CN104298976A - License plate detection method based on convolutional neural network - Google Patents
License plate detection method based on convolutional neural network Download PDFInfo
- Publication number
- CN104298976A CN104298976A CN201410549315.1A CN201410549315A CN104298976A CN 104298976 A CN104298976 A CN 104298976A CN 201410549315 A CN201410549315 A CN 201410549315A CN 104298976 A CN104298976 A CN 104298976A
- Authority
- CN
- China
- Prior art keywords
- car plate
- license plate
- convolutional neural
- neural networks
- candidate region
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Character Input (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a license plate detection method based on a convolutional neural network. The method specifically includes the steps that an Adaboost license plate detector based on Haar characteristics detects license plate images to be detected, license plate roughing regions are acquired, a convolutional neural network complete license plate recognition model recognizes the license plate roughing regions, a final license plate candidate region is acquired, the final license plate candidate region is segmented through a multi-threshold segmentation algorithm, license plate Chinese characters, letters and numbers are acquired, a Chinese character, letter and number convolutional neural network recognition model recognizes the license plate Chinese characters, letters and numbers, and then a license plate recognition result is acquired. License plate images under different conditions can be accurately recognized through the Adaboost license plate detector based on the Haar characteristics and the convolutional neural network complete license plate recognition model, meanwhile, characters are segmented through the multi-threshold segmentation algorithm, character images can be more easily and conveniently segmented, and the good effect is achieved in engineering application.
Description
Technical field
The invention belongs to Computer Vision Recognition technical field, particularly relate to a kind of detection method of license plate based on convolutional neural networks.
Background technology
Intelligent transportation system is as a part for smart city, digital city, be mainly used in magnitude of traffic flow monitoring, vehicle monitoring, freeway toll station management, cell intelligent management, parking lot management, traffic police etc., wherein car plate detects is the most key part of intelligent transportation system, being realize the intelligentized important step of traffic administration, is computer vision and the important subject of mode identification technology in intelligent transportation system.License plate recognition technology is passed by one's way at vehicle, to pass a bridge full-automatic non-parking charge, the measurement of magnitude of traffic flow Con trolling index, automatic vehicle identification, accident automatic telemetry on highway, onstream inspection, vehicle location, automobile burglar, inspection and tracking rule-breaking vehicle, illegal activities, safeguard traffic safety and urban public security, prevent traffic jam, improve the service rate of charge road and bridge, the aspect such as to relieve traffic density all has positive effect, main manifestations is: in social security and social management, intelligent transportation system is analyzed the video information transmitting monitoring scene from each crossing, obtain the information of vehicles in video, the running orbit of target vehicle is followed the tracks of, in charge station, parking lot etc., intelligent transportation system can realize extracting vehicle license plate, and then realizes the automatic management to vehicle, raises the efficiency and promptness, in high-grade community, intelligent transportation system can also manage the vehicle of community, identifies district vehicles information, for community increases chip safely.Complete license plate recognition technology comprises license plate image acquisition, license plate area location, License Plate Character Segmentation, Recognition of License Plate Characters etc., the car plate technology used both at home and abroad at present mainly comprises the localization methods such as the vehicle license location technique based on car plate textural characteristics, the vehicle license location technique based on car plate color characteristic, the vehicle license location technique based on car plate edge feature, the vehicle license location technique based on genetic algorithm, based on the cutting techniques such as Character segmentation technology of projecting method, based on the character recognition technologies etc. of character stroke.The main stream approach that car plate testing product uses both at home and abroad at present remains computer vision algorithms make and DSP technology combines.Due to the difference of domestic and international different regions car plate, so the portability of Related product is poor, so the car plate detection and indentification technology that exploitation is applicable to Continental Area is current problem urgently to be resolved hurrily.The car plate detection and indentification method that product on existing market adopts mainly faces three large problems: 1) when body color is close with car plate color, accurately can not determine license plate area; 2) what current Recognition of License Plate Characters adopted is comparatively loaded down with trivial details based on methods such as character strokes, and accuracy rate is not high.3) characters on license plate can not be split preferably when license plate image uneven illumination is even.The invention " license plate locating method " of the people such as the Deng Haifeng of Tencent Technology (Shenzhen) Co., Ltd. is applied for a patent to China national Department of Intellectual Property and gets the Green Light on May 7th, 2012, open on March 27th, 2013, publication number is: CN102999753A.A kind of license plate locating method of this disclosure of the invention, comprises the following steps: carry out color detection to the panorama sketch gathered, and extract designated color region as license plate candidate area; Extract the edge feature of license plate candidate area, utilize edge feature to split license plate candidate area, generate license plate candidate area collection; The license plate candidate area that license plate candidate area is concentrated is merged according to the merging condition preset, generates new license plate candidate area.Slant Rectify is carried out to the car plate in license plate candidate area; The candidate region that candidate region is concentrated is verified, rejects non-license plate area.This invention carries out car plate coarse positioning by utilizing the color characteristic of license plate area, then utilizes textural characteristics to screen to the car plate after coarse positioning, finally carries out correction process to car plate, obtain candidate license plate region.But in fact this invention has following three shortcomings: it is poor that 1) simple dependent color and texture space feature position anti-interference to car plate; 2) when using color space HSV to License Plate, to the bad assurance of Color characteristics parameters numerical value of being correlated with during License Plate under nonideality; 3) when carrying out textural characteristics screening to license plate area, utilize the transition times screening license plate area of every a line, when license plate area uneven illumination, not there is good effect.Can not better be adapted in this way the location of all license plate areas, when especially car plate color is close with body color, can not be well used in engineering reality.The invention " a kind of licence plate recognition method " of the people such as the fertile spark of South China Science & Engineering University is applied for a patent to China national Department of Intellectual Property and gets the Green Light on August 30th, 2012, and open on January 16th, 2013, publication number is: CN102880859A.A kind of licence plate recognition method of this disclosure of the invention, key step is as next time: obtain license plate image; Image semantic classification; License Plate; Car plate super-resolution rebuilding; Character segmentation; Character recognition.This invention achieve when not fogging clear accurately can identify car plate, can to the identification problem under car plate defect, ambiguity, and not by the restriction of image background, body color etc.But in fact this invention comparatively bothers when super-resolution rebuilding, on the other hand to the recognition methods based on character stroke that the identification of character adopts industry to commonly use, recognition methods based on stroke comparatively bothers, and when particularly, image not high to characters on license plate cut quality has a pollution, recognition effect can not ensure, such as in the identification of confusable character.
Summary of the invention
In order to overcome the above problems, the present invention proposes a kind of detection method of license plate based on convolutional neural networks, can realize easier, more accurately identifying the license plate image under various condition.
Content of the present invention for convenience of description, first makes an explanation to following term:
Term 1:Adaboost sorter
Adaboost sorter trains different Weak Classifiers for same training set, then these weak classifier set got up formation strong classifier.
Term 2:BP algorithm
BP algorithm is a kind of back-propagation algorithm, be divided into forward process, backward process two parts, wherein forward process refers to the process entering data in network and obtain net result, reverse procedure refers to using the difference of forward process and sample actual numerical value as error, to the process that network weight upgrades.
Term 3: multi-threshold segmentation algorithm
Multi-threshold segmentation algorithm may there is the problem such as shade, uneven illumination for license plate image, carries out binaryzation all characters can not be split preferably in binary image and propose a kind of multi-threshold segmentation algorithm with some specific threshold to license plate image.It occupies the percentage of license plate area as index using white point in license plate image after binaryzation, solve corresponding threshold value, then the ratio of white point is constantly reduced, and then by all character zone place UNICOM regional partition a kind of methods out, finally all UNICOM territories merged, sort and determine its position relationship, wherein to the UNICOM territory pixel number zero setting before next threshold value is split out of each Threshold segmentation.
Technical scheme of the present invention is: a kind of detection method of license plate based on convolutional neural networks, comprises the following steps:
S1. prepare the positive sample of car plate and negative sample, train the Adaboost car plate detecting device based on Haar feature;
S2. treat detected image and carry out gray processing process, and utilize the Adaboost car plate detecting device based on Haar feature in step S1 to carry out moving window detection, region roughly selected by the car plate obtained in image to be detected;
S3. the positive sample of car load board and negative sample is finished, the complete car plate model of cognition of training convolutional neural networks;
S4. utilize the complete car plate model of cognition of the convolutional neural networks in step S3 to roughly select region to the car plate in step S2 to screen, obtain the final candidate region of car plate, and preserve car plate final candidate region copy;
S5. utilize local horizontal projecting method to carry out the detection of horizontal tilt angle to the final candidate region of the car plate in step S4, and utilize rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy;
S6. the final candidate region of car plate after utilizing multi-threshold segmentation algorithm to correct horizontal tilt and car plate final candidate region copy process, and obtain characters on license plate correspondence image;
S7. Chinese character, letter and number sample is prepared, training Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition;
S8. utilize the Chinese character convolutional neural networks model of cognition in step S7, letter and number convolutional neural networks model of cognition carries out Recognition of License Plate Characters to characters on license plate correspondence image in step S6, and add up recognition result.
Further, described step S1 prepares the positive sample of car plate and negative sample, train the Adaboost car plate detecting device based on Haar feature, be specially: positive for the car plate of preparation sample and negative sample are normalized, generate the Adaboost car plate detecting device based on Haar feature; Wherein, the positive sample of car plate comprises the license plate grey level image sample under different illumination conditions, and car plate negative sample comprises non-license plate area gray level image sample.
Further, described step S3 finishes the positive sample of car load board and negative sample, the complete car plate model of cognition of training convolutional neural networks, be specially: the positive sample of complete car plate prepared and negative sample are normalized and name process, and design complete car plate convolutional neural networks structure, generate the complete car plate model of cognition of convolutional neural networks; Wherein, the positive sample of complete car plate comprises entire vehicle board coloured image sample, and complete car plate negative sample comprises non-license plate area coloured image sample.
Further, described step S5 utilizes local horizontal projecting method to carry out the detection of horizontal tilt angle to the final candidate region of car plate, and utilize rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy, be specially: choose the final candidate region of car plate and carry out gray processing process, and binary conversion treatment is carried out to the imagery exploitation Da-Jin algorithm after gray processing; Again the image after binaryzation is rotated within the scope of set angle, and solve postrotational image level projection values; Finally choose 5 groups of numerical value larger in the horizontal projection numerical value under each angle to sue for peace, summation numerical value relatively under different angles using the angle of inclination of angle corresponding for maximum summation numerical value as car plate, and according to the license plate sloped angle obtained, rotation correction is carried out to the final candidate region of car plate and car plate final candidate region copy.
Further, the final candidate region of car plate after described step S6 utilizes multi-threshold segmentation algorithm to correct horizontal tilt and car plate final candidate region copy process, obtain characters on license plate correspondence image, be specially: gray processing process is carried out to the final candidate region of car plate, and add up the ratio that the pixel count being greater than a certain gray values accounts for the total number of pixels of license plate area image; Determine one group of candidate thresholds, and to car plate final candidate region image, binary conversion treatment is carried out successively to each threshold value wherein; Solve UNICOM territory to the car plate final candidate region image after binary conversion treatment each time, and merged in all overlapping UNICOM territories, sorting in the UNICOM territory after being combined, obtains the UNICOM territory with characters on license plate sequence consensus; The car plate of positional information to correspondence final candidate region duplicate pictures according to the UNICOM territory obtained is split, and obtains characters on license plate correspondence image.
Further, described step S7 prepares Chinese character, letter and number sample, training Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition, be specially: the Chinese character prepared, letter and number sample are normalized, and design the convolutional neural networks structure of Chinese character, letter and number respectively; Adopt BP algorithm to train Chinese character convolutional neural networks, letter and number convolutional neural networks, generate Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition.
The invention has the beneficial effects as follows: the detection method of license plate based on convolutional neural networks of the present invention utilizes the Adaboost car plate detecting device based on Haar feature to treat detection license plate image and roughly selects, then the complete car plate model of cognition of the convolutional neural networks of proposition is utilized to screen license plate area, obtain final license plate area, realize the accurate identification to the license plate image under different condition; Recycling multi-threshold segmentation algorithm carries out Character segmentation to final license plate image, realizes easier identifying character; Chinese character convolutional neural networks model of cognition, Chinese character and alphabetical convolutional neural networks model of cognition is finally utilized to identify corresponding character picture; The present invention has good effect on engineer applied.
Accompanying drawing explanation
Fig. 1 is the detection method of license plate schematic flow sheet based on convolutional neural networks of the present invention.
Fig. 2 is complete car plate convolutional neural networks structural representation of the present invention.
Fig. 3 is the car plate figure before horizontal tilt of the present invention corrects.
Fig. 4 is the car plate figure after horizontal tilt of the present invention corrects.
Fig. 5 is multi-threshold segmentation algorithm schematic flow sheet of the present invention.
Fig. 6 is the design sketch of multi-threshold segmentation algorithm of the present invention.
Fig. 7 is Chinese character of the present invention, letter and number convolutional neural networks structural representation.
Fig. 8 is the mapping to be checked in the embodiment of the present invention 1.
Fig. 9 is the license plate area testing result figure in the embodiment of the present invention 1.
Figure 10 is the Character segmentation result figure in the embodiment of the present invention 1.
Figure 11 is the recognition result figure in the embodiment of the present invention 1.
Figure 12 is the mapping to be checked in the embodiment of the present invention 2.
Figure 13 is the license plate area testing result figure in the embodiment of the present invention 2.
Figure 14 is the Character segmentation result figure in the embodiment of the present invention 2.
Figure 15 is the recognition result figure in the embodiment of the present invention 2.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.
As shown in Figure 1, for of the present invention based on the detection method of license plate schematic flow sheet of convolutional neural networks.Based on a detection method of license plate for convolutional neural networks, comprise the following steps:
S1. prepare the positive sample of car plate and negative sample, train the Adaboost car plate detecting device based on Haar feature.
In order to positioning licence plate candidate region from license plate image to be detected, the present invention prepares 2000 positive samples and 5000 negative samples for training the Adaboost car plate detecting device based on Haar feature.Wherein, the positive sample of car plate comprises the license plate grey level image sample under different illumination conditions, and car plate negative sample comprises the non-license plate area gray level image samples such as vehicle body, windshield, exhaust screen.Image training progression in the positive sample set of car plate and car plate negative sample collection is 15 grades.Positive for car plate sample and car plate negative sample are all normalized to 45 × 15 pixel sizes, generate the Adaboost car plate detecting device based on Haar feature.
S2. treat detected image and carry out gray processing process, and utilize the Adaboost car plate detecting device based on Haar feature in step S1 to carry out moving window detection, region roughly selected by the car plate obtained in image to be detected.
Here image to be detected is the image that the needs obtained in advance carry out detecting.Car plate is roughly selected region and is generally comprised license plate area and deep bead etc. and easily obscure region.Detected by gray processing process and moving window, license plate area can be identified accurately for car plate color situation close to body color.
S3. the positive sample of car load board and negative sample is finished, the complete car plate model of cognition of training convolutional neural networks.
Isolate complete license plate image to roughly select in region from car plate, the present invention prepares 10000 positive samples and 10000 negative samples for the complete car plate model of cognition of training convolutional neural networks.Wherein, the positive sample of complete car plate comprises entire vehicle board coloured image sample, and complete car plate negative sample comprises non-license plate area coloured image sample.Positive sample and negative sample are normalized to 48 × 16 pixel sizes, and carry out name process; By designing complete car plate convolutional neural networks structure, training obtains the complete car plate model of cognition of convolutional neural networks.Here positive sample is named with 1 beginning, by negative sample with 0 beginning name, realize determining according to the name of positive sample and negative sample the correct label that sample is corresponding.According to the difference of label and network output valve, BP algorithm is utilized to upgrade network weight.As shown in Figure 2, be complete car plate convolutional neural networks structural representation of the present invention.Wherein, to be convolution operation layer 1, conv2 be conv1 that convolution operation layer 2, pool1 is pond operation layer 1, pool2 be pond operation layer 2, feature map is the characteristic pattern obtained after convolution operation, and fc is the full articulamentum in neural network; Conv1 layer has 4 feature map, and conv2 layer has 4 feature map, is full connection between input layer and conv1 layer, between pool1 layer and conv2 layer, between pool2 layer and full connection fc layer; Be that local connects between conv1 layer and pool1 layer, between conv2 layer and pool2 layer.
S4. utilize the complete car plate model of cognition of the convolutional neural networks in step S3 to roughly select region to the car plate in step S2 to screen, obtain the final candidate region of car plate, and preserve car plate final candidate region copy.
Utilize the complete car plate model of cognition of convolutional neural networks to roughly select region to car plate to screen, remove pseudo-region, obtain the final candidate region of car plate, and preserve car plate final candidate region copy.Character is cut after car plate final candidate region copy is used for determining character particular location.
S5. utilize local horizontal projecting method to carry out the detection of horizontal tilt angle to the final candidate region of the car plate in step S4, and utilize rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy.
May occur inaccurate problem for the final candidate region of car plate, the present invention adopts local horizontal sciagraphy and rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy respectively.Here partial projection method is for detecting the angle of inclination of license plate candidate area; Rotation correction method is used for, according to the license plate candidate area angular slope angle detected, carrying out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy.As shown in Figure 3, be the car plate figure before horizontal tilt correction of the present invention.As shown in Figure 4, be the car plate figure after horizontal tilt correction of the present invention.0.8 × 0.8 pixel region choosing center, car plate final candidate region carries out gray processing process, and utilizes large law to carry out binary conversion treatment to the area image after gray processing; Again the image after binaryzation is rotated respectively each angle in (-30 °, 30 °) angular range, and solve the horizontal projection numerical value of postrotational image; Finally choose 5 groups of numerical value larger in the horizontal projection numerical value under each angle to sue for peace, summation numerical value relatively under different angles using the angle of inclination of angle corresponding for maximum summation numerical value as car plate, and according to the license plate sloped angle obtained, rotation correction is carried out to the final candidate region of car plate and car plate final candidate region copy.
S6. the final candidate region of car plate after utilizing multi-threshold segmentation algorithm to correct horizontal tilt and car plate final candidate region copy process, and obtain characters on license plate correspondence image.
The shade that may occur for license plate image or the problem of uneven illumination, the present invention utilizes multi-threshold segmentation algorithm to carry out Character segmentation to car plate final candidate region image, can realize Character segmentation more easily.As shown in Figure 5, be multi-threshold segmentation algorithm schematic flow sheet of the present invention.First gray processing process is carried out to the final candidate region of the car plate after slant correction image, add up the pixel number that each gray-scale value is corresponding, and determine one group of threshold value according to the white point accounting after each gray-scale value is carried out binaryzation.When carrying out binaryzation operation, the grey scale pixel value being greater than a certain gray-scale value is revised as 255, i.e. white point; The gray-scale value being less than a certain threshold value is revised as 0, i.e. black color dots.For some gray-scale value after carrying out binaryzation operation, image effect is close, and namely white point accounting difference is less than a certain numerical value, then only need choose wherein a certain gray-scale value.Each group threshold value is utilized to carry out binary conversion treatment to car plate final candidate region image successively, because the sets of threshold values determined is according to order arrangement from small to large, so the image after binaryzation is gradual manner.As shown in Figure 6, be the design sketch of multi-threshold segmentation algorithm of the present invention.Solve UNICOM territory to the car plate final candidate region image after binary conversion treatment each time, and merged in all overlapping UNICOM territories, sorting based on x coordinate in the UNICOM territory after being combined, obtains the UNICOM territory with characters on license plate sequence consensus.Here theoretical according to Computer Image Processing, the horizontal respective coordinates in image is x-axis, i.e. Width; In image, longitudinal respective coordinates is y-axis, i.e. short transverse; Car plate is corresponding x-axis laterally, and longitudinally corresponding y-axis, utilizes the x coordinate size in each upper left corner, UNICOM territory to determine the left-right relation in each UNICOM territory.Again according to transition times, candidate region size, the candidate region aspect ratio value information in image outline region, determine whether profile is character; If profile is character, then keep character information; If profile is not character, then continue to utilize next group threshold value to carry out binary conversion treatment to car plate final candidate region image successively.After preserving character information, by the profile territory zero setting of final for car plate candidate region image, being namely revised as black, this region can not being detected again when utilizing next group threshold value to carry out binaryzation operation like this; Continue to utilize next group threshold value to carry out binary conversion treatment to car plate final candidate region image successively.The car plate of positional information to correspondence final candidate region duplicate pictures according to the UNICOM territory obtained is split, and obtains the correspondence image of car plate 7 characters.
S7. Chinese character, letter and number sample is prepared, training Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition.
The present invention utilizes Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition identifies characters on license plate correspondence image, reduces identification difficulty, also reduces the requirement of the quality to character simultaneously.The present invention prepares for training Chinese character convolutional neural networks model of cognition, the sample of letter and number convolutional neural networks model of cognition comprises 8000 each provinces and is called for short Hanzi specimen image, wherein Chinese character totally 31; 8000 letter and number sample images, wherein letter comprises A, B ... Z, numeral comprises 0, and 1 ... 9; Especially, the present invention has also prepared port, Australia, police, sample image.First be 16 × 32 pixel sizes to Chinese character, letter and number samples normalization respectively; Design corresponding Chinese character convolutional neural networks structure, letter and number convolutional neural networks structure more respectively; BP algorithm is used to train Chinese character convolutional neural networks structure, letter and number convolutional neural networks structure respectively, here Chinese character convolutional neural networks structure is identical with letter and number convolutional neural networks structure, finally generates Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition.As shown in Figure 7, be Chinese character of the present invention, letter and number convolutional neural networks structural representation.Conv1 layer, conv2 layer all have 16 feature map, input layer and conv1 layer, pool1 layer and conv2 layer, are full connection between pool2 layer and fc layer, conv1 layer and pool1 layer, are that local connects between conv2 layer and pool2 layer.
S8. utilize the Chinese character convolutional neural networks model of cognition in step S7, letter and number convolutional neural networks model of cognition carries out Recognition of License Plate Characters to characters on license plate correspondence image in step S6, and add up recognition result.
Respectively car plate 7 character correspondence image are detected, obtain final recognition result.
The present invention gathers image to be detected respectively at different conditions, and utilizes the detection method of license plate based on convolutional neural networks of the present invention to identify the image to be detected under 2 kinds of different conditions respectively.As shown in Figure 8, be the mapping to be checked in the embodiment of the present invention 1.As shown in Figure 9, be the license plate area testing result figure in the embodiment of the present invention 1.As shown in Figure 10, be the Character segmentation result figure in the embodiment of the present invention 1.As shown in figure 11, be the recognition result figure in the embodiment of the present invention 1.As shown in figure 12, be the car plate figure to be detected in the embodiment of the present invention 2.As shown in figure 13, be the license plate area testing result figure in the embodiment of the present invention 2.As shown in figure 14, be the Character segmentation result figure in the embodiment of the present invention 2.As shown in figure 15, be the recognition result figure in the embodiment of the present invention 2.As can be seen from above-mentioned 2 embodiments, in the present invention's practical implementation at different conditions, there is good recognition effect.
Those of ordinary skill in the art will appreciate that, embodiment described here is to help reader understanding's principle of the present invention, should be understood to that protection scope of the present invention is not limited to so special statement and embodiment.Those of ordinary skill in the art can make various other various concrete distortion and combination of not departing from essence of the present invention according to these technology enlightenment disclosed by the invention, and these distortion and combination are still in protection scope of the present invention.
Claims (6)
1. based on a detection method of license plate for convolutional neural networks, it is characterized in that, comprise the following steps:
S1. prepare the positive sample of car plate and negative sample, train the Adaboost car plate detecting device based on Haar feature;
S2. treat detected image and carry out gray processing process, and utilize the Adaboost car plate detecting device based on Haar feature in step S1 to carry out moving window detection, region roughly selected by the car plate obtained in image to be detected;
S3. the positive sample of car load board and negative sample is finished, the complete car plate model of cognition of training convolutional neural networks;
S4. utilize the complete car plate model of cognition of the convolutional neural networks in step S3 to roughly select region to the car plate in step S2 to screen, obtain the final candidate region of car plate, and preserve car plate final candidate region copy;
S5. utilize local horizontal projecting method to carry out the detection of horizontal tilt angle to the final candidate region of the car plate in step S4, and utilize rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy;
S6. the final candidate region of car plate after utilizing multi-threshold segmentation algorithm to correct horizontal tilt and car plate final candidate region copy process, and obtain characters on license plate correspondence image;
S7. Chinese character, letter and number sample is prepared, training Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition;
S8. utilize the Chinese character convolutional neural networks model of cognition in step S7, letter and number convolutional neural networks model of cognition carries out Recognition of License Plate Characters to characters on license plate correspondence image in step S6, and add up recognition result.
2. as claimed in claim 1 based on the detection method of license plate of convolutional neural networks, it is characterized in that, described step S1 prepares the positive sample of car plate and negative sample, train the Adaboost car plate detecting device based on Haar feature, be specially: positive for the car plate of preparation sample and negative sample are normalized, generate the Adaboost car plate detecting device based on Haar feature; Wherein, the positive sample of car plate comprises the license plate grey level image sample under different illumination conditions, and car plate negative sample comprises non-license plate area gray level image sample.
3. as claimed in claim 1 based on the detection method of license plate of convolutional neural networks, it is characterized in that, described step S3 finishes the positive sample of car load board and negative sample, the complete car plate model of cognition of training convolutional neural networks, be specially: the positive sample of complete car plate prepared and negative sample are normalized and name process, and design complete car plate convolutional neural networks structure, generate the complete car plate model of cognition of convolutional neural networks; Wherein, the positive sample of complete car plate comprises entire vehicle board coloured image sample, and complete car plate negative sample comprises non-license plate area coloured image sample.
4. as claimed in claim 1 based on the detection method of license plate of convolutional neural networks, it is characterized in that, described step S5 utilizes local horizontal projecting method to carry out the detection of horizontal tilt angle to the final candidate region of car plate, and utilize rotation correction method to carry out horizontal tilt correction to the final candidate region of car plate and car plate final candidate region copy, be specially: choose the final candidate region of car plate and carry out gray processing process, and binary conversion treatment is carried out to the imagery exploitation Da-Jin algorithm after gray processing; Again the image after binaryzation is rotated within the scope of set angle, and solve postrotational image level projection values; Finally choose 5 groups of numerical value larger in the horizontal projection numerical value under each angle to sue for peace, summation numerical value relatively under different angles using the angle of inclination of angle corresponding for maximum summation numerical value as car plate, and according to the license plate sloped angle obtained, rotation correction is carried out to the final candidate region of car plate and car plate final candidate region copy.
5. as claimed in claim 1 based on the detection method of license plate of convolutional neural networks, it is characterized in that, the final candidate region of car plate after described step S6 utilizes multi-threshold segmentation algorithm to correct horizontal tilt and car plate final candidate region copy process, obtain characters on license plate correspondence image, be specially: gray processing process is carried out to the final candidate region of car plate, and add up the ratio that the pixel count being greater than a certain gray values accounts for the total number of pixels of license plate area image; Determine one group of candidate thresholds, and to car plate final candidate region image, binary conversion treatment is carried out successively to each threshold value wherein; Solve UNICOM territory to the car plate final candidate region image after binary conversion treatment each time, and merged in all overlapping UNICOM territories, sorting in the UNICOM territory after being combined, obtains the UNICOM territory with characters on license plate sequence consensus; The car plate of positional information to correspondence final candidate region duplicate pictures according to the UNICOM territory obtained is split, and obtains characters on license plate correspondence image.
6. as claimed in claim 1 based on the detection method of license plate of convolutional neural networks, it is characterized in that, described step S7 prepares Chinese character, letter and number sample, training Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition, be specially: the Chinese character prepared, letter and number sample are normalized, and design the convolutional neural networks structure of Chinese character, letter and number respectively; Adopt BP algorithm to train Chinese character convolutional neural networks, letter and number convolutional neural networks, generate Chinese character convolutional neural networks model of cognition, letter and number convolutional neural networks model of cognition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410549315.1A CN104298976B (en) | 2014-10-16 | 2014-10-16 | Detection method of license plate based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410549315.1A CN104298976B (en) | 2014-10-16 | 2014-10-16 | Detection method of license plate based on convolutional neural networks |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104298976A true CN104298976A (en) | 2015-01-21 |
| CN104298976B CN104298976B (en) | 2017-09-26 |
Family
ID=52318697
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410549315.1A Expired - Fee Related CN104298976B (en) | 2014-10-16 | 2014-10-16 | Detection method of license plate based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104298976B (en) |
Cited By (86)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636309A (en) * | 2015-02-06 | 2015-05-20 | 南京理工大学 | Matrix calculator based on machine vision and matrix identification method |
| CN104680149A (en) * | 2015-03-10 | 2015-06-03 | 苏州科达科技股份有限公司 | Method and system for recognizing object type |
| CN104933420A (en) * | 2015-07-03 | 2015-09-23 | 北京旷视科技有限公司 | A scene image identification method and a scene image identification device |
| CN105279488A (en) * | 2015-10-16 | 2016-01-27 | 西南交通大学 | Mileage signboard automatic recognition method for road video routing inspection |
| CN105279556A (en) * | 2015-11-05 | 2016-01-27 | 国家卫星海洋应用中心 | Enteromorpha detection method and enteromorpha detection device |
| CN105335710A (en) * | 2015-10-22 | 2016-02-17 | 合肥工业大学 | Fine vehicle model identification method based on multi-stage classifier |
| CN105335743A (en) * | 2015-10-28 | 2016-02-17 | 重庆邮电大学 | Vehicle license plate recognition method |
| CN105354572A (en) * | 2015-12-10 | 2016-02-24 | 苏州大学 | Automatic identification system of number plate on the basis of simplified convolutional neural network |
| CN105469047A (en) * | 2015-11-23 | 2016-04-06 | 上海交通大学 | Chinese detection method based on unsupervised learning and deep learning network and system thereof |
| CN105528607A (en) * | 2015-10-30 | 2016-04-27 | 小米科技有限责任公司 | Region extraction method and model training method and device |
| CN105590102A (en) * | 2015-12-30 | 2016-05-18 | 中通服公众信息产业股份有限公司 | Front car face identification method based on deep learning |
| CN105608454A (en) * | 2015-12-21 | 2016-05-25 | 上海交通大学 | Text structure part detection neural network based text detection method and system |
| CN106156767A (en) * | 2016-03-02 | 2016-11-23 | 平安科技（深圳）有限公司 | Driving license effect duration extraction method, server and terminal |
| CN106407981A (en) * | 2016-11-24 | 2017-02-15 | 北京文安智能技术股份有限公司 | License plate recognition method, device and system |
| CN106446784A (en) * | 2016-08-30 | 2017-02-22 | 东软集团股份有限公司 | Image detection method and apparatus |
| CN106469304A (en) * | 2016-09-22 | 2017-03-01 | 西安理工大学 | Handwritten signature location positioning method in bill based on depth convolutional neural networks |
| CN106548155A (en) * | 2016-10-28 | 2017-03-29 | 安徽四创电子股份有限公司 | A kind of detection method of license plate based on depth belief network |
| CN106557747A (en) * | 2016-11-15 | 2017-04-05 | 平安科技（深圳）有限公司 | The method and device of identification insurance single numbers |
| CN106650748A (en) * | 2016-11-16 | 2017-05-10 | 武汉工程大学 | Chinese character recognition method based on convolution neural network |
| CN106778659A (en) * | 2016-12-28 | 2017-05-31 | 深圳市捷顺科技实业股份有限公司 | A kind of licence plate recognition method and device |
| CN106778737A (en) * | 2016-11-24 | 2017-05-31 | 北京文安智能技术股份有限公司 | A kind of car plate antidote, device and a kind of video acquisition device |
| CN106778835A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | The airport target by using remote sensing image recognition methods of fusion scene information and depth characteristic |
| CN106778650A (en) * | 2016-12-26 | 2017-05-31 | 深圳极视角科技有限公司 | Scene adaptive pedestrian detection method and system based on polymorphic type information fusion |
| CN106778745A (en) * | 2016-12-23 | 2017-05-31 | 深圳先进技术研究院 | A kind of licence plate recognition method and device, user equipment |
| CN106845487A (en) * | 2016-12-30 | 2017-06-13 | 佳都新太科技股份有限公司 | A kind of licence plate recognition method end to end |
| CN106874902A (en) * | 2017-01-19 | 2017-06-20 | 博康智能信息技术有限公司北京海淀分公司 | A kind of license board information recognition methods and device |
| CN106874907A (en) * | 2017-01-19 | 2017-06-20 | 博康智能信息技术有限公司北京海淀分公司 | A kind of method and device for setting up Car license recognition model |
| CN106886778A (en) * | 2017-04-25 | 2017-06-23 | 福州大学 | A kind of car plate segmentation of the characters and their identification method under monitoring scene |
| CN106897770A (en) * | 2017-01-19 | 2017-06-27 | 博康智能信息技术有限公司北京海淀分公司 | A kind of method and device for setting up Car license recognition model |
| CN106935060A (en) * | 2017-04-11 | 2017-07-07 | 厦门路桥信息股份有限公司 | A kind of machinery garage seeks car system and method |
| CN107025452A (en) * | 2016-01-29 | 2017-08-08 | 富士通株式会社 | Image-recognizing method and image recognition apparatus |
| CN107169954A (en) * | 2017-04-18 | 2017-09-15 | 华南理工大学 | A kind of image significance detection method based on parallel-convolution neutral net |
| CN107203754A (en) * | 2017-05-26 | 2017-09-26 | 北京邮电大学 | A kind of license plate locating method and device based on deep learning |
| CN107392201A (en) * | 2017-06-09 | 2017-11-24 | 中国科学院自动化研究所 | The pillar recognition methods of catenary mast, storage medium, processing equipment |
| CN107506765A (en) * | 2017-10-13 | 2017-12-22 | 厦门大学 | A kind of method of the license plate sloped correction based on neutral net |
| WO2017220032A1 (en) * | 2016-06-24 | 2017-12-28 | 平安科技（深圳）有限公司 | Vehicle license plate classification method and system based on deep learning, electronic apparatus, and storage medium |
| CN107590498A (en) * | 2017-09-27 | 2018-01-16 | 哈尔滨工业大学 | A kind of self-adapted car instrument detecting method based on Character segmentation level di- grader |
| CN107667382A (en) * | 2015-06-26 | 2018-02-06 | 莱克斯真株式会社 | Vehicle number code recognition device and its method |
| CN107705323A (en) * | 2017-10-13 | 2018-02-16 | 北京理工大学 | A kind of level set target tracking method based on convolutional neural networks |
| CN107886082A (en) * | 2017-11-24 | 2018-04-06 | 腾讯科技（深圳）有限公司 | Mathematical formulae detection method, device, computer equipment and storage medium in image |
| CN108109414A (en) * | 2017-12-19 | 2018-06-01 | 武汉极目智能技术有限公司 | A kind of preventing road monitoring system and method based on vehicle assistant drive |
| CN108121984A (en) * | 2016-11-30 | 2018-06-05 | 杭州海康威视数字技术股份有限公司 | A kind of character identifying method and device |
| WO2018112900A1 (en) * | 2016-12-23 | 2018-06-28 | 深圳先进技术研究院 | License plate recognition method and apparatus, and user equipment |
| CN108229463A (en) * | 2018-02-07 | 2018-06-29 | 众安信息技术服务有限公司 | Character recognition method based on image |
| CN108288037A (en) * | 2018-01-19 | 2018-07-17 | 深圳禾思众成科技有限公司 | A kind of tire coding identifying system |
| CN108304752A (en) * | 2017-01-13 | 2018-07-20 | 天津工业大学 | It is detected based on the traffic signboard of HOG features and BCNN |
| CN108416348A (en) * | 2018-01-29 | 2018-08-17 | 重庆邮电大学 | Plate location recognition method based on support vector machines and convolutional neural networks |
| CN108515967A (en) * | 2018-04-25 | 2018-09-11 | 深圳卡安兴科技有限公司 | A kind of anti-collision early warning method and device based on Car license recognition |
| CN108681729A (en) * | 2018-05-08 | 2018-10-19 | 腾讯科技（深圳）有限公司 | Text image antidote, device, storage medium and equipment |
| CN108875819A (en) * | 2018-06-08 | 2018-11-23 | 浙江大学 | A kind of object and component associated detecting method based on shot and long term memory network |
| CN108898137A (en) * | 2018-05-25 | 2018-11-27 | 黄凯 | A kind of natural image character identifying method and system based on deep neural network |
| CN108932453A (en) * | 2017-05-23 | 2018-12-04 | 杭州海康威视数字技术股份有限公司 | A kind of vehicle spare tyre detection method and device |
| CN109034166A (en) * | 2017-06-08 | 2018-12-18 | 北京君正集成电路股份有限公司 | Confusable character identification model training method and device |
| CN104809443B (en) * | 2015-05-05 | 2018-12-28 | 上海交通大学 | Detection method of license plate and system based on convolutional neural networks |
| CN109145927A (en) * | 2017-06-16 | 2019-01-04 | 杭州海康威视数字技术股份有限公司 | The target identification method and device of a kind of pair of strain image |
| CN109409272A (en) * | 2018-10-17 | 2019-03-01 | 北京空间技术研制试验中心 | Cable Acceptance Test System and method based on machine vision |
| CN109447030A (en) * | 2018-11-12 | 2019-03-08 | 重庆知遨科技有限公司 | A kind of fire-fighting robot movement real-time instruction algorithm for fire scenario |
| CN109543673A (en) * | 2018-10-18 | 2019-03-29 | 浙江理工大学 | A kind of low contrast punching press character recognition algorithm based on Interactive Segmentation |
| CN109657682A (en) * | 2018-11-29 | 2019-04-19 | 国网河北省电力有限公司电力科学研究院 | A kind of electric energy meter registration recognition methods based on deep neural network and the soft cutting of multi-threshold |
| CN109711407A (en) * | 2018-12-28 | 2019-05-03 | 深圳市捷顺科技实业股份有限公司 | A kind of method and relevant apparatus of Car license recognition |
| CN109740603A (en) * | 2019-01-21 | 2019-05-10 | 闽江学院 | Based on the vehicle character identifying method under CNN convolutional neural networks |
| US10296794B2 (en) | 2016-12-20 | 2019-05-21 | Jayant Rtti | On-demand artificial intelligence and roadway stewardship system |
| CN109829460A (en) * | 2019-04-02 | 2019-05-31 | 安徽省安瑞机电科技有限公司 | A kind of License Plate Character Segmentation recognition methods and its system applied to parking management system |
| CN110009929A (en) * | 2019-03-15 | 2019-07-12 | 北京筑梦园科技有限公司 | A kind of Vehicle berth management method, equipment and system |
| WO2019175686A1 (en) | 2018-03-12 | 2019-09-19 | Ratti Jayant | On-demand artificial intelligence and roadway stewardship system |
| CN110288626A (en) * | 2019-06-13 | 2019-09-27 | 北京云测信息技术有限公司 | The method and apparatus for detecting the text in primary electronic image |
| CN110619327A (en) * | 2018-06-20 | 2019-12-27 | 湖南省瞬渺通信技术有限公司 | Real-time license plate recognition method based on deep learning in complex scene |
| CN110717575A (en) * | 2018-07-13 | 2020-01-21 | 奇景光电股份有限公司 | Frame buffer free convolutional neural network system and method |
| CN110866430A (en) * | 2018-08-28 | 2020-03-06 | 上海富瀚微电子股份有限公司 | License plate recognition method and device |
| CN110992725A (en) * | 2019-10-24 | 2020-04-10 | 合肥讯图信息科技有限公司 | Method, system and storage medium for detecting traffic signal lamp fault |
| CN111046746A (en) * | 2019-11-22 | 2020-04-21 | 北京精英路通科技有限公司 | License plate detection method and device |
| CN111160338A (en) * | 2019-12-24 | 2020-05-15 | 苏州奥易克斯汽车电子有限公司 | Fuzzy license plate recognition method |
| CN111178359A (en) * | 2018-11-09 | 2020-05-19 | 中移（杭州）信息技术有限公司 | License plate number recognition method, device and equipment and computer storage medium |
| CN111259886A (en) * | 2020-01-08 | 2020-06-09 | 上海眼控科技股份有限公司 | License plate screw detection method, electronic device, computer equipment and storage medium |
| CN111275043A (en) * | 2020-01-22 | 2020-06-12 | 西北师范大学 | Paper numbered musical notation electronization play device based on PCNN handles |
| CN111310754A (en) * | 2019-12-31 | 2020-06-19 | 创泽智能机器人集团股份有限公司 | Method for segmenting license plate characters |
| US10706330B2 (en) | 2015-10-01 | 2020-07-07 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| CN111444911A (en) * | 2019-12-13 | 2020-07-24 | 珠海大横琴科技发展有限公司 | Training method and device of license plate recognition model and license plate recognition method and device |
| CN111489560A (en) * | 2020-04-13 | 2020-08-04 | 深圳市海圳汽车技术有限公司 | Detection method and control method for detecting front vehicle starting based on shallow convolutional neural network |
| CN111523416A (en) * | 2020-04-13 | 2020-08-11 | 北京巨视科技有限公司 | Vehicle early warning method and device based on highway ETC portal |
| CN111611863A (en) * | 2020-04-22 | 2020-09-01 | 浙江大华技术股份有限公司 | License plate image quality evaluation method and device and computer equipment |
| CN111814852A (en) * | 2020-06-24 | 2020-10-23 | 理光软件研究所（北京）有限公司 | Image detection method, image detection device, electronic equipment and computer-readable storage medium |
| WO2021044203A1 (en) * | 2019-09-05 | 2021-03-11 | 商汤国际私人有限公司 | Method and apparatus for identifying sequence in image, and electronic device and storage medium |
| CN112997190A (en) * | 2020-12-29 | 2021-06-18 | 深圳市锐明技术股份有限公司 | License plate recognition method and device and electronic equipment |
| CN113313143A (en) * | 2021-04-29 | 2021-08-27 | 浙江大华技术股份有限公司 | License plate detection method and device and computer storage medium |
| US11587327B2 (en) | 2015-10-01 | 2023-02-21 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100278436A1 (en) * | 2009-04-30 | 2010-11-04 | Industrial Technology Research Institute | Method and system for image identification and identification result output |
| CN103605953A (en) * | 2013-10-31 | 2014-02-26 | 电子科技大学 | Vehicle interest target detection method based on sliding window search |
| CN103679209A (en) * | 2013-11-29 | 2014-03-26 | 广东领域安防有限公司 | Sparse theory based character recognition method |
| CN103761531A (en) * | 2014-01-20 | 2014-04-30 | 西安理工大学 | Sparse-coding license plate character recognition method based on shape and contour features |
| CN103824091A (en) * | 2014-02-27 | 2014-05-28 | 中国石油大学(华东) | Vehicle license plate recognition method for intelligent transportation system |
-
2014
- 2014-10-16 CN CN201410549315.1A patent/CN104298976B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100278436A1 (en) * | 2009-04-30 | 2010-11-04 | Industrial Technology Research Institute | Method and system for image identification and identification result output |
| CN103605953A (en) * | 2013-10-31 | 2014-02-26 | 电子科技大学 | Vehicle interest target detection method based on sliding window search |
| CN103679209A (en) * | 2013-11-29 | 2014-03-26 | 广东领域安防有限公司 | Sparse theory based character recognition method |
| CN103761531A (en) * | 2014-01-20 | 2014-04-30 | 西安理工大学 | Sparse-coding license plate character recognition method based on shape and contour features |
| CN103824091A (en) * | 2014-02-27 | 2014-05-28 | 中国石油大学(华东) | Vehicle license plate recognition method for intelligent transportation system |
Non-Patent Citations (3)
| Title |
|---|
| "卷积神经网络的研究及其在车牌识别系统中的应用";陆璐;《中国优秀博硕士学位论文全文数据库 (硕士) 信息科技辑》;20060915;论文正文第15页第4段-第16页第3段、第27页第3段-第36页第4段，附图4.5-4.15，表4.1-4.5 * |
| "运动模糊车牌自动识别算法研究与实现";高巍;《中国优秀硕士学位论文全文数据库》;20120415;论文正文第28页第4段-第37页第3段、第46页第3段-第48页第2段，附图3.12-3.15 * |
| "面向车牌检测Adaboost算法的FPGA实现";王亮;《中国优秀硕士学位论文全文数据库 信息科技辑》;20130315;论文正文第9页第2段-第12页第2段、第15页第1段-第27页第5段，附图2.1-2.5、附图3.2、3.12 * |
Cited By (129)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636309B (en) * | 2015-02-06 | 2018-01-05 | 南京理工大学 | Matrix recognition methods based on machine vision |
| CN104636309A (en) * | 2015-02-06 | 2015-05-20 | 南京理工大学 | Matrix calculator based on machine vision and matrix identification method |
| CN104680149A (en) * | 2015-03-10 | 2015-06-03 | 苏州科达科技股份有限公司 | Method and system for recognizing object type |
| CN104680149B (en) * | 2015-03-10 | 2018-07-03 | 苏州科达科技股份有限公司 | A kind of object type recognition methods and system |
| CN104809443B (en) * | 2015-05-05 | 2018-12-28 | 上海交通大学 | Detection method of license plate and system based on convolutional neural networks |
| CN107667382A (en) * | 2015-06-26 | 2018-02-06 | 莱克斯真株式会社 | Vehicle number code recognition device and its method |
| CN104933420B (en) * | 2015-07-03 | 2018-11-27 | 北京旷视科技有限公司 | A kind of scene image recognition methods and scene image identify equipment |
| CN104933420A (en) * | 2015-07-03 | 2015-09-23 | 北京旷视科技有限公司 | A scene image identification method and a scene image identification device |
| US11587327B2 (en) | 2015-10-01 | 2023-02-21 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| US10706330B2 (en) | 2015-10-01 | 2020-07-07 | Intellivision Technologies Corp | Methods and systems for accurately recognizing vehicle license plates |
| CN105279488A (en) * | 2015-10-16 | 2016-01-27 | 西南交通大学 | Mileage signboard automatic recognition method for road video routing inspection |
| CN105279488B (en) * | 2015-10-16 | 2018-12-25 | 西南交通大学 | A kind of mileage Sign Board automatic identifying method of highway video routing inspection |
| CN105335710A (en) * | 2015-10-22 | 2016-02-17 | 合肥工业大学 | Fine vehicle model identification method based on multi-stage classifier |
| CN105335743A (en) * | 2015-10-28 | 2016-02-17 | 重庆邮电大学 | Vehicle license plate recognition method |
| WO2017071064A1 (en) * | 2015-10-30 | 2017-05-04 | 小米科技有限责任公司 | Area extraction method, and model training method and apparatus |
| CN105528607B (en) * | 2015-10-30 | 2019-02-15 | 小米科技有限责任公司 | Method for extracting region, model training method and device |
| CN105528607A (en) * | 2015-10-30 | 2016-04-27 | 小米科技有限责任公司 | Region extraction method and model training method and device |
| CN105279556A (en) * | 2015-11-05 | 2016-01-27 | 国家卫星海洋应用中心 | Enteromorpha detection method and enteromorpha detection device |
| CN105279556B (en) * | 2015-11-05 | 2017-11-07 | 国家卫星海洋应用中心 | A kind of Enteromorpha detection method and device |
| CN105469047B (en) * | 2015-11-23 | 2019-02-22 | 上海交通大学 | Chinese detection method and system based on unsupervised learning deep learning network |
| CN105469047A (en) * | 2015-11-23 | 2016-04-06 | 上海交通大学 | Chinese detection method based on unsupervised learning and deep learning network and system thereof |
| CN105354572A (en) * | 2015-12-10 | 2016-02-24 | 苏州大学 | Automatic identification system of number plate on the basis of simplified convolutional neural network |
| CN105354572B (en) * | 2015-12-10 | 2018-10-12 | 苏州大学 | A kind of automatic license plate identification system based on simplified convolutional neural networks |
| CN105608454A (en) * | 2015-12-21 | 2016-05-25 | 上海交通大学 | Text structure part detection neural network based text detection method and system |
| CN105608454B (en) * | 2015-12-21 | 2019-08-09 | 上海交通大学 | Character detecting method and system based on text structure component detection neural network |
| CN105590102A (en) * | 2015-12-30 | 2016-05-18 | 中通服公众信息产业股份有限公司 | Front car face identification method based on deep learning |
| CN107025452A (en) * | 2016-01-29 | 2017-08-08 | 富士通株式会社 | Image-recognizing method and image recognition apparatus |
| US10089553B1 (en) | 2016-03-02 | 2018-10-02 | Ping An Technology (Shenzhen) Co., Ltd. | Method, device, system, and storage medium for automatically extracting a validity period of a driving license |
| CN106156767A (en) * | 2016-03-02 | 2016-11-23 | 平安科技（深圳）有限公司 | Driving license effect duration extraction method, server and terminal |
| WO2017220032A1 (en) * | 2016-06-24 | 2017-12-28 | 平安科技（深圳）有限公司 | Vehicle license plate classification method and system based on deep learning, electronic apparatus, and storage medium |
| US10528841B2 (en) | 2016-06-24 | 2020-01-07 | Ping An Technology (Shenzhen) Co., Ltd. | Method, system, electronic device, and medium for classifying license plates based on deep learning |
| CN106446784A (en) * | 2016-08-30 | 2017-02-22 | 东软集团股份有限公司 | Image detection method and apparatus |
| CN106469304A (en) * | 2016-09-22 | 2017-03-01 | 西安理工大学 | Handwritten signature location positioning method in bill based on depth convolutional neural networks |
| CN106548155A (en) * | 2016-10-28 | 2017-03-29 | 安徽四创电子股份有限公司 | A kind of detection method of license plate based on depth belief network |
| CN106557747B (en) * | 2016-11-15 | 2018-06-22 | 平安科技（深圳）有限公司 | The method and device of identification insurance single numbers |
| CN106557747A (en) * | 2016-11-15 | 2017-04-05 | 平安科技（深圳）有限公司 | The method and device of identification insurance single numbers |
| CN106650748A (en) * | 2016-11-16 | 2017-05-10 | 武汉工程大学 | Chinese character recognition method based on convolution neural network |
| CN106778737B (en) * | 2016-11-24 | 2019-06-18 | 北京文安智能技术股份有限公司 | A kind of license plate antidote, device and a kind of video acquisition device |
| CN106407981A (en) * | 2016-11-24 | 2017-02-15 | 北京文安智能技术股份有限公司 | License plate recognition method, device and system |
| CN106778737A (en) * | 2016-11-24 | 2017-05-31 | 北京文安智能技术股份有限公司 | A kind of car plate antidote, device and a kind of video acquisition device |
| CN106778835B (en) * | 2016-11-29 | 2020-03-24 | 武汉大学 | Remote sensing image airport target identification method fusing scene information and depth features |
| CN106778835A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | The airport target by using remote sensing image recognition methods of fusion scene information and depth characteristic |
| CN108121984B (en) * | 2016-11-30 | 2021-09-21 | 杭州海康威视数字技术股份有限公司 | Character recognition method and device |
| CN108121984A (en) * | 2016-11-30 | 2018-06-05 | 杭州海康威视数字技术股份有限公司 | A kind of character identifying method and device |
| WO2018099194A1 (en) * | 2016-11-30 | 2018-06-07 | 杭州海康威视数字技术股份有限公司 | Character identification method and device |
| US11003941B2 (en) | 2016-11-30 | 2021-05-11 | Hangzhou Hikvision Digital Technology Co., Ltd. | Character identification method and device |
| US10296794B2 (en) | 2016-12-20 | 2019-05-21 | Jayant Rtti | On-demand artificial intelligence and roadway stewardship system |
| CN106778745A (en) * | 2016-12-23 | 2017-05-31 | 深圳先进技术研究院 | A kind of licence plate recognition method and device, user equipment |
| WO2018112900A1 (en) * | 2016-12-23 | 2018-06-28 | 深圳先进技术研究院 | License plate recognition method and apparatus, and user equipment |
| US10984289B2 (en) | 2016-12-23 | 2021-04-20 | Shenzhen Institute Of Advanced Technology | License plate recognition method, device thereof, and user equipment |
| CN106778650A (en) * | 2016-12-26 | 2017-05-31 | 深圳极视角科技有限公司 | Scene adaptive pedestrian detection method and system based on polymorphic type information fusion |
| CN106778659B (en) * | 2016-12-28 | 2020-10-27 | 深圳市捷顺科技实业股份有限公司 | License plate recognition method and device |
| CN106778659A (en) * | 2016-12-28 | 2017-05-31 | 深圳市捷顺科技实业股份有限公司 | A kind of licence plate recognition method and device |
| CN106845487A (en) * | 2016-12-30 | 2017-06-13 | 佳都新太科技股份有限公司 | A kind of licence plate recognition method end to end |
| CN106845487B (en) * | 2016-12-30 | 2020-04-17 | 佳都新太科技股份有限公司 | End-to-end license plate identification method |
| CN108304752A (en) * | 2017-01-13 | 2018-07-20 | 天津工业大学 | It is detected based on the traffic signboard of HOG features and BCNN |
| CN106874902B (en) * | 2017-01-19 | 2020-07-17 | 博康智能信息技术有限公司北京海淀分公司 | License plate information identification method and device |
| CN106874907B (en) * | 2017-01-19 | 2019-08-27 | 博康智能信息技术有限公司北京海淀分公司 | A kind of method and device for establishing Car license recognition model |
| CN106874907A (en) * | 2017-01-19 | 2017-06-20 | 博康智能信息技术有限公司北京海淀分公司 | A kind of method and device for setting up Car license recognition model |
| CN106897770A (en) * | 2017-01-19 | 2017-06-27 | 博康智能信息技术有限公司北京海淀分公司 | A kind of method and device for setting up Car license recognition model |
| CN106874902A (en) * | 2017-01-19 | 2017-06-20 | 博康智能信息技术有限公司北京海淀分公司 | A kind of license board information recognition methods and device |
| CN106935060A (en) * | 2017-04-11 | 2017-07-07 | 厦门路桥信息股份有限公司 | A kind of machinery garage seeks car system and method |
| CN107169954A (en) * | 2017-04-18 | 2017-09-15 | 华南理工大学 | A kind of image significance detection method based on parallel-convolution neutral net |
| CN106886778B (en) * | 2017-04-25 | 2020-02-07 | 福州大学 | License plate character segmentation and recognition method in monitoring scene |
| CN106886778A (en) * | 2017-04-25 | 2017-06-23 | 福州大学 | A kind of car plate segmentation of the characters and their identification method under monitoring scene |
| CN108932453A (en) * | 2017-05-23 | 2018-12-04 | 杭州海康威视数字技术股份有限公司 | A kind of vehicle spare tyre detection method and device |
| CN107203754A (en) * | 2017-05-26 | 2017-09-26 | 北京邮电大学 | A kind of license plate locating method and device based on deep learning |
| CN107203754B (en) * | 2017-05-26 | 2019-04-12 | 北京邮电大学 | A kind of license plate locating method and device based on deep learning |
| CN109034166A (en) * | 2017-06-08 | 2018-12-18 | 北京君正集成电路股份有限公司 | Confusable character identification model training method and device |
| CN109034166B (en) * | 2017-06-08 | 2021-09-24 | 北京君正集成电路股份有限公司 | Confusable character recognition model training method and device |
| CN107392201A (en) * | 2017-06-09 | 2017-11-24 | 中国科学院自动化研究所 | The pillar recognition methods of catenary mast, storage medium, processing equipment |
| CN109145927A (en) * | 2017-06-16 | 2019-01-04 | 杭州海康威视数字技术股份有限公司 | The target identification method and device of a kind of pair of strain image |
| US11126888B2 (en) | 2017-06-16 | 2021-09-21 | Hangzhou Hikvision Digital Technology Co., Ltd | Target recognition method and apparatus for a deformed image |
| CN107590498B (en) * | 2017-09-27 | 2020-09-01 | 哈尔滨工业大学 | Self-adaptive automobile instrument detection method based on character segmentation cascade two classifiers |
| CN107590498A (en) * | 2017-09-27 | 2018-01-16 | 哈尔滨工业大学 | A kind of self-adapted car instrument detecting method based on Character segmentation level di- grader |
| CN107506765A (en) * | 2017-10-13 | 2017-12-22 | 厦门大学 | A kind of method of the license plate sloped correction based on neutral net |
| CN107705323A (en) * | 2017-10-13 | 2018-02-16 | 北京理工大学 | A kind of level set target tracking method based on convolutional neural networks |
| CN107506765B (en) * | 2017-10-13 | 2020-09-01 | 厦门大学 | License plate inclination correction method based on neural network |
| CN107886082A (en) * | 2017-11-24 | 2018-04-06 | 腾讯科技（深圳）有限公司 | Mathematical formulae detection method, device, computer equipment and storage medium in image |
| CN108109414A (en) * | 2017-12-19 | 2018-06-01 | 武汉极目智能技术有限公司 | A kind of preventing road monitoring system and method based on vehicle assistant drive |
| CN108288037A (en) * | 2018-01-19 | 2018-07-17 | 深圳禾思众成科技有限公司 | A kind of tire coding identifying system |
| CN108288037B (en) * | 2018-01-19 | 2021-08-06 | 深圳禾思众成科技有限公司 | Tire code identification system |
| CN108416348A (en) * | 2018-01-29 | 2018-08-17 | 重庆邮电大学 | Plate location recognition method based on support vector machines and convolutional neural networks |
| CN108229463A (en) * | 2018-02-07 | 2018-06-29 | 众安信息技术服务有限公司 | Character recognition method based on image |
| WO2019175686A1 (en) | 2018-03-12 | 2019-09-19 | Ratti Jayant | On-demand artificial intelligence and roadway stewardship system |
| CN108515967A (en) * | 2018-04-25 | 2018-09-11 | 深圳卡安兴科技有限公司 | A kind of anti-collision early warning method and device based on Car license recognition |
| CN108681729A (en) * | 2018-05-08 | 2018-10-19 | 腾讯科技（深圳）有限公司 | Text image antidote, device, storage medium and equipment |
| CN108898137A (en) * | 2018-05-25 | 2018-11-27 | 黄凯 | A kind of natural image character identifying method and system based on deep neural network |
| CN108898137B (en) * | 2018-05-25 | 2022-04-12 | 黄凯 | Natural image character recognition method and system based on deep neural network |
| CN108875819B (en) * | 2018-06-08 | 2020-10-27 | 浙江大学 | Object and component joint detection method based on long-term and short-term memory network |
| CN108875819A (en) * | 2018-06-08 | 2018-11-23 | 浙江大学 | A kind of object and component associated detecting method based on shot and long term memory network |
| CN110619327A (en) * | 2018-06-20 | 2019-12-27 | 湖南省瞬渺通信技术有限公司 | Real-time license plate recognition method based on deep learning in complex scene |
| CN110717575A (en) * | 2018-07-13 | 2020-01-21 | 奇景光电股份有限公司 | Frame buffer free convolutional neural network system and method |
| CN110717575B (en) * | 2018-07-13 | 2022-07-26 | 奇景光电股份有限公司 | Frame buffer free convolutional neural network system and method |
| CN110866430A (en) * | 2018-08-28 | 2020-03-06 | 上海富瀚微电子股份有限公司 | License plate recognition method and device |
| CN110866430B (en) * | 2018-08-28 | 2022-07-01 | 上海富瀚微电子股份有限公司 | License plate recognition method and device |
| CN109409272A (en) * | 2018-10-17 | 2019-03-01 | 北京空间技术研制试验中心 | Cable Acceptance Test System and method based on machine vision |
| CN109543673A (en) * | 2018-10-18 | 2019-03-29 | 浙江理工大学 | A kind of low contrast punching press character recognition algorithm based on Interactive Segmentation |
| CN111178359A (en) * | 2018-11-09 | 2020-05-19 | 中移（杭州）信息技术有限公司 | License plate number recognition method, device and equipment and computer storage medium |
| CN109447030A (en) * | 2018-11-12 | 2019-03-08 | 重庆知遨科技有限公司 | A kind of fire-fighting robot movement real-time instruction algorithm for fire scenario |
| CN109657682B (en) * | 2018-11-29 | 2022-02-11 | 国网河北省电力有限公司电力科学研究院 | Electric energy representation number identification method based on deep neural network and multi-threshold soft segmentation |
| CN109657682A (en) * | 2018-11-29 | 2019-04-19 | 国网河北省电力有限公司电力科学研究院 | A kind of electric energy meter registration recognition methods based on deep neural network and the soft cutting of multi-threshold |
| CN109711407B (en) * | 2018-12-28 | 2023-02-28 | 深圳市捷顺科技实业股份有限公司 | License plate recognition method and related device |
| CN109711407A (en) * | 2018-12-28 | 2019-05-03 | 深圳市捷顺科技实业股份有限公司 | A kind of method and relevant apparatus of Car license recognition |
| CN109740603A (en) * | 2019-01-21 | 2019-05-10 | 闽江学院 | Based on the vehicle character identifying method under CNN convolutional neural networks |
| CN109740603B (en) * | 2019-01-21 | 2020-12-01 | 闽江学院 | Vehicle character recognition method based on CNN convolutional neural network |
| CN110009929A (en) * | 2019-03-15 | 2019-07-12 | 北京筑梦园科技有限公司 | A kind of Vehicle berth management method, equipment and system |
| CN109829460A (en) * | 2019-04-02 | 2019-05-31 | 安徽省安瑞机电科技有限公司 | A kind of License Plate Character Segmentation recognition methods and its system applied to parking management system |
| CN110288626B (en) * | 2019-06-13 | 2021-05-25 | 北京云测信息技术有限公司 | Method and apparatus for detecting text in native electronic images |
| CN110288626A (en) * | 2019-06-13 | 2019-09-27 | 北京云测信息技术有限公司 | The method and apparatus for detecting the text in primary electronic image |
| WO2021044203A1 (en) * | 2019-09-05 | 2021-03-11 | 商汤国际私人有限公司 | Method and apparatus for identifying sequence in image, and electronic device and storage medium |
| AU2019455811B2 (en) * | 2019-09-05 | 2022-06-02 | Sensetime International Pte. Ltd. | Method and apparatus for recognizing sequence in image, electronic device, and storage medium |
| US11308351B2 (en) | 2019-09-05 | 2022-04-19 | Sensetime International Pte. Ltd. | Method and apparatus for recognizing sequence in image, electronic device, and storage medium |
| CN110992725A (en) * | 2019-10-24 | 2020-04-10 | 合肥讯图信息科技有限公司 | Method, system and storage medium for detecting traffic signal lamp fault |
| CN111046746A (en) * | 2019-11-22 | 2020-04-21 | 北京精英路通科技有限公司 | License plate detection method and device |
| CN111444911A (en) * | 2019-12-13 | 2020-07-24 | 珠海大横琴科技发展有限公司 | Training method and device of license plate recognition model and license plate recognition method and device |
| CN111160338A (en) * | 2019-12-24 | 2020-05-15 | 苏州奥易克斯汽车电子有限公司 | Fuzzy license plate recognition method |
| CN111310754A (en) * | 2019-12-31 | 2020-06-19 | 创泽智能机器人集团股份有限公司 | Method for segmenting license plate characters |
| CN111259886A (en) * | 2020-01-08 | 2020-06-09 | 上海眼控科技股份有限公司 | License plate screw detection method, electronic device, computer equipment and storage medium |
| CN111275043B (en) * | 2020-01-22 | 2021-08-20 | 西北师范大学 | Paper numbered musical notation electronization play device based on PCNN handles |
| CN111275043A (en) * | 2020-01-22 | 2020-06-12 | 西北师范大学 | Paper numbered musical notation electronization play device based on PCNN handles |
| CN111489560A (en) * | 2020-04-13 | 2020-08-04 | 深圳市海圳汽车技术有限公司 | Detection method and control method for detecting front vehicle starting based on shallow convolutional neural network |
| CN111523416A (en) * | 2020-04-13 | 2020-08-11 | 北京巨视科技有限公司 | Vehicle early warning method and device based on highway ETC portal |
| CN111611863A (en) * | 2020-04-22 | 2020-09-01 | 浙江大华技术股份有限公司 | License plate image quality evaluation method and device and computer equipment |
| CN111814852A (en) * | 2020-06-24 | 2020-10-23 | 理光软件研究所（北京）有限公司 | Image detection method, image detection device, electronic equipment and computer-readable storage medium |
| CN112997190A (en) * | 2020-12-29 | 2021-06-18 | 深圳市锐明技术股份有限公司 | License plate recognition method and device and electronic equipment |
| CN112997190B (en) * | 2020-12-29 | 2024-01-12 | 深圳市锐明技术股份有限公司 | License plate recognition method and device and electronic equipment |
| CN113313143A (en) * | 2021-04-29 | 2021-08-27 | 浙江大华技术股份有限公司 | License plate detection method and device and computer storage medium |
| CN113313143B (en) * | 2021-04-29 | 2022-08-09 | 浙江大华技术股份有限公司 | License plate detection method and device and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104298976B (en) | 2017-09-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104298976B (en) | Detection method of license plate based on convolutional neural networks | |
| CN110619750B (en) | Intelligent aerial photography identification method and system for illegal parking vehicle | |
| CN101334836B (en) | License plate positioning method incorporating color, size and texture characteristic | |
| CN102708356B (en) | Automatic license plate positioning and recognition method based on complex background | |
| CN109948416A (en) | A kind of illegal occupancy bus zone automatic auditing method based on deep learning | |
| CN102364496B (en) | Method and system for identifying automobile license plates automatically based on image analysis | |
| CN103136528B (en) | A kind of licence plate recognition method based on dual edge detection | |
| CN109255350B (en) | New energy license plate detection method based on video monitoring | |
| CN104392212A (en) | Method for detecting road information and identifying forward vehicles based on vision | |
| CN104036262B (en) | A kind of method and system of LPR car plates screening identification | |
| CN106934374B (en) | Method and system for identifying traffic signboard in haze scene | |
| CN103116751A (en) | Automatic license plate character recognition method | |
| CN103065138A (en) | Recognition method of license plate number of motor vehicle | |
| CN109726717A (en) | A kind of vehicle comprehensive information detection system | |
| CN104700072A (en) | Lane line historical frame recognition method | |
| CN100385452C (en) | Registration number character dividing method | |
| CN104966049A (en) | Lorry detection method based on images | |
| CN103902981A (en) | Method and system for identifying license plate characters based on character fusion features | |
| CN105184301B (en) | A kind of method that vehicle heading is differentiated using four-axle aircraft | |
| CN104102909A (en) | Vehicle characteristic positioning and matching method based on multiple-visual information | |
| CN101246553B (en) | Cab model recognizing method based on triangulation transformation | |
| CN111242002A (en) | Shared bicycle standardized parking judgment method based on computer vision | |
| CN113837094A (en) | Road condition rapid analysis method based on full-color high-resolution remote sensing image | |
| CN106650752A (en) | Vehicle body color recognition method | |
| Wang et al. | License plate location algorithm based on edge detection and morphology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170926 Termination date: 20181016 |