CN111860496A - License plate recognition method, device, equipment and computer readable storage medium - Google Patents
License plate recognition method, device, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN111860496A CN111860496A CN202010575439.2A CN202010575439A CN111860496A CN 111860496 A CN111860496 A CN 111860496A CN 202010575439 A CN202010575439 A CN 202010575439A CN 111860496 A CN111860496 A CN 111860496A
- Authority
- CN
- China
- Prior art keywords
- license plate
- image
- target
- target image
- image area
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012545 processing Methods 0.000 claims abstract description 85
- 230000007246 mechanism Effects 0.000 claims abstract description 39
- 238000001514 detection method Methods 0.000 claims abstract description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 34
- 230000009191 jumping Effects 0.000 claims description 22
- 238000013527 convolutional neural network Methods 0.000 claims description 21
- 238000012795 verification Methods 0.000 claims description 21
- 239000011159 matrix material Substances 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 16
- 238000010586 diagram Methods 0.000 claims description 14
- 238000007781 pre-processing Methods 0.000 claims description 14
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 230000004807 localization Effects 0.000 claims description 5
- 238000005260 corrosion Methods 0.000 claims description 4
- 230000007797 corrosion Effects 0.000 claims description 4
- 230000005764 inhibitory process Effects 0.000 claims description 4
- 230000001629 suppression Effects 0.000 claims description 4
- 238000010276 construction Methods 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 10
- 238000012549 training Methods 0.000 description 8
- 238000011176 pooling Methods 0.000 description 6
- 230000009466 transformation Effects 0.000 description 5
- 238000005070 sampling Methods 0.000 description 4
- 230000007704 transition Effects 0.000 description 4
- 235000008694 Humulus lupulus Nutrition 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- 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
- G06V20/63—Scene text, e.g. street names
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The application provides a license plate recognition method, a license plate recognition device, license plate recognition equipment and a computer readable storage medium, wherein the method comprises the following steps: acquiring a target image containing a license plate, and determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST; converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image; and processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image. The application relates to image processing, can improve the rate of accuracy of license plate discernment, and this application can also be applied to fields such as wisdom security protection, wisdom community or intelligent city management to promote the construction in wisdom city.
Description
Technical Field
The present application relates to the field of image processing technologies, and in particular, to a license plate recognition method, apparatus, device, and computer-readable storage medium.
Background
At present, the license plate recognition technology is widely applied to automatic snapshot and recognition of license plates of vehicles in parking lots, urban roads, expressways and other areas, and in practical application, when the license plates are snapshot and recognized, due to insufficient illumination of the environment where the vehicles are located, severe weather and other reasons, the license plate images obtained through shooting are not clear, and the license plates cannot be recognized accurately.
In order to solve the problems, training data are subjected to data enhancement in different degrees, such as cutting, turning, Gaussian noise and the like, through a deep learning mode, and then a convolutional neural network is trained based on the enhanced training data, so that a license plate recognition model is obtained. Therefore, how to improve the accuracy of license plate recognition is a problem to be solved urgently at present.
Disclosure of Invention
The present application mainly aims to provide a license plate recognition method, device, equipment and computer readable storage medium, and aims to improve the accuracy of license plate recognition.
In a first aspect, the present application provides a license plate recognition method, including:
Acquiring a target image containing a license plate, and determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
and processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
In a second aspect, the present application further provides a license plate recognition apparatus, including:
the acquisition module is used for acquiring a target image containing a license plate;
the license plate positioning module is used for determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
the preprocessing module is used for converting the first license plate image area into a license plate gray image and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
and the license plate recognition module is used for processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
In a third aspect, the present application further provides a computer device comprising a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of the license plate recognition method as described above.
In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of the license plate recognition method as described above.
The application provides a license plate recognition method, a license plate recognition device, a license plate recognition equipment and a computer readable storage medium, wherein a license plate positioning model based on a scene text detection framework EAST is used for determining a license plate image area in a target image, converting the first license plate image area into a license plate gray image, performing contrast enhancement processing on the license plate gray image to obtain a target license plate image, inputting the target license plate image into an end-to-end license plate recognition model based on a two-dimensional attention mechanism, so that license plate information can be accurately recognized from the target license plate image, the accuracy of license plate recognition is greatly improved, and the scheme can be further applied to the fields of intelligent security, intelligent communities or intelligent city management and the like, and the construction of intelligent cities is promoted.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flow chart of a license plate recognition method according to an embodiment of the present disclosure;
fig. 2 is a schematic flowchart of another license plate recognition method according to an embodiment of the present disclosure;
fig. 3 is a schematic block diagram of a license plate recognition device according to an embodiment of the present disclosure;
FIG. 4 is a schematic block diagram of another license plate recognition device provided in an embodiment of the present application;
fig. 5 is a schematic block diagram of a structure of a computer device according to an embodiment of the present application.
The implementation, functional features and advantages of the objectives of the present application will be further explained with reference to the accompanying drawings.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. 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 application.
The flow diagrams depicted in the figures are merely illustrative and do not necessarily include all of the elements and operations/steps, nor do they necessarily have to be performed in the order depicted. For example, some operations/steps may be decomposed, combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.
The embodiment of the application provides a license plate identification method and device, computer equipment and a computer readable storage medium. The license plate identification method can be applied to terminal equipment, the terminal equipment can be electronic equipment such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant and wearable equipment, the license plate identification method can also be applied to a server, and the server can be a single server or a server cluster consisting of a plurality of servers.
Some embodiments of the present application will be described in detail below with reference to the accompanying drawings. The embodiments described below and the features of the embodiments can be combined with each other without conflict.
Referring to fig. 1, fig. 1 is a schematic flow chart of a license plate recognition method according to an embodiment of the present disclosure. As shown in fig. 1, the license plate recognition method includes steps S101 to S103.
Step S101, a target image containing a license plate is obtained, and a first license plate image area in the target image is determined through a license plate positioning model realized based on a scene text detection framework EAST.
The license plate recognition method is applied to terminal equipment or a server, wherein the terminal equipment is provided with a shooting device, the shooting device comprises a monocular shooting device and a binocular shooting device, the terminal equipment comprises a smart phone, a tablet personal computer, a PC (personal computer), a notebook computer and the like, the terminal equipment shoots a license plate of a vehicle through the shooting device to obtain a target image containing the license plate or obtain the target image containing the license plate from a local map library; the method comprises the steps of determining a first license plate image area in a target image through a license plate positioning model realized based on an Efficient and accurate Scene Text detection framework (EAST), or uploading a target image containing a license plate to a server by a terminal device, acquiring the target image uploaded by the terminal device by the server, and determining the first license plate image area in the target image through the license plate positioning model realized based on the EAST.
In some embodiments, the EAST-based license plate location model includes a full convolutional neural network (FCN) layer and a Non Maximum Suppression (NMS) layer, and the determining the first license plate image region in the target image by the EAST-based license plate location model is specifically: inputting the target image into a full convolution neural network layer for processing to obtain a target image comprising a plurality of rectangular frames; inputting a target image containing a plurality of rectangular frames to the non-maximum value inhibition layer for processing to obtain a target rectangular frame; and extracting the image area where the target rectangular frame is located from the target image to obtain a first license plate image area.
In some embodiments, the license plate location model implemented based on EAST is trained in advance, and the training process is as follows: collecting a large number of pictures containing license plates, and marking license plate image areas in the pictures so as to construct a sample data set; and obtaining an initial EAST model, performing iterative training on the initial EAST model based on the constructed sample data set, and continuously optimizing the EAST model until the EAST model converges, thereby obtaining a license plate positioning model realized based on EAST.
It can be understood that the first license plate image region in the target image may also be determined by using a license plate location algorithm based on edge detection, a license plate location algorithm based on color segmentation, a license plate location algorithm based on wavelet transformation, a license plate location algorithm based on genetic algorithm, a license plate location algorithm based on mathematical morphology, a license plate location algorithm based on gray-scale image texture feature analysis, or a license plate location algorithm based on edge, color segmentation and contour, which is not specifically limited in this application.
And S102, converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image.
After a first license plate image area of a target image is determined, cutting the first license plate image area from the target image, converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain the target license plate image. The license plate gray level image can be subjected to contrast enhancement processing based on contrast enhancement algorithms such as histogram stretching, histogram equalization, exponential transformation, logarithmic transformation, gray level stretching or linear stretching and the like.
In some embodiments, the manner of converting the first license plate image area into the license plate grayscale image is specifically: acquiring gray value matrixes corresponding to an R channel, a G channel and a B channel in RGB three channels of a first license plate image area, and determining a target gray value matrix based on the gray value matrixes corresponding to the R channel, the G channel and the B channel; and respectively inputting the target gray value matrix into the RGB three channels, thereby obtaining the license plate gray image. The target grayscale matrix may be determined by any one of a floating-point algorithm (Gray ═ R × 0.3+ G × 0.59+ B × 0.11), an integer algorithm (Gray ═ R × 30+ G × 59+ B × 11)/100), an average value method (Gray ═ R + G + B)/3), a shift method (Gray ═ R77 + G151 + B28) > >8), and taking only green (Gray ═ G), where Gray is the target grayscale matrix, and R, G and B are grayscale matrices corresponding to the R channel, the G channel, and the B channel, respectively.
It can be understood that the contrast enhancement processing may also be performed on the first license plate image area, and then the first license plate image area after the contrast enhancement processing is converted into a gray image, so as to obtain the target license plate image, which is not specifically limited in this application.
And S103, processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
The two-dimensional attention mechanism-based end-to-end license plate recognition model is a pre-trained end-to-end character recognition model based on the two-dimensional attention mechanism, the end-to-end character recognition model based on the two-dimensional attention mechanism comprises a convolutional neural network CNN, a two-way long-time memory network LSTM based on the two-dimensional attention mechanism, a full connection layer and a time sequence classification (CTC) algorithm layer, and the convolutional neural network comprises a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer, a second pooling layer and a fourth convolutional layer.
The following describes in detail the training process of an end-to-end text recognition model based on a two-dimensional attention mechanism. Specifically, license plate images shot in different scenes are collected to obtain a large number of sample license plate images, and license plate numbers in each sample license plate image are marked to obtain a sample license plate image set; converting each sample license plate image in the sample license plate image set into a license plate gray image, and performing contrast enhancement processing on each license plate gray image to obtain a target sample data set; and training the end-to-end character recognition model based on the two-dimensional attention mechanism based on the target sample data set until the end-to-end character recognition model based on the two-dimensional attention mechanism converges, thereby obtaining the end-to-end license plate recognition model based on the two-dimensional attention mechanism.
In some embodiments, the license plate recognition model includes a convolutional neural network layer, a two-way long-and-short-term memory network layer based on a two-dimensional attention mechanism, a full connection layer, and a time-sequence classification CTC algorithm layer, and the mode of processing the target license plate image by the end-to-end license plate recognition model based on the two-dimensional attention mechanism is specifically as follows: inputting the target license plate image into a convolutional neural network layer for processing to obtain a plurality of first characteristic graphs of the target license plate image; inputting the plurality of first characteristic graphs into a two-way long-short time memory network layer based on a two-dimensional attention mechanism for processing to obtain a plurality of second characteristic graphs; inputting the plurality of second feature maps into a full-connection layer for processing to obtain a feature vector matrix of the target image; and inputting the characteristic vector matrix into a CTC algorithm layer for processing to obtain a license plate identification result of the target image.
Exemplarily, the convolutional neural network comprises a first convolutional layer, a second convolutional layer, a first pooling layer, a third convolutional layer, a second pooling layer and a fourth convolutional layer, the target license plate image is input into the first convolutional layer, and 64 80 × 16 feature maps are output; inputting 64 feature maps of 80 × 16 into the second convolution layer, and outputting 128 feature maps of 80 × 16; inputting 128 80 × 16 feature maps into the first pooling layer, and outputting 128 40 × 8 feature maps; inputting 128 feature maps of 40 × 8 into the third convolutional layer, and outputting 256 feature maps of 40 × 8; inputting 256 40 × 8 feature maps into the second pooling layer, and outputting 256 40 × 4 feature maps; inputting 256 feature maps of 40 × 4 into the fourth convolutional layer, and outputting 512 feature maps of 20 × 1 (first feature map); inputting 512 20 × 1 feature maps into a two-way long-short time memory network LSTM based on a two-dimensional attention mechanism, and outputting 512 20 × 1 feature maps (second feature maps); inputting 512 20 × 1 feature maps into a full connection layer, and outputting a 20 × 37 matrix; and inputting the 20 x 37 matrix into a CTC algorithm layer, decoding the 20 x 37 matrix, and outputting a license plate recognition result.
In some embodiments, a plurality of target images containing license plates of the same vehicle are acquired, and license plate image areas in each target image are determined through a license plate positioning model realized based on EAST; extracting the license plate image area from each target image to obtain a plurality of license plate images, and performing contrast enhancement and gray level processing on each license plate image to obtain a plurality of target license plate images; processing each target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a plurality of license plate recognition results; and determining the similarity between every two license plate recognition results, determining whether the similarity between every two license plate recognition results is greater than a preset similarity, and if the similarity between every two license plate recognition results is greater than the preset similarity, taking any license plate recognition result as a target license plate recognition result of the vehicle.
According to the license plate recognition method provided by the embodiment, the license plate image area in the target image is determined through the license plate positioning model realized based on the efficient and accurate scene text detection framework EAST, the first license plate image area is converted into the license plate gray image, the contrast enhancement processing is carried out on the license plate gray image to obtain the target license plate image, then the target license plate image is input into the end-to-end license plate recognition model based on the two-dimensional attention mechanism, the license plate information can be accurately recognized from the target license plate image, the accuracy of license plate recognition is greatly improved, the scheme can be further applied to the fields of intelligent security, intelligent communities or intelligent city management and the like, and therefore the construction of the intelligent city is promoted.
Referring to fig. 2, fig. 2 is a schematic flow chart of another license plate recognition method according to an embodiment of the present disclosure. As shown in fig. 2, the license plate recognition method includes steps S201 to S205.
Step S201, a target image containing a license plate is obtained, and a first license plate image area in the target image is determined through a license plate positioning model realized based on a scene text detection framework EAST.
The terminal equipment shoots the license plate of the vehicle through the shooting device to obtain a target image containing the license plate or obtain the target image containing the license plate from a local image library; the method comprises the steps of determining a first license plate image area in a target image through a license plate positioning model realized based on an efficient and accurate Scene Text detection framework (EAST), or uploading a target image containing a license plate to a server by a terminal device, acquiring the target image uploaded by the terminal device by the server, and determining the first license plate image area in the target image through the license plate positioning model realized based on the EAST.
Step S202, determining a second license plate image area in the target image through a license plate positioning algorithm based on image texture characteristics.
Specifically, preprocessing a target image, wherein the preprocessing comprises maximum inter-class binarization processing, mathematical morphology corrosion processing and edge enhancement processing; performing texture feature scanning on the preprocessed target image in a line scanning mode to determine the upper and lower boundaries of the license plate in the target image; performing texture feature scanning on the preprocessed target image in a column scanning mode to determine left and right boundaries of a license plate in the target image; and determining a second license plate image area in the target image according to the upper and lower boundaries and the left and right boundaries of the license plate in the target image. The license plate image area in the target image can be accurately determined through a license plate positioning algorithm based on image texture characteristics.
Because the image after the license plate binarization has the texture characteristic of black and white two-pixel jumping, the upper and lower boundaries of the license plate can be determined according to the continuous jumping times of the texture characteristic during line scanning, and because the license plate image area has 7 continuous characters and the distance between the characters is within a certain range, the jump from the character to the background or from the background to the character is defined. Specifically, a texture feature scanning mode is used for scanning the preprocessed target image from top to bottom to obtain the jumping times of the texture feature of each line and the pixel distance between every two jumping positions; determining jump starting points and jump end points of a plurality of target lines according to the jump times of the texture features of each line and the pixel distance between every two jump positions; taking a line segment between a jump starting point and a jump finishing point of a first target line in the plurality of target lines as an upper boundary of a license plate in a target image; and taking a line segment between the jump starting point and the jump ending point of the last target line in the plurality of target lines as the lower boundary of the license plate in the target image.
In some embodiments, the texture feature scanning is performed on the preprocessed target image in a column scanning manner, and the manner of determining the left and right boundaries of the license plate in the target image is specifically as follows: performing texture feature scanning on the preprocessed target image from left to right in a column scanning mode to obtain the jumping times of the texture feature of each column and the pixel distance between each two jumping positions; determining jump starting points and jump finishing points of a plurality of target columns according to the jump times of the texture features of each column and the pixel distance between every two jump positions; taking a line segment between a jump starting point and a jump finishing point of a first target column in the plurality of target columns as a left boundary of a license plate in a target image; and taking a line segment between the jump starting point and the jump ending point of the last target column in the plurality of target columns as the right boundary of the license plate in the target image.
In some embodiments, the determining the transition start point and the transition end point of the multiple target rows/columns according to the transition times of the texture features of each row/column and the pixel distance between each two transition positions is specifically as follows: determining whether the jumping times of the textural features of each row/column are greater than preset times, and determining whether the pixel distance between every two jumping positions of the textural features of each row/column is in a preset distance range; and taking the lines of which the jumping times of the texture features are greater than the preset times and the pixel distance between every two jumping positions is within the preset distance range as the lines/columns in the license plate image area, thereby obtaining a plurality of target lines/columns and obtaining the jumping starting points and the jumping end points of the target lines/columns. The jumping starting point is a jumping point detected for the first time in a row/column, and the jumping end point is a jumping point detected for the last time in the row/column. It can be understood that the number of hops of the license plate image region is more than that of other non-license plate image regions, and the pixel distance between two adjacent hops is within a preset distance range, the number of hops in the license plate image region is necessarily greater than the preset number of times, and the preset number of times and the preset distance range may be set based on an actual situation, for example, the preset number of times is 18 or 15.
And S203, verifying the first license plate image area according to the second license plate image area.
After determining the second license plate image area, the first license plate image area is verified based on the second license plate image area. Specifically, determining the similarity between a first license plate image area and a second license plate image area, and determining whether the similarity is greater than or equal to a preset similarity; if the similarity is greater than or equal to the preset similarity, determining that the first license plate image area passes the verification; and if the similarity is smaller than the preset similarity, determining that the first license plate image area does not pass the verification. The preset similarity may be set based on an actual situation, which is not specifically limited in the present application.
And S204, when the first license plate image area passes the verification, converting the first license plate image area into a first license plate gray image, and performing contrast enhancement processing on the first license plate gray image to obtain a target license plate image.
When the first license plate image area passes the verification, cutting out the first license plate image area from the target image, converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain the target license plate image. The license plate gray level image can be subjected to contrast enhancement processing based on contrast enhancement algorithms such as histogram stretching, histogram equalization, exponential transformation, logarithmic transformation, gray level stretching or linear stretching and the like.
And S205, processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
The license plate recognition model comprises a convolutional neural network layer, a two-way long-time memory network layer based on a two-dimensional attention mechanism, a full connection layer and a time sequence classification (CTC) algorithm layer, and the mode of processing a target license plate image through an end-to-end license plate recognition model based on the two-dimensional attention mechanism specifically comprises the following steps: inputting the target license plate image into a convolutional neural network layer for processing to obtain a plurality of first characteristic graphs of the target license plate image; inputting the plurality of first characteristic graphs into a two-way long-short time memory network layer based on a two-dimensional attention mechanism for processing to obtain a plurality of second characteristic graphs; inputting the plurality of second feature maps into a full-connection layer for processing to obtain a feature vector matrix of the target image; and inputting the characteristic vector matrix into a CTC algorithm layer for processing to obtain a license plate identification result of the target image.
In some embodiments, a first license plate image region in the target image is determined by a license plate location model implemented based on EAST, and a second license plate image region in the target image is determined by a license plate location algorithm based on image texture features; converting the first license plate image area into a first license plate gray scale image, and converting the second license plate image area into a second license plate gray scale image; carrying out contrast enhancement processing on the first license plate gray level image to obtain a first target license plate image, and carrying out contrast enhancement processing on the second license plate gray level image to obtain a second target license plate image; processing a first target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a first license plate recognition result, and performing character recognition on a second target license plate image based on a preset license plate recognition model to obtain a second license plate recognition result; checking the first license plate recognition result according to the second license plate recognition result, and taking the first license plate recognition result as a target license plate recognition result when the first license plate recognition result passes the checking; and if the first license plate identification result does not pass the verification, re-identifying or feeding back prompt information.
The preset license plate recognition model is a pre-trained convolutional neural network, the convolutional neural network comprises a first convolutional layer, a first sampling layer, a second convolutional layer, a second sampling layer, a vector connection layer and an output layer, and the training process of the convolutional neural network is described in detail below. Specifically, license plate images shot in different scenes are collected to obtain a large number of sample license plate images, and license plate numbers in each sample license plate image are marked to obtain a sample license plate image set; converting each sample license plate image in the sample license plate image set into a license plate gray image, and performing contrast enhancement processing on each license plate gray image to obtain a target sample data set; and training the convolutional neural network type based on the target sample data set until the convolutional neural network converges, thereby obtaining a preset license plate recognition model.
Specifically, inputting a second target license plate image into the first convolution layer, and outputting 4 characteristic graphs of 32 multiplied by 24; inputting 4 32 × 24 feature maps into a first sampling layer, and outputting 4 16 × 12 feature maps; inputting 4 16 × 12 feature maps into the second convolutional layer, and outputting 12 × 8 feature maps; inputting 12 feature maps of 12 × 8 into a second sampling layer, and outputting 12 feature maps of 6 × 4; transforming and converting the 12 feature maps of 6 × 4 into a vector which comprises 288 neurons of 12 × 6 × 4, inputting the vector into an output layer, and outputting a second card recognition result.
Specifically, the similarity between the first license plate recognition result and the second license plate recognition result is determined, whether the similarity between the first license plate recognition result and the second license plate recognition result is greater than or equal to a preset similarity is determined, if the similarity between the first license plate recognition result and the second license plate recognition result is greater than or equal to the preset similarity, the first license plate recognition result is determined to pass the verification, and if the similarity between the first license plate recognition result and the second license plate recognition result is less than the preset similarity, the first license plate recognition result is determined not to pass the verification. The preset similarity may be set based on an actual situation, which is not specifically limited in the present application.
The license plate recognition method provided by the embodiment determines a first license plate image area in a target image through a license plate positioning model realized based on an efficient and accurate scene text detection framework EAST, then determines a second license plate image area in the target image through a license plate positioning algorithm based on image texture characteristics, verifies the first license plate image area based on the second license plate image area, converts the first license plate image area into a license plate gray image when the first license plate image area passes the verification, performs contrast enhancement processing on the license plate gray image to obtain the target license plate image, can improve the positioning accuracy of the license plate, then inputs the target license plate image into an end-to-end license plate recognition model based on a two-dimensional attention machine, can accurately recognize license plate information from the target license plate image, and greatly improves the accuracy of license plate recognition, this scheme can also be applied to in fields such as wisdom security protection, wisdom community or intelligent city management to promote the construction in wisdom city.
Referring to fig. 3, fig. 3 is a schematic block diagram of a license plate recognition device according to an embodiment of the present disclosure.
As shown in fig. 3, the license plate recognition device 300 includes: the license plate recognition system comprises an acquisition module 301, a license plate positioning module 302, a preprocessing module 303 and a license plate recognition module 304, wherein:
the acquisition module 301 is configured to acquire a target image including a license plate;
the license plate positioning module 302 is configured to determine a first license plate image area in the target image through a license plate positioning model implemented based on a scene text detection framework EAST;
the preprocessing module 303 is configured to convert the first license plate image area into a license plate grayscale image, and perform contrast enhancement processing on the license plate grayscale image to obtain a target license plate image;
the license plate recognition module 304 is configured to process the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism, so as to obtain a license plate recognition result of the target image.
In some embodiments, the license plate location model includes a full convolution neural network layer and a non-maxima suppression layer, the license plate location module 302 is further configured to:
inputting the target image into the full convolution neural network layer for processing to obtain a target image containing a plurality of rectangular frames;
Inputting a target image containing a plurality of rectangular frames to the non-maximum value inhibition layer for processing to obtain a target rectangular frame;
and extracting the image area where the target rectangular frame is located from the target image to obtain a first license plate image area.
In some embodiments, the license plate recognition model includes a convolutional neural network layer, a two-way long-and-short-term memory network layer based on a two-dimensional attention mechanism, a full-connected layer, and a time-series classification CTC algorithm layer, and the license plate recognition module 304 is further configured to:
inputting the target license plate image into the convolutional neural network layer for processing to obtain a plurality of first characteristic graphs of the target license plate image;
inputting the plurality of first characteristic diagrams to the two-way long-short time memory network layer based on the two-dimensional attention mechanism for processing to obtain a plurality of second characteristic diagrams;
inputting the plurality of second feature maps into the full-connection layer for processing to obtain a feature vector matrix of the target image;
and inputting the characteristic vector matrix into the CTC algorithm layer for processing to obtain a license plate identification result of the target image.
Referring to fig. 4, fig. 4 is a schematic block diagram of another license plate recognition device according to an embodiment of the present disclosure.
As shown in fig. 4, the license plate recognition device 400 includes: the license plate recognition system comprises an acquisition module 401, a first license plate positioning module 402, a second license plate positioning module 403, a verification module 404, a preprocessing module 405 and a license plate recognition module 406, wherein:
the obtaining module 401 is configured to obtain a target image including a license plate;
the first license plate positioning module 402 is configured to determine a first license plate image area in the target image through a license plate positioning model implemented based on a scene text detection framework EAST;
the second license plate positioning module 403 is configured to determine a second license plate image area in the target image through a license plate positioning algorithm based on image texture features;
the verification module 404 is configured to verify the first license plate image area according to the second license plate image area;
the preprocessing module 405 is configured to convert the first license plate image area into a license plate gray image when the first license plate image area passes verification, and perform contrast enhancement processing on the license plate gray image to obtain a target license plate image;
the license plate recognition module 406 is configured to process the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism, so as to obtain a license plate recognition result of the target image.
In some embodiments, the second card locating module 403 is further configured to:
preprocessing the target image, wherein the preprocessing comprises maximum inter-class binarization processing, mathematical morphology corrosion processing and edge enhancement processing;
performing texture feature scanning on the preprocessed target image in a line scanning mode to determine the upper and lower boundaries of the license plate in the target image;
performing texture feature scanning on the preprocessed target image in a column scanning mode to determine left and right boundaries of a license plate in the target image;
and determining a second license plate image area in the target image according to the upper and lower boundaries and the left and right boundaries of the license plate in the target image.
In some embodiments, the second card locating module 403 is further configured to:
performing texture feature scanning on the preprocessed target image in a line scanning mode to obtain the jumping times of the texture feature of each line and the pixel distance between every two jumping positions;
determining jump starting points and jump end points of a plurality of target lines according to the jump times of the texture features of each line and the pixel distance between every two jump positions,
Taking a line segment between a jump starting point and a jump finishing point of a first target line in the plurality of target lines as an upper boundary of a license plate in the target image;
and taking a line segment between the jump starting point and the jump ending point of the last target line in the plurality of target lines as the lower boundary of the license plate in the target image.
In some embodiments, the verification module 404 is further configured to:
determining a similarity between the first license plate image area and the second license plate image area, and determining whether the similarity is greater than or equal to a preset similarity;
if the similarity is greater than or equal to a preset similarity, determining that the first license plate image area passes verification;
and if the similarity is smaller than the preset similarity, determining that the first license plate image area does not pass the verification.
It should be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatus and each module and unit may refer to the corresponding processes in the foregoing embodiments of the license plate recognition method, and are not described herein again.
The apparatus provided by the above embodiments may be implemented in the form of a computer program that can be run on a computer device as shown in fig. 5.
Referring to fig. 5, fig. 5 is a schematic block diagram of a computer device according to an embodiment of the present disclosure. The computer device may be a server or a terminal.
As shown in fig. 5, the computer device includes a processor, a memory, and a network interface connected by a system bus, wherein the memory may include a nonvolatile storage medium and an internal memory.
The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause a processor to perform any of the license plate recognition methods.
The processor is used for providing calculation and control capability and supporting the operation of the whole computer equipment.
The internal memory provides an environment for running a computer program in the non-volatile storage medium, which when executed by the processor, causes the processor to perform any of the license plate recognition methods.
The network interface is used for network communication, such as sending assigned tasks and the like. Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
It should be understood that the Processor may be a Central Processing Unit (CPU), and the Processor may be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, etc. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Wherein, in some embodiments, the processor is configured to execute a computer program stored in the memory to implement the steps of:
acquiring a target image containing a license plate, and determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
and processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
In some embodiments, the license plate localization model includes a full convolution neural network layer and a non-maxima suppression layer, the processor, in implementing the license plate localization model implemented by the scene text detection framework EAST based, in determining the first license plate image region in the target image, is to implement:
inputting the target image into the full convolution neural network layer for processing to obtain a target image containing a plurality of rectangular frames;
inputting a target image containing a plurality of rectangular frames to the non-maximum value inhibition layer for processing to obtain a target rectangular frame;
and extracting the image area where the target rectangular frame is located from the target image to obtain a first license plate image area.
In some embodiments, the license plate recognition model includes a convolutional neural network layer, a two-way long-and-short-term memory network layer based on a two-dimensional attention mechanism, a full connection layer, and a time-series classification CTC algorithm layer, and the processor is configured to, when processing the target license plate image through an end-to-end license plate recognition model based on the two-dimensional attention mechanism to obtain a license plate recognition result of the target image, implement:
inputting the target license plate image into the convolutional neural network layer for processing to obtain a plurality of first characteristic graphs of the target license plate image;
Inputting the plurality of first characteristic diagrams to the two-way long-short time memory network layer based on the two-dimensional attention mechanism for processing to obtain a plurality of second characteristic diagrams;
inputting the plurality of second feature maps into the full-connection layer for processing to obtain a feature vector matrix of the target image;
and inputting the characteristic vector matrix into the CTC algorithm layer for processing to obtain a license plate identification result of the target image.
In other embodiments, the processor is configured to execute a computer program stored in the memory to perform the steps of:
acquiring a target image containing a license plate, and determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
determining a second license plate image area in the target image through a license plate positioning algorithm based on image texture characteristics;
according to the second license plate image area, checking the first license plate image area;
when the first license plate image area passes the verification, converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
And processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
In some embodiments, the processor, in implementing determining a second license plate image region in the target image by a license plate localization algorithm based on image texture features, is configured to implement:
preprocessing the target image, wherein the preprocessing comprises maximum inter-class binarization processing, mathematical morphology corrosion processing and edge enhancement processing;
performing texture feature scanning on the preprocessed target image in a line scanning mode to determine the upper and lower boundaries of the license plate in the target image;
performing texture feature scanning on the preprocessed target image in a column scanning mode to determine left and right boundaries of a license plate in the target image;
and determining a second license plate image area in the target image according to the upper and lower boundaries and the left and right boundaries of the license plate in the target image.
In some embodiments, the processor, when implementing line scanning to perform texture feature scanning on the preprocessed target image to determine the upper and lower boundaries of the license plate in the target image, is configured to implement:
Performing texture feature scanning on the preprocessed target image in a line scanning mode to obtain the jumping times of the texture feature of each line and the pixel distance between every two jumping positions;
determining jump starting points and jump end points of a plurality of target lines according to the jump times of the texture features of each line and the pixel distance between every two jump positions,
taking a line segment between a jump starting point and a jump finishing point of a first target line in the plurality of target lines as an upper boundary of a license plate in the target image;
and taking a line segment between the jump starting point and the jump ending point of the last target line in the plurality of target lines as the lower boundary of the license plate in the target image.
In some embodiments, the processor, in effecting verification of the first license plate image area from the second license plate image area, is configured to effect:
determining a similarity between the first license plate image area and the second license plate image area, and determining whether the similarity is greater than or equal to a preset similarity;
if the similarity is greater than or equal to a preset similarity, determining that the first license plate image area passes verification;
And if the similarity is smaller than the preset similarity, determining that the first license plate image area does not pass the verification.
The embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, where the computer program includes program instructions, and a method implemented when the program instructions are executed may refer to various embodiments of the license plate identification method of the present application.
Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, for example, a hard disk or a memory of the computer device. The computer readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the computer device.
It is to be understood that the terminology used in the description of the present application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the specification of the present application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should also be understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments. While the invention has been described with reference to specific embodiments, the scope of the invention is not limited thereto, and those skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the invention. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A license plate recognition method is characterized by comprising the following steps:
acquiring a target image containing a license plate, and determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
converting the first license plate image area into a license plate gray image, and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
and processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
2. The license plate recognition method of claim 1, wherein the license plate localization model comprises a full convolution neural network layer and a non-maximum suppression layer, and the determining the first license plate image region in the target image through the license plate localization model implemented based on the scene text detection framework EAST comprises:
Inputting the target image into the full convolution neural network layer for processing to obtain a target image containing a plurality of rectangular frames;
inputting a target image containing a plurality of rectangular frames to the non-maximum value inhibition layer for processing to obtain a target rectangular frame;
and extracting the image area where the target rectangular frame is located from the target image to obtain a first license plate image area.
3. The license plate recognition method of claim 1, wherein the license plate recognition model comprises a convolutional neural network layer, a two-way long-and-short-term memory network layer based on a two-dimensional attention mechanism, a full connection layer and a time-sequence classification (CTC) algorithm layer, and the license plate recognition result of the target image is obtained by processing the target license plate image through an end-to-end license plate recognition model based on the two-dimensional attention mechanism, and the method comprises the following steps:
inputting the target license plate image into the convolutional neural network layer for processing to obtain a plurality of first characteristic graphs of the target license plate image;
inputting the plurality of first characteristic diagrams to the two-way long-short time memory network layer based on the two-dimensional attention mechanism for processing to obtain a plurality of second characteristic diagrams;
Inputting the plurality of second feature maps into the full-connection layer for processing to obtain a feature vector matrix of the target image;
and inputting the characteristic vector matrix into the CTC algorithm layer for processing to obtain a license plate identification result of the target image.
4. The license plate recognition method of any one of claims 1-3, wherein before converting the first license plate image region into a license plate grayscale image and performing contrast enhancement processing on the license plate grayscale image to obtain a target license plate image, the method further comprises:
determining a second license plate image area in the target image through a license plate positioning algorithm based on image texture characteristics;
according to the second license plate image area, checking the first license plate image area;
and when the first license plate image area passes the verification, converting the first license plate image area into a first license plate gray level image, and performing contrast enhancement processing on the first license plate gray level image to obtain a target license plate image.
5. The license plate recognition method of claim 4, wherein the determining a second license plate image area in the target image by a license plate location algorithm based on image texture features comprises:
Preprocessing the target image, wherein the preprocessing comprises maximum inter-class binarization processing, mathematical morphology corrosion processing and edge enhancement processing;
performing texture feature scanning on the preprocessed target image in a line scanning mode to determine the upper and lower boundaries of the license plate in the target image;
performing texture feature scanning on the preprocessed target image in a column scanning mode to determine left and right boundaries of a license plate in the target image;
and determining a second license plate image area in the target image according to the upper and lower boundaries and the left and right boundaries of the license plate in the target image.
6. The license plate recognition method of claim 5, wherein the performing texture feature scanning on the preprocessed target image in a line scanning manner to determine upper and lower boundaries of a license plate in the target image comprises:
performing texture feature scanning on the preprocessed target image in a line scanning mode to obtain the jumping times of the texture feature of each line and the pixel distance between every two jumping positions;
determining jump starting points and jump end points of a plurality of target lines according to the jump times of the texture features of each line and the pixel distance between every two jump positions,
Taking a line segment between a jump starting point and a jump finishing point of a first target line in the plurality of target lines as an upper boundary of a license plate in the target image;
and taking a line segment between the jump starting point and the jump ending point of the last target line in the plurality of target lines as the lower boundary of the license plate in the target image.
7. The license plate recognition method of claim 4, wherein the verifying the first license plate image area based on the second license plate image area comprises:
determining a similarity between the first license plate image area and the second license plate image area, and determining whether the similarity is greater than or equal to a preset similarity;
if the similarity is greater than or equal to a preset similarity, determining that the first license plate image area passes verification;
and if the similarity is smaller than the preset similarity, determining that the first license plate image area does not pass the verification.
8. A license plate recognition device, characterized in that the license plate recognition device comprises:
the acquisition module is used for acquiring a target image containing a license plate;
the license plate positioning module is used for determining a first license plate image area in the target image through a license plate positioning model realized based on a scene text detection framework EAST;
The preprocessing module is used for converting the first license plate image area into a license plate gray image and performing contrast enhancement processing on the license plate gray image to obtain a target license plate image;
and the license plate recognition module is used for processing the target license plate image through an end-to-end license plate recognition model based on a two-dimensional attention mechanism to obtain a license plate recognition result of the target image.
9. A computer arrangement, characterized in that the computer arrangement comprises a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein the computer program, when executed by the processor, carries out the steps of the license plate recognition method according to any one of claims 1 to 7.
10. A computer-readable storage medium, having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the license plate recognition method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010575439.2A CN111860496A (en) | 2020-06-22 | 2020-06-22 | License plate recognition method, device, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010575439.2A CN111860496A (en) | 2020-06-22 | 2020-06-22 | License plate recognition method, device, equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111860496A true CN111860496A (en) | 2020-10-30 |
Family
ID=72988021
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010575439.2A Pending CN111860496A (en) | 2020-06-22 | 2020-06-22 | License plate recognition method, device, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111860496A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112488136A (en) * | 2021-01-22 | 2021-03-12 | 山东商业职业技术学院 | Image recognition system and image recognition device |
| CN112801103A (en) * | 2021-01-19 | 2021-05-14 | 网易（杭州）网络有限公司 | Text direction recognition and text direction recognition model training method and device |
| CN113642577A (en) * | 2021-10-14 | 2021-11-12 | 深圳市爱深盈通信息技术有限公司 | Low-contrast license plate recognition method, system, equipment and storage medium |
| CN114399693A (en) * | 2021-12-29 | 2022-04-26 | 国能铁路装备有限责任公司 | Method and device for estimating service life of bearing saddle |
| CN115050016A (en) * | 2022-08-15 | 2022-09-13 | 深圳市爱深盈通信息技术有限公司 | License plate detection method and device, equipment terminal and readable storage medium |
| CN116168376A (en) * | 2023-04-18 | 2023-05-26 | 苏州大学 | Voice broadcast medicine box recognition device and medicine box recognition method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105354570A (en) * | 2015-10-15 | 2016-02-24 | 深圳市捷顺科技实业股份有限公司 | Method and system for precisely locating left and right boundaries of license plate |
| CN109472262A (en) * | 2018-09-25 | 2019-03-15 | 平安科技（深圳）有限公司 | Licence plate recognition method, device, computer equipment and storage medium |
| CN110414451A (en) * | 2019-07-31 | 2019-11-05 | 深圳市捷顺科技实业股份有限公司 | It is a kind of based on end-to-end licence plate recognition method, device, equipment and storage medium |
| US20200090506A1 (en) * | 2018-09-19 | 2020-03-19 | National Chung-Shan Institute Of Science And Technology | License plate recognition system and license plate recognition method |
-
2020
- 2020-06-22 CN CN202010575439.2A patent/CN111860496A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105354570A (en) * | 2015-10-15 | 2016-02-24 | 深圳市捷顺科技实业股份有限公司 | Method and system for precisely locating left and right boundaries of license plate |
| US20200090506A1 (en) * | 2018-09-19 | 2020-03-19 | National Chung-Shan Institute Of Science And Technology | License plate recognition system and license plate recognition method |
| CN109472262A (en) * | 2018-09-25 | 2019-03-15 | 平安科技（深圳）有限公司 | Licence plate recognition method, device, computer equipment and storage medium |
| CN110414451A (en) * | 2019-07-31 | 2019-11-05 | 深圳市捷顺科技实业股份有限公司 | It is a kind of based on end-to-end licence plate recognition method, device, equipment and storage medium |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112801103A (en) * | 2021-01-19 | 2021-05-14 | 网易（杭州）网络有限公司 | Text direction recognition and text direction recognition model training method and device |
| CN112801103B (en) * | 2021-01-19 | 2024-02-27 | 网易（杭州）网络有限公司 | Text direction recognition and text direction recognition model training method and device |
| CN112488136A (en) * | 2021-01-22 | 2021-03-12 | 山东商业职业技术学院 | Image recognition system and image recognition device |
| CN113642577A (en) * | 2021-10-14 | 2021-11-12 | 深圳市爱深盈通信息技术有限公司 | Low-contrast license plate recognition method, system, equipment and storage medium |
| CN114399693A (en) * | 2021-12-29 | 2022-04-26 | 国能铁路装备有限责任公司 | Method and device for estimating service life of bearing saddle |
| CN115050016A (en) * | 2022-08-15 | 2022-09-13 | 深圳市爱深盈通信息技术有限公司 | License plate detection method and device, equipment terminal and readable storage medium |
| CN115050016B (en) * | 2022-08-15 | 2023-01-17 | 深圳市爱深盈通信息技术有限公司 | License plate detection method and device, equipment terminal and readable storage medium |
| CN116168376A (en) * | 2023-04-18 | 2023-05-26 | 苏州大学 | Voice broadcast medicine box recognition device and medicine box recognition method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111860496A (en) | License plate recognition method, device, equipment and computer readable storage medium | |
| CN110414507B (en) | License plate recognition method and device, computer equipment and storage medium | |
| CN108268867B (en) | License plate positioning method and device | |
| CN110781756A (en) | Urban road extraction method and device based on remote sensing image | |
| CN113673530B (en) | Remote sensing image semantic segmentation method, device, computer equipment and storage medium | |
| CN111340077B (en) | Attention mechanism-based disparity map acquisition method and device | |
| CN110689043A (en) | Vehicle fine granularity identification method and device based on multiple attention mechanism | |
| CN110084238B (en) | Finger vein image segmentation method and device based on LadderNet network and storage medium | |
| CN114581887B (en) | Method, device, equipment and computer readable storage medium for detecting lane line | |
| Xiang et al. | Lightweight fully convolutional network for license plate detection | |
| CN110852311A (en) | Three-dimensional human hand key point positioning method and device | |
| CN114038004A (en) | Certificate information extraction method, device, equipment and storage medium | |
| CN113792741A (en) | Character recognition method, device, equipment and storage medium | |
| CN105243154A (en) | Remote sensing image retrieval method and system based on significant point characteristics and spare self-encodings | |
| CN114444565B (en) | Image tampering detection method, terminal equipment and storage medium | |
| CN111507337A (en) | License plate recognition method based on hybrid neural network | |
| CN111160169A (en) | Face detection method, device, equipment and computer readable storage medium | |
| CN113221601B (en) | Character recognition method, device and computer readable storage medium | |
| CN113052170A (en) | Small target license plate recognition method under unconstrained scene | |
| CN108509826B (en) | Road identification method and system for remote sensing image | |
| Fan et al. | Road vanishing point detection using weber adaptive local filter and salient‐block‐wise weighted soft voting | |
| CN115578590A (en) | Image identification method and device based on convolutional neural network model and terminal equipment | |
| CN109359542B (en) | Vehicle damage level determining method based on neural network and terminal equipment | |
| CN112330619B (en) | Method, device, equipment and storage medium for detecting target area | |
| CN116229406B (en) | Lane line detection method, system, electronic equipment and storage medium |
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 |