CN116665321A - Parking lot vehicle management method based on edge nano-tube technology - Google Patents
Parking lot vehicle management method based on edge nano-tube technology Download PDFInfo
- Publication number
- CN116665321A CN116665321A CN202310508871.3A CN202310508871A CN116665321A CN 116665321 A CN116665321 A CN 116665321A CN 202310508871 A CN202310508871 A CN 202310508871A CN 116665321 A CN116665321 A CN 116665321A
- Authority
- CN
- China
- Prior art keywords
- license plate
- image information
- edge
- result
- parking lot
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000002071 nanotube Substances 0.000 title claims abstract description 43
- 238000005516 engineering process Methods 0.000 title claims abstract description 35
- 238000007726 management method Methods 0.000 title abstract description 47
- 238000000034 method Methods 0.000 claims abstract description 52
- 238000001514 detection method Methods 0.000 claims abstract description 25
- 230000011218 segmentation Effects 0.000 claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims description 71
- 238000007781 pre-processing Methods 0.000 claims description 29
- 239000011159 matrix material Substances 0.000 claims description 23
- 238000012545 processing Methods 0.000 claims description 20
- 238000003708 edge detection Methods 0.000 claims description 19
- 230000009466 transformation Effects 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 7
- 230000000877 morphologic effect Effects 0.000 claims description 5
- 238000009499 grossing Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 230000006872 improvement Effects 0.000 abstract description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 239000000463 material Substances 0.000 abstract 1
- 239000013598 vector Substances 0.000 description 16
- 230000008569 process Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 8
- 230000004044 response Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000005265 energy consumption Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 229910052704 radon Inorganic materials 0.000 description 1
- SYUHGPGVQRZVTB-UHFFFAOYSA-N radon atom Chemical compound [Rn] SYUHGPGVQRZVTB-UHFFFAOYSA-N 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07B—TICKET-ISSUING APPARATUS; FARE-REGISTERING APPARATUS; FRANKING APPARATUS
- G07B15/00—Arrangements or apparatus for collecting fares, tolls or entrance fees at one or more control points
- G07B15/02—Arrangements or apparatus for collecting fares, tolls or entrance fees at one or more control points taking into account a variable factor such as distance or time, e.g. for passenger transport, parking systems or car rental systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Finance (AREA)
- Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Traffic Control Systems (AREA)
- Devices For Checking Fares Or Tickets At Control Points (AREA)
Abstract
The disclosure discloses a parking lot vehicle management method based on an edge nano-tube technology. According to the method, the license plate is identified by collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot and utilizing the license plate identification model, the entrance time stamp of the vehicle is recorded, and the parking cost is calculated. Meanwhile, the method realizes automatic identification and charging of license plate parts through detection and character segmentation technology of license plate areas, and improves management efficiency of a parking lot. The method can realize quick, accurate and automatic vehicle management and has the beneficial effects of high efficiency, manpower and material resource saving and management level improvement.
Description
Technical Field
The present disclosure relates to the field of vehicle management technologies, and in particular, to a method for managing vehicles in a parking lot based on an edge nanotube technology.
Background
With the continuous development of cities, the popularization rate of private cars is continuously improved, and parking difficulty becomes a ubiquitous problem. In order to solve this problem, parking lots in various places have been developed. However, in the running parking lot, there are some problems, such as people's feelings, remote control opening gate, entering and leaving, cash or personal WeChat collection, illegal making of profit, etc., which seriously affect the normal operation of the parking lot, resulting in the loss of the charging record of the parking lot, reduced income, and inaccurate checking of the bill. Therefore, there is a need to enhance the management of parking lot charges and the supervision of charging behavior to improve the efficiency and transparency of parking lot management.
In the prior art, management and billing of parking lots mainly relies on manual billing by parking lot management personnel. In this case, human errors such as calculation errors or missing fee records are liable to occur. In addition, a manager of the parking lot spends a lot of time and effort to collect, sort and check the parking records, which may greatly waste human resources. In addition, the defect of the parking lot billing management software is also a common problem, such as the fact that the vehicle type and the license plate number cannot be identified correctly, which results in inaccurate billing records in the parking lot and inaccurate billing. The existence of these problems results in inefficient management of the parking lot, affecting the user experience.
For these problems, edge nanotube technology has evolved. The edge nano-tube technology is an emerging technology, and integrates the functions of calculation, storage, network communication and the like into the edge of the equipment, so that the equipment has the intelligent and autonomous decision-making capability. In parking lots, edge-based nanotubes technology can be used for automatic identification of vehicles, lane control, and billing. Specifically, the edge nano-tube technology can automatically identify the type and license plate number of the vehicle through the camera and the vehicle identification technology, the vehicle is controlled to enter and exit through the lane control system, the parking cost is automatically calculated through the charging system, and the charging record is uploaded to the cloud server, so that parking lot management personnel can check and check at any time.
Compared with the traditional parking lot management mode, the parking lot vehicle management method based on the edge nano-tube technology has the following advantages: firstly, the management efficiency and the transparency of the parking lot can be greatly improved, and the occurrence of human errors is reduced; secondly, the manpower resources can be greatly saved, and the operation efficiency of the parking lot is improved; thirdly, the safety and the accuracy of the parking lot can be improved, and the cost of parking lot management is reduced. These advantages make the parking lot vehicle management method based on the edge nano-tube technology a trend of future parking lot development.
In the prior art, although some parking lot management systems have employed automation technology, there are still some problems. For example, in some parking lots, illegal activities such as illegally making a profit and opening a gate still exist, and these activities may seriously affect the normal operation of the parking lot. In addition, in the parking lot, the number of vehicles is numerous, if the system fails, the charging record of the parking lot is lost, and the bill cannot be checked accurately. To solve these problems, a parking lot vehicle management method based on an edge nanotube technology has been developed.
Disclosure of Invention
The disclosure provides a parking lot vehicle management method based on an edge nano-tube technology.
In order to solve the problems, the technical scheme of the invention is realized as follows:
a method for managing vehicles in a parking lot based on an edge nano-tube technology, the method performing the steps of:
step S1: collecting entrance image information of a vehicle entering a parking lot and exit image information of the vehicle leaving the parking lot; the entrance image information and the exit image information at least comprise license plate parts of vehicles;
step S2: uploading the acquired entrance image information to an edge nano tube cloud platform; the edge nano tube cloud platform uses a preset license plate recognition model to recognize the entering image information, a first recognition result is obtained, the first recognition result is stored in a memory of the edge nano tube cloud platform, and meanwhile, the vehicle entering time stamp is recorded;
step S3: uploading the acquired out-of-field image information to an edge nano tube cloud platform; the edge nano-tube cloud platform uses a preset license plate recognition model to recognize the outgoing image information, and a second recognition result is obtained; based on the second recognition results, carrying out matching query on all the stored first recognition results in the memory to obtain matching query results;
step S4: if the matching inquiry result is true, calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the arrival time; if the matching inquiry result is false, calling a preset matching decision model, and carrying out similarity calculation on all the stored first recognition results in a memory based on the second recognition result to obtain a similarity calculation result; the similarity calculation result is a percentage result and is used for representing the similarity between the second identification result and all the first identification results stored in the memory; finding out a first identification result corresponding to the maximum similarity calculation result, obtaining a corresponding vehicle entrance time stamp, and calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the entrance time.
Further, in step S1, the high definition recognition camera is used to collect the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot.
Further, the step S1 further includes a step of performing image processing on the entrance image information and the exit image information after collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot, and specifically includes: image preprocessing is carried out on the acquired entrance image information and exit image information, and a preprocessing result is obtained; carrying out license plate detection on the preprocessing result to determine a license plate region, and carrying out license plate segmentation on the basis of the determined license plate region to obtain a license plate part of the entering image information or the exiting image information and other image parts except the license plate part; the license plate part is used as the entrance image information or the exit image information, and other image parts except the license plate part are added as the additional information.
Further, the image preprocessing sequentially performs the following steps: image denoising, image enhancement and gray level conversion to obtain a preprocessing result.
Further, the method for detecting the license plate of the preprocessing result to determine the license plate area comprises the following steps: performing edge detection on the preprocessing result by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain all possible straight line candidate sets; randomly extracting two straight lines as edges of the license plate, and calculating intersection coordinates of the two straight lines as candidate points of the upper left corner of the license plate; in a surrounding area formed by 200x100 pixels of a candidate point of the upper left corner of the license plate, performing edge detection by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain a possible straight line candidate set in the area; randomly extracting a group of straight lines from the straight line candidate set, and calculating the intersection point coordinates of the straight lines as candidate points of the lower right corner of the license plate; taking candidate points of the upper left corner and the lower right corner of the license plate as 4 vertexes to form a quadrangle of the license plate; calculating the score of the quadrangle, wherein the higher the score is, the more the quadrangle looks like a license plate, and the formula for calculating the score is as follows:
score＝w1*ratio+w2*compactness+w3*anglediff；
Wherein ratio represents the aspect ratio of the license plate, compatibility represents the compactness of the quadrangle, the area ratio of the interior area to the exterior area of the quadrangle is defined, angle represents the angle difference of the quadrangle boundary, and the maximum difference value between four angles is defined; score is a quadrilateral score; w1, w2 and w3 are weight coefficients and are set values; and repeatedly executing all the steps for N times, and finally selecting the quadrangle with the highest score as the determined license plate area.
Further, the method for obtaining the license plate part of the entrance image information or the exit image information and other image parts except the license plate part by carrying out license plate segmentation based on the determined license plate region comprises the following steps: converting the gray level image of the license plate region into a binary image; performing morphological processing on the binary image, removing unnecessary noise and details, and reserving a communication area on a license plate; edge detection is carried out on the connected region on the license plate, and the edge of the license plate region is obtained; extracting and describing the characteristics of the edge to obtain the coordinate information of the starting point and the end point of the edge and the direction information of the edge; constructing an adjacent matrix of the edge, and judging an adjacent relation according to the direction information of the edge; traversing the adjacent matrix to obtain all strong connected components. Screening and processing the strong connected components to obtain edge information of each character; according to the edge information of each character, the license plate is divided into a plurality of character areas, and the character areas are used as license plate parts of the entrance image information or the exit image information.
Further, the method for identifying the entrance image information by the license plate identification model or identifying the exit image information by the license plate identification model comprises the following steps: smoothing the license plate part by using a Gaussian filter; detecting key points by using the improved Hessian matrix; computing Haar wavelet features in a region around each keypoint; generating descriptors using Haar wavelet features; and then based on the descriptors, the pre-established recognition templates are used for recognition.
Further, the modified Hessian matrix is defined as:
wherein I is a license plate part, and x and y are coordinates in the license plate part; eigenvalue lambda of improved Hessian matrix 1 And lambda (lambda) 2 Used for judging whether the point is an extreme point or not.
Further, in the step S4, a preset matching decision model is called, and based on the second recognition result, similarity calculation is performed on all the stored first recognition results in the memory, so as to obtain a similarity calculation result, where the method includes: based on the second recognition result, calling the corresponding additional information, and performing similarity calculation on all the stored first recognition results in a memory to obtain a similarity calculation result; firstly, based on a second recognition result, performing similarity calculation on all stored first recognition results in a storage mode to obtain a first similarity calculation result; then, invoking additional information in the outgoing image information corresponding to the second identification result to perform similarity calculation with additional information in the incoming image information corresponding to the first identification result stored in the memory, so as to obtain a second similarity calculation result; and carrying out weighted average operation on the first similarity calculation result and the second similarity calculation result by using a preset weighting coefficient to obtain a final similarity calculation result.
Further, the additional information at least includes: vehicle brand, vehicle color, and vehicle type.
The parking lot vehicle management method based on the edge nano-tube technology has the following beneficial effects: the invention aims to solve a series of problems existing in the traditional parking lot management mode. In the traditional parking lot management mode, the charging mainly depends on manual charging of parking lot management personnel, and the problems of human error, calculation error or missing charging records and the like are easy to occur. In addition, a manager of the parking lot spends a lot of time and effort collecting, sorting, and checking the parking records, which wastes human resources. Meanwhile, the traditional parking lot management mode also has the problems of inaccurate identification of vehicle types and license plate numbers, and the like, so that the charging record of the parking lot is inaccurate, and the bill cannot be checked accurately. The existence of these problems results in inefficient management of the parking lot, affecting the user experience.
The technical scheme of the intelligent parking lot and the autonomous decision making capability are realized through functions of automatic vehicle identification, lane control, charging system and the like based on the edge nano-tube technology. Specifically, the technical scheme of the patent comprises the following aspects:
Firstly, when a vehicle enters a parking lot, using a high-definition identification camera to collect entrance image information of the vehicle, performing image processing and license plate detection on the entrance image information to determine a license plate area, and finally obtaining a license plate part of the entrance image information and other image parts except the license plate part. When the vehicle leaves the parking lot, the high-definition recognition camera is also used for collecting the outgoing image information of the vehicle, similar image processing and license plate detection are carried out on the outgoing image information so as to determine a license plate area, and finally, a license plate part and other image parts except the license plate part of the outgoing image information are obtained. Then, the license plate part is identified by using the license plate identification model, and information such as a license plate number is obtained and stored in a memory.
Secondly, aiming at illegal actions such as illegal break-out and the like which possibly occur when a vehicle leaves a parking lot, the technical scheme of the patent adopts a matching decision model of two recognition results to process. Specifically, when the vehicle leaves the parking lot, the license plate part of the outgoing image information is identified by using the license plate identification model, and a second identification result is obtained. And then, based on the second recognition results, performing similarity calculation on all the stored first recognition results in the memory to obtain similarity calculation results. Finally, a preset matching decision model is used for judging and processing the similarity calculation result, so that illegal making of a brake is reduced, and in addition, the safety and accuracy of the parking lot can be improved by the parking lot vehicle management method based on the edge nano tube technology. In the traditional parking lot management mode, manual inspection and management are often required by parking lot management personnel, and certain safety risks can be brought. The parking lot vehicle management method based on the edge nano-tube technology can realize automatic identification and access control of vehicles, so that the chance of manual intervention is reduced, and the safety of the parking lot is improved. In addition, the method adopts a vehicle identification technology and a charging system, so that the accuracy of the charging record can be ensured, the artificial errors such as calculation errors, missing of the charging record and the like are avoided, and the accuracy of charging is improved.
Finally, the parking lot vehicle management method based on the edge nano-tube technology can also reduce the cost of parking lot management. In the conventional parking lot management method, a large amount of human resources are required for management and maintenance, and a large amount of funds are required to be invested to purchase hardware equipment and a maintenance system. The parking lot vehicle management method based on the edge nano-tube technology can realize automatic identification and access control of vehicles through the edge equipment such as the camera, the lane control system and the like, and reduces the input of human resources. In addition, the method can upload the charging records to the cloud server for management and verification, so that the cost of parking lot management is reduced.
In summary, the parking lot vehicle management method based on the edge nano-tube technology has the advantages of high efficiency, accuracy, safety, cost saving and the like, can greatly improve the management efficiency and transparency of the parking lot, saves manpower resources, reduces the cost of parking lot management, and is a trend of future parking lot development.
Drawings
Fig. 1 is a schematic flow chart of a parking lot vehicle management method based on an edge nano-tube technology according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a processing flow of a parking lot vehicle management method based on an edge nanotube technology when a vehicle enters a field according to an embodiment of the present invention;
Fig. 3 is a schematic diagram of a processing flow of a parking lot vehicle management method based on an edge nanotube technology when a vehicle enters a field according to an embodiment of the present invention.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects to be solved by the present disclosure more clear and obvious, the present disclosure is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present disclosure and are not intended to limit the present disclosure. .
Example 1
Referring to fig. 1, a method for managing vehicles in a parking lot based on an edge nano tube technology, the method performs the steps of:
step S1: collecting entrance image information of a vehicle entering a parking lot and exit image information of the vehicle leaving the parking lot; the entrance image information and the exit image information at least comprise license plate parts of vehicles;
when a vehicle enters a parking lot, the method can acquire the entrance image information of the vehicle and identify the license plate part of the vehicle. Likewise, when the vehicle leaves the parking lot, the method collects outgoing image information of the vehicle and identifies a license plate portion of the vehicle. License plate recognition is realized through a preset license plate recognition model, and can be optimized and adjusted according to actual needs.
Step S2: uploading the acquired entrance image information to an edge nano tube cloud platform; the edge nano tube cloud platform uses a preset license plate recognition model to recognize the entering image information, a first recognition result is obtained, the first recognition result is stored in a memory of the edge nano tube cloud platform, and meanwhile, the vehicle entering time stamp is recorded;
step S3: uploading the acquired out-of-field image information to an edge nano tube cloud platform; the edge nano-tube cloud platform uses a preset license plate recognition model to recognize the outgoing image information, and a second recognition result is obtained; based on the second recognition results, carrying out matching query on all the stored first recognition results in the memory to obtain matching query results;
step S4: if the matching inquiry result is true, calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the arrival time; if the matching inquiry result is false, calling a preset matching decision model, and carrying out similarity calculation on all the stored first recognition results in a memory based on the second recognition result to obtain a similarity calculation result; the similarity calculation result is a percentage result and is used for representing the similarity between the second identification result and all the first identification results stored in the memory; finding out a first identification result corresponding to the maximum similarity calculation result, obtaining a corresponding vehicle entrance time stamp, and calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the entrance time.
Specifically, when a vehicle enters a parking lot, the camera is used for collecting the entrance image information of the vehicle, a preset license plate recognition model is used for recognizing the license plate of the vehicle, a first recognition result is stored in a memory of the edge nano tube cloud platform, and the entrance time stamp of the vehicle is recorded.
When the vehicle leaves the parking lot, acquiring the departure image information of the vehicle through a camera, identifying the license plate of the vehicle by using a preset license plate identification model, carrying out matching inquiry on a second identification result and a first identification result stored in a memory, and if the matching is successful, calculating parking cost according to the arrival time stamp of the vehicle and the current departure time; if the matching fails, calculating the similarity between the second recognition result and all the first recognition results by using a preset matching decision model, finding the first recognition result corresponding to the maximum similarity, and calculating the parking cost according to the entry time stamp of the result.
In addition, the method also uses the edge nano tube technology, namely, the acquired entrance image information and the acquired exit image information are uploaded to the edge nano tube cloud platform for processing, so that the burden of local equipment is reduced, and real-time processing and analysis can be realized. In addition, the preset license plate recognition model and the matching decision model can be optimized and adjusted according to actual needs, and accuracy and performance of the system are improved.
Referring to fig. 2 and 3, when a vehicle enters a parking lot, an entrance recognition camera first recognizes a snap-shot license plate picture, judges whether the vehicle enters the parking lot, if not, continues to monitor, if not, the entrance counter counts one more, meanwhile, the snap-shot picture is uploaded to a cloud server, OCR license plate recognition is performed in the cloud server, features are extracted by using a feature extractor, and then the vehicle model, the vehicle brand, the license plate number and the vehicle feature vector are obtained. Finally, the entering time, license plate number and entering times are stored.
When the vehicle leaves the parking lot, the exit recognition camera firstly recognizes the snapshot license plate picture, then judges whether the vehicle leaves the parking lot, if not, continues to monitor, if not, the exit counter counts one more, and meanwhile uploads the snapshot picture to the cloud server for OCR license plate recognition and feature extraction by using the feature extractor, and the vehicle model, the vehicle brand, the vehicle number and the vehicle feature vector are obtained. Finally, judging whether the vehicle is a fixed vehicle, if so, charging 0 yuan, and storing the departure time, license plate number, parking cost and departure times; if not, performing entrance record matching on all present vehicle data of the parking lot, performing accurate matching and partial accurate matching according to the matching result, and selecting record charging with highest confidence coefficient in the partial accurate matching.
Example 2
On the basis of the above embodiment, the high-definition recognition camera is used in step S1 to collect the entry image information of the vehicle at the entrance into the parking lot and the exit image information at the exit from the parking lot.
In the step S1, the high-definition recognition camera is used for collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot, which are important bases for realizing license plate recognition and vehicle management, and the high-definition recognition camera is used for ensuring the definition and quality of the images and improving the accuracy of license plate recognition.
In practical application, the high-definition recognition camera generally has the characteristics of high resolution, high frame rate, high dynamic range and the like, and can acquire clear images under different illumination conditions. In addition, some high-end high-definition identification cameras also have the functions of self-adaptive exposure, automatic focusing and the like, and can automatically adjust camera parameters to adapt to different shooting scenes, so that the accuracy and the robustness of license plate identification are further improved.
It is worth noting that the image acquisition by using the high-definition recognition camera has a certain influence on the cost and energy consumption of the system, and needs to be fully considered in the system design. In addition, the high-definition recognition camera is required to be reasonably installed and debugged so as to ensure the quality and stability of image acquisition.
Example 3
On the basis of the above embodiment, the step S1 further includes the step of performing image processing on the entrance image information and the exit image information after collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot, and specifically includes: image preprocessing is carried out on the acquired entrance image information and exit image information, and a preprocessing result is obtained; carrying out license plate detection on the preprocessing result to determine a license plate region, and carrying out license plate segmentation on the basis of the determined license plate region to obtain a license plate part of the entering image information or the exiting image information and other image parts except the license plate part; the license plate part is used as the entrance image information or the exit image information, and other image parts except the license plate part are added as the additional information.
Step S1, after collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot, further comprises the step of carrying out image processing on the entrance image information and the exit image information. The method aims at improving the accuracy of license plate recognition, separating license plate parts from other parts in an image through operations such as image preprocessing, license plate detection, license plate segmentation and the like, taking the license plate parts as part of entrance image information or exit image information, and simultaneously taking the other parts as additional information to provide more information support for subsequent vehicle management and charging.
Specifically, the image preprocessing includes denoising, graying, sharpening, and the like, on the acquired in-field image information and out-field image information, so as to improve the image quality and contrast. The license plate detection is to locate the position and the size of the license plate in the image aiming at the preprocessed image so as to carry out subsequent license plate segmentation. The license plate segmentation is to separate the license plate region to obtain the image information of the license plate part and other parts.
In practical application, image preprocessing, license plate detection and license plate segmentation are core links in license plate recognition technology, and can directly influence the accuracy and speed of license plate recognition. For different image acquisition scenes and license plate types, proper algorithms and parameters are required to be selected for optimization and adjustment so as to achieve the optimal recognition effect. In addition, the implementation of the license plate recognition algorithm also needs to fully consider the resources and energy consumption of the system so as to realize efficient edge calculation.
Example 4
On the basis of the above embodiment, the image preprocessing sequentially performs the following steps: image denoising, image enhancement and gray level conversion to obtain a preprocessing result.
Specifically, image denoising is the first step of preprocessing, and the purpose of the image denoising is to eliminate noise in the image, so that a license plate area is clearer. Image denoising can employ various algorithms, such as median filtering, gaussian filtering, and the like. These algorithms may be selected and optimized for different noise types and strengths.
Image enhancement is the second step of preprocessing, and the purpose of the image enhancement is to enhance the contrast of the license plate area so that the license plate area is clearer. Image enhancement may employ various algorithms such as histogram equalization, contrast stretching, etc. The algorithms can be selected and optimized according to the characteristics of license plate images and illumination conditions.
The gray level conversion is the last step of preprocessing, and the purpose of the gray level conversion is to convert the license plate image into a gray level image, so that the subsequent license plate detection and license plate segmentation are facilitated. The gray scale conversion may employ various algorithms such as an average method, a weighted average method, and the like. The algorithms can be selected and optimized according to the characteristics and requirements of license plate images.
It should be noted that the effect of image preprocessing directly affects the accuracy and robustness of subsequent license plate detection and license plate segmentation, so that sufficient experiments and optimization are required in practical application. Meanwhile, the image preprocessing also needs to consider factors such as system resources and energy consumption and the like so as to realize efficient edge calculation.
Example 5
On the basis of the above embodiment, the method for detecting the license plate of the preprocessing result to determine the license plate area includes: performing edge detection on the preprocessing result by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain all possible straight line candidate sets; randomly extracting two straight lines as edges of the license plate, and calculating intersection coordinates of the two straight lines as candidate points of the upper left corner of the license plate; in a surrounding area formed by 200x100 pixels of a candidate point of the upper left corner of the license plate, performing edge detection by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain a possible straight line candidate set in the area; randomly extracting a group of straight lines from the straight line candidate set, and calculating the intersection point coordinates of the straight lines as candidate points of the lower right corner of the license plate; taking candidate points of the upper left corner and the lower right corner of the license plate as 4 vertexes to form a quadrangle of the license plate; calculating the score of the quadrangle, wherein the higher the score is, the more the quadrangle looks like a license plate, and the formula for calculating the score is as follows:
score＝w1*ratio+w2*compactness+w3*anglediff；
Wherein ratio represents the aspect ratio of the license plate, compatibility represents the compactness of the quadrangle, the area ratio of the interior area to the exterior area of the quadrangle is defined, angle represents the angle difference of the quadrangle boundary, and the maximum difference between four angles is defined; score is a quadrilateral score; w1, w2 and w3 are weight coefficients and are set values; and repeatedly executing all the steps for N times, and finally selecting the quadrangle with the highest score as the determined license plate area.
Specifically, when license plate detection is performed, it is generally necessary to use an edge detection algorithm to extract edge information in an image. The Sobel operator is a commonly used edge detection algorithm, and can extract horizontal and vertical edge information in an image.
Specifically, the specific process of edge detection using the Sobel operator is as follows:
and carrying out gray level conversion on the pretreatment result to obtain a gray level image. The Sobel operator is defined, typically a 3x3 matrix, sx and Sy, respectively, corresponding to the edge detection in the horizontal and vertical directions, respectively. The Sx and Sy matrices are generally as follows:
carrying out convolution operation on the gray level image, and respectively carrying out convolution by using Sx and Sy operators to obtain a gradient image G in the horizontal direction and a gradient image G in the vertical direction x And G y 。
Calculating the gradient amplitude and direction, wherein the formula is as follows:θ＝arctan(G y /G x ) Where G is the gradient magnitude and θ is the gradient direction.
The gradient amplitude is binarized, all pixels below the set threshold are set to 0, and all pixels above the set threshold are set to 255.
And obtaining an edge detection image.
In practical application, the Sobel operator can be optimized and adjusted as required to achieve the best edge detection effect. Meanwhile, the Sobel operator can be combined with other edge detection algorithms, such as Canny operators, so that accuracy and robustness of license plate detection are further improved.
The process of straight line detection using hough transform is as follows:
and performing edge detection on the image to obtain an edge binarization image.
A set of parameter spaces is defined for describing the likelihood of a straight line in an image. For two-dimensional line detection, the parameter space is typically two parameters: a straight line slope k and an intercept b.
For each edge point, the probability line for that edge point is incremented in parameter space, i.e. a counter is incremented at the corresponding position of k and b.
And counting all possible straight lines in the parameter space, and finding out the straight line with the highest counter value, namely the straight line with the highest probability. Whether a straight line exists or not may be determined according to a set threshold value.
For each line detected, the position and direction of the line can be determined by calculating their intersection with the image edge.
In license plate detection, hough transform is typically used to detect edge lines of the license plate. When determining the candidate point of the upper left corner of the license plate, generally, straight line detection in the horizontal direction can be performed first, and then two straight lines are randomly extracted from the straight line candidate set to serve as the left edge and the right edge of the license plate. When the candidate point of the lower right corner of the license plate is determined, straight line detection in the vertical direction can be performed in the area determined by the left edge and the right edge, and then a straight line is randomly extracted from the straight line candidate set to serve as the lower edge of the license plate.
Example 6
On the basis of the above embodiment, the method for obtaining the license plate part of the entrance image information or the exit image information and other image parts except the license plate part by carrying out license plate segmentation based on the determined license plate region includes: converting the gray level image of the license plate region into a binary image; performing morphological processing on the binary image, removing unnecessary noise and details, and reserving a communication area on a license plate; edge detection is carried out on the connected region on the license plate, and the edge of the license plate region is obtained; extracting and describing the characteristics of the edge to obtain the coordinate information of the starting point and the end point of the edge and the direction information of the edge; constructing an adjacent matrix of the edge, and judging an adjacent relation according to the direction information of the edge; traversing the adjacent matrix to obtain all strong connected components. Screening and processing the strong connected components to obtain edge information of each character; according to the edge information of each character, the license plate is divided into a plurality of character areas, and the character areas are used as license plate parts of the entrance image information or the exit image information.
Specifically, after the license plate region is determined, the license plate region needs to be further divided, and the license plate is divided into a plurality of character regions so as to be conveniently identified. The specific license plate segmentation process is as follows:
converting the gray level image of the license plate area into a binary image, and dividing all pixels in the license plate area into two types: foreground pixels and background pixels on the license plate.
Morphological processing is carried out on the binary image, mainly unnecessary noise and details are removed, and a communication area on a license plate is reserved. The morphological treatment generally comprises two steps of corrosion and expansion, and parameters can be adjusted according to actual conditions so as to realize the optimal license plate segmentation effect.
And carrying out edge detection on the connected region on the license plate to obtain the edge of the license plate region. The edges can be obtained by edge detection algorithms such as Sobel operators.
And extracting and describing the characteristics of the edge to obtain the coordinate information of the starting point and the ending point of the edge and the direction information of the edge. Common feature extraction algorithms include Hough transform and Radon transform.
And constructing an adjacent matrix of the edge, and judging an adjacent relation according to the direction information of the edge. The adjacency matrix can represent adjacency relations among edges, and can be optimized and adjusted according to actual needs.
Traversing the adjacent matrix to obtain all strong connected components. A strongly connected component refers to a collection of edges where connectivity exists on the license plate, and may typically represent a character or a symbol.
And screening and processing the strong connected components to obtain the edge information of each character. Character edges may be screened and judged by geometric features such as width, height, aspect ratio, etc.
According to the edge information of each character, the license plate is divided into a plurality of character areas, and the character areas are used as license plate parts of the entrance image information or the exit image information.
Example 7
On the basis of the above embodiment, the method for identifying the entrance image information by the license plate identification model or identifying the exit image information by the license plate identification model includes: smoothing the license plate part by using a Gaussian filter; detecting key points by using the improved Hessian matrix; computing Haar wavelet features in a region around each keypoint; generating descriptors using Haar wavelet features; and then based on the descriptors, the pre-established recognition templates are used for recognition.
The specific license plate recognition process is as follows:
and smoothing the license plate part by using a Gaussian filter to remove noise and details, and keeping important features on the license plate.
The keypoints are then detected using a modified Hessian matrix. The Hessian matrix is a second derivative operator that can detect local extremum points in an image. The improved Hessian matrix refers to improvement and optimization of the original Hessian matrix so as to improve the detection effect of key points of the license plate.
Haar wavelet features are computed in the region around each keypoint. Haar wavelet features are a method for image feature extraction that can characterize an image by computing wavelet responses of different scales and directions in the image.
The descriptor is generated using Haar wavelet features. The descriptor refers to encoding license plate features into fixed length vectors for comparison and matching. Common descriptive sub-algorithms include SIFT, SURF, ORB, etc.
And then based on the descriptors, the pre-established recognition templates are used for recognition. The recognition template is a license plate number sample library which is collected and established in advance, and can convert license plate numbers into character sequences according to the characteristics and rules of license plates and store the character sequences in the template library. During recognition, the best matching sample can be found by comparing the descriptors with the samples in the template library, and a recognition result is output.
The process of computing Haar wavelet features is generally divided into the following steps:
The license plate partial image is divided into a number of small blocks, typically 4x4 or 8x8 sized blocks, each of which is a rectangular area.
And carrying out Haar wavelet transformation on the pixel values in each small block to obtain response values in the horizontal, vertical and diagonal directions. The image is typically transformed using a 2x2 Haar wavelet kernel, which means that each block will get 3 response values.
From the results of the Haar wavelet transform, the Haar wavelet feature vectors within each patch are calculated. Haar wavelet feature vectors are typically composed of a plurality of response values of different directions and scales, such as a vector containing 9 elements, the first three of which represent response values in the horizontal direction, the next three representing response values in the vertical direction, and the last three representing response values in the diagonal direction.
And combining the Haar wavelet feature vectors in all the small blocks to obtain the Haar wavelet feature vector of the whole license plate image for the subsequent recognition process.
It should be noted that, computing Haar wavelet features requires performing wavelet transformation and vector computation multiple times, and thus, the computation amount is large, so that a computation acceleration technique, such as GPU acceleration, multi-threaded computation, etc., is generally required to improve the computation efficiency.
Example 8
On the basis of the above embodiment, the modified Hessian matrix is defined as:
wherein I is a license plate part, and x and y are coordinates in the license plate part; eigenvalue lambda of improved Hessian matrix 1 And lambda (lambda) 2 Used for judging whether the point is an extreme point or not.
Example 9
On the basis of the above embodiment, the step S4 of invoking a preset matching decision model, and performing similarity calculation on all the stored first recognition results in the memory based on the second recognition results, to obtain a similarity calculation result includes: based on the second recognition result, calling the corresponding additional information, and performing similarity calculation on all the stored first recognition results in a memory to obtain a similarity calculation result; firstly, based on a second recognition result, performing similarity calculation on all stored first recognition results in a storage mode to obtain a first similarity calculation result; then, invoking additional information in the outgoing image information corresponding to the second identification result to perform similarity calculation with additional information in the incoming image information corresponding to the first identification result stored in the memory, so as to obtain a second similarity calculation result; and carrying out weighted average operation on the first similarity calculation result and the second similarity calculation result by using a preset weighting coefficient to obtain a final similarity calculation result.
Specifically, in step S4, the similarity calculation is performed based on the second recognition result to find the first recognition result most similar to the second recognition result, thereby determining the entrance time stamp of the vehicle and calculating the parking fee. The specific similarity calculation method comprises the following steps:
firstly, based on the second recognition result, calling the corresponding additional information, and performing similarity calculation on all the stored first recognition results in a memory to obtain a first similarity calculation result. The similarity calculation generally adopts cosine similarity or Euclidean distance of feature vectors and other methods, and is used for measuring the similarity or distance between two feature vectors. Specifically, the Haar wavelet feature vector obtained in the license plate recognition model may be used as a feature vector, and the similarity or distance between the two feature vectors may be calculated.
And secondly, invoking additional information in the outgoing image information corresponding to the second identification result to perform similarity calculation with additional information in the incoming image information corresponding to the first identification result stored in the memory, so as to obtain a second similarity calculation result. The similarity calculation method may also adopt a cosine similarity or euclidean distance of the feature vector.
And finally, carrying out weighted average operation on the first similarity calculation result and the second similarity calculation result by using a preset weighting coefficient to obtain a final similarity calculation result. The weighting coefficient is usually set according to practical situations, for example, the weighting coefficient can be weighted according to indexes such as accuracy, stability, calculation time and the like of the two similarity calculation methods.
Example 10
On the basis of the above embodiment, the additional information includes at least: vehicle brand, vehicle color, and vehicle type.
Specifically, in step S1, the acquired entry image information and exit image information include at least a license plate portion of the vehicle. In addition to the license plate portion, other information of the vehicle, such as the vehicle brand, vehicle color, and vehicle type, etc., may be obtained, which may be stored as additional information in the memory for subsequent vehicle identification and billing operations.
In the vehicle identification process, if the second identification result is successfully matched with a certain first identification result in the memory, parking cost is directly calculated according to an entrance time stamp corresponding to the first identification result; otherwise, a first recognition result which is the most similar to the second recognition result is found through similarity calculation, and the entrance time stamp corresponding to the first recognition result is acquired to calculate the parking cost. In the similarity calculation, besides the calculation of the Haar wavelet feature vector obtained by using the license plate recognition model, the similarity of the additional information can be considered, so that the matching precision and the robustness can be improved. For example, similarity calculation may be performed using the vehicle brand, the vehicle color, the vehicle type, and the like in the additional information to further improve accuracy and reliability of the matching.
It should be noted that the apparatus (device) embodiments and the readable storage medium embodiments and the method embodiments described above belong to the same concept, and specific implementation processes thereof are detailed in the method embodiments. The technical features in the method embodiment are applicable to the device embodiment correspondingly, and are not described herein.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof.
In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
The preferred embodiments of the present disclosure have been described above with reference to the accompanying drawings, and are not thereby limiting the scope of the claims of the present disclosure. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and spirit of the present disclosure shall fall within the scope of the claims of the present disclosure.
Claims (10)
1. A method for managing vehicles in a parking lot based on an edge nano-tube technology, characterized in that the method performs the following steps:
step S1: collecting entrance image information of a vehicle entering a parking lot and exit image information of the vehicle leaving the parking lot; the entrance image information and the exit image information at least comprise license plate parts of vehicles;
step S2: uploading the acquired entrance image information to an edge nano tube cloud platform; the edge nano tube cloud platform uses a preset license plate recognition model to recognize the entering image information, a first recognition result is obtained, the first recognition result is stored in a memory of the edge nano tube cloud platform, and meanwhile, the vehicle entering time stamp is recorded;
step S3: uploading the acquired out-of-field image information to an edge nano tube cloud platform; the edge nano-tube cloud platform uses a preset license plate recognition model to recognize the outgoing image information, and a second recognition result is obtained; based on the second recognition results, carrying out matching query on all the stored first recognition results in the memory to obtain matching query results;
Step S4: if the matching inquiry result is true, calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the arrival time; if the matching inquiry result is false, calling a preset matching decision model, and carrying out similarity calculation on all the stored first recognition results in a memory based on the second recognition result to obtain a similarity calculation result; the similarity calculation result is a percentage result and is used for representing the similarity between the second identification result and all the first identification results stored in the memory; finding out a first identification result corresponding to the maximum similarity calculation result, obtaining a corresponding vehicle entrance time stamp, and calling a billing engine to calculate the parking cost of the vehicle according to the current departure time and the entrance time.
2. The method of claim 1, wherein the high definition recognition camera is used to collect the entry image information of the vehicle at the entrance to the parking lot and the exit image information at the exit from the parking lot in step S1.
3. The method according to claim 1, wherein the step S1 further includes the step of performing image processing on the entrance image information and the exit image information after collecting the entrance image information of the vehicle entering the parking lot and the exit image information leaving the parking lot, specifically including: image preprocessing is carried out on the acquired entrance image information and exit image information, and a preprocessing result is obtained; carrying out license plate detection on the preprocessing result to determine a license plate region, and carrying out license plate segmentation on the basis of the determined license plate region to obtain a license plate part of the entering image information or the exiting image information and other image parts except the license plate part; the license plate part is used as the entrance image information or the exit image information, and other image parts except the license plate part are added as the additional information.
4. A method according to claim 3, wherein the image preprocessing sequentially performs the steps of: image denoising, image enhancement and gray level conversion to obtain a preprocessing result.
5. The method of claim 4, wherein the step of detecting the license plate from the pre-processing result to determine the license plate region comprises: performing edge detection on the preprocessing result by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain all possible straight line candidate sets; randomly extracting two straight lines as edges of the license plate, and calculating intersection coordinates of the two straight lines as candidate points of the upper left corner of the license plate; in a surrounding area formed by 200x100 pixels of a candidate point of the upper left corner of the license plate, performing edge detection by using a Sobel operator, and performing straight line detection by using Hough transformation to obtain a possible straight line candidate set in the area; randomly extracting a group of straight lines from the straight line candidate set, and calculating the intersection point coordinates of the straight lines as candidate points of the lower right corner of the license plate; taking candidate points of the upper left corner and the lower right corner of the license plate as 4 vertexes to form a quadrangle of the license plate; calculating the score of the quadrangle, wherein the higher the score is, the more the quadrangle looks like a license plate, and the formula for calculating the score is as follows:
score＝w1*ratio+w2*compactness+w3*anglediff；
Wherein ratio represents the aspect ratio of the license plate, compatibility represents the compactness of the quadrangle, the area ratio of the interior area to the exterior area of the quadrangle is defined, angle represents the angle difference of the quadrangle boundary, and the maximum difference value between four angles is defined; score is a quadrilateral score; w1, w2 and w3 are weight coefficients and are set values; and repeatedly executing all the steps for N times, and finally selecting the quadrangle with the highest score as the determined license plate area.
6. The method of claim 5, wherein the method for performing license plate segmentation based on the determined license plate region to obtain a license plate portion of the in-field image information or the out-field image information and other image portions except the license plate portion comprises: converting the gray level image of the license plate region into a binary image; performing morphological processing on the binary image, removing unnecessary noise and details, and reserving a communication area on a license plate; edge detection is carried out on the connected region on the license plate, and the edge of the license plate region is obtained; extracting and describing the characteristics of the edge to obtain the coordinate information of the starting point and the end point of the edge and the direction information of the edge; constructing an adjacent matrix of the edge, and judging an adjacent relation according to the direction information of the edge; traversing the adjacent matrix to obtain all strong connected components. Screening and processing the strong connected components to obtain edge information of each character; according to the edge information of each character, the license plate is divided into a plurality of character areas, and the character areas are used as license plate parts of the entrance image information or the exit image information.
7. The method of claim 6, wherein the license plate recognition model recognizes the incoming image information or the license plate recognition model recognizes the outgoing image information comprises: smoothing the license plate part by using a Gaussian filter; detecting key points by using the improved Hessian matrix; computing Haar wavelet features in a region around each keypoint; generating descriptors using Haar wavelet features; and then based on the descriptors, the pre-established recognition templates are used for recognition.
8. The method of claim 7, wherein the modified Hessian matrix is defined as:
wherein I is a license plate part, and x and y are coordinates in the license plate part; eigenvalue lambda of improved Hessian matrix 1 And lambda (lambda) 2 Used for judging whether the point is an extreme point or not.
9. The method of claim 8, wherein the step S4 of calling a preset matching decision model, and performing similarity calculation on all the stored first recognition results in the memory based on the second recognition results, to obtain a similarity calculation result comprises: based on the second recognition result, calling the corresponding additional information, and performing similarity calculation on all the stored first recognition results in a memory to obtain a similarity calculation result; firstly, based on a second recognition result, performing similarity calculation on all stored first recognition results in a storage mode to obtain a first similarity calculation result; then, invoking additional information in the outgoing image information corresponding to the second identification result to perform similarity calculation with additional information in the incoming image information corresponding to the first identification result stored in the memory, so as to obtain a second similarity calculation result; and carrying out weighted average operation on the first similarity calculation result and the second similarity calculation result by using a preset weighting coefficient to obtain a final similarity calculation result.
10. The method of claim 9, wherein the additional information comprises at least: vehicle brand, vehicle color, and vehicle type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310508871.3A CN116665321A (en) | 2023-05-08 | 2023-05-08 | Parking lot vehicle management method based on edge nano-tube technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310508871.3A CN116665321A (en) | 2023-05-08 | 2023-05-08 | Parking lot vehicle management method based on edge nano-tube technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116665321A true CN116665321A (en) | 2023-08-29 |
Family
ID=87708865
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310508871.3A Pending CN116665321A (en) | 2023-05-08 | 2023-05-08 | Parking lot vehicle management method based on edge nano-tube technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116665321A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117496607A (en) * | 2023-11-07 | 2024-02-02 | 武汉无线飞翔科技有限公司 | ETC (electronic toll collection) -based intelligent parking lot management method and system |
-
2023
- 2023-05-08 CN CN202310508871.3A patent/CN116665321A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117496607A (en) * | 2023-11-07 | 2024-02-02 | 武汉无线飞翔科技有限公司 | ETC (electronic toll collection) -based intelligent parking lot management method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wei et al. | Multi-target defect identification for railway track line based on image processing and improved YOLOv3 model | |
| CN111178291B (en) | Parking payment system and parking payment method | |
| WO2020173022A1 (en) | Vehicle violation identifying method, server and storage medium | |
| Comelli et al. | Optical recognition of motor vehicle license plates | |
| US8509526B2 (en) | Detection of objects in digital images | |
| CN112686812B (en) | Bank card inclination correction detection method and device, readable storage medium and terminal | |
| CN103699905B (en) | Method and device for positioning license plate | |
| CN112270310A (en) | Cross-camera pedestrian multi-target tracking method and device based on deep learning | |
| Li et al. | An anti-fraud system for car insurance claim based on visual evidence | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| CN114549981A (en) | Intelligent inspection pointer type instrument recognition and reading method based on deep learning | |
| Azad et al. | New method for optimization of license plate recognition system with use of edge detection and connected component | |
| CN114387591A (en) | License plate recognition method, system, equipment and storage medium | |
| CN116665321A (en) | Parking lot vehicle management method based on edge nano-tube technology | |
| CN112836683A (en) | License plate recognition method, device, equipment and medium for portable camera equipment | |
| Hendryli et al. | Automatic license plate recognition for parking system using convolutional neural networks | |
| Ganapathy et al. | A Malaysian vehicle license plate localization and recognition system | |
| da Silva et al. | ALPRs-A new approach for license plate recognition using the SIFT algorithm | |
| CN113112479A (en) | Progressive target detection method and device based on key block extraction | |
| CN109325487B (en) | Full-category license plate recognition method based on target detection | |
| Sadique et al. | A comparative study of license plate detection and recognition techniques | |
| CN116152758A (en) | Intelligent real-time accident detection and vehicle tracking method | |
| Satish et al. | Edge assisted fast binarization scheme for improved vehicle license plate recognition | |
| CN112418262A (en) | Vehicle re-identification method, client and system | |
| CN111046876A (en) | License plate character rapid recognition method and system based on texture detection technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 518000, Building 502, Nanhai Yiku, Xinghua Road, Shekou, Shuiwan Community, Merchants Street, Nanshan District, Shenzhen, Guangdong Province Applicant after: Shenzhen easy Alliance Technology Co.,Ltd. Address before: 413, Building 2, Nanhai Yiku, Xinghua Road, Shekou, Shuiwan Community, Merchants Street, Nanshan District, Shenzhen City, Guangdong Province, 518000 Applicant before: Shenzhen easy Alliance Technology Co.,Ltd. |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |