KR102008630B1 - Apparatus and method for increasing image recognition rate - Google Patents

Abstract

The present invention relates to an apparatus for increasing an image recognition rate, and to a method thereof. The apparatus for increasing a recognition rate of a vehicle license plate comprises: an image preprocessing unit performing image preprocessing to remove noise included in an image; a vehicle license plate detection unit detecting a vehicle license plate in the preprocessed image; a vehicle license plate image extraction unit extracting only an image of the detected vehicle license plate; an image restoring unit restoring the extracted image of the license plate to an image having a predetermined resolution or more using an SRGAN algorithm; a character area detection unit detecting at least one area of a number area, a symbol area, and a character area in the image of the vehicle license plate restored to a predetermined resolution or more; a character reading unit reading by using an optical character reading device detecting at least one of the detected number area, the detected symbol area, and the detected character area; and a control unit controlling the overall operation.

Description

Apparatus and method for increasing image recognition rate

The present invention relates to an apparatus and method for increasing image recognition rate, and more particularly, to an apparatus and method for increasing recognition rate of a license plate for increasing a resolution of a low-resolution vehicle license plate into a high-resolution vehicle license plate through SRGAN learning.

Various vehicles and motorcycles used as a means of transportation for modern people are required to operate by attaching license plates in accordance with related laws. The license plate of the vehicle is not only used to identify the vehicle but also variously used to identify the license plate of a vehicle captured by a black box or CCTV and refer to the investigation data or recognize the license plate of a vehicle such as an automatic parking lot.

However, in order to recognize a conventional license plate, a professional recognition device or the like must be provided separately, and there is an economic problem because it is not easily compatible with existing equipment.

Accordingly, it is urgent to develop a vehicle recognition rate increasing device and method that can recognize a license plate of a vehicle at a sharper resolution than a professional recognition device and can be conveniently compatible without upgrading existing equipment.

[Related Technical Documents]

1. Vehicle number recognition system and method for improving the sharpness of vehicle license plate image using the system (Patent Application No. 10-2005-0030669)

The problem to be solved by the present invention is to use the algorithm that performed at least one of the training of SRGAN and SRCNN based on a database in which a reduced image matching the original image is stored together to obtain a higher resolution clear image. An apparatus and method for increasing image recognition rate may be provided.

Another problem to be solved by the present invention is to provide an apparatus and method for increasing image recognition rate, which can be used to restore a license plate of a contaminated vehicle according to a set algorithm, which can be usefully used for a business or investigative agency that recognizes a license plate. .

Another problem to be solved by the present invention can be configured in the form of a removable disk or provided in the form of a cloud-based application, can be used without upgrading the existing equipment, which can improve the economics or convenience of the user The present invention provides an apparatus and a method for increasing the recognition rate of a vehicle license plate.

The objects of the present invention are not limited to the above-mentioned objects, and other objects that are not mentioned will be clearly understood by those skilled in the art from the following description.

In order to solve the above problems, the vehicle recognition rate increasing apparatus according to an embodiment of the present invention, the image pre-processing unit to remove the noise included in the image by performing image pre-processing; A vehicle license plate detection unit for detecting a vehicle license plate in the preprocessed image; A vehicle license plate image extraction unit for extracting only an image of the detected vehicle license plate; An image restoring unit for restoring the extracted image of the license plate to an image having a predetermined resolution or more using an SRGAN algorithm; A character area detector for detecting at least one of a number, a symbol, and a text area in an image of the vehicle license plate restored to a set resolution or more; A character reading unit that reads at least one of the detected numbers, symbols, and character areas using an optical character reading device; And a control unit for controlling the overall operation.

Another problem to be solved by the present invention is the SRGAN algorithm, the SRGAN learning is completed based on the stored database.

Another problem to be solved by the present invention further comprises a database for comparing and storing the original image consisting of at least one of the numbers, symbols, and letters constituting the license plate and a reduced image having a resolution equal to or smaller than the set resolution for the same image. Include.

Another problem to be solved by the present invention is a text extraction unit for extracting at least one of the read numbers, symbols and characters as text; And a storage unit which stores the text extracted by the text extraction unit.

In order to solve the above problems, the vehicle recognition rate increase method according to an embodiment of the present invention, the image pre-processing unit, performing the image pre-processing, to remove the noise contained in the image; Detecting, by the vehicle license plate detection unit, the vehicle license plate in the preprocessed image; Extracting only the image of the detected vehicle license plate by the vehicle license plate image extractor; Restoring, by the image restoration unit, the extracted image of the license plate to an image having a predetermined resolution or more by using the SRGAN algorithm; Detecting, by the character area detection unit, at least one of a number, a symbol, and a character area in an image of the vehicle license plate restored to a set resolution or more; In the character reading unit, reading at least one of the detected numbers, symbols, and character areas by using the optical character reading device; And controlling, by the controller, the overall operation.

The present invention uses an algorithm that performs at least one of SRGAN and SRCNN learning based on a database in which a reduced image matching the original image is stored together. An apparatus and method for increasing recognition rate may be provided.

The present invention can restore the license plate of a contaminated vehicle according to a set algorithm, it is possible to provide an apparatus and method for increasing the recognition rate of a vehicle license plate that can be usefully used in businesses or investigation agencies that recognize the license plate.

Another object of the present invention is to be configured in the form of a removable disk or provided in the form of a cloud-based application, can be used without upgrading the existing equipment, it can improve the economics or convenience of the user .

The effects according to the invention are not limited by the contents exemplified above, and more various effects are included in the present specification.

1A to 1E are block diagrams schematically illustrating a configuration of an apparatus for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention, and a diagram for describing contents of recognizing a contaminated license plate.
2 is a diagram for describing contents of generating a learning database according to an exemplary embodiment.
3 is a view for explaining the content of the SRGAN learning in the apparatus for increasing the recognition rate of the vehicle license plate according to an embodiment of the present invention.
4 is a flowchart illustrating an operation sequence of an apparatus for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention.

Advantages and features of the present invention, and methods for achieving them will be apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only the embodiments are to make the disclosure of the present invention complete, and those skilled in the art to which the present invention pertains. It is provided to fully inform the scope of the invention, and the invention is defined only by the scope of the claims.

Combinations of each block of the accompanying block diagrams and respective steps of the flowcharts may be performed by algorithms or computer program instructions, consisting of firmware, software, or hardware. These algorithms or computer program instructions may be embedded in a processor of a general purpose computer, special purpose computer, or other programmable digital signal processing device, so that instructions executed through a processor of a computer or other programmable data processing equipment. These will create means for performing the functions described in each block of the block diagram or in each step of the flowchart. These algorithms or computer program instructions may be stored in a computer usable or computer readable memory that can be directed to a computer or other programmable data processing equipment to implement functionality in a particular manner, and thus the computer available or computer readable. It is also possible for the instructions stored in the capable memory to produce an article of manufacture containing instructions means for performing the functions described in each block or flow chart step of the block diagram. Computer program instructions may also be mounted on a computer or other programmable data processing equipment, such that a series of operating steps may be performed on the computer or other programmable data processing equipment to create a computer-implemented process to create a computer or other programmable data. Instructions that perform processing equipment may also provide steps for performing the functions described in each block of the block diagram and in each step of the flowchart.

In addition, each block or step may represent a portion of a module, segment or code that includes one or more executable instructions for executing a specified logical function (s). It should also be noted that in some alternative embodiments, the functions noted in the blocks or steps may occur out of order. For example, the two blocks or steps shown in succession may in fact be executed substantially concurrently or the blocks or steps may sometimes be performed in the reverse order, depending on the functionality involved.

Like reference numerals refer to like elements throughout.

Each of the features of the various embodiments of the present invention may be combined or combined with each other, partly or wholly, various technically interlocking and driving as can be understood by those skilled in the art, each of the embodiments may be independently implemented with respect to each other It may be possible to carry out together in an association.

Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

1A to 1E are block diagrams schematically illustrating a configuration of an apparatus for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention, and a diagram for describing contents of recognizing a contaminated license plate.

1A is a block diagram schematically illustrating a configuration of an apparatus for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention. First, as shown in FIG. 1A, the apparatus 100 for increasing the recognition rate of a vehicle license plate according to an embodiment of the present invention may include an image preprocessor 110, a vehicle license plate detector 111, and a vehicle license plate image extractor 112. The image reconstructor 113, the character region detector 114, the character reader 115, the text extractor 116, the storage 117, the database 118, and the controller 120 are included.

The apparatus 100 for increasing the recognition rate of a license plate may be various electronic devices such as a smartphone, a tablet PC, a notebook, a desktop, and the like, and may be configured in the form of a mobile storage disk such as a USB, and the like. The deep learning machine may be connected to various electronic devices such as the electronic device and may be immediately used without upgrading the electronic device.

In addition, the apparatus 100 for increasing the recognition rate of a vehicle license plate according to an embodiment of the present disclosure has been described as an example in the form of a device, but may provide a cloud-based application regardless of the type of device.

The image preprocessing unit 110 performs image preprocessing to remove noise included in the image.

The vehicle license plate detector 111 detects the vehicle license plate from the preprocessed image. Specifically, the vehicle license plate detection unit 111 detects only the vehicle license plate from the entire preprocessed image.

The vehicle license plate image extraction unit 112 extracts only the image of the detected vehicle license plate. Specifically, the vehicle license plate image extraction unit 112 extracts an image of only the vehicle license plate from the vehicle license plate detected by the vehicle license plate detection unit 111.

The image restoration unit 113 restores the extracted image of the license plate to an image having a predetermined resolution or more using a Super Resolution Generation Adversarial Network (SRGAN) algorithm.

More specifically, the image restoration unit 113 restores the image of the extracted vehicle license plate to the image of the high-resolution vehicle license plate using the SRGAN algorithm.

Here, the SRGAN algorithm is an algorithm based on GAN (Generative Adversarial Networks), and is an algorithm that applies a loss function having superior characteristics to MES (Manufacturing Execution System).

The text area detector 114 detects at least one of numbers, symbols, and text areas in the image of the vehicle license plate restored to a set resolution or more. In detail, the character area detector 114 detects at least one of the numbers, symbols, and character areas included in the image of the vehicle license plate restored in high resolution.

For example, when the image of the high-resolution restored license plate is "00 is 1234", the character area detection unit 114 is "0", "0", "ga", "1", "2", "3" And numerals " 4 ", respectively.

The character reading unit 115 reads the detected character area by using an optical character reader. For example, in the above-described example, the character reading unit 115 characterizes the numbers and characters “0”, “0”, “ga”, “1”, “2”, “3”, and “4”. Using the reading apparatus, the license plate of the vehicle can read "00ga1234".

The text extraction unit 116 extracts at least one of the read numbers, symbols, and characters as text.

The storage unit 117 stores the text extracted by the text extraction unit 116.

The database 118 compares and stores an original image composed of at least one of numbers, symbols, and letters constituting the vehicle license plate and a reduced image having a resolution equal to or smaller than the resolution set for the same image as the original image.

Here, the reduced image is the same image as the original image, and only the resolution is an image falling by the set resolution compared to the original image.

For example, when the license plate displayed on the original image having a resolution higher than the set resolution is “00 is 1234”, the database 118 reduces the number and the license plate of “00 is 1234” displayed on the original image under the control of the controller 120. Each car number plate “00 is 1234” displayed on the image can be compared and stored.

That is, the database 118 stores an original image having a high resolution and a reduced image having a low resolution.

The controller 120 controls the apparatus 100 for increasing the recognition rate of the vehicle license plate as a whole. In an embodiment of the present invention, the components 110 to 118 and the controller 120 are distinguished from each other and displayed. However, all the operations may be performed by the controller 120.

FIG. 1B is a block diagram schematically showing a detailed configuration of the controller 120 included in the apparatus 100 for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention. FIG. 1C is a block diagram illustrating an embodiment of the present invention. It is a situation diagram for explaining the contaminated vehicle license plate. Hereinafter, for convenience of description, FIGS. 1B and 1C will be described together.

First, as shown in Figure 1b, the control unit 120 of the recognition rate increasing apparatus 100 of the vehicle license plate according to an embodiment of the present invention is a pollution license plate recognition unit 121, pollution license plate estimation unit 122, the vehicle License plate inquiry unit 123 and result calculation unit 124 is included.

The pollution license plate recognition unit 121 determines whether the vehicle license plate is a dirty license plate. Here, the contaminated license plate is a license plate of a vehicle in which the license plate of the vehicle cannot be recognized by various substances or external shocks.

That is, when any one of the numbers, symbols, and letters included in the license plate is not completely recognized, the pollution license plate recognition unit 121 determines that the license plate is dirty.

For example, as illustrated in FIG. 1C, when the license plate recognition unit 121 does not recognize the third number 131 located after the letter “a” included in the license plate 130 of the vehicle to be recognized, It is determined that the license plate 130 of the vehicle is contaminated.

If the license plate recognition unit 121 determines that the vehicle license plate is contaminated, the pollution license plate recognition unit 121 recognizes that the license plate 130 of the vehicle to be recognized is a dirty license plate.

Thereafter, the contamination license plate estimator 122 estimates one or more numbers, symbols, and characters that are not recognized among the numbers, symbols, and characters included in the contaminated license plate. For example, the third number 131 located after the letter “a” included in the license plate 130 of the vehicle to be recognized shown in FIG.

At this time, the pollution license plate estimator 122 compares the original image and the reduced image stored in the database 118 described above, and estimates the number before the contamination number is contaminated.

For example, the contamination plate estimator 122 may determine whether the polluted number is “0”, “6” or “8” by looking at the shape displayed on the uncontaminated area among the areas of the contaminated number 131. Can be estimated.

Here, the pollution license plate estimator 122 may estimate each estimated number, symbol, and character in a percentage format. In the above example, the contamination plate estimator 122 may estimate 100 percent that the contaminated number 131 is any one of “0”, “6” and “8”, respectively, 30 percent, 50 percent and 20 percent. It can also be estimated as a percentage.

The vehicle license plate inquiry unit 123 is configured based on the vehicle model and the finally recognized license plate when any one of the numbers, symbols, and letters included in the license plate of the vehicle to be recognized is not completely recognized due to contamination. Inquiries with registered vehicle license plates stored in the database.

For example, when the pollution license plate estimator 122 identifies the pollution license plate, it is estimated that "00 is 12X4", the number indicated by "X" is not recognized, and the model of the license plate to be recognized is "A". Let me explain about the case.

In the above-described example, the vehicle license plate inquiry unit 123 may acquire information that the vehicle license plate issued to the vehicle model "A" is "00ga1234" based on a database stored in the server of the set engine.

The result calculator 124 receives data from the pollution license plate estimator 122 or the license plate inquiry unit 123 to calculate a recognition result of the license plate. For example, the result calculator 124 may display the vehicle license plate estimated by the pollution license plate estimator 122 and the percentage of the estimated vehicle license plate together, and display the vehicle license plate inquiry unit 123. Numbers, symbols, and letters included in the license plate may also be displayed.

Although not described in an embodiment of the present disclosure, the apparatus for increasing the recognition rate of the vehicle license plate may generate a tool for learning unstructured data based on deep learning. For example, the license plate recognition device of the license plate generates tools for learning unstructured data based on deep learning, and displays various distortion images of the same image through semi-transparency, contrast, image filter, perspective expression, and vertical rotation. You can also include it in deep learning.

Although not described in an embodiment of the present invention, the apparatus for increasing the recognition rate of a vehicle license plate recognizes an irregular image reconstructed through a dirty license plate image filter on a collected license plate image with an atypical license plate image learner to flexibly recognize image distortion. You can also

FIG. 1D is a block diagram schematically showing a detailed configuration of the controller 120 included in the apparatus 100 for increasing a recognition rate of a vehicle license plate according to another embodiment of the present invention, and FIG. 1E is another embodiment of the present invention. FIG. 9 is a diagram illustrating the contents of determining region similarity by the region similarity determining unit according to an example. Hereinafter, for convenience of description, FIGS. 1D and 1E will be described together.

First, as shown in FIG. 1D, the controller 120 of the apparatus 100 for increasing a recognition rate of a vehicle license plate according to another exemplary embodiment of the present invention may include a vehicle model checking unit 125, a recognition position checking unit 126, and a recognition unit. And a time zone checker 127, an area similarity determiner 128, a weight determiner 129, and a vehicle license plate calculator 129-1.

If the vehicle model identification unit 125 is a contaminated license plate in which numbers, letters, and symbols included in the license plate of the vehicle to be recognized are not recognized, the vehicle model confirmation unit 125 checks the vehicle model of the contaminated license plate vehicle.

The recognition location checking unit 126 checks the detailed location of the area where the vehicle of the currently polluted license plate is located. For example, the recognition location checking unit 126 is a GPS module configured in the recognition rate increasing device 100 of the license plate, and the detailed location of the current vehicle location such as a specific restaurant, sauna, apartment, public institution, and building. You can check it.

The recognition time zone checker 127 checks a specific time when the vehicle of the currently polluted license plate is recognized. For example, the recognition time zone checker 127 checks whether the time zones for recognizing the vehicle of the currently polluted license plate are AM and PM, together with a specific time.

If the license plate of the vehicle to be recognized is a contaminated license plate, the area similarity determination unit 128 determines the similarity of the area of the license plate in comparison with the license plate of the vehicle which is completely recognized.

Specifically, in the case of a contaminated license plate in which some of numbers, letters, and symbols included in the license plate of the vehicle to be recognized are not recognized, the area similarity determination unit 128 may store the storage unit 117 or the database 118. Compared with the stored license plate of the vehicle which is completely recognized, it is determined whether the similarity is recognized among the set areas.

For example, as shown in FIG. 1E, the currently contaminated region among the regions of the license plate 130 of the contaminated vehicle divided into the first to third regions 132 to 134 is the first region 132. We will now describe the case where data for a fully recognized license plate is stored.

In the above-described example, when the second and third regions 133 and 134 except for the contaminated region, the first region 132 are the same as the previously recognized license plate, the region similarity determination unit 128 Can be found to be 66.6 percent similar to the previously recognized license plate.

In the present exemplary embodiment, an example of dividing the license plate of the vehicle into the first to third regions 132 to 134 has been described. .

In addition, the similarity determination unit 128 may further determine whether or not the unique region of the contaminated license plate and the previously recognized license plate match, thereby increasing the probability of determining the similarity.

Here, the unique area is an area which can be distinguished from other license plates of the vehicle, and is a unique area only for the license plate of the vehicle such as, for example, scratches, wrinkles, discoloration, and unique marks.

Also, the similarity determination unit 128 may be driven by dualizing the full range matching for matching the entire area of the stored license plate with the dirty license plate and the partial area matching for matching the partial area. In detail, the similarity determination unit 128 includes a full range matching for matching the entire area of the stored license plate and the dirty license plate, and a partial area for matching some areas of the stored license plate and some areas that are not recognized among the dirty license plates. It can be driven by matching to binary.

Accordingly, according to the similarity determination unit of the present invention, the recognition rate of the vehicle license plate may not only match the stored license plate with the entire area of the contaminated license plate but also may match only a specific area according to the user's setting or the position of the contaminated license plate. There is an advantage to reduce the overall load of the augmentation device.

In addition, according to the similarity determination unit of the present invention, since the portion of the license plate that is not recognized due to contamination of the vehicle plate may be mainly recognized, there is an advantage of more precisely recognizing only the area of the contamination plate.

The weight determining unit 129 weights specific data based on the results output from the vehicle model checking unit 125, the recognition position checking unit 126, the recognition time zone checking unit 127, and the area similarity determining unit 128. Give it differentiation.

For example, the number "X" in the license plate of the vehicle to be recognized is contaminated, so that the license plate of the vehicle is incompletely recognized as "00 is 123X", and the vehicle model confirmed by the vehicle model identification unit 125 is identified as the A model. If the license plate of the previously recognized vehicle is "00 is 1234" and the vehicle model is A, it will be described.

In the above-described example, the weight determining unit 129 not only matches all numbers except the “X” number among the numbers and letters of the previously recognized vehicle license plate with respect to the license plate of the incompletely recognized vehicle, but also the vehicle type Since it is matched with, it is possible to give a weight as much as the percentage set in the data for the vehicle model.

As another example, the number "X" in the license plate of the vehicle to be recognized is contaminated, so that the license plate of the vehicle is incompletely recognized as "00 is 123X", and the position confirmed by the recognition position checking unit 126 is A I will explain the case where the license plate of the vehicle that has been identified as an apartment and was previously recognized near the apartment A is "00 is 1234".

In the above-described example, the weight determining unit 129 not only matches the numbers of the previously recognized vehicle license plates and letters except for the “X” number with respect to the license plate of the incompletely recognized vehicle, but also the position of the vehicle. Since it is similar to the A apartment, it is possible to give a weight as much as a percentage set to the data for the recognition position.

As another example, the number "X" in the license plate of the vehicle to be recognized is contaminated, and the license plate of the vehicle is incompletely recognized as "00 is 123X", and the time zone checked by the recognition time zone checker 127 is AM. It will be described at 10 o'clock, and the license plate of the vehicle which was recognized around 10 am is “00 is 1234”.

In the above-described example, the weight determining unit 129 not only matches the numbers of the previously recognized number plates and letters of the vehicle number plate except for the “X” number, but also the time zone of the vehicle. Since it is similar to 10:00 am, it is possible to give more weight by the percentage set to the data for the recognition time zone.

As another example, the number "X" in the license plate of the vehicle to be recognized is contaminated, the license plate of the vehicle is incompletely recognized as "00 is 123X", and the unique area is detected by the area similarity determination unit 127. The case where the license plate of the vehicle having the same unique area is "00 is 1234" will be described.

In the above-described example, the weight determining unit 129 not only matches the numbers on the license plate of the incompletely recognized vehicle license plate but all numbers except the “X” number of letters of the previously recognized vehicle license plate. Since the included inherent regions also match, weights may be further given by the percentages set in the data for region similarity.

The vehicle license plate calculator 129-1 finally calculates a contaminated vehicle license plate based on the weight determined by the weight determiner 129. Specifically, the vehicle license plate calculator 129-1 finally calculates the contaminated vehicle license plate based on a weight of the vehicle model, the recognition location, the recognition time zone, and the license plate area similarity of the vehicle possessing the contaminated vehicle license plate. .

2 is a diagram for describing contents of generating a learning database according to an exemplary embodiment. First, as shown in FIG. 2, the license plate database according to an embodiment of the present invention stores an original license plate image and a license plate reduced image, respectively.

Here, the original image is an image having a set resolution or more that can identify the license plate of the vehicle, and the reduced image is the same image as the original image, and the resolution is an image that is lowered by the set resolution compared to the original image.

In detail, the license plate database stores and compares the original image and the reduced image matching the original image. That is, in the license plate database, the original image and the reduced image corresponding thereto are stored, respectively, to generate a learning database.

In the vehicle license plate data collection method according to an embodiment of the present invention, license plates of each country may be stored. US and Chinese license plates which are difficult to collect in Korea may be collected primarily as a black box image. At this time, there is a constraint of collecting data of the license plate on the rear of the vehicle and a constraint of low resolution, but it may be helpful for learning to be described later.

In addition, the license plate database may automatically collect images based on numbers, alphabets, Korean, and Chinese characters.

3 is a view for explaining the content of the SRGAN learning in the apparatus for increasing the recognition rate of the vehicle license plate according to an embodiment of the present invention. First, as shown in FIG. 3, when performing SRGAN learning in the apparatus for increasing the recognition rate of a vehicle license plate according to an embodiment of the present disclosure, the apparatus for increasing the recognition rate of a vehicle license plate may perform SRGAN learning based on the generated database. To perform.

Here, the apparatus for increasing the recognition rate of the license plate performs SRGAN learning based on a database generated based on a car database, a road database, and an object database on a road, which are photographed and stored in a black box of a vehicle.

That is, the apparatus for increasing a license plate recognition rate has an advantage of obtaining a clearer image having a higher resolution because SRGAN learning is performed based on a database in which an original image and a reduced image matching the original image are stored together.

In an embodiment of the present invention, the SRGAN learning is described based on the database generated by the apparatus for increasing the recognition rate of the license plate, but the present invention is not limited to the SRGAN learning.

Specifically, the type of deep learning learning-based super-resolution (SR) is not limited to SRGAN, but may be a super-resolution convolutional neural network (SRCNN). That is, in the present invention, it is also possible to perform learning with an SR learning algorithm based on CNN (Convolutional Neural Network).

Therefore, since the SRGAN learning or the SRCNN learning is performed based on the database generated by the apparatus for increasing the recognition rate of the vehicle license plate, it is possible to obtain a clearer image having a higher resolution.

4 is a flowchart illustrating an operation sequence of an apparatus for increasing a recognition rate of a vehicle license plate according to an embodiment of the present invention. First, as shown in FIG. 4, the image preprocessing unit performs image preprocessing to remove noise included in an image (S410).

Thereafter, the vehicle license plate detection unit detects the vehicle license plate from the preprocessed image (S420).

Thereafter, the vehicle license plate image extractor extracts only an image of the detected vehicle license plate (S430).

Thereafter, the image restoring unit restores an image having a predetermined resolution or more by using the SRGAN algorithm from the extracted image of the license plate (S440).

Thereafter, the character area detection unit detects the character area in the image of the vehicle license plate restored to the set resolution or more (S450).

Thereafter, the character reading unit reads the detected character area using the optical character reading apparatus (S460).

Thereafter, the text extracting unit extracts the read text as text (S470).

Thereafter, the storage unit stores the text extracted by the text extraction unit (S480).

In this specification, each block or each step may represent a portion of a module, segment or code containing one or more executable instructions for executing a specified logical function (s). It should also be noted that in some alternative embodiments, the functions noted in the blocks or steps may occur out of order. For example, the two blocks or steps shown in succession may in fact be executed substantially concurrently or the blocks or steps may sometimes be performed in the reverse order, depending on the functionality involved.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be implemented directly in hardware, a software module or a combination of the two executed by a processor. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, which can read information from and write information to the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside within an application specific integrated circuit (ASIC). The ASIC may reside in a user terminal. In the alternative, the processor and the storage medium may reside as discrete components in a user terminal.

Although the embodiments of the present invention have been described in more detail with reference to the accompanying drawings, the present invention is not necessarily limited to these embodiments, and various modifications can be made without departing from the spirit of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention but to describe the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

Device for increasing recognition rate of 100 license plate
110 Image Preprocessor
112 license plate image extraction unit
113 Image Restoration Unit
114 character area detection unit
115 character reader
116 Text Extraction Unit
117 storage
118 databases
120 control unit
121 Pollution Plate Recognition Unit
122 Pollution Plate Estimator
123 license plate tracking
124 result calculator
125 model confirmation
126 Recognition Position Confirmation Unit
127 Time Zone Confirmation Section
128 region similarity determining unit
129 weight determination unit
129-1 license plate calculator
130 Dirty Plates
131 Contaminated Number
132 First Zone
133 Second Area
134 Third Area

Claims (1)

An image preprocessing unit performing image preprocessing to remove noise included in an image;
A vehicle license plate detector detecting a vehicle license plate in the preprocessed image;
A vehicle license plate image extraction unit for extracting only the image of the detected vehicle license plate;
An image restoring unit for restoring the extracted image of the license plate to an image having a predetermined resolution or more using a Super Resolution Generation Adversarial Network (SRGAN) algorithm;
A text area detector for detecting at least one of a number, a sign, and a text area in an image of the license plate restored to at least the set resolution;
A character reader configured to read at least one of the detected numbers, symbols, and character areas using an optical character reader;
A controller for controlling the overall operation;
A database for comparing and storing an original image consisting of at least one of numbers, symbols, and letters constituting a vehicle license plate, and a reduced image having a resolution lower than or equal to a set resolution with respect to the same image as the original image;
A text extraction unit for extracting at least one of the read numbers, symbols, and characters as text; And
Including a storage unit for storing the text extracted by the text extraction unit,
The control unit
A pollution license plate recognition unit determining whether the license plate is dirty;
A pollution license plate estimator for estimating the contents before pollution in a percentage form based on the type displayed in the non-contaminated area among the contaminated numbers, symbols, or letters;
If any of the numbers, symbols, and letters included in the license plate of the vehicle to be recognized are not fully recognized due to contamination, compared with the registered vehicle license plate stored in the database of the established institution, based on the vehicle model and the finally recognized license plate Vehicle license plate inquiry unit to query by;
A result calculator for receiving data from at least one of the pollution license plate estimator and the license plate inquiry unit, and calculating a recognition result of the license plate;
A vehicle model identification unit for identifying a vehicle model for a contaminated license plate in which at least one of numbers, letters and symbols included in the license plate of the vehicle to be recognized is not recognized;
Recognition location check unit for confirming the detailed location of the area where the vehicle of the dirty plate is located;
A recognition time zone checking unit for identifying a time zone in which the vehicle of the contaminated license plate is recognized;
If the license plate of the vehicle to be recognized is a contaminated license plate, the license plate of the vehicle is compared with the previously recognized license plate of the vehicle, and based on judging whether or not the unique area corresponding to the scratches, wrinkles, discolorations, and marks is identical. A similarity determination unit that determines the similarity of the license plate area;
A weight determination unit for assigning a weighted percentage to the matched data among the result data output from the vehicle type identification unit, the recognition position identification unit, the recognition time zone identification unit, and the similarity determination unit;
And a vehicle license plate calculator configured to calculate a final license plate of the contaminated vehicle based on the weight determined by the weight determiner.
The SRGAN algorithm is
An apparatus for increasing image recognition rate, which is an algorithm for which SRGAN learning is completed based on a stored database.

Images