CN111723800A - License plate calibration and identification method and system based on convolutional neural network and electronic equipment - Google Patents
License plate calibration and identification method and system based on convolutional neural network and electronic equipment Download PDFInfo
- Publication number
- CN111723800A CN111723800A CN202010572630.1A CN202010572630A CN111723800A CN 111723800 A CN111723800 A CN 111723800A CN 202010572630 A CN202010572630 A CN 202010572630A CN 111723800 A CN111723800 A CN 111723800A
- Authority
- CN
- China
- Prior art keywords
- license plate
- results
- vehicle
- neural network
- requirements
- 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.)
- Withdrawn
Links
Images
Classifications
-
- 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
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a license plate calibration identification method, a license plate calibration identification system and electronic equipment based on a convolutional neural network, wherein the license plate calibration identification method comprises the steps of obtaining vehicle contour information; acquiring a multi-frame image of a certain vehicle, and identifying the license plate number of the multi-frame image; comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results; and (4) checking a plurality of consistency results to determine the license plate information of a certain vehicle. When the vehicle is in a fast condition, the image can be selected clearly, so that the optimal image is obtained to identify the correct license plate number, and in addition, if different results are identified from the license plates in the continuous images, the convolutional neural network is combined to perform networking query on the outline information of the vehicle, so that the license plate identification error can be effectively avoided.
Description
Technical Field
The invention relates to the technical field of license plate recognition, in particular to a license plate calibration recognition method and system based on a convolutional neural network and an electronic device.
Background
The license plate recognition technology requires that the license plate of the moving automobile can be extracted and recognized from a complex background, and the information such as the license plate number and the color of the automobile can be recognized through the technologies such as license plate extraction, image preprocessing, feature extraction, license plate character recognition and the like.
In parking lot management, the license plate recognition technology is also a main means for recognizing the identity of a vehicle.
The license plate recognition technology is combined with an Electronic Toll Collection (ETC) system to recognize vehicles, and the vehicles can be automatically recognized and automatically charged without stopping when passing through a road junction. In the management of the parking lot, in order to improve the passing efficiency of vehicles at an entrance and an exit, the license plate recognition aims at the vehicles (such as a lunar truck and internal free passing vehicles) which do not need to collect parking fees, an unattended fast passage is built, the entrance and exit experience of card taking and non-stop is avoided, and the management mode of entering and exiting the parking lot is changed.
Due to the fact that the driving speeds of the automobiles are inconsistent, when the vehicles are subjected to license plate recognition of different vehicles, the images are often blurred when the vehicles are in a quick state, so that the correct license plate number cannot be recognized, and in addition, if the license plates in the continuous images are recognized to obtain different results, the existing technology cannot judge.
Disclosure of Invention
Aiming at the problems, the invention provides a license plate calibration and identification method and a license plate calibration and identification system based on a convolutional neural network, which can be used for carrying out clear selection on images when a vehicle is in a fast state, so that the optimal images can be obtained to identify the correct license plate number, and in addition, if different results are identified from the license plates in continuous images, the convolutional neural network is combined to carry out networking query on the contour information of the vehicle, so that the error of license plate identification can be effectively avoided, and the problems in the background technology can be effectively solved.
In order to achieve the purpose, the invention provides the following technical scheme: a license plate calibration and identification method based on a convolutional neural network is applied to electronic equipment and comprises the steps of
Acquiring vehicle contour information;
acquiring a multi-frame image of a certain vehicle, and identifying the license plate number of the multi-frame image;
comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results;
and (4) checking a plurality of consistency results to determine the license plate information of a certain vehicle.
As a preferable technical solution of the present invention, the obtaining of the vehicle contour information includes obtaining the contour information of the vehicle, where the contour information includes
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
As a preferred embodiment of the present invention, the method for acquiring a plurality of frames of images of a certain vehicle includes
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
As a preferred technical solution of the present invention, the process of identifying the license plate number of the plurality of frame images includes:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
checking whether the identification results of the license plates meeting the requirements are consistent or not, and if the identification results of the license plates do not meet the requirements, pairing any identification license plate with the maximum similarity in pairs to define adjacent frames;
and if the requirements are met, selecting two frames of images close in time as adjacent frames.
As a preferred technical solution of the present invention, the method for comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results comprises:
identifying license plate numbers of adjacent frame images respectively, identifying inconsistent positions if license plate number results of adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol, and inputting the special symbol into the networking information;
and inquiring other same symbols except the special symbols in the networking information, and determining final license plate number information by combining the contour information to obtain a license plate consistency result.
A license plate calibration and recognition system based on a convolutional neural network is applied to electronic equipment and comprises a vehicle positioning module, a vehicle identification module and a license plate recognition module, wherein the vehicle positioning module is used for positioning a vehicle needing license plate recognition and acquiring a clear outline;
the contour determining module is used for acquiring vehicle contour information;
the image acquisition module and the license plate recognition module are used for acquiring multi-frame images of a certain vehicle and recognizing the license plate number of the multi-frame images;
the result comparison module is used for comparing license plate number identification results of adjacent frames to obtain a plurality of license plate consistency results;
and the result determining module is used for checking a plurality of consistency results and determining the license plate information of a certain vehicle.
As a preferable technical solution of the present invention, the contour determining module acquires vehicle contour information, wherein the contour information includes
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
As a preferred technical solution of the present invention, the method for acquiring the multi-frame image of a certain vehicle by the image acquisition module includes
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
As a preferred technical solution of the present invention, the process of the license plate recognition module recognizing the license plate number of the multiple frames of images includes:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
the result comparison module is used for checking whether the license plate number identification results meeting the requirements have consistency, and if the license plate number identification results do not meet the requirements, two pairs of the identification license plate numbers with the maximum similarity are carried out to define as adjacent frames;
if the requirements are met, selecting two frames of images with close time as adjacent frames;
the method for comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results comprises the following steps:
identifying the license plate numbers of the adjacent frame images respectively, identifying inconsistent positions if the license plate number results of the adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol, and inputting the special symbol into the networking information;
and the result determining module inquires the other same symbols except the special symbols in the networking information, and determines the final license plate number information by combining the contour information to obtain a license plate consistency result.
An electronic device comprising a convolutional neural network-based license plate calibration recognition system as described in any one of the above.
Compared with the prior art, the invention has the beneficial effects that:
comparing license plate number identification results of adjacent frames, identifying the license plate numbers of the images of the adjacent frames respectively, identifying inconsistent positions if the license plate number results of the adjacent frames are inconsistent, and outputting respective results; defining the symbol at the inconsistent position as a special symbol, marking the special symbol and inputting the special symbol into the networking information; the same symbols except for the special symbols are queried in the networking information, for example, if there is an adjacent frame image recognition result of 12345678 and 12346578 respectively, then the front 1234 and the rear two 78 are consistent, then the group of data can be represented as 1234 × 78, and 1234 × 78 is input into the networking information for viewing, and the final license plate number information is determined by combining the contour information of the vehicle which is initially recognized, so as to obtain a license plate consistency result. When the vehicle is in a fast condition, the image can be selected clearly, so that the optimal image is obtained to identify the correct license plate number, and in addition, if different results are identified from the license plates in the continuous images, the network query is carried out on the contour information of the vehicle by combining the convolutional neural network, so that the error of license plate identification can be effectively avoided.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention;
FIG. 2 is a schematic flow chart illustrating the process of recognizing the license plate number of a plurality of frames of images according to the method of the present invention;
FIG. 3 is a schematic flow chart of a plurality of license plate consistency results obtained by the method of the present invention;
FIG. 4 is a schematic diagram of the general system of the present invention;
FIG. 5 is a schematic view of the wheel contour recognition of the present invention;
FIG. 6 is a schematic diagram of the convolution process of the present invention;
fig. 7 is a schematic diagram of CNN network identification numbers according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example (b):
referring to fig. 1 to 7, the present invention provides a technical solution: a license plate calibration and identification method based on a convolutional neural network is applied to electronic equipment and comprises the steps of
Acquiring vehicle contour information;
acquiring a multi-frame image of a certain vehicle, and identifying the license plate number of the multi-frame image;
comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results;
and (4) checking a plurality of consistency results to determine the license plate information of a certain vehicle.
Preferably, the vehicle contour information is acquired, wherein the contour information includes
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
Preferably, the method for acquiring the multiframe images of a certain vehicle comprises the following steps
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
Preferably, the process of identifying the license plate number of the plurality of frames of images includes:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
checking whether the identification results of the license plates meeting the requirements are consistent or not, and if the identification results of the license plates do not meet the requirements, pairing any identification license plate with the maximum similarity in pairs to define adjacent frames;
and if the requirements are met, selecting two frames of images close in time as adjacent frames.
Preferably, the method for comparing the license plate number recognition results of the adjacent frames to obtain a plurality of license plate consistency results comprises:
identifying license plate numbers of adjacent frame images respectively, identifying inconsistent positions if license plate number results of adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol, and inputting the special symbol into the networking information;
and inquiring other same symbols except the special symbols in the networking information, and determining final license plate number information by combining the contour information to obtain a license plate consistency result.
Wherein, the identification process aiming at the outline information and the license plate number information of the vehicle is obtained by training by adopting a convolution neural network,
the convolutional neural network is a CNN network, the comparison result of the processed image information and the trained image data is obtained, the outline of the wheel and the license plate of the wheel are identified, and the output result is similar to other neural network algorithms. Specifically, before inputting the learning data into the convolutional neural network, normalization is performed on the input data in a channel or time/frequency dimension, if the input data is a pixel, the original pixel values distributed in [0, 255] can also be normalized to the [0, 1] interval, after the obtained image is obtained, a multilayer network is adopted for supervised learning, the difference of the classification result between preset values is judged,
when the convolution kernel works, input characteristics are regularly swept, matrix element multiplication summation is carried out on the input characteristics, and deviation amount is superposed:
the summation part in the equation is equivalent to solving a cross-correlation (cross-correlation). b is the amount of deviation, ZIAnd ZI+1Convolutional inputs and outputs of layer L +1 of the table, also called feature map, LI+1Is ZI+1The feature pattern length and width are assumed to be the same. z (i, j) corresponds to the pixel of the feature map, K is the channel number of the feature map, f, s0And p is the convolutional layer parameter, the convolutional kernel size, convolutional step length (stride), and number of padding (padding) layers;
taking the contour as an example, the CNN algorithm shown in fig. 5 inputs an image of a vehicle into a network, acquires and identifies an edge image of the vehicle in a convolution manner, and determines contour information of the vehicle, where the contour information includes a color, a size, and a model corresponding to the vehicle, and for example, it can be identified that one vehicle is a car, the color is red, and the type of car; one vehicle is a big car, white, truck type; one vehicle is of the medium and small car, gray, minibus type.
The first convolution can extract low-level features.
The second convolution can extract mid-level features.
The third convolution can extract high-level features.
The features are extracted and compressed continuously, and higher-level features can be obtained finally, in short, the original features are concentrated step by step, and the finally obtained features are more reliable. Various tasks can be done using the last layer of features: such as classification, regression, etc.
For the identification of the car license plate number, as shown in fig. 7, similarly identifying a 3 number, for example, a 6x6 picture is convolved by a 3x3 filter (which can be regarded as a window), the 3x3 filter is convolved with the 3x3 matrix at the leftmost corner of the 6x6 picture to obtain the result, and then the convolution is continued by moving one step to the right (window sliding) until the whole picture is filtered, a 4x4 matrix is output, the depth of the output layer picture can be increased by increasing the number of the filters, and the number of the filters also determines the depth of the output layer picture (the two are equal)
The convolution process is shown in fig. 6, where three large matrices in the left area are the input of the original image, and three channels of RGB are represented by three matrices, with the size of 7 × 3.
Filter W0 represents 1 Filter assistant, size 3x3, depth 3 (three matrices); filter W1 also represents 1 Filter assistant. Because we used 2 filters in the convolution, the output depth of the convolutional layer result is 2 (there are 2 green matrices).
Bias b0 is the Bias term for Filter W0 and Bias b1 is the Bias term for Filter W1.
OutPut is the convolved OutPut, with a size of 3x3 and a depth of 2.
The input is fixed and the filter is specified, so the calculation is how to get the green matrix. Firstly, a sliding window with the same size as the filter is arranged on the input matrix, and then the part of the input matrix in the sliding window is multiplied by the corresponding position of the filter matrix: summing the results from the 3 matrices and adding the offset term, i.e., 0+2+0+1 to 3, thus yielding 3 in the upper left corner of the output matrix;
the sliding step length is denoted as S. The smaller S, the more features are extracted, but S generally does not take 1, mainly considering the problem of time efficiency. S cannot be too large, otherwise information on the image will be missed. In this application, the value of S is 2. Because the side length of the filter is greater than S, an intersection part exists after the sliding window is moved every time, the intersection part means that features are extracted for many times, and particularly, the extraction times of the middle area of the image are more, the extraction times of the edge part are less, and one circle of 0 is added on the periphery of the image.
In convolutional neural networks, there is a very important property: and sharing the weight value.
The weight sharing means that, for an input picture, a filter is used to scan the graph, the number in the filter is called the weight, each position in the graph is scanned by the same filter, and therefore the weight is the same, i.e. sharing. Pooling is the compression of features on the feature map, and pooling is also called down-sampling. By selecting max or mean of a region instead of that region, the whole is concentrated.
Thereby obtaining the license plate information of each different frame image.
A license plate calibration and recognition system based on a convolutional neural network is applied to electronic equipment and comprises a vehicle positioning module, a vehicle identification module and a license plate recognition module, wherein the vehicle positioning module is used for positioning a vehicle needing license plate recognition and acquiring a clear outline;
the contour determining module is used for acquiring vehicle contour information;
the image acquisition module and the license plate recognition module are used for acquiring multi-frame images of a certain vehicle and recognizing the license plate number of the multi-frame images;
the result comparison module is used for comparing license plate number identification results of adjacent frames to obtain a plurality of license plate consistency results;
and the result determining module is used for checking a plurality of consistency results and determining the license plate information of a certain vehicle.
Preferably, the contour determination module acquires vehicle contour information, wherein the contour information includes
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
Preferably, the method for acquiring the multi-frame image of the certain vehicle by the image acquisition module comprises the following steps
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
Preferably, the process of recognizing the license plate number of the multiple frames of images by the license plate recognition module includes:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
the result comparison module is used for checking whether the license plate number identification results meeting the requirements have consistency, and if the license plate number identification results do not meet the requirements, two pairs of the identification license plate numbers with the maximum similarity are carried out to define as adjacent frames;
if the requirements are met, selecting two frames of images with close time as adjacent frames;
the method for comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results comprises the following steps:
identifying the license plate numbers of the adjacent frame images respectively, identifying inconsistent positions if the license plate number results of the adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol, and inputting the special symbol into the networking information;
and the result determining module inquires the other same symbols except the special symbols in the networking information, and determines the final license plate number information by combining the contour information to obtain a license plate consistency result.
An electronic device comprising a convolutional neural network-based license plate calibration recognition system as described in any one of the above.
When the invention works, a clear threshold value is set based on time sequence, images which do not meet the requirements are removed, and license plates in the images which meet the requirements are identified; wherein the sharpness threshold is adaptively adjustable according to the human eye; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment; acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements; whether the identification results of the license plate numbers meeting the requirements are consistent or not is checked, if the identification results of the license plate numbers do not meet the requirements, the identification license plate numbers with the maximum similarity are paired pairwise to define adjacent frames; and if the requirements are met, selecting two frames of images close in time as adjacent frames. Comparing license plate number recognition results of adjacent frames, respectively recognizing license plate numbers of images of the adjacent frames, if license plate number results of the adjacent frames are inconsistent, recognizing inconsistent positions, and outputting respective results; defining the symbol at the inconsistent position as a special symbol, marking the special symbol and inputting the special symbol into the networking information; the method includes the steps that other same symbols except for special symbols are inquired in networking information, for example, the recognition results of adjacent frame images are 12345678 and 12346578 respectively, then the front 1234 and the rear two 78 are identical, the group of data can be represented as 1234 x 78, 1234 x 78 is input into the networking information to be checked, the final license plate number information is determined by combining the outline information of the vehicle which is initially recognized, the image can be clearly selected when the vehicle is fast, the optimal image is obtained to recognize the correct license plate number, in addition, if the license plates in continuous images are recognized to be different results, networking inquiry is conducted on the outline information of the vehicle by combining a convolutional neural network, license plate recognition errors can be effectively avoided, and license plate consistency results are obtained.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A license plate calibration and identification method based on a convolutional neural network is applied to electronic equipment, and is characterized in that: comprises that
Acquiring vehicle contour information;
acquiring a multi-frame image of a certain vehicle, and identifying the license plate number of the multi-frame image;
comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results;
and (4) checking a plurality of consistency results to determine the license plate information of a certain vehicle.
2. The convolutional neural network-based license plate calibration identification method of claim 1, wherein: the method comprises the steps of obtaining vehicle contour information, wherein the contour information comprises
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
3. The convolutional neural network-based license plate calibration identification method of claim 1, wherein: the method for acquiring the multiframe images of a certain vehicle comprises the following steps
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
4. The convolutional neural network-based license plate calibration identification method of claim 3, wherein: the process of identifying the license plate number of the multi-frame image comprises the following steps:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
whether the identification results of the license plates meeting the requirements are consistent or not is checked, if the identification results of the license plates do not meet the requirements, any identification license plate number with the maximum similarity is paired pairwise to define as an adjacent frame;
and if the requirements are met, selecting two frames of images close in time as adjacent frames.
5. The convolutional neural network-based license plate calibration identification method of claim 4, wherein: the method for comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results comprises the following steps:
identifying license plate numbers of adjacent frame images respectively, identifying inconsistent positions if license plate number results of adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol and inputting the special symbol into the networking information;
and inquiring other same symbols except the special symbols in the networking information, and determining final license plate number information by combining the contour information to obtain a license plate consistency result.
6. The utility model provides a license plate calibration identification system based on convolutional neural network, is applied to electronic equipment which characterized in that: comprises that
The vehicle positioning module is used for positioning the vehicle of which the license plate needs to be identified and acquiring a clear outline;
the contour determining module is used for acquiring vehicle contour information;
the image acquisition module and the license plate recognition module are used for acquiring multi-frame images of a certain vehicle and recognizing the license plate number of the multi-frame images;
the result comparison module is used for comparing license plate number identification results of adjacent frames to obtain a plurality of license plate consistency results;
and the result determining module is used for checking a plurality of consistency results and determining the license plate information of a certain vehicle.
7. The convolutional neural network-based license plate calibration recognition system of claim 6, wherein: the contour determination module obtains vehicle contour information, wherein the contour information comprises
And processing the vehicle image by adopting a CNN network to obtain the color, the size and the type information of different vehicles.
8. The convolutional neural network-based license plate calibration recognition system of claim 6, wherein:
the method for acquiring the multi-frame image of a certain vehicle by the image acquisition module comprises the following steps
The method comprises the steps of extracting a vehicle image needing to be observed, extracting continuous frames of the vehicle image, and sequencing the vehicle image in a multi-frame mode in a time sequence.
9. The convolutional neural network-based license plate calibration recognition system of claim 8, wherein:
the license plate recognition module carries out recognition on the license plate numbers of the multi-frame images, and the process comprises the following steps:
setting a definition threshold value based on time sequence, removing the images which do not meet the requirements, and identifying the license plate in the images which meet the requirements; wherein the definition threshold can be adaptively adjusted according to human eyes; when the clear threshold is selected, the clear threshold is automatically set for the first time, and the clear threshold can be changed according to manual adjustment;
acquiring license plates in the images meeting the requirements by using a convolutional neural network to obtain a plurality of license plate identification results meeting the requirements;
the result comparison module is used for checking whether the license plate number identification results meeting the requirements are consistent or not, if not, pairwise pairing is carried out on any identification license plate number with the maximum similarity, and the identification license plate number is defined as an adjacent frame;
if the requirements are met, selecting two frames of images with close time as adjacent frames;
the method for comparing license plate number recognition results of adjacent frames to obtain a plurality of license plate consistency results comprises the following steps:
identifying the license plate numbers of the adjacent frame images respectively, identifying inconsistent positions if the license plate number results of the adjacent frames are inconsistent, and outputting respective results;
defining the symbol at the inconsistent position as a special symbol, marking the special symbol and inputting the special symbol into the networking information;
and the result determining module inquires the other same symbols except the special symbols in the networking information, and determines the final license plate number information by combining the contour information to obtain a license plate consistency result.
10. An electronic device comprising a convolutional neural network based license plate calibration recognition system as claimed in any one of claims 6-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010572630.1A CN111723800A (en) | 2020-06-22 | 2020-06-22 | License plate calibration and identification method and system based on convolutional neural network and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010572630.1A CN111723800A (en) | 2020-06-22 | 2020-06-22 | License plate calibration and identification method and system based on convolutional neural network and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111723800A true CN111723800A (en) | 2020-09-29 |
Family
ID=72569879
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010572630.1A Withdrawn CN111723800A (en) | 2020-06-22 | 2020-06-22 | License plate calibration and identification method and system based on convolutional neural network and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111723800A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112115939A (en) * | 2020-08-26 | 2020-12-22 | 深圳市金溢科技股份有限公司 | Vehicle license plate recognition method and device |
| CN113610770A (en) * | 2021-07-15 | 2021-11-05 | 浙江大华技术股份有限公司 | License plate recognition method, device and equipment |
-
2020
- 2020-06-22 CN CN202010572630.1A patent/CN111723800A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112115939A (en) * | 2020-08-26 | 2020-12-22 | 深圳市金溢科技股份有限公司 | Vehicle license plate recognition method and device |
| CN112115939B (en) * | 2020-08-26 | 2024-06-04 | 深圳市金溢科技股份有限公司 | Vehicle license plate recognition method and device |
| CN113610770A (en) * | 2021-07-15 | 2021-11-05 | 浙江大华技术股份有限公司 | License plate recognition method, device and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107545239B (en) | Fake plate detection method based on license plate recognition and vehicle characteristic matching | |
| CN110414507B (en) | License plate recognition method and device, computer equipment and storage medium | |
| CN109145928B (en) | Method and device for identifying vehicle head orientation based on image | |
| Jolly et al. | Vehicle segmentation and classification using deformable templates | |
| CN109635656A (en) | Vehicle attribute recognition methods, device, equipment and medium neural network based | |
| CN107239778B (en) | Efficient and accurate license plate recognition method | |
| CN108205649B (en) | Method and device for recognizing state of driver for calling and answering | |
| Wang et al. | An effective method for plate number recognition | |
| CN111444917A (en) | License plate character recognition method and device, electronic equipment and storage medium | |
| CN107506765B (en) | License plate inclination correction method based on neural network | |
| CN106570439B (en) | Vehicle detection method and device | |
| CN112488046B (en) | Lane line extraction method based on high-resolution images of unmanned aerial vehicle | |
| CN111723800A (en) | License plate calibration and identification method and system based on convolutional neural network and electronic equipment | |
| CN111832461A (en) | Non-motor vehicle riding personnel helmet wearing detection method based on video stream | |
| CN111428735B (en) | Truck brand classification method based on migration learning deep network fusion model | |
| CN109784171A (en) | Car damage identification method for screening images, device, readable storage medium storing program for executing and server | |
| Yamashita et al. | Multiple skip connections of dilated convolution network for semantic segmentation | |
| CN109635637A (en) | A kind of licence plate recognition method, device and calculate equipment | |
| Cretu et al. | Biologically-inspired visual attention features for a vehicle classification task | |
| Musaddid et al. | Improvement of character segmentation for indonesian license plate recognition algorithm using CNN | |
| CN111160338B (en) | Fuzzy license plate recognition method | |
| Nguwi et al. | Number plate recognition in noisy image | |
| CN112733851A (en) | License plate recognition method for optimizing grain warehouse truck based on convolutional neural network | |
| Shima | Extraction of number plate images based on image category classification using deep learning | |
| CN112686247A (en) | Identification card number detection method and device, readable storage medium and terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200929 |
|
| WW01 | Invention patent application withdrawn after publication |