CN108090429A - Face bayonet model recognizing method before a kind of classification - Google Patents
Face bayonet model recognizing method before a kind of classification Download PDFInfo
- Publication number
- CN108090429A CN108090429A CN201711293529.7A CN201711293529A CN108090429A CN 108090429 A CN108090429 A CN 108090429A CN 201711293529 A CN201711293529 A CN 201711293529A CN 108090429 A CN108090429 A CN 108090429A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- brand
- result
- classification
- confidence level
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses face bayonet model recognizing methods before a kind of classification.The present invention is using front face detection location technology, vehicle-logo location technology as foundation stone, using multiple depth convolutional neural networks integrated studies, realize that (wherein vehicle year money includes logo and vehicle for the identification of the brand logo of vehicle, vehicle, vehicle year money these three tasks respectively, vehicle includes logo), the similar vehicle confusion matrix and the inclusion relation of three ranks come out then in conjunction with output confidence level, the priori of each rank Study strategies and methods exports the vehicle cab recognition result of robust.The vehicle that the present invention identifies can be specific to the brand, model and time of vehicle；By the integrated of three graders, vehicle vehicle precision can be further promoted.
Description
Technical field
The invention belongs to technical field of video monitoring, are related to face bayonet model recognizing method before a kind of classification.
Background technology
With the continuous development of Chinese national economy, traffic administration has become to get in the economy of people, social activities
Come more important.People also proposed more and more new demands to the problems such as level, quality of traffic administration.Therefore, to intelligence
The further investigation of traffic becomes more significant.
Specific objective vehicle is found in substantial amounts of traffic block port video data, is had for traffic administration and video investigation
Important meaning.Traditional technological means is retrieved and positioned by identifying vehicle identification number, although Car license recognition accuracy rate is higher,
But under the conditions of deck, unlicensed, license plate shading etc., such technology substantially can not be practical.Therefore, the vehicle cab recognition based on image
Technology is more worth and practicability.
Patent based on such technology has《Model recognizing method and device -201410381923.6》、《A kind of vehicle is known
Other method and system -201410313009.8》、《Model recognizing method -201410009098.7 based on front face feature》And
《A kind of model recognizing method -201510071919.4 based on convolutional neural networks》.
The defects of such technology：
Some uses manual features extracting method in these technologies, but this category feature is highly dependent upon the priori of people, robust
Property is poor；Some extracts robust features by deep learning, then identifies vehicle using the method for template matches, the method is deposited
In template limitation, matching precision difference and inefficiency problem；Some learns off-line training using convolutional neural networks, but only
Can existing brand and vehicle in recognition training storehouse, it is helpless for the new model that does not stop listing.
The content of the invention
In view of the deficiencies of the prior art, the present invention provides face bayonet model recognizing methods before a kind of classification.
The present invention is using front face detection location technology, vehicle-logo location technology as foundation stone, using multiple depth convolutional Neural nets
Network integrated study realizes the identification of the brand logo of vehicle, vehicle, vehicle year money these three tasks, then in conjunction at different levels respectively
The similar vehicle confusion matrix and three including for rank that output confidence level, the priori of other Study strategies and methods come out close
System exports the vehicle cab recognition result of robust.
The method of the present invention mainly includes the following steps that：
1st step：Analyze traffic block port image, using based on the method that depth convolutional neural networks are classified to the front face in image
Do coarse positioning.
2nd step：To the front face region that the 1st step is oriented, the feature based on depth convolutional neural networks detection front face
Point position, amounts at 11 points including logo, car light, windshield, car plate, Chinese herbaceous peony bumper lower edge.
3rd step：11 points oriented according to the 2nd step become more meticulous take front face bounding box and by priori respectively
Relative position relation takes out rough car mark region.
4th step：Car mark region and front face bounding box zoom to fixed size respectively, and car mark region is input to vehicle respectively
Mark grader, front face bounding box are input in vehicle and vehicle year money grader, obtain the brand, vehicle, vehicle of this vehicle
Year money ranking results and corresponding confidence level.
5th step：From top to bottom merge three classifier results；According in advance come out confidence level high-low threshold value and
Similar vehicle confusion matrix merges three one consistent vehicle result of output.
Beneficial effects of the present invention：The vehicle that the present invention identifies can be specific to the brand, model and time of vehicle；Pass through
Three graders it is integrated, can further promote vehicle vehicle precision.
Description of the drawings
Fig. 1 is the method for the present invention flow chart.
Fig. 2 merges flow chart for classification results.
Specific embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, clear, complete description is carried out to the technical solution in the embodiment of the present invention, it is clear that described embodiment is only
Only it is part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiment of the present invention, ordinary skill people
The every other embodiment that member is obtained without making creative work, belongs to the scope of protection of the invention.
Fig. 1 illustrates model recognizing method flow diagram provided in an embodiment of the present invention, the vehicle cab recognition of the present embodiment
Method is included with lower part：
The first step：A kind of front face rough localization method of base depth convolutional neural networks classification；Specially：
(1) zooms to fixed size size (the present embodiment uses 224*244 sizes), using 16 layers of convolution to input picture
Neutral net extracts the feature of image, this feature is used for doing subsequent candidate frame extraction and classification (vehicle and Fei Che).
(2) candidate frames extract；Candidate frame is generated according to the feature extracted in (1).
(3) integrates the feature in candidate frame and (1), generates candidate frame characteristic pattern.
(4) the candidate frame feature in (3) is input to a sorter network by, and it is that vehicle is also non-vehicle to judge it.
Second step：A kind of vehicle characteristics independent positioning method of base depth convolutional neural networks classification；Specially：
(1) each car detected in the first step is taken out and zooms to fixed size (the present embodiment use by from artwork
40*40 sizes).
(2) returns the picture taken out 5 layers of convolutional neural networks of input, orients 11 characteristic points of vehicle.
3rd step：The front face that becomes more meticulous position and take car mark region；Specially：
(1) extracts bounding box, and expands 10% as the tight of vehicle using wide height is each to 11 characteristic points oriented in second step
Cause bounding box.
(2) takes two headlight points each with the half of logo point average distance as wide respectively centered on logo point
And height, obtain rough car mark region.
4th step：Brand, vehicle (including brand and vehicle), vehicle year money (including brand, vehicle and year money) identification；Tool
Body is：
(1) car mark region in the 3rd step is zoomed to fixed size (the present embodiment uses 64*64 sizes) by, is then input to
Classify in one 5 layers of convolutional neural networks, obtain brand result.
(2) the bounding box region of compacting of the front face in the 3rd step is taken out by, and zooms to fixed size (this implementation
Example uses 112*112 sizes), it inputs into 10 layers of convolutional neural networks of a multitask, respectively obtains vehicle and vehicle year money
And confidence level as a result.
5th step：Final vehicle result is merged to obtain to vehicle brand, vehicle, vehicle year money (see Fig. 2)；Specially：
(1) judges whether vehicle year money confidence level (probability for belonging to current class of grader output) is more than or equal to high threshold
(the present embodiment is set to 0.65) if it is directly exports vehicle year money and terminates, otherwise into next step.
(2) judges whether vehicle year money confidence level is more than or equal to Low threshold (the present embodiment is set to 0.14), if it is
To brand, vehicle and the money corresponding brand ballot of vehicle year, see whether two tickets and more than brand, if there is then confirming product
Board does not confirm then；If vehicle year money confidence level is less than Low threshold, brand is not also confirmed.
(3) if (2) step can not confirm brand, directly current brand is provided by logo classifier result.
(4) after the brand that is confirmed, in vehicle classification device result, find maximum probability under this brand as a result,
It observes whether its confidence level is more than or equal to threshold value (the present embodiment is set to 0.05), confirms vehicle if meeting, if being unsatisfactory for not
Confirm.
(5) is according to the result (brand or brand and vehicle) currently having confirmed that, is selected in vehicle year money classifier result
Selecting the maximum probability under known results, (the present embodiment is set as a result, whether observation maximum probability result confidence level is more than or equal to threshold value
0.05), to export complete result (brand, vehicle and year money) if meeting, only being exported if being unsatisfactory for and have confirmed that knot
Fruit.
Above-mentioned threshold value in the sample of each grader of training all by coming out.
To sum up, the present invention can robustly identify the brand and vehicle of unknown newly-increased vehicle, and recognize this rank and had
There is important practical value.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention, should
Understand, the present invention is not limited to implementation as described herein, the purpose of these implementations description is to help this field
In technical staff put into practice the present invention.
Claims (5)
1. face bayonet model recognizing method before a kind of classification, it is characterised in that this method comprises the following steps：
1st step：Analyze traffic block port image, using based on the method that depth convolutional neural networks are classified to the front face in image
Do coarse positioning；
2nd step：To the front face region that the 1st step is oriented, the characteristic point position based on depth convolutional neural networks detection front face
It puts；
3rd step：According to the characteristic point position of the 2nd step, become more meticulous take front face bounding box and the opposite position by priori respectively
The relation of putting takes out rough car mark region；
4th step：Car mark region and front face bounding box zoom to fixed size respectively, and car mark region is input to logo point respectively
Class device, front face bounding box are input in vehicle and vehicle year money grader, obtain the brand, vehicle, vehicle year money of this vehicle
Ranking results and corresponding confidence level；
5th step：From top to bottom merge three classifier results；According to coming out confidence level high-low threshold value and similar in advance
Vehicle confusion matrix merges one consistent vehicle of output as a result, being specifically to three：
(1) judges whether vehicle year money confidence level is more than or equal to high threshold, and if it is directly output vehicle year money terminates, no
Then into next step；
(2) judges whether vehicle year money confidence level is more than or equal to Low threshold, if it is to brand, vehicle and vehicle year money pair
Answer brand ballot, see whether two tickets and more than brand, if there is then confirming brand, do not confirm then；If vehicle
Type year money confidence level be less than Low threshold, then also do not confirm brand；
(3) if (2) step can not confirm brand, directly current brand is provided by logo classifier result；
(4) after the brand that is confirmed, in vehicle classification device result, the maximum probability under this brand is found as a result, observing it
Whether confidence level is more than or equal to the 3rd threshold value, confirms vehicle if meeting, is unsatisfactory for, does not confirm；
(5) according to currently having confirmed that as a result, selecting the maximum probability knot under known results in vehicle year money classifier result
Fruit, whether observation maximum probability result confidence level is more than or equal to the 4th threshold value, if exported if meeting complete as a result, not
Meet then only to export and have confirmed that result.
2. face bayonet model recognizing method before a kind of classification according to claim 1, it is characterised in that：1st step is specifically：
(1-1) zooms to fixed size size to input picture, and the feature of image is extracted using 16 layers of convolutional neural networks,
This feature is used for doing subsequent candidate frame extraction and classification；
(1-2) candidate frames extract；Candidate frame is generated according to the feature extracted in (1-1)；
(1-3) integrates the feature of extraction in candidate frame and (1-1), generates candidate frame characteristic pattern；
Candidate frame characteristic pattern in (1-3) is input to a sorter network by (1-4), and it is that vehicle is also non-vehicle to judge it.
3. face bayonet model recognizing method before a kind of classification according to claim 1, it is characterised in that：2nd step is specifically：
The each car detected in 1st step is taken out and zooms to fixed size by (2-1) from artwork；
(2-2) returns the picture taken out 5 layers of convolutional neural networks of input, orients 11 characteristic points of vehicle, wraps
Include logo, car light, windshield, car plate, Chinese herbaceous peony bumper lower edge.
4. face bayonet model recognizing method before a kind of classification according to claim 1, it is characterised in that：3rd step is specifically：
(3-1) extracts bounding box, and expands 10% encirclement of compacting as vehicle using wide height is each to the characteristic point determined in the 2nd step
Box；
(3-2) centered on logo point, take respectively two headlight points each with the half of logo point average distance as wide and
Height obtains rough car mark region.
5. face bayonet model recognizing method before a kind of classification according to claim 1, it is characterised in that：4th step is specifically：
Car mark region in 3rd step is zoomed to fixed size by (4-1), is then input in 5 layers of convolutional neural networks
Classification, obtains brand result；
Front face bounding box region in 3rd step is taken out by (4-2), and zooms to fixed size, is inputted to more than one
In 10 layers of convolutional neural networks of task, vehicle and vehicle year money result and confidence level are respectively obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711293529.7A CN108090429B (en) | 2017-12-08 | 2017-12-08 | Vehicle type recognition method for graded front face bayonet |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711293529.7A CN108090429B (en) | 2017-12-08 | 2017-12-08 | Vehicle type recognition method for graded front face bayonet |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108090429A true CN108090429A (en) | 2018-05-29 |
| CN108090429B CN108090429B (en) | 2020-07-24 |
Family
ID=62174810
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711293529.7A Expired - Fee Related CN108090429B (en) | 2017-12-08 | 2017-12-08 | Vehicle type recognition method for graded front face bayonet |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108090429B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190639A (en) * | 2018-08-16 | 2019-01-11 | 新智数字科技有限公司 | A kind of vehicle color identification method, apparatus and system |
| CN109948610A (en) * | 2019-03-14 | 2019-06-28 | 西南交通大学 | A kind of vehicle fine grit classification method in the video based on deep learning |
| CN110246336A (en) * | 2019-06-26 | 2019-09-17 | 武汉万集信息技术有限公司 | The determination method and system of information of vehicles |
| CN110705719A (en) * | 2018-06-21 | 2020-01-17 | 第四范式（北京）技术有限公司 | Method and apparatus for performing automatic machine learning |
| CN111062821A (en) * | 2019-12-29 | 2020-04-24 | 北京车与车科技有限公司 | Rule-based automatic anchoring method for annual vehicle payment |
| CN111259885A (en) * | 2020-02-27 | 2020-06-09 | 深圳成子科技有限公司 | Wrist watch identification method and equipment |
| CN111259777A (en) * | 2020-01-13 | 2020-06-09 | 天地伟业技术有限公司 | End-to-end multitask vehicle brand identification method |
| CN111325231A (en) * | 2018-12-14 | 2020-06-23 | 财团法人工业技术研究院 | Neural network model fusion method and electronic device applying same |
| CN112101246A (en) * | 2020-09-18 | 2020-12-18 | 济南博观智能科技有限公司 | Vehicle identification method, device, equipment and medium |
| CN112417922A (en) * | 2019-08-20 | 2021-02-26 | 杭州海康威视数字技术股份有限公司 | Target identification method and device |
| CN112966709A (en) * | 2021-01-27 | 2021-06-15 | 中国电子进出口有限公司 | Deep learning-based fine vehicle type identification method and system |
| CN112990240A (en) * | 2019-12-13 | 2021-06-18 | 搜狗（杭州）智能科技有限公司 | Method for determining vehicle type and related device |
| CN113312945A (en) * | 2020-02-27 | 2021-08-27 | 浙江大搜车软件技术有限公司 | Vehicle identification method and device, electronic equipment and readable storage medium |
| WO2021174863A1 (en) * | 2020-03-05 | 2021-09-10 | 苏州科达科技股份有限公司 | Method for training vehicle model-year recognition model and method for recognizing vehicle model year |
| CN113408559A (en) * | 2020-03-17 | 2021-09-17 | 杭州萤石软件有限公司 | Vehicle brand identification method and device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196979A (en) * | 2006-12-22 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for recognizing vehicle type by digital picture processing technology |
| WO2012017436A1 (en) * | 2010-08-05 | 2012-02-09 | Hi-Tech Solutions Ltd. | Method and system for collecting information relating to identity parameters of a vehicle |
| CN104361343A (en) * | 2014-10-30 | 2015-02-18 | 深圳市中盟科技股份有限公司 | Method and device for identifying vehicle types |
| US20160217319A1 (en) * | 2012-10-01 | 2016-07-28 | The Regents Of The University Of California | Unified face representation for individual recognition in surveillance videos and vehicle logo super-resolution system |
| US9405988B2 (en) * | 2013-08-13 | 2016-08-02 | James Alves | License plate recognition |
| EP2701098B1 (en) * | 2012-08-23 | 2017-01-04 | Xerox Corporation | Region refocusing for data-driven object localization |
| CN106650550A (en) * | 2015-10-28 | 2017-05-10 | 中通服公众信息产业股份有限公司 | Vehicle model identification method and vehicle model identification system through image characteristics of vehicle mark and vehicle head |
-
2017
- 2017-12-08 CN CN201711293529.7A patent/CN108090429B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101196979A (en) * | 2006-12-22 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for recognizing vehicle type by digital picture processing technology |
| WO2012017436A1 (en) * | 2010-08-05 | 2012-02-09 | Hi-Tech Solutions Ltd. | Method and system for collecting information relating to identity parameters of a vehicle |
| EP2701098B1 (en) * | 2012-08-23 | 2017-01-04 | Xerox Corporation | Region refocusing for data-driven object localization |
| US20160217319A1 (en) * | 2012-10-01 | 2016-07-28 | The Regents Of The University Of California | Unified face representation for individual recognition in surveillance videos and vehicle logo super-resolution system |
| US9405988B2 (en) * | 2013-08-13 | 2016-08-02 | James Alves | License plate recognition |
| CN104361343A (en) * | 2014-10-30 | 2015-02-18 | 深圳市中盟科技股份有限公司 | Method and device for identifying vehicle types |
| CN106650550A (en) * | 2015-10-28 | 2017-05-10 | 中通服公众信息产业股份有限公司 | Vehicle model identification method and vehicle model identification system through image characteristics of vehicle mark and vehicle head |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110705719A (en) * | 2018-06-21 | 2020-01-17 | 第四范式（北京）技术有限公司 | Method and apparatus for performing automatic machine learning |
| CN109190639A (en) * | 2018-08-16 | 2019-01-11 | 新智数字科技有限公司 | A kind of vehicle color identification method, apparatus and system |
| CN111325231A (en) * | 2018-12-14 | 2020-06-23 | 财团法人工业技术研究院 | Neural network model fusion method and electronic device applying same |
| CN111325231B (en) * | 2018-12-14 | 2023-08-15 | 财团法人工业技术研究院 | Neural network model fusion method and electronic device applying same |
| CN109948610A (en) * | 2019-03-14 | 2019-06-28 | 西南交通大学 | A kind of vehicle fine grit classification method in the video based on deep learning |
| CN110246336B (en) * | 2019-06-26 | 2022-02-22 | 武汉万集信息技术有限公司 | Method and system for determining vehicle information |
| CN110246336A (en) * | 2019-06-26 | 2019-09-17 | 武汉万集信息技术有限公司 | The determination method and system of information of vehicles |
| CN112417922A (en) * | 2019-08-20 | 2021-02-26 | 杭州海康威视数字技术股份有限公司 | Target identification method and device |
| CN112990240B (en) * | 2019-12-13 | 2024-05-03 | 北京搜狗科技发展有限公司 | Method and related device for determining vehicle type |
| CN112990240A (en) * | 2019-12-13 | 2021-06-18 | 搜狗（杭州）智能科技有限公司 | Method for determining vehicle type and related device |
| CN111062821A (en) * | 2019-12-29 | 2020-04-24 | 北京车与车科技有限公司 | Rule-based automatic anchoring method for annual vehicle payment |
| CN111259777A (en) * | 2020-01-13 | 2020-06-09 | 天地伟业技术有限公司 | End-to-end multitask vehicle brand identification method |
| CN111259885A (en) * | 2020-02-27 | 2020-06-09 | 深圳成子科技有限公司 | Wrist watch identification method and equipment |
| CN113312945A (en) * | 2020-02-27 | 2021-08-27 | 浙江大搜车软件技术有限公司 | Vehicle identification method and device, electronic equipment and readable storage medium |
| CN111259885B (en) * | 2020-02-27 | 2023-11-24 | 深圳成子科技有限公司 | Wristwatch identification method and device |
| WO2021174863A1 (en) * | 2020-03-05 | 2021-09-10 | 苏州科达科技股份有限公司 | Method for training vehicle model-year recognition model and method for recognizing vehicle model year |
| CN113408559A (en) * | 2020-03-17 | 2021-09-17 | 杭州萤石软件有限公司 | Vehicle brand identification method and device and storage medium |
| CN112101246A (en) * | 2020-09-18 | 2020-12-18 | 济南博观智能科技有限公司 | Vehicle identification method, device, equipment and medium |
| CN112966709B (en) * | 2021-01-27 | 2022-09-23 | 中国电子进出口有限公司 | Deep learning-based fine vehicle type identification method and system |
| CN112966709A (en) * | 2021-01-27 | 2021-06-15 | 中国电子进出口有限公司 | Deep learning-based fine vehicle type identification method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108090429B (en) | 2020-07-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108090429A (en) | Face bayonet model recognizing method before a kind of classification | |
| CN104021574B (en) | Pavement disease automatic identifying method | |
| CN104933710B (en) | Based on the shop stream of people track intelligent analysis method under monitor video | |
| CN102842045B (en) | A kind of pedestrian detection method based on assemblage characteristic | |
| CN104408449B (en) | Intelligent mobile terminal scene literal processing method | |
| CN108629319B (en) | Image detection method and system | |
| CN105335702B (en) | A kind of bayonet model recognizing method based on statistical learning | |
| CN106022285A (en) | Vehicle type identification method and vehicle type identification device based on convolutional neural network | |
| CN106127747A (en) | Car surface damage classifying method and device based on degree of depth study | |
| CN103903005A (en) | License plate image identification system and method | |
| CN104299009B (en) | License plate character recognition method based on multi-feature fusion | |
| CN106650553A (en) | License plate recognition method and system | |
| CN103886760B (en) | Real-time vehicle detecting system based on traffic video | |
| CN102915432B (en) | A kind of vehicle-mounted microcomputer image/video data extraction method and device | |
| CN104794491A (en) | Fuzzy clustering steel plate surface defect detection method based on pre classification | |
| CN103077384A (en) | Method and system for positioning and recognizing vehicle logo | |
| CN102043945A (en) | License plate character recognition method based on real-time vehicle tracking and binary index classification | |
| CN104616664A (en) | Method for recognizing audio based on spectrogram significance test | |
| CN104463138B (en) | The text positioning method and system of view-based access control model structure attribute | |
| CN105354533B (en) | A kind of unlicensed vehicle model recognizing method of bayonet based on bag of words | |
| CN105893946A (en) | Front face image detection method | |
| CN104217217A (en) | Vehicle logo detection method and system based on two-layer classification | |
| CN103136528A (en) | Double-edge detection based vehicle license plate identification method | |
| CN103413145A (en) | Articulation point positioning method based on depth image | |
| CN107180230B (en) | Universal license plate recognition method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200724 Termination date: 20201208 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |