CN109447076A - A kind of vehicle VIN code recognition detection method for vehicle annual test - Google Patents
A kind of vehicle VIN code recognition detection method for vehicle annual test Download PDFInfo
- Publication number
- CN109447076A CN109447076A CN201811104960.7A CN201811104960A CN109447076A CN 109447076 A CN109447076 A CN 109447076A CN 201811104960 A CN201811104960 A CN 201811104960A CN 109447076 A CN109447076 A CN 109447076A
- Authority
- CN
- China
- Prior art keywords
- vin code
- vehicle
- vehicle vin
- character
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Discrimination (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of vehicle VIN code recognition detection methods for vehicle annual test, comprising the following steps: obtains the VIN code model answer in the vehicle VIN code image and server to be detected in vehicle annual test；VIN code region is detected, judges that target whether there is；The VIN code recognition result that image to be detected application Character segmentation model is obtained judges whether VIN code recognition result is consistent with server VIN code model answer content.The present invention realizes the automatic identification verification of VIN code in vehicle annual test, and existing manual examination and verification mode is substituted, has saved manpower, accelerates audit speed, and identification VIN code can be effectively detected, and ensure that the disclosure of examination, just.
Description
Technical field
The present invention relates to the artificial intelligence judgment technology fields of automotive vehicle annual test, in particular to a kind of to be used for vehicle year
The vehicle VIN code recognition detection method of inspection.
Background technique
Constantly improve with living standards of the people with the continuous social and economic development, Urban vehicles poputation rapidly increases
It is long.The workload of automotive vehicle annual test also increases rapidly therewith.Vehicle VIN code recognition detection is main in traditional vehicle annual test
It is by artificial detection, thus often there is desk checking is at high cost, fatiguability, the easily drawbacks such as carelessness influence the standard of detection
True rate and efficiency.
Summary of the invention
The purpose of the present invention is: propose a kind of vehicle VIN code recognition detection method for vehicle annual test, automatic audit is known
Other vehicle VIN code, to meet nowadays the needs of to vehicle annual test working efficiency, accuracy rate.
The technical solution adopted by the present invention to solve the technical problems is:
1. a kind of vehicle VIN code recognition detection method for vehicle annual test, comprising the following steps:
S1, vehicle VIN code image to be detected is obtained；
S2, using the target detection model inspection VIN code image based on deep learning, position and judge that VIN code region exists
It whether there is in image, then recording this mark if it exists is 0, and extracts VIN code region simultaneously；This is then recorded if it does not exist
Mark is 1, and saves picture concerned, into statistical analysis process；
S3, using the Character segmentation model inspection VIN code region based on deep learning, VIN code is divided into single character,
Whether the number for judging character is 17, and it is 0 that this mark is recorded if setting up, and extracts each character picture；If invalid
Recording this mark is 1, and saves picture concerned, into statistical analysis process；
It is S4, for statistical analysis to the result of the action of whole process, flag bit all 0 is recorded, then VIN code recognition detection
Pass through, if it exists mark 1, then the identification of VIN code changes detection and do not pass through；Meanwhile it is obstructed according to the position acquisition verification that mark 1 occurs
The reason of crossing and problem picture.
Judge that VIN code region whether there is in the step S2 and use following method:
The classification information of vehicle VIN code image object to be detected is obtained using Softmax, uses bounding box
Regression obtains the location information of vehicle VIN code image object to be detected.
Target detection model obtaining step in the step S2 based on deep learning is as follows:
S1, training data prepare: the real vehicles VIN code image of different shooting conditions are obtained, so that the number that training prepares
According to being more bonded actual application scenarios；
S2, data mark: the different vehicle VIN code region obtained in S1 is marked in the picture using rectangle frame, rectangle
Frame region domestic demand completely includes vehicle VIN code, while recording the corresponding coordinate of rectangle frame, and finally with the preservation of xml document format
The corresponding coordinate of rectangle frame；
S3, model training: being based on deep learning SSD network, be trained in deep learning frame, while utilizing one
Pre-training model is finely adjusted.
Described be finely adjusted includes the following steps:
S231. the mean value file of vehicle VIN code data set is calculated；
S232. the last output of modification SSD frame；
Basic studies rate is adjusted to 0.0001, weight_decay and is adjusted to 0.0005, learned by S233. regularized learning algorithm rate
Habit rate strategy is set as " multistep ", and gamma is set as 0.1, momentum and is set as 0.9；
S234. load grounding model is finely adjusted.
The vehicle VIN code object detection unit uses this kind of convolutional neural networks of SSD, and is loaded by CAFFE frame.
Character segmentation model obtaining step in the step S3 based on deep learning is as follows:
S31, training data prepare: the vehicle VIN code character image of different shooting conditions are obtained, under the conditions of obtaining nature
Truthful data；
S32, data mark: the single character of S1 vehicle VIN code is marked in the picture using the method that character retouches side, rectangle
Frame region domestic demand completely includes the single character zone of vehicle VIN code, and labeled data is finally stored as xml format；
S33, model training: being based on deep learning SSD network, be trained in deep learning frame, while utilizing one
A pre-training model is finely adjusted.
Described be finely adjusted includes the following steps:
S231. the mean value file of vehicle VIN code data set is calculated；
S232. the last output of modification SSD frame；
Basic studies rate is adjusted to 0.0001, weight_decay and is adjusted to 0.0005, learned by S233. regularized learning algorithm rate
Habit rate strategy is set as " multistep ", and gamma is set as 0.1, momentum and is set as 0.9；
S234. load grounding model is finely adjusted.
The vehicle VIN code object detection unit uses SSD Character segmentation algorithm model, and is loaded by CAFFE frame.
The beneficial effects of the present invention are: present invention is mainly applied to vehicle VIN code recognition detection in automotive vehicle annual test,
Its automatic identification verification for realizing VIN code, while unsanctioned verification image and reason can be passed back to server preservation and stayed
Wait collect evidence.Both manpower has been saved, has in turn ensured the just, openly of verifying work.
Detailed description of the invention
Fig. 1 is vehicle VIN code recognition detection flow chart of the invention.
Fig. 2 is structural schematic diagram of the invention.
Fig. 3 is the structural schematic diagram of object detection unit of the present invention.
Specific embodiment
Below in conjunction with attached drawing, the present invention will be further described.
Present invention is primarily based on module of target detection and determination module.
As shown in Fig. 2, module of target detection is made of object detection unit, Character segmentation unit.Firstly, image is passed to
Object detection unit, application target detection model, obtains vehicle VIN code area image on the image.Character segmentation mould is used again
VIN code is divided into single character by type.Module of target detection detects vehicle VIN code region first, then in vehicle VIN code area
Character segmentation is carried out in domain, this substep detection means can be effectively avoided because of vehicle VIN code areas case complexity bring
Erroneous detection influence, improve vehicle VIN code zone location and Character segmentation accuracy rate, further improve vehicle VIN code detect and
The accuracy of comparison.
The specific detection method of object detection unit include: as shown in figure 3,
The code area image of VIN containing vehicle is input to the SSD target detection based on deep learning first and calculated by S1, detection module
In method model.Wherein SSD (single shot multibox detection) can only by single deep neural network,
Target area in image detected by rectangle frame.Meanwhile compared to Faster RCNN, Mask RCNN algorithm, speed
Degree with it is more outstanding in efficiency
The area image of S2, VIN containing vehicle code will be loaded SSD target detection in S1 and be calculated by the CAFFE frame of open source
Method model is compiled decoding to image automatically, obtained output the result is that N number of one-dimension array [class, x, y, width,
height].Wherein, first element class of each one-dimension array represents object type, is vehicle VIN code then for 1, is not then
It is 0, rectangular area, x, y represent rectangle upper left angular coordinate where four element characterization target objects after array, and width is represented
Rectangle width, height represent rectangular elevation.
S3, N number of array obtained correspond to a region.In last steps in decision-making, using region rectangle frame face
The strategy of product size building region distance information, is exported using the maximum array of rectangle frame area as detection module, is then passed through
Rectangle frame location information extracts vehicle VIN code region from image.The method may finally effectively pick out other dry in background
Disturb region.
Target detection model acquisition methods are as follows:
S21, training data prepare: obtaining the real vehicles VIN code image of different shooting conditions (such as illumination, angle), make
The data prepared must be trained more to be bonded actual application scenarios；
S22, data mark: the different vehicle VIN code region obtained in S1 is marked in the picture using rectangle frame, rectangle
Frame region domestic demand completely includes vehicle VIN code, while recording the corresponding coordinate of rectangle frame, and finally with the preservation of xml document format
The corresponding coordinate of rectangle frame；
S23, model training: being based on deep learning SSD (single shot multibox detection) network,
It is trained in CAFFE deep learning frame, while being finely adjusted using a pre-training model.Wherein pre-training model is
The CAFFE model obtained is trained based on ILSVRC image data base using VGG16 network.We by S2 mark data with
And input data of the VIN code image as deep learning in S1, it is trained.Pass through the pre-training model of VGG16, SSD net
Network can carry out transfer learning based on the image features learnt, allow to carry out quick using a small amount of data
Fine tuning training, to reach high accuracy.Wherein, the selection of fine tuning can increase learning rate with let us, and improve depth
The hyper parameter of habit, in this way, the step of passing through fine tuning, we can will improve 40% the training time, precision improvement 4%.Fine tuning
Specific steps are as follows: firstly, the mean value file of vehicle VIN code data set is calculated, because of the mean value file of vehicle VIN code data set
It is not quite alike with the mean value file of ImageNet data set.Then, the last output of modification SSD frame, ImageNet is one
The classification task of 1000 classes, and this model only has 2 classes, is background and vehicle VIN code respectively.It is followed by regularized learning algorithm rate, is learnt
Rate is very big corresponding to the performance influence of neural network, but by experience and can only test to obtain suitable value, by testing,
Basic studies rate is adjusted to 0.0001, weight_decay and is adjusted to 0.0005 by this patent, and learning rate strategy is set as
" multistep ", gamma is set as 0.1, momentum and is set as 0.9.Finally, load grounding model is finely adjusted.It is logical
Above-mentioned setting is crossed, model can restrain quickly and loss drops to 1.8 or so, improves the efficiency and precision of model training.Finally, logical
The hyper parameter in adjustment learning rate and deep learning is crossed, multiple experiment and test are carried out, obtains the highest target inspection of accuracy
Survey model；
The specific detection method of Character segmentation unit include: as shown in figure 3,
Vehicle VIN code area image is input to the SSD Character segmentation algorithm based on deep learning by S1, detection module first
In model.Wherein SSD (single shot multibox detection) can be only by single deep neural network, will
Target area in image detected by rectangle frame.Meanwhile compared to Faster RCNN, Mask RCNN algorithm, speed
With it is more outstanding in efficiency
S2, image will be loaded character partitioning algorithm model, be compiled automatically to image by the CAFFE frame of open source
It deciphers code to obtain N number of one-dimension array [class, x, y, width, height], first element of array represents object type, vehicle
Character in VIN code has " 0~9 ", " A~N ", " P " and " R~Z " and totally 34 kinds, indicates class, array with 0~33 respectively
Rectangular area, x, y represent rectangle upper left angular coordinate where four element characterization target objects afterwards, and width represents rectangle width,
Height represents rectangular elevation.
S3, wherein N number of array correspond to a region, extract vehicle VIN code from image by rectangle frame location information
Single character position.The method can effectively pick out other interference regions in background.
Character segmentation model acquisition methods are as follows:
S31, training data prepare: the vehicle VIN code character image of different shooting conditions (such as illumination, angle) is obtained, with
Obtain the truthful data under the conditions of nature；
S32, data mark: the single character of S1 vehicle VIN code is marked in the picture using the method that character retouches side, rectangle
Frame region domestic demand completely includes the single character zone of vehicle VIN code, and labeled data is finally stored as xml format；
S33, model training: being based on deep learning SSD (single shot multibox detection) network,
It is trained in CAFFE deep learning frame, while being finely adjusted using a pre-training model.Wherein pre-training model is
The CAFFE model obtained is trained based on ILSVRC image data base using VGG16 network.We by S2 mark data with
And input data of the VIN code character image as deep learning in S1, it is trained.By the pre-training model of VGG16,
SSD network can carry out transfer learning based on the image features learnt, allow to using a small amount of data, quickly
It is finely adjusted training, to reach high accuracy.The specific steps of fine tuning are as follows: firstly, calculating vehicle VIN code character data set
Mean value file, because the mean value file of vehicle VIN code character data set and the mean value file of ImageNet data set are different.
Then, the last output of modification SSD frame, ImageNet is the classification task of 1000 classes, and this model only has 2 classes, point
It is not background and vehicle VIN code character.It is followed by regularized learning algorithm rate, the performance influence that learning rate corresponds to neural network is very big,
But can only be by experience and experiment to obtain suitable value, by experiment, basic studies rate is adjusted to 0.0001 by this patent,
Weight_decay is adjusted to 0.0008, and learning rate strategy is set as " multistep ", and gamma is set as 0.1, momentum
It is set as 0.9.Finally, load grounding model is finely adjusted.By above-mentioned setting, model can restrain quickly and loss is dropped to
1.5 or so, improve the efficiency and precision of model training.Meanwhile learning rate and hyper parameter by adjusting deep learning, it carries out
Test and experiment, obtain optimal Character segmentation model；
Then the VIN code of image to be detected is obtained using Character segmentation model, and is answered with the standard of the VIN code in server
Case judges whether content is consistent.
Vehicle VIN code recognition detection standard of the invention is as follows: vehicle VIN code region whether there is in image to be detected；
Whether VIN code character number is 17；The character standard answer of the character identification result and archival image of image to be detected whether one
It causes；The present invention indicates verification state using an one-dimension array [x1, x2, x3], and initial value is [0,0,0], and flag bit x1 is represented
Vehicle VIN code region whether there is, and then x1 is 0 if it exists, and then x1 is 1 if it does not exist；Flag bit x2 represents VIN code character number
It whether is 17, if then x2 is 0, if not being 1 in then x2；Flag bit x3 represents the character identification result of image to be detected and deposits
Whether the model answer of the character of shelves image is consistent, and x3 is 0 if consistent, if inconsistent x3 is 1.Finally, statistical mark position
State, if mark is is 0, verification passes through, if it exists 1, then it verifies and does not pass through.It can be obtained according to the position that state 1 occurs
To the unsanctioned reason of verification.If x1 is 1, vehicle VIN code region may be not present in image or shooting angle does not meet rule
It is fixed；If x2 is 1, possible reason is that image taking is imperfect；If x3 is 1, the character and archive photo of image to be detected
Character do not correspond to, it is understood that there may be the case where distorting.
Determination module judges whether vehicle VIN code recognition detection passes through according to verification standard, and school is directly returned to if passing through
Success flag is tested, it is careful to remain the later period for the position back-checking failure cause and corresponding picture for being 1 according to flag bit if not passing through
Verify card.
Implementation detailed process of the invention is as shown in Figure 1, a kind of vehicle VIN code tampering detection side for vehicle annual test
Method includes the following steps:
S1, vehicle VIN code image to be detected is obtained；
S2, using the target detection model inspection VIN code image based on deep learning, judge that VIN code region whether there is,
Then recording this mark if it exists is 0, extracts VIN code region；Then recording this mark if it does not exist is 1, and saves related figure
Piece, into statistical analysis process；
S3, using the Character segmentation model inspection VIN code region based on deep learning, VIN code is divided into single character,
Whether the number for judging character is 17, and it is 0 that this mark is recorded if setting up, and extracts each character picture；If invalid
Recording this mark is 1, and saves picture concerned, into statistical analysis process；
It is S4, for statistical analysis to the result of the action of whole process, flag bit all 0 is recorded, then VIN code recognition detection
Pass through, if it exists mark 1, then the identification of VIN code changes detection and do not pass through；Meanwhile it is obstructed according to the position acquisition verification that mark 1 occurs
The reason of crossing and problem picture；
The advantages of basic principles and main features and this programme of this programme have been shown and described above.The technology of the industry
Personnel are it should be appreciated that this programme is not restricted to the described embodiments, and the above embodiments and description only describe this
The principle of scheme, under the premise of not departing from this programme spirit and scope, this programme be will also have various changes and improvements, these changes
Change and improvement is both fallen within the scope of claimed this programme.This programme be claimed range by appended claims and its
Equivalent thereof.
Claims (8)
1. a kind of vehicle VIN code recognition detection method for vehicle annual test, which comprises the following steps:
S1, vehicle VIN code image to be detected is obtained；
S2, using the target detection model inspection VIN code image based on deep learning, position and judge VIN code region in image
In whether there is, record this mark then if it exists as 0, and extract VIN code region simultaneously；This mark is then recorded if it does not exist
It is 1, and saves picture concerned, into statistical analysis process；
S3, using the Character segmentation model inspection VIN code region based on deep learning, VIN code is divided into single character, judge
Whether the number of character is 17, and it is 0 that this mark is recorded if setting up, and extracts each character picture；It is recorded if invalid
This mark is 1, and saves picture concerned, into statistical analysis process；
It is S4, for statistical analysis to the result of the action of whole process, flag bit all 0 is recorded, then VIN code recognition detection leads to
It crosses, if it exists mark 1, then the identification of VIN code changes detection and do not pass through；Meanwhile the position acquisition verification occurred according to mark 1 does not pass through
The reason of and problem picture.
2. a kind of vehicle VIN code recognition detection method for vehicle annual test as described in claim 1, which is characterized in that institute
It states and judges that VIN code region whether there is using following method in step S2:
The classification information of vehicle VIN code image object to be detected is obtained using Softmax, uses bounding box
Regression obtains the location information of vehicle VIN code image object to be detected.
3. a kind of vehicle VIN code recognition detection method for vehicle annual test as described in claim 1, which is characterized in that institute
The target detection model obtaining step stated in step S2 based on deep learning is as follows:
S1, training data prepare: the real vehicles VIN code image of different shooting conditions are obtained, so that the data that training prepares are more
It is bonded actual application scenarios；
S2, data mark: the different vehicle VIN code region obtained in S1 is marked in the picture using rectangle frame, rectangle frame area
Domain domestic demand completely includes vehicle VIN code, while recording the corresponding coordinate of rectangle frame, and finally saves rectangle with xml document format
The corresponding coordinate of frame；
S3, model training: being based on deep learning SSD network, be trained in deep learning frame, while utilizing a pre- instruction
Practice model to be finely adjusted.
4. a kind of vehicle VIN code recognition detection method for vehicle annual test as claimed in claim 3, which is characterized in that institute
It states to be finely adjusted and include the following steps:
S231. the mean value file of vehicle VIN code data set is calculated；
S232. the last output of modification SSD frame；
Basic studies rate is adjusted to 0.0001, weight_decay and is adjusted to 0.0005 by S233. regularized learning algorithm rate, learning rate
Strategy is set as " multistep ", and gamma is set as 0.1, momentum and is set as 0.9；
S234. load grounding model is finely adjusted.
5. a kind of vehicle VIN code recognition detection method for vehicle annual test as claimed in claim 1 or 2 or 3 or 4, special
Sign is that the vehicle VIN code object detection unit uses this kind of convolutional neural networks of SSD, and is loaded by CAFFE frame.
6. a kind of vehicle VIN code recognition detection method for vehicle annual test as described in claim 1, which is characterized in that institute
The Character segmentation model obtaining step stated in step S3 based on deep learning is as follows:
S31, training data prepare: the vehicle VIN code character image of different shooting conditions is obtained, it is true under the conditions of nature to obtain
Real data；
S32, data mark: the single character of S1 vehicle VIN code is marked in the picture using the method that character retouches side, rectangle frame area
Domain domestic demand completely includes the single character zone of vehicle VIN code, and labeled data is finally stored as xml format；
S33, model training: being based on deep learning SSD network, be trained in deep learning frame, while pre- using one
Training pattern is finely adjusted.
7. a kind of vehicle VIN code recognition detection method for vehicle annual test as claimed in claim 6, which is characterized in that institute
It states to be finely adjusted and include the following steps:
S231. the mean value file of vehicle VIN code data set is calculated；
S232. the last output of modification SSD frame；
Basic studies rate is adjusted to 0.0001, weight_decay and is adjusted to 0.0005 by S233. regularized learning algorithm rate, learning rate
Strategy is set as " multistep ", and gamma is set as 0.1, momentum and is set as 0.9；
S234. load grounding model is finely adjusted.
8. a kind of vehicle VIN code recognition detection method for vehicle annual test as claimed in claims 6 or 7, which is characterized in that
The vehicle VIN code object detection unit uses SSD Character segmentation algorithm model, and is loaded by CAFFE frame.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811104960.7A CN109447076A (en) | 2018-09-20 | 2018-09-20 | A kind of vehicle VIN code recognition detection method for vehicle annual test |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811104960.7A CN109447076A (en) | 2018-09-20 | 2018-09-20 | A kind of vehicle VIN code recognition detection method for vehicle annual test |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109447076A true CN109447076A (en) | 2019-03-08 |
Family
ID=65533084
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811104960.7A Pending CN109447076A (en) | 2018-09-20 | 2018-09-20 | A kind of vehicle VIN code recognition detection method for vehicle annual test |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109447076A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110245583A (en) * | 2019-05-27 | 2019-09-17 | 上海眼控科技股份有限公司 | A kind of intelligent identification Method of Vehicular exhaust survey report |
| CN110276295A (en) * | 2019-06-20 | 2019-09-24 | 上海眼控科技股份有限公司 | Vehicle identification number detection recognition method and equipment |
| CN110348360A (en) * | 2019-07-04 | 2019-10-18 | 上海眼控科技股份有限公司 | A kind of examining report recognition methods and equipment |
| CN110458070A (en) * | 2019-08-01 | 2019-11-15 | 上海眼控科技股份有限公司 | Method and system based on motor vehicle annual test check table picture recognition amount of testing |
| CN110503102A (en) * | 2019-08-27 | 2019-11-26 | 上海眼控科技股份有限公司 | Vehicle identification code detection method, device, computer equipment and storage medium |
| CN110598687A (en) * | 2019-09-18 | 2019-12-20 | 上海眼控科技股份有限公司 | Vehicle identification code detection method and device and computer equipment |
| CN110619330A (en) * | 2019-09-18 | 2019-12-27 | 上海眼控科技股份有限公司 | Recognition model training method and device, computer equipment and recognition method |
| CN111027532A (en) * | 2019-12-11 | 2020-04-17 | 上海眼控科技股份有限公司 | System and method for identifying tax amount of insurance policy vehicle and ship for forced insurance |
| CN111104934A (en) * | 2019-12-22 | 2020-05-05 | 上海眼控科技股份有限公司 | Engine label detection method, electronic device and computer readable storage medium |
| CN111126043A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code multiple checking method, VIN code identification method and device |
| CN111126044A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code multiple checking method using confidence coefficient, VIN code identification method and device |
| CN111126042A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN (vehicle identification number) code verification method by using confidence coefficient, and VIN code identification method and device |
| CN111126041A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code dictionary base checking method, VIN code identification method and device |
| CN111160342A (en) * | 2019-12-30 | 2020-05-15 | 上海眼控科技股份有限公司 | Vehicle reflective mark verification method and device, computer equipment and storage medium |
| CN111401362A (en) * | 2020-03-06 | 2020-07-10 | 上海眼控科技股份有限公司 | Tampering detection method, device, equipment and storage medium for vehicle VIN code |
| CN111507332A (en) * | 2020-04-17 | 2020-08-07 | 上海眼控科技股份有限公司 | Vehicle VIN code detection method and equipment |
| CN111814576A (en) * | 2020-06-12 | 2020-10-23 | 上海品览数据科技有限公司 | Shopping receipt picture identification method based on deep learning |
| CN113610083A (en) * | 2021-08-13 | 2021-11-05 | 天津大学 | Character recognition and character engraving depth detection system and detection method for vehicle VIN code |
| CN113642556A (en) * | 2021-08-04 | 2021-11-12 | 五八有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN114359201A (en) * | 2021-12-29 | 2022-04-15 | 丽台(上海)信息科技有限公司 | Method for detecting quality of engraving identifier on automobile production line |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120277949A1 (en) * | 2011-04-29 | 2012-11-01 | Toyota Motor Engin. & Manufact. N.A.(TEMA) | Collaborative multi-agent vehicle fault diagnostic system & associated methodology |
| CN106611162A (en) * | 2016-12-20 | 2017-05-03 | 西安电子科技大学 | Method for real-time detection of road vehicle based on deep learning SSD frame |
| CN107451511A (en) * | 2016-05-31 | 2017-12-08 | 优信拍（北京）信息科技有限公司 | A kind of system, method based on vehicle information extraction OBD data |
| CN107818322A (en) * | 2017-10-13 | 2018-03-20 | 上海眼控科技股份有限公司 | A kind of vehicle VIN code tampering detection system and methods for vehicle annual test |
| WO2018072233A1 (en) * | 2016-10-20 | 2018-04-26 | 中山大学 | Method and system for vehicle tag detection and recognition based on selective search algorithm |
| CN108133023A (en) * | 2017-12-27 | 2018-06-08 | 上海唯链信息科技有限公司 | A kind of automotive check information system based on block chain technology |
-
2018
- 2018-09-20 CN CN201811104960.7A patent/CN109447076A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120277949A1 (en) * | 2011-04-29 | 2012-11-01 | Toyota Motor Engin. & Manufact. N.A.(TEMA) | Collaborative multi-agent vehicle fault diagnostic system & associated methodology |
| CN107451511A (en) * | 2016-05-31 | 2017-12-08 | 优信拍（北京）信息科技有限公司 | A kind of system, method based on vehicle information extraction OBD data |
| WO2018072233A1 (en) * | 2016-10-20 | 2018-04-26 | 中山大学 | Method and system for vehicle tag detection and recognition based on selective search algorithm |
| CN106611162A (en) * | 2016-12-20 | 2017-05-03 | 西安电子科技大学 | Method for real-time detection of road vehicle based on deep learning SSD frame |
| CN107818322A (en) * | 2017-10-13 | 2018-03-20 | 上海眼控科技股份有限公司 | A kind of vehicle VIN code tampering detection system and methods for vehicle annual test |
| CN108133023A (en) * | 2017-12-27 | 2018-06-08 | 上海唯链信息科技有限公司 | A kind of automotive check information system based on block chain technology |
Non-Patent Citations (2)
| Title |
|---|
| 戚永军等: "机动车VIN码信息管理系统的设计与实现", 《北华航天工业学院学报》 * |
| 阮航等: "基于Faster R-CNN的车辆多属性识别", 《计算机技术与发展》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110245583A (en) * | 2019-05-27 | 2019-09-17 | 上海眼控科技股份有限公司 | A kind of intelligent identification Method of Vehicular exhaust survey report |
| CN110276295A (en) * | 2019-06-20 | 2019-09-24 | 上海眼控科技股份有限公司 | Vehicle identification number detection recognition method and equipment |
| CN110348360A (en) * | 2019-07-04 | 2019-10-18 | 上海眼控科技股份有限公司 | A kind of examining report recognition methods and equipment |
| CN110348360B (en) * | 2019-07-04 | 2020-11-24 | 上海眼控科技股份有限公司 | Detection report identification method and equipment |
| CN110458070A (en) * | 2019-08-01 | 2019-11-15 | 上海眼控科技股份有限公司 | Method and system based on motor vehicle annual test check table picture recognition amount of testing |
| CN110503102A (en) * | 2019-08-27 | 2019-11-26 | 上海眼控科技股份有限公司 | Vehicle identification code detection method, device, computer equipment and storage medium |
| CN110598687A (en) * | 2019-09-18 | 2019-12-20 | 上海眼控科技股份有限公司 | Vehicle identification code detection method and device and computer equipment |
| CN110619330A (en) * | 2019-09-18 | 2019-12-27 | 上海眼控科技股份有限公司 | Recognition model training method and device, computer equipment and recognition method |
| CN111027532A (en) * | 2019-12-11 | 2020-04-17 | 上海眼控科技股份有限公司 | System and method for identifying tax amount of insurance policy vehicle and ship for forced insurance |
| CN111104934A (en) * | 2019-12-22 | 2020-05-05 | 上海眼控科技股份有限公司 | Engine label detection method, electronic device and computer readable storage medium |
| CN111126041A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code dictionary base checking method, VIN code identification method and device |
| CN111126043A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code multiple checking method, VIN code identification method and device |
| CN111126044A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN code multiple checking method using confidence coefficient, VIN code identification method and device |
| CN111126042A (en) * | 2019-12-24 | 2020-05-08 | 北京中安未来科技有限公司 | VIN (vehicle identification number) code verification method by using confidence coefficient, and VIN code identification method and device |
| CN111126043B (en) * | 2019-12-24 | 2023-04-11 | 北京中安未来科技有限公司 | VIN code multiple checking method, VIN code identification method and device |
| CN111126041B (en) * | 2019-12-24 | 2023-04-11 | 北京中安未来科技有限公司 | VIN code dictionary base checking method, VIN code identification method and device |
| CN111160342A (en) * | 2019-12-30 | 2020-05-15 | 上海眼控科技股份有限公司 | Vehicle reflective mark verification method and device, computer equipment and storage medium |
| CN111401362A (en) * | 2020-03-06 | 2020-07-10 | 上海眼控科技股份有限公司 | Tampering detection method, device, equipment and storage medium for vehicle VIN code |
| CN111507332A (en) * | 2020-04-17 | 2020-08-07 | 上海眼控科技股份有限公司 | Vehicle VIN code detection method and equipment |
| CN111814576A (en) * | 2020-06-12 | 2020-10-23 | 上海品览数据科技有限公司 | Shopping receipt picture identification method based on deep learning |
| CN113642556A (en) * | 2021-08-04 | 2021-11-12 | 五八有限公司 | Image processing method and device, electronic equipment and storage medium |
| CN113610083A (en) * | 2021-08-13 | 2021-11-05 | 天津大学 | Character recognition and character engraving depth detection system and detection method for vehicle VIN code |
| CN113610083B (en) * | 2021-08-13 | 2023-07-25 | 天津大学 | Character recognition and character engraving depth detection system and method for vehicle VIN code |
| CN114359201A (en) * | 2021-12-29 | 2022-04-15 | 丽台(上海)信息科技有限公司 | Method for detecting quality of engraving identifier on automobile production line |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109447076A (en) | A kind of vehicle VIN code recognition detection method for vehicle annual test | |
| CN109344835A (en) | Altering detecting method based on vehicle VIN code character position | |
| CN109784326A (en) | A kind of vehicle chassis detection method based on deep learning | |
| CN103258213B (en) | A kind of for the dynamic vehicle model recognizing method in intelligent transportation system | |
| CN108875600A (en) | A kind of information of vehicles detection and tracking method, apparatus and computer storage medium based on YOLO | |
| CN107818322A (en) | A kind of vehicle VIN code tampering detection system and methods for vehicle annual test | |
| CN105574550A (en) | Vehicle identification method and device | |
| Zhang et al. | A GANs-based deep learning framework for automatic subsurface object recognition from ground penetrating radar data | |
| CN109858414A (en) | A kind of invoice piecemeal detection method | |
| CN103258214A (en) | Remote sensing image classification method based on image block active learning | |
| CN104166841A (en) | Rapid detection identification method for specified pedestrian or vehicle in video monitoring network | |
| CN104298992B (en) | A kind of adaptive scale pedestrian recognition methods again based on data-driven | |
| CN111368690A (en) | Deep learning-based video image ship detection method and system under influence of sea waves | |
| CN103208008A (en) | Fast adaptation method for traffic video monitoring target detection based on machine vision | |
| CN108734210A (en) | A kind of method for checking object based on cross-module state multi-scale feature fusion | |
| Yuan et al. | Learning to count buildings in diverse aerial scenes | |
| CN108399378A (en) | A kind of natural scene image recognition methods based on VGG depth convolutional networks | |
| CN103530590A (en) | DPM (direct part mark) two-dimensional code recognition system | |
| CN110288006A (en) | A kind of license plate number automatic Verification method and system | |
| CN104777176A (en) | PCB detection method and device | |
| CN108776777A (en) | The recognition methods of spatial relationship between a kind of remote sensing image object based on Faster RCNN | |
| CN107958200A (en) | A kind of intelligent checking system and method for car repacking exhaust fan | |
| CN110245583A (en) | A kind of intelligent identification Method of Vehicular exhaust survey report | |
| Han et al. | Vision-based crack detection of asphalt pavement using deep convolutional neural network | |
| CN109376738A (en) | Altering detecting method based on vehicle VIN code start-stop symbol type |
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 | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20230728 |
|
| AD01 | Patent right deemed abandoned |