CN113392839B - License plate recognition method and device for non-motor vehicle, computer equipment and storage medium - Google Patents
License plate recognition method and device for non-motor vehicle, computer equipment and storage medium Download PDFInfo
- Publication number
- CN113392839B CN113392839B CN202110540159.2A CN202110540159A CN113392839B CN 113392839 B CN113392839 B CN 113392839B CN 202110540159 A CN202110540159 A CN 202110540159A CN 113392839 B CN113392839 B CN 113392839B
- Authority
- CN
- China
- Prior art keywords
- license plate
- area
- character
- image
- target
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The utility model relates to a non-motor vehicle license plate recognition method, a device, computer equipment and a storage medium, through obtaining the license plate image and the license plate type of a target license plate, the license plate image is segmented according to gray jump information in the license plate image, character areas of the obtained license plate image are obtained, when the license plate type is a non-single-layer license plate, a first area and a second area which are positioned above the character areas are obtained, character types of the first area are obtained, under the condition that the character types of the first area are configurable characters, registration information of the target license plate is matched for the first area according to the license plate standard type of the target license plate, and finally, a preset number recognition model is input into the second area to obtain the license plate number of the target license plate, and finally, the recognition result of the target license plate is obtained according to the registration information and the license plate number, so that the universality of non-motor vehicle license plate recognition of different areas can be increased, and the accuracy and the non-motor vehicle license plate recognition efficiency can be improved.
Description
Technical Field
The present disclosure relates to the field of image processing, and in particular, to a method and apparatus for identifying a license plate of a non-motor vehicle, a computer device, and a storage medium.
Background
With the development of deep learning in the traffic field, the application of deep learning in the non-motor vehicle management field has emerged. In the process of processing illegal snapshots of a non-motor vehicle by deep learning, identifying license plate information of the non-motor vehicle is a very important ring. Because the license plates of the non-motor vehicles do not have uniform license plate systems, the license plates of the non-motor vehicles with specific systems in different areas contain names of administrative areas or cities to which the license plates belong, and additional information which does not need to be identified, such as anti-theft record numbers, are added to some license plates. All the above information can increase the difficulty of identifying the license plate of the non-motor vehicle. Therefore, a general non-motor vehicle license plate recognition scheme capable of being applied to different areas does not exist at present.
Aiming at the problem that the non-motor vehicle license plate recognition scheme in the related technology cannot be universally used in different areas, no effective solution is proposed at present.
Disclosure of Invention
In this embodiment, a method, an apparatus, a computer device, and a storage medium for identifying a license plate of a non-motor vehicle are provided to solve the problem that the license plate identification scheme of a non-motor vehicle in the related art cannot be commonly used in different areas.
In a first aspect, in this embodiment, there is provided a license plate recognition method of a non-motor vehicle, including the steps of:
obtaining a license plate image and license plate types of a target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
dividing the license plate image according to gray jump information in the license plate image to obtain a character area of the license plate image;
when the license plate type is a non-single-layer license plate, acquiring a first area of the character area, which is positioned above, and a second area of the character area, which is positioned below; acquiring the character type of the first area, matching registration information of the target license plate for the first area according to the license plate type of the target license plate under the condition that the character type of the first area is a configurable character, and inputting the second area into a preset number recognition model to obtain the license plate number of the target license plate;
and obtaining the recognition result of the target license plate according to the registration information and the license plate number.
In some embodiments, the acquiring the license plate image and the license plate type of the target license plate includes the following steps:
acquiring a target non-motor vehicle image;
And inputting the target non-motor vehicle image into a preset detection model to obtain the license plate image and the license plate type.
In some embodiments, after obtaining the license plate image and the license plate type of the target license plate, before dividing the license plate image according to the gray jump information in the license plate image to obtain the character area of the license plate image, the method further comprises the following steps:
carrying out graying treatment on the license plate image;
and inputting the grey license plate image into a preset correction model, and correcting the shape and angle of the license plate image.
In some embodiments, the dividing the license plate image according to the gray jump information in the license plate image to obtain the character area of the license plate image includes the following steps:
acquiring gray level wave peak values, gray level wave trough values and jump times of a character area of the license plate image;
obtaining gray adjacent combinations according to the gray wave peak value, the gray wave trough value and the jump times;
and determining a segmentation threshold value of the license plate image by using the gray adjacent combination, and segmenting the license plate image according to the segmentation threshold value.
In some embodiments, the matching, according to the license plate type of the target license plate, the registration information of the target license plate for the first area includes the following steps:
if the license plate type is one type, acquiring preset license plate configuration information, and acquiring the registration information from the license plate configuration information.
In some embodiments, the matching, according to the license plate type of the target license plate, the registration information of the target license plate for the first area further includes the following steps:
if the types of license plate systems are more than one type, inputting the first region into a first classification model with complete training, and determining the license plate system corresponding to the target license plate;
and acquiring preset license plate configuration information, and matching corresponding registration information for the first area according to the mapping relation between the license plate system and the license plate configuration information.
In some embodiments, when the license plate type is a non-single layer license plate and the character type of the first area is a recognizable character, the method further comprises the following steps:
and inputting the spliced first area and the spliced second area into a preset license plate recognition model to obtain a recognition result of the target license plate.
In some embodiments, when the license plate type is a single-layer license plate, the method comprises the following steps:
and inputting the character area into the number recognition model to obtain a recognition result of the target license plate.
In some embodiments, the obtaining the character class of the first region includes:
and inputting the first region into a complete training second classification model to obtain the character category of the first region.
In a second aspect, in this embodiment, there is provided a license plate recognition device for a non-motor vehicle, including: the device comprises an image acquisition module, a segmentation module, an identification module and a result acquisition module, wherein:
the image acquisition module is used for acquiring license plate images and license plate types of the target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
the segmentation module is used for segmenting the license plate image according to the gray jump information in the license plate image to obtain a character area of the license plate image;
the identification module is used for acquiring a first area, above which the character area is located, and a second area, below which the character area is located, when the license plate type is a non-single-layer license plate; under the condition that the character type of the first area is a configurable character, matching registration information for the first area according to the license plate type of the target license plate, and inputting the second area into a preset number recognition model to obtain a license plate number;
The result acquisition module is used for acquiring the recognition result of the target license plate according to the registration information and the license plate number.
In a third aspect, in this embodiment there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method of the first aspect described above when the computer program is executed.
In a fourth aspect, in this embodiment there is provided a storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method of the first aspect described above.
According to the method, the device, the computer equipment and the storage medium for identifying the non-motor vehicle license plate, the license plate image and the license plate type of the target license plate are obtained, the license plate image is segmented according to the gray jump information in the license plate image, the character area of the obtained license plate image is obtained, when the license plate type is a non-single-layer license plate, a first area with the character area located above and a second area with the character area located below are obtained, the character type of the first area is obtained, when the character type of the first area is a configurable character, registration information of the target license plate is matched for the first area according to the license plate type of the target license plate, the second area is input into a preset number identification model to obtain the number of the target license plate, and finally, the identification result of the target license plate is obtained according to the registration information and the license plate number, so that the universality of non-motor vehicle license plate identification of different areas can be increased, and the accuracy of the non-motor vehicle license plate identification can be improved.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the other features, objects, and advantages of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
FIG. 1 is an application scenario diagram of a non-motor vehicle license plate recognition method according to an embodiment of the present application;
FIG. 2 is a flowchart one of a method for identifying a license plate of a non-motor vehicle according to an embodiment of the present application;
FIG. 3a is a schematic illustration of a license plate of a non-motor vehicle according to an embodiment of the present application;
FIG. 3b is a schematic illustration of another non-motor vehicle license plate according to an embodiment of the present application;
FIG. 3c is a schematic illustration of yet another non-motor vehicle license plate according to an embodiment of the present application;
FIG. 4a is a schematic illustration of yet another non-motor vehicle license plate according to an embodiment of the present application;
FIG. 4b is a schematic illustration of yet another non-motor vehicle license plate according to an embodiment of the present application;
FIG. 4c is a schematic illustration of yet another non-motor vehicle license plate according to an embodiment of the present application;
FIG. 5 is a second flowchart of a method for identifying a license plate of a non-motor vehicle according to an embodiment of the present application;
FIG. 6 is a schematic structural view of a license plate recognition device for a non-motor vehicle according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a computer device according to an embodiment of the present application.
Detailed Description
For a clearer understanding of the objects, technical solutions and advantages of the present application, the present application is described and illustrated below with reference to the accompanying drawings and examples.
Unless defined otherwise, technical or scientific terms used herein shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terms "a," "an," "the," "these," and the like in this application are not intended to be limiting in number, but rather are singular or plural. The terms "comprising," "including," "having," and any variations thereof, as used in the present application, are intended to cover a non-exclusive inclusion; for example, a process, method, and system, article, or apparatus that comprises a list of steps or modules (units) is not limited to the list of steps or modules (units), but may include other steps or modules (units) not listed or inherent to such process, method, article, or apparatus. The terms "connected," "coupled," and the like in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Reference to "a plurality" in this application means two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., "a and/or B" may mean: a exists alone, A and B exist together, and B exists alone. Typically, the character "/" indicates that the associated object is an "or" relationship. The terms "first," "second," "third," and the like, as referred to in this application, merely distinguish similar objects and do not represent a particular ordering of objects.
Fig. 1 is an application scenario diagram of a non-motor vehicle license plate recognition method in one embodiment. As shown in fig. 1, data transmission between the server 101 and the monitor terminal 102 can be performed through a network. The monitoring terminal 102 is configured to acquire image information of a non-motor vehicle, and transmit the image information to the server 101. After the server 101 receives the image information of the non-motor vehicle, a license plate image and a license plate type of a target license plate are obtained from the image information of the non-motor vehicle, the license plate image is segmented according to gray jump information in the license plate image to obtain a character area of the license plate image, and finally the license plate of the non-motor vehicle is identified according to the character area, so that license plate information of the non-motor vehicle is obtained. The server 101 may be implemented by a stand-alone server or a server cluster formed by a plurality of servers, and the monitoring terminal 102 may be any kind of camera.
In this embodiment, a method for identifying a license plate of a non-motor vehicle is provided, fig. 2 is a flowchart of a remote access method of this embodiment, and as shown in fig. 2, the flowchart includes the following steps:
step S210, obtaining a license plate image and license plate types of a target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates.
Because of the diversity of license plate systems of the non-motor vehicles, the same license plate recognition scheme of the non-motor vehicles cannot be shared. And when the license plate of the non-motor vehicle is identified, other character information on the license plate can be utilized to further obtain other license plate information of the license plate, such as the region to which the license plate belongs, besides the license plate number of the license plate of the non-motor vehicle, so that the identification of the license plate of the non-motor vehicle is perfected. Therefore, the license plate content format in the target license plate can be targeted to classify the non-motor license plate, and the corresponding license plate recognition scheme is selected for the non-motor license plate according to different license plate types.
Specifically, the license plate image can be obtained after the target non-motor vehicle image is obtained through image processing, or can be directly provided by other processing equipment. When the target non-motor vehicle image is processed, the non-motor vehicle license plate coordinate and license plate type analysis can be performed by using a non-motor vehicle license plate classification model. The license plate region can be obtained from the target non-motor vehicle image through the license plate coordinates, and the non-motor vehicle license plate can be determined to belong to a single-layer license plate or a non-single-layer license plate through the license plate type. The single-layer license plate is a license plate containing single-layer characters, the non-single-layer license plate is a license plate containing multi-layer characters, and the non-motor vehicle license plate detection model can be any one of a SSD (Single Shot MultiBox Detector) model, a YOLO (You Only Look Once) model and a fasterRCNN (faster Regions with CNN Features) detection model.
Further, a license plate image and a license plate type of the target license plate are obtained through a non-motor vehicle license plate detection model and are used for providing parameters for the identification of subsequent license plate information. Specifically, the single-layer license plate can be a license plate which is identified and does not contain characters and only contains a license plate number. The non-single-layer license plate can be a license plate containing characters and license plate numbers, wherein the characters in the non-single-layer license plate can comprise Chinese character and letter information of provinces, cities and administrative areas to which the non-motor belongs, and can also comprise license plate additional information which does not need to be identified, such as anti-theft record numbers and XX electric bicycles. After the license plate type of the target license plate is obtained, a corresponding recognition scheme can be selected for the target license plate according to the license plate type so as to obtain license plate information. For example, for a single-layer license plate, the character area of the license plate can be directly identified, while for a non-single-layer license plate, the character part and the license plate number part in the license plate need to be respectively identified.
Step S220, dividing the license plate image according to the gray jump information in the license plate image to obtain a character area of the license plate image.
In order to improve the robustness of the non-motor vehicle license plate recognition, the acquired license plate image of the target license plate is subjected to image preprocessing, so that the noise interference is reduced. The license plate for the acquired target can be a graying-processed image. In the license plate image after the graying treatment, the gray level change exists in the character area and the non-character area, so that gray level jump information of the license plate image can be obtained, and the license plate image is segmented by utilizing the gray level jump information, so that the character area of the license plate image is obtained. Specifically, the gray jump information may include a peak value and a trough value of a pixel gray value in the license plate image and a number of jumps between the peak value and the trough value, and the gray jump information may determine a segmentation threshold of the license plate image, so that a character region given to the license plate image is obtained by using the segmentation threshold.
Further, the segmentation threshold may be a gray value, and the single-layer license plate and the non-single-layer license plate may be segmented by using the segmentation threshold to obtain the character region. When the single-layer license plate is segmented, the single-layer license plate is subjected to projection segmentation and connected domain segmentation by utilizing a segmentation threshold value, so that a character region is obtained. And obtaining the character area after obtaining the boundary of the character area. In addition, when the non-single-layer license plate is divided, the lower layer region of the non-single-layer license plate including the license plate number and the upper layer region of other text information can be divided respectively. Similarly, the segmentation threshold is utilized to carry out projection segmentation and connected domain segmentation so as to obtain the character boundary of the lower layer region and the character region of the upper layer region of the non-single-layer license plate
By acquiring the character area of the license plate image, the efficiency of acquiring the license plate information of the non-motor vehicle can be improved. The character area of the non-single-layer license plate is segmented, so that the license plate number area and other text information in the non-single-layer license plate can be conveniently and respectively identified, and the accuracy of license plate identification is improved.
Step S230, when the license plate type is a non-single-layer license plate, acquiring a first area with a character area positioned above and a second area positioned below; acquiring character types of a first area, matching registration information of a target license plate for the first area according to license plate type of the target license plate under the condition that the character types of the first area are configurable characters, and inputting a preset number recognition model into a second area to obtain the license plate number of the target license plate.
The first area is other text information, and may be registration information of a license plate. The second area is a license plate number. Because registration information of the license plate is usually text, the license plate number is a combination of numbers and letters, and font formats in the two areas are different, different identification can be carried out on the two areas.
Specifically, for the second area including the license plate number information, the identification can be directly performed by using a preset number identification model. The number recognition model may be any text recognition model, such as CRNN (Convolutional Recurrent Neural Network) recognition model. Alternatively, for the first region including the license plate registration information, after the character type of the first region is acquired, corresponding processing may be performed according to the character type of the first region. Wherein the character category of the first region may include recognizable characters, configurable characters, and other characters. Specifically, the recognizable character may be registration information composed of provincial abbreviations and letters, and as shown in fig. 3a, the information may be classified into the recognizable character because the content of the text contained in the information is small, and the recognizable character is directly recognized by using a recognition model. The configurable characters may be registration information composed of place names, as shown in fig. 3b, the registration information contains full text information, and text fonts and spaces in different areas are different, so that the configurable characters are not suitable for directly identifying the characters, and the characters can be configured by means of external configuration information. Additionally, fig. 3c is a non-single-layer license plate with other characters in the character category, as shown in fig. 3c, the other characters are characters in the license plate which do not need to be recognized, i.e. the characters do not represent the identity information of the non-motor vehicle, so that the characters do not need to be recognized.
Further, since the target license plate is designed according to a certain license plate system, registration information contained in the first area of the target license plate can be determined through the license plate system. Therefore, when the character type of the first area is the configurable character, the license plate type predetermined by the target license plate can be obtained. The license plate type can be the license plate type of the non-motor vehicle in the area where the non-motor vehicle picture belongs. For example, when the illegal snapshot is performed on the non-motor vehicle in the area a, the license plate type of the non-motor vehicle in the area a can be correspondingly obtained, so that the registration information of the corresponding target license plate is matched for the first area according to the license plate type.
According to the registration information of the license plate of the first region matching target by using the license plate type, a mode of character recognition on the first region is replaced, training of a recognition model and updating of a character library are not needed for characters in the first region, meanwhile, recognition errors caused by pixel problems can be reduced, and therefore accuracy and efficiency of acquiring registration information of the license plate can be improved.
Step S240, according to the registration information and the license plate number, the recognition result of the target license plate is obtained.
In the steps S210 to S240, the license plate image and the license plate type of the target license plate are obtained, the license plate image is segmented according to the gray jump information in the license plate image, the character region of the obtained license plate image is obtained, when the license plate type is a non-single-layer license plate, the first region of the character region located above and the second region of the character region located below are obtained, under the condition that the character type of the first region is a configurable character, the registration information of the target license plate is matched for the first region according to the license plate standard type of the target license plate, the second region is input into a preset number identification model to obtain the license plate number of the target license plate, and finally the identification result of the target license plate is obtained according to the registration information and the license plate number, so that the universality of the non-motor vehicle license plate identification of different regions can be increased, and the accuracy and the efficiency of the non-motor vehicle license plate identification can be improved.
Further, based on the step S210, a license plate image and a license plate type of the target license plate are obtained, including the following steps:
in step S211, a target non-motor vehicle image is acquired.
Specifically, the target non-motor vehicle image may be obtained by video capturing by a monitoring device provided near the non-motor vehicle lane.
Step S212, inputting the target non-motor vehicle image into a preset detection model to obtain a license plate image and a license plate type.
For the target non-motor vehicle image, the license plate image contains the identity information of the target non-motor vehicle, and the license plate image occupies only a part of the area in the target non-motor vehicle image. Therefore, after the target non-motor vehicle image is acquired, processing is further required to acquire the license plate image of the target non-motor vehicle. Specifically, the detection model may be any image detection model, such as an SSD model, a YOLO model, and a fastercnn model. The coordinate of the license plate image in the target non-motor vehicle image can be obtained through the detection model, and the license plate type can also be obtained.
Further, in one embodiment, after obtaining the license plate image and the license plate type of the target license plate, before dividing the license plate image according to the gray jump information in the license plate image to obtain the character area of the obtained license plate image, the method comprises the following steps:
step S213, gray scale processing is performed on the license plate image.
Step S214, inputting the grey license plate image into a preset correction model, and correcting the shape and angle of the license plate image.
Due to the movement state of the non-motor vehicle itself, the license plate may have a plurality of different forms, such as a tilt at a certain angle, in the non-motor vehicle image captured by the monitoring device. Therefore, in order to improve the accuracy and efficiency of the subsequent recognition of the license plate image, the license plate image can be corrected by using a correction model before recognition. The correction model may be an STN model based on deep learning, or may be a method for processing problems such as image distortion and angle inclination in the field of image processing, which is not limited herein.
Additionally, in one embodiment, based on the step S220, the license plate image is segmented according to the gray jump information in the license plate image, so as to obtain a character area end of the license plate image, which includes the following steps:
step S221, obtaining the gray wave peak value, the gray wave trough value and the jump times of the character area of the license plate image.
Before the gray wave peak value, the gray wave trough value and the jump times are obtained, the license plate image can be cut in advance based on priori knowledge. For example, for a non-single-layer license plate, the upper layer region often contains only text characters, and the lower layer region contains a license plate number, so that when the license plate number of the lower layer region is divided, the corrected license plate image can be cut according to 1/2 of the height of the license plate image, and the cut lower half image is selected for processing.
Because the character area and the background area in the license plate image after the graying treatment have obvious difference in gray value, the gray jump information in the license plate image can be utilized to determine the segmentation threshold. Specifically, the gray-scale peak value may be a peak value generated during gray-scale jump in a middle row of pixels in the clipping region, and the gray-scale trough value is a trough value generated during gray-scale jump in a middle row of pixels in the clipping region. The peak and valley values may be recorded as a set of arrays. The gray level jump number can be determined by a preset jump threshold, namely when the difference value between the peak value and the trough value in a certain section of curve is higher than the preset jump threshold, the gray level change from the peak to the trough in the section of curve is recorded as one jump, otherwise, no jump exists in the section of curve. In addition, a hopping frequency threshold value can be preset, and when the hopping frequency calculated under the preset hopping frequency threshold value is smaller than the hopping frequency threshold value, a hopping threshold value is additionally set to determine the set of peak values, trough values and hopping frequency.
Further, a sequence of gray level differences between the peak and trough values, e.g., a descending order of gray level differences, may be recorded below the predetermined transition threshold. And acquiring the smallest gray level difference value in the sequence, for example, the last bit in descending order, and updating the set of peak values and trough values according to the magnitude relation between the jump threshold value and the smallest gray level difference value. For example, when the jump threshold is greater than 3/4 times the minimum gray level difference, the subsequent segmentation threshold is calculated using the determined peak value, trough value and number of jumps below the jump threshold. Otherwise, taking 3/4 times of the minimum gray level difference value as a new jump threshold value, and updating the set of peak values and trough values by using the new jump threshold value.
For example, under a preset jump threshold T, a set of determined valley values is a set of grayA, a set of peak values is a set of grayB, and the number of jumps is JmpNum. The pixel X-axis coordinates corresponding to each trough constitute an array idxA, and the pixel X-axis coordinates corresponding to each peak constitute an array idxB. The sequence obtained after descending order of the peak-trough difference values is a group MaxDIff, if the length of the group is 7, the last bit is MaxDIff [6]. And if the T_new is smaller than T, continuing to calculate a subsequent segmentation threshold, otherwise, taking the T_new as a new jump threshold T, and updating the set of peak values, trough values and jump times until the current T_new is smaller than the current T.
Step S222, according to the gray wave peak value, the gray wave trough value and the jump times, the gray adjacent combination is obtained.
Specifically, a current hopping index is preset to JmpIdx, and the initial value of the current hopping index is set to 0, i.e., jmpidx=0. A hopping index k is predetermined according to the hopping number JmpNum. For example, when the number of hops JmpNum is 10, k is set to 4. And obtaining an array diffA= |grada [ JmpIdx ] -grada [ k ] | formed by the difference value of the trough and the trough between the two jumps. And obtaining an array diffB= |gradb [ JmpIdx ] -gradb [ k ] | formed by the difference value between the peaks of the two hops. And obtaining the difference cheight=grayB [ JmpIdx ] -grayA [ JmpIdx ] of the current jump peak and the current jump trough. The difference value nhight=grayb [ k ] -grayA [ k ] of the next jump peak and trough is obtained, and the difference value diffidxb=idxb [ k ] -idxB [ JmpIdx ] of the coordinates between the two peaks. Taking the larger of diffA and diffB as diffMax, taking the smaller of chight and nfight as HeightMin, setting a fixed comparison threshold, for example, the comparison threshold is 50, if diffMax is 4< = HeightMin and diffIdxB < = 50, determining the index of the gray adjacent combination corresponding to the current jump index as next friend = jmpidx+k, that is, when the current jump index is 0, the index of the gray adjacent combination is 4. Let k=k-1 later, if diffMax 4< =heightmin and diffIdxB < =50 still holds at this time, update the next friend until k <1. When the index of the gray adjacent combination corresponding to the JmpIdx is determined to be 0, the jmpidx=jmpidx+1 is made, and the index of the gray adjacent combination corresponding to the JmpIdx at this time is calculated by using the same step until jmpidx=jmpnum.
After a group of current jump indexes JMPIdx and corresponding gray adjacent combination indexes next friend are obtained, respectively calculating the corresponding combination numbers slopcnt under the jump times JMPIdx from 0 to JMPNum. Table 1 shows a set of the relationships between JMPIdx and nextFriend, and the table 1 is taken as an example to determine the number of the combinations slopeCNt corresponding to each JMPIdx.
TABLE 1
| JmpIdx | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| nextFriend | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 |
As shown in table 1, the initial value of the slopcnt corresponding to each JmpIdx is first set to 0. When JmpIdx is 0, it is determined whether or not the corresponding next friend is 0, and if not, the sloppcnt=sloppcnt+1. And continuously judging whether the next friend corresponding to the next friend is 0, namely when the jmphdx is 0, the corresponding next friend is 1, the slopcnt=0+1, and when the next friend is 1, the corresponding next friend is 2, the slopcnt=0+1+1 is sequentially accumulated until the corresponding next friend is 0. It can thus be determined from Table 1 that when JMPIdx is 0, it corresponds to a slopeCNt of 9. Finally, based on table 1, the corresponding relation between all current jump indexes and the number of combinations can be obtained as shown in table 2. And finding out a group of combinations with the largest number of combinations in the corresponding relation, namely, the JmpIdx is 0, and the sloppC is 9.
TABLE 2
| JmpIdx | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| slopeCnt | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
The value 9 with the largest number of combinations at this time is taken as the new number of hops Num for this combination. Judging whether the new jump number Num is larger than a preset minimum jump frame number N, if so, calculating a segmentation threshold at the moment according to a formula (1). Otherwise, let n=n-1 and recalculate the new segmentation threshold according to the procedure described above.
Step S223, determining a segmentation threshold of the license plate image by using the gray level adjacent combination, and segmenting the license plate image according to the segmentation threshold.
Through the steps S221 to S223, the accuracy of segmentation can be improved under the conditions of unclear license plate and low contrast caused by the existence of interference factors, such as rainwater or muddy water shielding, in the target license plate, so that the robustness of character region segmentation is increased.
Additionally, in one embodiment, based on the step S230, the matching of the registration information of the target license plate for the first area according to the license plate format type of the target license plate includes the following steps:
step S231, if the license plate type is one, acquiring preset license plate configuration information, and obtaining registration information from the license plate configuration information.
For example, when the license plate system of the area from which the target non-motor vehicle picture originates is one type, for example, in all the non-motor vehicle license plates in the area a, other license plate registration information except the license plate number is "area a", license plate configuration information preset in the area a, that is, a character string with content of "area a", is directly obtained, and the content of the character string is used as registration information corresponding to the first area of the target license plate.
Additionally, in one embodiment, based on the step S230, the registration information of the target license plate is matched for the first area according to the license plate format type of the target license plate, and the method further includes the following steps:
step S232, if the type of license plate system is more than one type, inputting the first region into a first classification model with complete training, and determining the license plate system corresponding to the target license plate.
The license plate configuration information can be license plate systems correspondingly configured according to different regions, and comprises registration information corresponding to various license plate systems in the regions. When the number plate system is more than one type, for example, three number plate systems with different character contents of administrative district C, administrative district D and administrative district E exist under city B. For example, the license plate of the non-motor vehicle in Taizhou city includes various types, and as shown in fig. 4a, 4b and 4c, the license plate of the non-motor vehicle in Taizhou city includes three types, namely Wen Ling, spiced river and jade ring, so that the types of the license plate of the non-motor vehicle in the area can be set to be three. When the number plate type is more than one, inputting the first area into a first classification model, and determining which of the three character contents the character content contained in the first area belongs to, so as to replace direct recognition of the character content in the first area. The first classification model may specifically be any classification model based on machine learning. Additionally, before classifying by using the first classification model, the first classification model may be trained by using the first regions of the plurality of non-single-layer license plates as a training set to obtain a first classification model with complete training.
Step S233, obtaining preset license plate configuration information, and matching corresponding registration information for the first area according to the mapping relation between license plate system and license plate configuration information.
After the license plate system corresponding to the first area is determined, registration information corresponding to the license plate system can be obtained from the pre-configured license plate configuration information.
In one embodiment, when the license plate type is a non-single layer license plate and the character type of the first area is a recognizable character, the method further comprises the following steps:
step S250, after the first area and the second area are spliced, a preset license plate recognition model is input, and a recognition result of the target license plate is obtained.
For example, if the character content of the first area is a combination of a letter of a province abbreviation and a letter of a city administrative level, the character content can be directly identified through an identification model in the image processing field or the deep learning field, so that after the first area is spliced with a second area containing a license plate number, the first area is input into a license plate identification model, and an identification result of the target license plate is determined. Specifically, the license plate recognition model may be a pre-trained CRNN model.
In one embodiment, when the license plate type is a single-layer license plate, the method further comprises the following steps:
Step S260, inputting the character area into a number recognition model to obtain a recognition result of the target license plate.
In particular, the numbered recognition model may be a pre-trained CRNN model.
In one embodiment, based on the step S230, the obtaining the character class of the first area specifically includes the following steps:
s231, inputting the first region into a complete training second classification model to obtain the character class of the first region.
In particular, due to the diversity of non-motorized license plates, for non-single layer license plates, the character category of the first region may contain multiple types. For example, the first area of the license plate of the non-motor vehicle in the first area is registered information composed of the province abbreviation and the letters of the first area, and the information contains less text content, so that the identification of the detection model is facilitated, and the information can be attributed to the identifiable characters. The first area of the license plate of the non-motor vehicle in the second area is a full text character area formed by the place names of the second area, and the condition that the character recognition is performed by the input detection model can cause the condition of low recognition efficiency, so that the characters can be configured by means of external configuration information. Additionally, a portion of the other characters contained in the first region of the non-motor vehicle license plate that may not contain key information for license plate registration, and therefore need not be identified.
In order to determine the character class to which the first region belongs, detection may be performed using a second classification model. The second classification model may be any one of the classification models of Alexnet, vggnet, resnet, which is not specifically limited in this embodiment.
Further, in one embodiment, as shown in fig. 5, there is provided a second flowchart of a method for identifying a license plate of a non-motor vehicle, including the steps of:
and step S10, acquiring a non-motor vehicle picture.
And S11, detecting license plates of the non-motor vehicles by using the detection model to obtain license plate images and license plate types.
And S12, correcting the license plate image.
And step S20, character region segmentation is carried out according to the gray level jump information.
Step S30, judging whether the license plate type is a single-layer license plate.
Step S31, if the license plate is a single-layer license plate, the character area is segmented by utilizing the texture boundary of the single-layer license plate.
In step S32, if the license plate is a non-single-layer license plate, a first area located above the non-single-layer license plate and a second area located below the non-single-layer license plate are respectively segmented.
Step S321, the first area is sent to the second classification model to judge the character type.
In step S322, if the character type of the first region is recognizable character of the combination of province abbreviation and letter, the first region and the second region are spliced.
In step S323, if the character type of the first region is other characters that do not need to be recognized, only the second region is selected.
In step S324, if the first region is a configurable character of the full text, the first region and the second region are selected respectively.
Step S325, judging whether the license plate system is one.
Step S3251, if the license plate system to which the target license plate belongs is one, acquiring registration information corresponding to the first area by using license plate configuration information.
Step S3252, if the license plate system of the target license plate is at least two, classifying the first region by using a first classification model, and matching the registration information in the license plate configuration information by using the license plate system corresponding to the first region.
In step S326, a second region is selected.
Step S40, the selected area is identified by using the CRNN identification model.
In the steps S210 to S260, the target non-motor vehicle image is processed by using the preset detection model to obtain the license plate image and the license plate type, and the license plate image is subjected to gray processing and correction processing, so that the accuracy of the subsequent license plate image recognition can be improved, the gray adjacent combination is determined according to the gray peak value, the gray trough value and the jump number of the license plate image, the segmentation threshold of the license plate image is determined by using the gray adjacent combination, the accuracy of the segmentation of the character area in the license plate image is improved, the influence of the interference factor of the license plate image is reduced, when the license plate type is a non-single-layer license plate, the registration information of the first area is determined according to the preset license plate configuration information, when the character type of the first area is a recognizable character, the first area and the second area are recognized by using the license plate recognition model, and when the type of the license plate is a single-layer license plate, the target license plate is recognized by using the number recognition model, the problem of low recognition model training efficiency and recognition accuracy caused by containing more characters in the target license plate is avoided, and the different license plate recognition modes can be selected according to different types, so that the universal license plate recognition of different vehicles can be processed, and the non-vehicle license plate type is increased.
In this embodiment, a license plate recognition device for a non-motor vehicle is further provided, and the device is used for implementing the foregoing embodiments and preferred embodiments, and is not described again. The terms "module," "unit," "sub-unit," and the like as used below may refer to a combination of software and/or hardware that performs a predetermined function. While the means described in the following embodiments are preferably implemented in software, implementations in hardware, or a combination of software and hardware, are also possible and contemplated.
Fig. 6 is a schematic structural diagram of a non-motor vehicle license plate recognition device 60 in the embodiment of the present application, and as shown in fig. 6, the non-motor vehicle license plate recognition device 60 includes: an image acquisition module 62, a segmentation module 64, an identification module 66, and a result acquisition module 68, wherein:
the image acquisition module 62 is used for acquiring a license plate image and license plate types of the target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
the segmentation module 64 is used for segmenting the license plate image according to the gray jump information in the license plate image to obtain a character area of the license plate image;
the recognition module 66 is configured to obtain a first area with a character area located above and a second area located below when the license plate type is a non-single-layer license plate; acquiring the character type of the first area, matching registration information for the first area according to the license plate type of the target license plate under the condition that the character type of the first area is a configurable character, and inputting a preset number recognition model into the second area to obtain a license plate number;
The result obtaining module 68 is configured to obtain a recognition result of the target license plate according to the registration information and the license plate number.
The above-mentioned non-motor vehicle license plate recognition device 60, through obtaining the license plate image and license plate type of the target license plate, cut apart the license plate image according to the gray jump information in the license plate image, obtain the character area of the license plate image, and when the license plate type is the non-monolayer license plate, obtain the first area that the character area is located above and the second area that is located below, obtain the character type of the first area, in the condition that the character type of the first area is the configurable character, according to the license plate standard type of the target license plate, match the registration information of the target license plate for the first area, and input the second area into the preset number recognition model to obtain the license plate number of the target license plate, finally according to registration information and license plate number, obtain the recognition result of the target license plate, thus can increase the commonality of the non-motor vehicle license plate recognition to different areas, and improve the accuracy and efficiency of the non-motor vehicle license plate recognition.
In one embodiment, the image obtaining module 62 is further configured to obtain a target non-motor vehicle image, and input the target non-motor vehicle image into a preset detection model to obtain a license plate image and a license plate type.
In one embodiment, the non-motor vehicle license plate recognition device 60 further includes a correction module, where the correction module is configured to perform grayscale processing on the license plate image, input the grayscale license plate image into a preset correction model, and correct the shape and angle of the license plate image.
In one embodiment, the segmentation module 64 is further configured to obtain a gray-scale peak value, a gray-scale trough value, and a number of hops of the character area of the license plate image, obtain a gray-scale adjacent combination according to the gray-scale peak value, the gray-scale trough value, and the number of hops, determine a segmentation threshold of the license plate image by using the gray-scale adjacent combination, and segment the license plate image according to the segmentation threshold.
In one embodiment, the identification module 66 is further configured to obtain preset license plate configuration information when the license plate system type is one, and obtain registration information from the license plate configuration information.
In one embodiment, the identification module 66 is further configured to input a first classification model with complete training into the first area when the number of license plate types is greater than one, determine the number of license plates corresponding to the target number of license plates, obtain preset number of license plates configuration information, and match corresponding registration information for the first area according to a mapping relationship between the number of license plates and the number of license plates configuration information.
In one embodiment, the recognition module 66 is further configured to splice the first area and the second area, and input a preset license plate recognition model to obtain a recognition result of the target license plate.
In one embodiment, the recognition module 66 is further configured to input the character area into a number recognition model to obtain a recognition result of the target license plate.
In one embodiment, the recognition model 66 is further configured to input the first region into a trained second classification model to obtain a character class for the first region.
The above-described respective modules may be functional modules or program modules, and may be implemented by software or hardware. For modules implemented in hardware, the various modules described above may be located in the same processor; or the above modules may be located in different processors in any combination.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing a preset configuration information set. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by the processor to implement the above-described non-motor vehicle license plate recognition method.
In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method for identifying a license plate of a non-motor vehicle. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those skilled in the art that the structure shown in fig. 7 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer readable storage medium is provided having a computer program stored thereon, which when executed by a processor, performs the steps of:
obtaining a license plate image and license plate types of a target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
dividing the license plate image according to the gray jump information in the license plate image to obtain a character area of the license plate image;
when the license plate type is a non-single-layer license plate, a first area with a character area positioned above and a second area positioned below are obtained; acquiring the character type of the first area, matching registration information of the target license plate for the first area according to the license plate type of the target license plate under the condition that the character type of the first area is configurable characters, and inputting the second area into a preset number recognition model to obtain the license plate number of the target license plate;
and obtaining the recognition result of the target license plate according to the registration information and the license plate number.
In one embodiment, the processor when executing the computer program further performs the steps of:
acquiring a target non-motor vehicle image;
and inputting the target non-motor vehicle image into a preset detection model to obtain a license plate image and a license plate type.
In one embodiment, the processor when executing the computer program further performs the steps of:
carrying out graying treatment on the license plate image;
and inputting the grey license plate image into a preset correction model, and correcting the shape and angle of the license plate image.
In one embodiment, the processor when executing the computer program further performs the steps of:
acquiring gray level wave peak values, gray level wave trough values and jump times of a character area of a license plate image;
obtaining gray adjacent combinations according to gray wave peak values, gray wave trough values and jump times;
and determining a segmentation threshold value of the license plate image by using the gray adjacent combination, and segmenting the license plate image according to the segmentation threshold value.
In one embodiment, the processor when executing the computer program further performs the steps of:
if the type of license plate system is one, acquiring preset license plate configuration information, and acquiring registration information from the license plate configuration information.
In one embodiment, the processor when executing the computer program further performs the steps of:
if the types of license plate systems are more than one type, inputting a first region into a first classification model with complete training, and determining a license plate system corresponding to a target license plate;
Acquiring preset license plate configuration information, and matching corresponding registration information for the first area according to the mapping relation between license plate system and license plate configuration information.
In one embodiment, the processor when executing the computer program further performs the steps of:
and after the first area and the second area are spliced, inputting a preset license plate recognition model to obtain a recognition result of the target license plate.
In one embodiment, the processor when executing the computer program further performs the steps of:
and inputting the character area into a number recognition model to obtain a recognition result of the target license plate.
In one embodiment, the processor when executing the computer program further performs the steps of:
and inputting the first region into a trained second classification model to obtain the character class of the first region.
According to the storage medium, the license plate image and the license plate type of the target license plate are obtained, the license plate image is segmented according to the gray jump information in the license plate image, the character area of the obtained license plate image is obtained, when the license plate type is a non-single-layer license plate, a first area of which the character area is located above and a second area of which the character area is located below are obtained, the character type of the first area is obtained, under the condition that the character type of the first area is a configurable character, registration information of the target license plate is matched for the first area according to the license plate standard type of the target license plate, the second area is input into a preset number recognition model to obtain the number of the target license plate, and finally, the recognition result of the target license plate is obtained according to the registration information and the number of the license plate, so that the universality of non-motor vehicle license plate recognition for different areas can be increased, and the accuracy and the recognition efficiency of the non-motor vehicle license plate can be improved.
It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to be limiting. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present application, are within the scope of the present application in light of the embodiments provided herein.
It is evident that the drawings are only examples or embodiments of the present application, from which the present application can also be adapted to other similar situations by a person skilled in the art without the inventive effort. In addition, it should be appreciated that while the development effort might be complex and lengthy, it would nevertheless be a routine undertaking of design, fabrication, or manufacture for those of ordinary skill having the benefit of this disclosure, and thus should not be construed as an admission of insufficient detail.
The term "embodiment" in this application means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive. It will be clear or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.
The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the patent. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.
Claims (11)
1. A method for identifying a license plate of a non-motor vehicle, comprising the steps of:
obtaining a license plate image and license plate types of a target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
dividing the license plate image according to gray jump information in the license plate image to obtain a character area of the license plate image;
when the license plate type is a non-single-layer license plate, acquiring a first area of the character area, which is positioned above, and a second area of the character area, which is positioned below; acquiring the character type of the first area, and if the character type of the first area is a configurable character, inputting the first area into a first classification model which is complete in training if the type of the license plate system of the target license plate is more than one type, and determining the license plate system corresponding to the target license plate; wherein the configurable character is registration information composed of place names;
Acquiring preset license plate configuration information, matching corresponding registration information for the first area according to the mapping relation between the license plate system and the license plate configuration information, and inputting the second area into a preset number identification model to obtain the license plate number of the target license plate;
and obtaining the recognition result of the target license plate according to the registration information and the license plate number.
2. The method of claim 1, wherein the obtaining a license plate image and license plate type of the target license plate comprises:
acquiring a target non-motor vehicle image;
and inputting the target non-motor vehicle image into a preset detection model to obtain the license plate image and the license plate type.
3. The method of claim 1, wherein after obtaining the license plate image and the license plate type of the target license plate, before dividing the license plate image according to the gray jump information in the license plate image to obtain the character area of the license plate image, further comprising:
carrying out graying treatment on the license plate image;
and inputting the grey license plate image into a preset correction model, and correcting the shape and angle of the license plate image.
4. The method of claim 1, wherein the dividing the license plate image according to the gray jump information in the license plate image to obtain the character area of the license plate image comprises:
acquiring gray level wave peak values, gray level wave trough values and jump times of a character area of the license plate image;
obtaining gray adjacent combinations according to the gray wave peak value, the gray wave trough value and the jump times;
and determining a segmentation threshold value of the license plate image by using the gray adjacent combination, and segmenting the license plate image according to the segmentation threshold value.
5. The method according to claim 1, wherein the method further comprises:
if the license plate type is one type, acquiring preset license plate configuration information, and acquiring the registration information from the license plate configuration information.
6. The method of claim 1, wherein when the license plate type is a non-single layer license plate and the character class of the first region is a recognizable character, the method further comprises:
the first area and the second area are spliced and then input into a preset license plate recognition model, and a recognition result of the target license plate is obtained; wherein, the recognizable character is registration information composed of province abbreviation and letters.
7. The method of claim 1, wherein when the license plate type is a single layer license plate, the method further comprises:
and inputting the character area into the number recognition model to obtain a recognition result of the target license plate.
8. The method of claim 1, wherein the obtaining the character class of the first region comprises:
and inputting the first region into a complete training second classification model to obtain the character category of the first region.
9. A non-motor vehicle license plate recognition device, comprising: the device comprises an image acquisition module, a segmentation module, an identification module and a result acquisition module, wherein:
the image acquisition module is used for acquiring license plate images and license plate types of the target license plate, wherein the license plate types comprise single-layer license plates and non-single-layer license plates;
the segmentation module is used for segmenting the license plate image according to the gray jump information in the license plate image to obtain a character area of the license plate image;
the identification module is used for acquiring a first area, above which the character area is located, and a second area, below which the character area is located, when the license plate type is a non-single-layer license plate; acquiring the character type of the first area, and if the character type of the first area is a configurable character, inputting the first area into a first classification model which is complete in training if the type of the license plate system of the target license plate is more than one type, and determining the license plate system corresponding to the target license plate; acquiring preset license plate configuration information, matching corresponding registration information for the first area according to the mapping relation between the license plate system and the license plate configuration information, and inputting the second area into a preset number identification model to obtain a license plate number; wherein the configurable character is registration information composed of place names;
The result acquisition module is used for acquiring the recognition result of the target license plate according to the registration information and the license plate number.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 8 when the computer program is executed by the processor.
11. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110540159.2A CN113392839B (en) | 2021-05-18 | 2021-05-18 | License plate recognition method and device for non-motor vehicle, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110540159.2A CN113392839B (en) | 2021-05-18 | 2021-05-18 | License plate recognition method and device for non-motor vehicle, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113392839A CN113392839A (en) | 2021-09-14 |
| CN113392839B true CN113392839B (en) | 2023-07-28 |
Family
ID=77617200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110540159.2A Active CN113392839B (en) | 2021-05-18 | 2021-05-18 | License plate recognition method and device for non-motor vehicle, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113392839B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114639092A (en) * | 2022-03-11 | 2022-06-17 | 阿里巴巴（中国）有限公司 | Image recognition method, device, equipment and storage medium |
| CN115542362A (en) * | 2022-12-01 | 2022-12-30 | 成都信息工程大学 | High-precision space positioning method, system, equipment and medium for electric power operation site |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9014432B2 (en) * | 2012-05-04 | 2015-04-21 | Xerox Corporation | License plate character segmentation using likelihood maximization |
| CN107729899B (en) * | 2016-08-11 | 2019-12-20 | 杭州海康威视数字技术股份有限公司 | License plate number recognition method and device |
| CN106384113A (en) * | 2016-11-07 | 2017-02-08 | 湖南源信光电科技有限公司 | Odd-and-even-line license plate character dividing method based on projection and template matching |
-
2021
- 2021-05-18 CN CN202110540159.2A patent/CN113392839B/en active Active
Non-Patent Citations (2)
| Title |
|---|
| license plate localization based on edge-geometrical features using morphological approach;Jinn-Li Tan；et al;2013 IEEE International Conference on Image Processing;4549-4553 * |
| 基于自适应阈值投影的双行车牌分割方法;马永杰，等;计算机工程与科学;第42卷(第9期);1616-1624 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113392839A (en) | 2021-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109740469B (en) | Lane line detection method, lane line detection device, computer device, and storage medium | |
| CN113392839B (en) | License plate recognition method and device for non-motor vehicle, computer equipment and storage medium | |
| Ntirogiannis et al. | Performance evaluation methodology for historical document image binarization | |
| CN107944450B (en) | License plate recognition method and device | |
| CN111027563A (en) | Text detection method, device and recognition system | |
| CN110020692B (en) | Handwriting separation and positioning method based on print template | |
| CN103577818A (en) | Method and device for recognizing image characters | |
| CN110866529A (en) | Character recognition method, character recognition device, electronic equipment and storage medium | |
| CN110852311A (en) | Three-dimensional human hand key point positioning method and device | |
| CN109858438B (en) | Lane line detection method based on model fitting | |
| CN111191649A (en) | Method and equipment for identifying bent multi-line text image | |
| CN103383732A (en) | Image processing method and device | |
| WO2023138538A1 (en) | Vehicle-mounted video image stabilization method and apparatus, vehicle and storage medium | |
| CN114038004A (en) | Certificate information extraction method, device, equipment and storage medium | |
| JP4128837B2 (en) | Road lane detection device | |
| CN111104941B (en) | Image direction correction method and device and electronic equipment | |
| CN110135428B (en) | Image segmentation processing method and device | |
| CN111898428A (en) | Unmanned aerial vehicle feature point matching method based on ORB | |
| CN114820679A (en) | Image annotation method and device, electronic equipment and storage medium | |
| CN110781195A (en) | System, method and device for updating point of interest information | |
| CN112560856B (en) | License plate detection and identification method, device, equipment and storage medium | |
| CN116704490B (en) | License plate recognition method, license plate recognition device and computer equipment | |
| CN110728276B (en) | License plate recognition method and device | |
| CN107944352B (en) | License plate recognition system and method based on picture retrieval method | |
| CN113344961B (en) | Image background segmentation method, device, computing 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |