degree discrimination formula is specially configured in the scheme. The formula compares the length of the color union from left to right with the length of the license plate from left to right, and measures the characteristic degree of the license plate background by calculating their ratio. The length of the color union from left to right refers to the total length of various color areas from left to right in the license plate image, and the length of the license plate from left to right is the length of the entire license plate in the horizontal direction. By setting such a formula, the characteristics of the license plate background can be quantitatively evaluated.
[0027] Step S222, according to the background degree discrimination formula, discriminant analysis is performed on the first vehicle license plate image to obtain the color with the maximum background degree, which is set as the license plate background color feature. Specifically, according to the background degree discrimination formula, discriminant analysis is performed on the first vehicle license plate image to obtain the color with the maximum background degree, which is set as the license plate background color feature. After the background degree discrimination formula is obtained, it is applied to the first vehicle license plate image, and the license plate image is analyzed pixel by pixel, and the distribution of different colors on the license plate is counted, and the background degree values of various colors are calculated according to the background degree discrimination formula. By comparing the background degree values of different colors, the color with the maximum background degree is found. The color is considered to be the main background color of the license plate because it occupies a large proportion in the license plate image or has a high feature significance. Finally, the color with the maximum background degree is set as the license plate background color feature, which provides key information for subsequent symbol segmentation and license plate recognition.
[0028] Step S300, compare the head k-digit number recognition result with the entry license plate number library to obtain a first-level entry matching license plate. Specifically, when the vehicle is ready to exit the gate, the license plate recognition software performs head k-digit number recognition on the first vehicle license plate image (7>k≥2) to obtain the head k-digit number recognition result, and then compares this result with the entry license plate number library that stores the license plate number information of the previous entry vehicle, and checks one by one whether the head k digits of each license plate number in the library are consistent with the recognition result of the current exit vehicle, and selects the license plate numbers that meet the conditions to form a first-level entry matching license plate, which provides a basis for the subsequent determination of the identity of the exit vehicle.
[0029] In a possible implementation, the first-level matching license plate is obtained by comparing the first k-digit number recognition result with the entry license plate number library, and step S300 further includes step S310, comparing the first digit of the first k-digit number recognition result with the entry license plate number library to obtain the first-digit entry matching license plate. Specifically, the first digit of the first k-digit number recognition result is compared with the entry license plate number library to obtain the first-digit entry matching license plate, extract the first digit number from the first k-digit number recognition result obtained by the license plate recognition software in the license plate image of the exit vehicle, and compare the first digit number with all the license plate numbers in the entry license plate number library one by one. Since only the first digit is compared, multiple license plate numbers may be screened out as preliminary matching results. The license plate numbers constitute the first-digit entry matching license plate set. By comparing only one digit, the scope of subsequent comparison is quickly narrowed.
[0030] Step S320, compare the first 2-digit number with the first digit number of the gate-matching license plate to obtain a 2-digit number-entry-gate-matching license plate. Specifically, compare the first 2-digit number with the first digit number of the gate-matching license plate to obtain a 2-digit number-entry-gate-matching license plate. After obtaining the first-digit number-entry-gate-matching license plate set, further use the first 2-digit number to compare, extract the first two digits from the first k-digit number recognition result, and then compare with each license plate number in the first-digit number-entry-gate-matching license plate set. Only the license plate whose first two digits of the license plate number are consistent with the first two digits of the head of the exiting vehicle will be retained to form a 2-digit number-entry-gate-matching license plate set, which further narrows the possible matching range on the basis of the first step and improves the matching accuracy.
[0031] Step S330, until the first-level gate-matching license plate is obtained by comparing the k-1-digit number-entry-matching license plate with the k-digit number-identification result of the head. Specifically, until the first-level gate-matching license plate is obtained by comparing the k-1-digit number-entry-matching license plate with the k-1-digit number-entry-matching license plate, according to the above-mentioned method, the number with more digits in the head is used in turn for layer-by-layer comparison. For example, after obtaining a set of 2-digit number-entry-matching license plates, the first 3-digit number is used to compare with the license plates in this set to obtain a set of 3-digit number-entry-matching license plates, and the first 4-digit number is used for comparison, and so on, until the k-1-digit number-entry-matching license plate is compared with the k-1-digit number-entry-matching license plate, and the final set of matching license plates is the first-level gate-matching license plate. The license plates in the set are the most likely to correspond to the current exiting vehicle after gradual screening and comparison, which provides an important basis for the subsequent determination of the identity of the exiting vehicle.
[0032] Step S400, when the number of the first-level entry matching license plates is equal to 1, configure the exit vehicle identification result according to the first-level entry matching license plates. Specifically, when the first k-digit number recognition of the exit vehicle license plate image is performed and compared with the entry license plate number library to determine the first-level entry matching license plate, if its number is 1, the information of the current exit vehicle can be determined based on this unique license plate, including the complete license plate number, the time of entry, the type of vehicle, etc., and then the exit vehicle identification result is generated, which is used for management operations such as controlling the gate opening, recording the exit time and updating the vehicle status, so as to efficiently and accurately complete the identification and management of the exit vehicle, and provide reliable data support for traffic management and safety monitoring.
[0033] In a possible implementation, when the number of the first-level entry-gate matching license plates is equal to 1, the exit-gate vehicle identification result is configured according to the first-level entry-gate matching license plates, and step S400 further includes step S410, when the number of the first-level entry-gate matching license plates is not equal to 1, the vehicle license plate image is subjected to head k+1-digit number recognition by the license plate recognition software to obtain the head k+1-digit number recognition result. Specifically, when the number of the first-level entry-gate matching license plates is not equal to 1, it means that after the head k-digit number recognition, multiple possible matching license plates are obtained, and the specific identity of the exit-gate vehicle cannot be determined. The vehicle license plate image is subjected to head k+1-digit number recognition by the license plate recognition software to obtain the head k+1-digit number recognition result, and the license plate recognition software analyzes the vehicle license plate image again, and this time the number of digits recognized is one more than before, so as to obtain more license plate information, thereby further narrowing the possible matching range.
[0034] Step S420, determine whether the k+1 digit is equal to 7. Specifically, determine whether the k+1 digit is equal to 7. A key determination is made here, because the license plate has a total of 7 digits. If the k+1 digit is equal to 7, it means that the complete license plate number has been recognized (except for the k digits at the head that have been determined before). If it does not reach 7 digits, the number of recognized digits can be further increased for more accurate matching.
[0035] Step S430, when k+1=7, compare the first-level gate-entry matching license plate with the first-level gate-entry matching license plate according to the head k+1 digit number recognition result to obtain the second-level gate-entry matching license plate, and configure the gate-exit vehicle recognition result according to the second-level gate-entry matching license plate. Specifically, when k+1=7, compare the first-level gate-entry matching license plate with the first-level gate-entry matching license plate according to the head k+1 digit number recognition result to obtain the second-level gate-entry matching license plate, and configure the gate-exit vehicle recognition result according to the second-level gate-entry matching license plate. In this case, since the complete license plate number (except the initial head k digits) has been recognized, this complete recognition result is compared with the first-level gate-entry matching license plate. After comparison, a more accurate second-level gate-entry matching license plate is screened out, and the identity of the gate-exiting vehicle is determined according to the second-level gate-entry matching license plate, and the gate-exiting vehicle recognition result is configured. The result can be used to control the opening of the gate and other subsequent operations.
[0036] Step S440, when k+1<7, compare the first-level gate-entry matching license plate with the first-level gate-entry matching license plate according to the head k+1-digit number recognition result to obtain the second-level gate-entry matching license plate, and when the number of the second-level gate-entry matching license plates is equal to 1, configure the gate-exit vehicle recognition result according to the second-level gate-entry matching license plate. Specifically, when k+1<7, compare the first-level gate-entry matching license plate with the first-level gate-entry matching license plate according to the head k+1-digit number recognition result to obtain the second-level gate-entry matching license plate. Although the complete license plate number is not recognized, by adding one recognition digit, it is compared with the first-level gate-entry matching license plate again to further narrow the matching range. When the number of the second-level gate-entry matching license plate is equal to 1, configure the gate-exit vehicle recognition result according to the second-level gate-entry matching license plate. If after this round of comparison, the number of the second-level gate-entry matching license plate is 1, then the identity of the gate-exiting vehicle can be determined. As in the previous steps, the gate-exiting vehicle recognition result is configured according to this unique second-level gate-entry matching license plate.
[0037] In the embodiment of the present application, when a vehicle exits the gate, the image acquisition hardware captures the license plate image, and the license plate recognition software recognizes the k-digit numbers at the head of the image (7>k≥2), and compares the recognition result with the entry license plate number library to obtain a first-level entry matching license plate. If the number of matching license plates is 1, the exit vehicle recognition result is configured according to the license plate. By analyzing the vehicle exit behavior, the partial number on the head of the license plate is first recognized, and then compared with the entry license plate number library, the computing power usage and comparison calculation amount are reduced, thereby achieving the technical effect of improving the vehicle exit efficiency.
[0038] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
Claims
1. A method for intelligent identification of vehicle entry and exit based on behavior analysis, characterized in that: include: When the vehicle is leaving the gate, the license plate image of the first vehicle is captured by the image acquisition hardware; Using license plate recognition software to identify the first k-digit number of the first vehicle license plate image, and obtain a recognition result of the first k-digit number, 7>k≥2, k is an integer; Compare the first k-digit number recognition result with the entry license plate number database to obtain a first-level entry matching license plate; When the number of the first-level entry matching license plates is equal to 1, the exit vehicle recognition result is configured according to the first-level entry matching license plates.
2. The method according to claim 1, characterized in that Also includes: When the number of matching license plates at the first-level gate is not equal to 1, the license plate recognition software is used to recognize the k+1 digits of the first part of the license plate image to obtain a recognition result of the k+1 digits of the first part; Determine whether the k+1 bit is equal to 7; When k+1=7, the first k+1 digit number recognition result is compared with the first-level entry matching license plate to obtain the second-level entry matching license plate, and the exit vehicle recognition result is configured according to the second-level entry matching license plate; When k+1<7, compare the recognition result of the k+1-digit number on the head with the first-level entry matching license plate to obtain the second-level entry matching license plate. When the number of the second-level entry matching license plates is equal to 1, configure the exit vehicle recognition result according to the second-level entry matching license plate.
3. The method according to claim 1, characterized in that Also includes: When the vehicle enters the gate, the license plate image of the second vehicle is captured by the image capture hardware; Perform 7-digit number recognition on the second vehicle license plate image by the license plate recognition software to obtain a vehicle license plate recognition result; The vehicle license plate recognition result and the gate entry timestamp are stored in the gate entry license plate number library.
4. The method according to claim 1, characterized in that The first vehicle license plate image is subjected to the first k-digit number recognition by the license plate recognition software to obtain the first k-digit number recognition result, including: The license plate recognition software includes a background color recognition network and a symbol segmentation network; Recognize the first vehicle license plate image through the background color recognition network to obtain a license plate background color feature; Sending the license plate background color feature to the symbol segmentation network to remove the background color of the first vehicle license plate image from left to right to obtain a segmentation symbol sequence; According to the segmentation symbol sequence, the head k+1-bit symbol is extracted and set as the head k-bit number recognition result.
5. The method according to claim 4, characterized in that Recognizing the first vehicle license plate image through the background color recognition network to obtain a license plate background color feature includes: Configure a background degree discrimination formula, wherein the background degree discrimination formula is the ratio of the length of the union of colors from left to right to the length of the license plate from left to right; According to the background degree discrimination formula, the first vehicle license plate image is discriminated and analyzed to obtain the color with the maximum background degree, which is set as the license plate background color feature.
6. The method according to claim 1, characterized in that The first k-digit number recognition result is compared with the entry license plate number database to obtain a first-level entry matching license plate, including: Compare the first digit of the k-digit number recognition result with the entry license plate number library to obtain the first digit entry matching license plate; Compare the first two-digit number with the first-digit entry-gate matching license plate to obtain a two-digit entry-gate matching license plate; Until the first k-digit number recognition result is compared with the k-1-digit number entry matching license plate to obtain the first-level entry matching license plate.
7. The method according to claim 1, characterized in that Also includes: The millimeter-wave radar is used to collect the vehicle motion positioning sequence of the preset range lane centered on the entrance and exit gate; According to the vehicle motion positioning sequence, obtaining the vehicle head direction; When the vehicle's front is facing the gate and the vehicle is inside the gate, the vehicle is exiting the gate; When the front of the vehicle is facing the gate and the vehicle is outside the gate, the vehicle is entering the gate.
8. The method according to claim 7, characterized in that Also includes: When the front of the vehicle is not facing the gate, the vehicle is not in or out of the gate.
Citation Information
Patent Citations
License plate recognition method, license plate recognition module, cloud server and parking lot management system
CN107680382A
Intelligent license plate recognition method
CN110991467A
Vehicle illegal driving detection method and device, computer equipment and storage medium
CN111539317A
License plate matching method and device, electronic equipment and readable storage medium
CN114120298A
