CN107977643A - A kind of officer's car monitoring method based on road camera - Google Patents
A kind of officer's car monitoring method based on road camera Download PDFInfo
- Publication number
- CN107977643A CN107977643A CN201711361709.4A CN201711361709A CN107977643A CN 107977643 A CN107977643 A CN 107977643A CN 201711361709 A CN201711361709 A CN 201711361709A CN 107977643 A CN107977643 A CN 107977643A
- Authority
- CN
- China
- Prior art keywords
- car
- car plate
- officer
- image
- plate
- 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
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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)
- Multimedia (AREA)
- Data Mining & Analysis (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)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
A kind of officer's car monitoring method based on road camera, the image photographed and shooting time, taking location information are passed back into monitoring system by internet in each camera of road or traffic block port, the relevant information of all departments' officer's car, including license plate number, vehicle and affiliated unit have been previously stored in background data base；To after monitoring system, system is first identified the car plate in image image transmitting, and the result of identification is compared with official vehicle data in database, judges whether the vehicle is officer's car.Find that the car is officer's car if comparing, the information such as license plate number, image and shooting time, camera site is saved in database；If result mismatches, the identification and matching of next pictures are carried out.The present invention strengthens the supervision of official vehicle, guides specification car.
Description
Technical field
The present invention relates to image recognition and vehicle monitoring field, especially a kind of method of officer's car identification monitoring.
Background technology
As camera the continuous of scale of taking pictures in current city expands, obtained in intelligent transportation and road safety monitoring field
Large-scale application.Target following based on camera is an important application in intelligent traffic monitoring field, is ground both at home and abroad
Study carefully personnel great amount of images recognizer and ancillary equipment are applied in target following technology and achieve certain achievement.
At the same time with the construction of expanding economy and equality justice society, disclosure, car for public affairs for three public funds
Management has become the hot issue of people's common concern.The scheduling of car for public affairs at present substantially manually exercises supervision,
It is lack of standardization to repeated with car due to being lacking in supervision.
The content of the invention
In order to which the automatic managerial ability in overcoming current officer's car to go on a journey is low, manual oversight is of high cost, asking of being lacking in supervision
Topic, the present invention combine city camera monitoring system, there is provided a kind of managerial ability is higher, supervision cost is relatively low automatically, supervision
Officer's car monitoring method with stronger dynamics based on road camera.
The technical solution adopted by the present invention to solve the technical problems is：
Each camera of a kind of officer's car monitoring method based on road camera, road or traffic block port will photograph
Image and shooting time and taking location information monitoring system is passed back to by internet, be previously stored in background data base
The relevant information of all departments' officer's car, including license plate number, vehicle and affiliated unit；For image transmitting to after monitoring system, system is first
First the car plate in image is identified, and the result of identification is compared with official vehicle data in database, judging should
Whether vehicle is officer's car；Find that the car is officer's car if comparing, license plate number, image and shooting time and camera site are believed
Breath is saved in database；If result mismatches, the identification and matching of next pictures are carried out.
Further, the officer's car monitoring method based on road camera comprises the following steps：
Step 1, road camera by forming picture library in the Internet transmission to system server, take the picture captured
Device program of being engaged in extracts picture successively from picture library and carries out image recognition processing, extracts license plate number；
Step 2, contrasted stored officer's car car plate in digital number plate and system database, and whether inquiry has
With successful officer's car.If if having by license plate number, image, shooting time and place latitude and longitude information storage into database, if
Do not match, return to the identifying processing that step 1 continues next pictures.
When step 3, relevant departments need to review the traveling record of officer's car, the car plate and row of officer's car are inputted in systems
Sail the period, the running section of officer's car can be checked in system, the multiple tracing points that can be also travelled are included in electronic map
On, detailed image and time can be checked by clicking on tracing point, ensure that the driving trace of vehicle has good grounds, so as to strengthen public affair
The supervision of vehicle, guides specification car.
Further, in the step 1, the process of image recognition license plate number is：
1.1) car plate rectangular area is obtained.Car plate rectangular area is obtained according to ROI acquisition algorithms, is arranged to region ABCD,
Respectively A (x1, y1), B (x1, y2), C (x2, y1), D (x2, y2)；
1.2) picture pre-processes.Gray processing processing and binary conversion treatment, the above are carried out to the rectangular area picture got
Two kinds of image processing methods are the standard preprocess method before image procossing, after above preprocess method, image at this time
In pixel there was only 0 and 255 two value, wherein the corresponding pixel value of characters on license plate is 255, and the corresponding pixel value of car plate background is
0；
1.3) Character segmentation.Character quantity is 8 (including round dots after first English alphabet) in car plate, according to car plate
The middle wide principle of character, is arranged to 8 parts by car plate is from left to right wide, can ensure at this time each two character using character midpoint as
Cut-off rule.According to the method, car plate can be separated into 8 orderly characters；
1.4) character match.The length and width of the single character handled at this time are all consistent, and the pixel value of length and width is both configured to fix
(being arranged to m*n).
In the step 1.4), the process of character match is as follows：
A) all characters in car plate will likely be appeared in advance to be put into the image that pixel is m*n sizes, respectively to figure
Piece carries out vertical scanning from left to right, horizontal sweep from up to down, the vertical scanning of right-to-left and bottom-up water
Simple scan.The transverse axis coordinate of the black pixel point of first appearance in the case of above-mentioned four kinds, ordinate of orthogonal axes, transverse axis is recorded respectively to sit
Mark and ordinate of orthogonal axes value, are denoted as T (p1, p2, p3, p4).The each character of empirical tests corresponds to a different characteristic value T, therefore
All characters being likely to occur can form a feature database T { } in car plate；
B) during character match, the image where treating glyphomancy symbol carries out vertical scanning from left to right, water from up to down
Simple scan, the vertical scanning of right-to-left and bottom-up horizontal sweep.First goes out in the case of recording above-mentioned four kinds respectively
Transverse axis coordinate, ordinate of orthogonal axes, transverse axis coordinate and the ordinate of orthogonal axes value of existing black pixel point, are denoted as T2 (p5, p6, p7, p8)；
C) the T2 coordinates measured and each characteristic value in feature database are compared, and deviation model is set to each characteristic value
Enclose, that is, whens 4 coordinate values for comparing T and T2 allows each coordinate value to have certain deviation range.When matching individual features value most
During close characteristic character, can determine current character to be measured why character.Circulate this step successively, when all characters all
With finishing, you can obtain the required number-plate number.
In the step 1.1), ROI acquisition algorithms obtain car plate rectangular area, determine to drive according to face recognition algorithms first
The region of personnel's face is sailed, and then license plate area is determined according to human face region, implementation method is：
1.1.1) recognition of face.Face recognition algorithms are more mature at present, and face area is identified in a pictures
Domain is relatively easy, we select Adaboost algorithm to carry out recognition of face, it may be determined that the human face region histogram of driver is wide
Spend for width, a height of height, and it is p (x, y) that the rectangle bottom right angular coordinate, which is manually set,；
1.1.2 car plate up-and-down boundary region) is obtained.Draw a conclusion by the experiment of a large amount of pictures, according to the vertical of face coordinate
Coordinate value moves down 6 times of height, obtains y1=y-6*height, using y1 as the coboundary ordinate of orthogonal axes of car plate rectangle
Value；Point y1 is moved down into 2/3 times of height again, point y2=y-20/3*height is obtained, using y2 as the following of car plate rectangle
The value of boundary's ordinate of orthogonal axes；
1.1.3 car plate right boundary region) is obtained.According to the textural characteristics of car plate, the vertical edge of car plate is compared with horizontal sides
Edge is concentrated, so making upright projection to the license plate image for having oriented up-and-down boundary, seven characters on car plate are in perspective view
Corresponding is the wave crest that distribution is continuous and concentrates, using upright projection change rate can positioning licence plate right boundary, from
And determine license plate area.
The deterministic process is as follows：
A) scanned from left to right in vertical projection diagram first, if projection change rate is more than maximum rate of change (according to warp
Test, compared with 10), left margin x1 of this row labeled as car plate；So obtain car plate rectangle top left corner apex A and lower-left
The coordinate of angular vertex B, i.e. A (x1, y1), B (x1, y2)；
B) continued to scan in vertical projection diagram, car is just labeled as when the row of projection value change rate maximum are arrived in scanning again
The right margin x2 of board；
C) it is 3 according to the ratio of width to height of car plate：1, if x2-x1 is more than or equal to 3 times that height is surveyed in car plate scanning, x1 to x2
For the right boundary of car plate, otherwise return and b) continue to scan to the right from x2, just marked when scanning to the row of car plate change rate maximum
For the right margin of car plate；So obtain the coordinate of car plate rectangle bottom right angular vertex C and upper right angular vertex D, i.e. C (x2, y1), D
(x2, y2)；
Therefore it is ABCD regions that car plate rectangle can be obtained.
In the present invention, each camera of road or traffic block port is by the image photographed and shooting time, camera site
Information passes back to monitoring system by internet, and the relevant information of all departments' officer's car has been previously stored in background data base, bag
Include license plate number, vehicle, affiliated unit etc..To after monitoring system, system is first identified the car plate in image image transmitting,
And the result of identification is compared with official vehicle data in database, judge whether the vehicle is officer's car.If compare hair
Now the car is officer's car, then the information such as license plate number, image and shooting time, camera site is saved in database；If result
Mismatch, then carry out the identification and matching of next pictures.Using the system, trip situation that relevant departments can be to each officer's car
It is monitored, checks the running section of officer's car, the multiple tracing points that can be also travelled is shown on the electronic map, click on rail
Mark point can check detailed image and time, ensure that the driving trace of vehicle has good grounds, so as to strengthen the prison of official vehicle
Superintend and direct, guide specification car.
Beneficial effects of the present invention are mainly manifested in：Official vehicle monitoring, nothing can be carried out using existing road camera
Extra cost need to be increased transformation and upgrade, while the effectively traveling feelings of identification official vehicle outside are carried out to officer's car built-in system
Condition, works well.
Brief description of the drawings
Fig. 1 is overall identification process figure；
Fig. 2 carries out the schematic diagram of recognition of face for Adaboost algorithm；
Fig. 3 is the ROI schematic diagrames of image.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings.
With reference to Fig. 1~Fig. 3, a kind of officer's car monitoring method based on road camera, road or traffic block port it is each
The image photographed and shooting time and taking location information are passed back to monitoring system, back-end data by camera by internet
The relevant information of all departments' officer's car, including license plate number, vehicle and affiliated unit have been previously stored in storehouse.Image transmitting to monitoring
After system, system is first identified the car plate in image, and by official vehicle data in the result of identification and database into
Row compares, and judges whether the vehicle is officer's car.If compare find the car be officer's car, by license plate number, image and shoot when
Between and taking location information be saved in database；If result mismatches, the identification and matching of next pictures are carried out.
Using the system, the trip situation that relevant departments can be to each officer's car is monitored, and checks the traveling road of officer's car
Section, the multiple tracing points that can be also travelled show that on the electronic map detailed image and time can be checked by clicking on tracing point,
Ensure that the driving trace of vehicle has good grounds, so as to strengthen the supervision of official vehicle, guide specification car.
Further, the officer's car monitoring method based on road camera, comprises the following steps：
Step 1, road camera by forming picture library in the Internet transmission to system server, take the picture captured
Device program of being engaged in extracts picture successively from picture library and carries out image recognition processing, extracts license plate number and forms digital number plate；
Step 2, contrasted stored officer's car car plate in digital number plate and system database, and whether inquiry has
With successful officer's car.If if having by license plate number, image, shooting time and place latitude and longitude information storage into database, if
Do not match, return to the identifying processing that step 1 continues next pictures；
When step 3, relevant departments need to review the traveling record of officer's car, the car plate and row of officer's car are inputted in systems
Sail the period, the running section of officer's car can be checked in system, the multiple tracing points that can be also travelled are included in electronic map
On, detailed image and time can be checked by clicking on tracing point, ensure that the driving trace of vehicle has good grounds, so as to strengthen public affair
The supervision of vehicle, guides specification car.
In the step 1, the process of image recognition license plate number is：
1.1) car plate rectangular area is obtained.Car plate rectangular area is obtained according to ROI acquisition algorithms, four angles of rectangle are set
It is respectively A (x1, y1), B (x1, y2), C (x2, y1), D (x2, y2) for ABCD；
1.2) picture pre-processes.Gray processing processing and binary conversion treatment, both the above are carried out to the car plate picture got
Image processing method is the standard preprocess method before image procossing, after above preprocess method, at this time in image
Pixel only has 0 and 255 two value, and the wherein corresponding pixel value of characters on license plate is 255, as black, the corresponding picture of car plate background
Element value is 0, is white；
1.3) Character segmentation.Character quantity is 8 (including round dots after first English alphabet) in car plate, according to car plate
The middle wide principle of character, is arranged to 8 parts by car plate is from left to right wide, can ensure at this time each two character using character midpoint as
Cut-off rule.According to the method, car plate can be separated into 8 orderly characters；
1.4) character match.The length and width of the single character handled at this time are all consistent, and the pixel value of length and width is both configured to m*n pictures
Plain size.
In the step 1.4), the process of character match is as follows：
A) standard character feature database is created.Will likely appear in advance in car plate all characters (as " Zhejiang ", " capital ",
" A ", " ", " 1 ", " 2 " etc.) it is put into the image of m*n pixel sizes and placed in the middle.Origin is used as using the lower-left angular vertex of image at this time
Rectangular coordinate system is established, then carries out vertical scanning from left to right, horizontal sweep from up to down to picture respectively, from dextrad
Left vertical scanning and bottom-up horizontal sweep, record the black pixel point of first appearance in the case of above-mentioned four kinds respectively
Transverse axis coordinate, ordinate of orthogonal axes, transverse axis coordinate and ordinate of orthogonal axes value, that is, it is minimum and maximum in rectangular coordinate system to obtain the character
Abscissa value, minimum and maximum ordinate value, be denoted as T (p1, p2, p3, p4), the signal in character feature storehouse is as shown in table 1.
| Character | Minimum abscissa | Maximum abscissa | Minimum ordinate | Maximum ordinate | Characteristic value |
| A | 20 | 60 | 10 | 70 | (20,70,60,10) |
| B | 30 | 60 | 10 | 60 | (30,60,60,10) |
| Zhejiang | 10 | 70 | 10 | 70 | (10,70,70,10) |
| 1 | 40 | 40 | 10 | 70 | (40,70,40,10) |
| 0 | 30 | 50 | 10 | 70 | (30,70,50,10) |
| ● | 40 | 40 | 40 | 40 | (40,40,40,40) |
| .... | .... | .... | .... | .... | .... |
Table 1
Through corresponding to a different characteristic value T experimental results demonstrate each character, therefore all in car plate it is likely to occur
Character can form a feature database T { }；
B) character feature storehouse to be measured is created.After obtaining character to be measured, according to create standard character feature when process, successively
Treat glyphomancy symbol where image carry out vertical scanning from left to right, horizontal sweep from up to down, right-to-left it is vertical
Scanning and bottom-up horizontal sweep.The transverse axis for recording the black pixel point of first appearance in the case of above-mentioned four kinds respectively is sat
Mark, ordinate of orthogonal axes, transverse axis coordinate and ordinate of orthogonal axes value, are denoted as T2 (p5, p6, p7, p8), the signal such as table in character feature storehouse to be measured
Shown in 2；
Table 2
C) standard character is compared with character feature storehouse to be measured., will be each in the T2 coordinates measured and feature database during character match
A characteristic value compares, and sets deviation range to each characteristic value, that is, whens 4 coordinate values for comparing T and T2 allows each coordinate
Value has certain deviation range.When matching the immediate characteristic character of individual features value, current character to be measured can be determined
Why character.This step is circulated successively, is finished when all characters all match, you can obtain the required number-plate number.
In the step 1.1), ROI acquisition algorithms obtain car plate rectangular area, determine to drive according to face recognition algorithms first
The region of personnel's face is sailed, and then license plate area is determined according to human face region, implementation method is：
1.1.1) recognition of face.Face recognition algorithms are more mature at present, and face area is identified in a pictures
Domain is relatively easy, we select Adaboost algorithm to carry out recognition of face, it may be determined that the human face region histogram of driver is wide
Spend for width, a height of height, and it is p (x, y) that the rectangle bottom right angular coordinate, which is manually set,.
1.1.2 car plate up-and-down boundary region) is obtained.Draw a conclusion by the experiment of a large amount of pictures, according to the vertical of face coordinate
Coordinate value moves down 6 times of height, obtains y1=y-6*height, using y1 as the coboundary ordinate of orthogonal axes of car plate rectangle
Value；Point y1 is moved down into 2/3 times of height again, point y2=y-20/3*height is obtained, using y2 as the following of car plate rectangle
The value of boundary's ordinate of orthogonal axes.
1.1.3 car plate right boundary region) is obtained.According to the textural characteristics of car plate, the vertical edge of car plate is compared with horizontal sides
Edge is concentrated, so making upright projection to the license plate image for having oriented up-and-down boundary, seven characters on car plate are in perspective view
Corresponding is the wave crest that distribution is continuous and concentrates, using upright projection change rate can positioning licence plate right boundary, from
And determine license plate area.
The deterministic process is as follows：
A) scanned from left to right in vertical projection diagram first, if projection change rate is more than maximum rate of change (according to warp
Test, compared with 10), left margin x1 of this row labeled as car plate；So obtain car plate rectangle top left corner apex A and lower-left
The coordinate of angular vertex B, i.e. A (x1, y1), B (x1, y2)；
B) continued to scan in vertical projection diagram, car is just labeled as when the row of projection value change rate maximum are arrived in scanning again
The right margin x2 of board；
C) it is 3 according to the ratio of width to height of car plate：1, if x2-x1 is more than or equal to 3 times that height is surveyed in car plate scanning, x1 to x2
For the right boundary of car plate, otherwise return and b) continue to scan to the right from x2, just marked when scanning to the row of car plate change rate maximum
For the right margin of car plate；So obtain the coordinate of car plate rectangle bottom right angular vertex C and upper right angular vertex D, i.e. C (x2, y1), D
(x2, y2)；
Therefore it is ABCD regions that car plate rectangle can be obtained.
In the present embodiment, the pixel size of each image residing for character is 80*80 in character match, image lower-left angular vertex
As coordinate origin, it is denoted as (0,0), pixel deviation range of the standard character compared with character feature storehouse to be measured is arranged to 5.
The character that server is corresponded in each car plate carries out circular treatment, and all characters can inquire about whether the car is database after having handled
In typing official vehicle.
Those of ordinary skill in the art is it should be appreciated that above content is intended merely to the explanation present invention, and simultaneously
Non- to be used as limitation of the invention, as long as in the spirit of the present invention, change, modification to above example all will
Fall in the range of claims of the present invention.
Claims (5)
- A kind of 1. officer's car monitoring method based on road camera, it is characterised in that：Each shooting of road or traffic block port The image photographed and shooting time, taking location information are passed back to monitoring system by head by internet, in background data base It has been previously stored the relevant information of all departments' officer's car, including license plate number, vehicle and affiliated unit；Image transmitting is to monitoring system Afterwards, system is first identified the car plate in image, and official vehicle data in the result of identification and database are compared It is right, judge whether the vehicle is officer's car；Find that the car is officer's car if comparing, by license plate number, image and shooting time and Taking location information is saved in database；If result mismatches, the identification and matching of next pictures are carried out.
- A kind of 2. officer's car monitoring method based on road camera as claimed in claim 1, it is characterised in that：The monitoring Method comprises the following steps：Step 1, road camera are by the picture captured by forming picture library, server in the Internet transmission to system server Program extracts picture successively from picture library and carries out image recognition processing, extracts license plate number；Step 2, contrasted stored officer's car car plate in digital number plate and system database, inquiry whether have matching into The officer's car of work(, if storing license plate number, image, shooting time and place latitude and longitude information into database if having, if not having Matching then returns to the identifying processing that step 1 continues next pictures；When step 3, relevant departments need to review the traveling record of officer's car, input in systems the car plate of officer's car with when driving Between section, the running section of officer's car can be checked in system, the multiple tracing points that can be also travelled are shown on the electronic map, point Detailed image and time can be checked by hitting tracing point, ensure that the driving trace of vehicle has good grounds, so as to strengthen official vehicle Supervision, guide specification car.
- A kind of 3. officer's car monitoring method based on road camera as claimed in claim 2, it is characterised in that：The step In 1, the process of described image identification license plate number is：1.1) car plate rectangular area is obtained：Car plate rectangular area is obtained according to ROI acquisition algorithms, is arranged to region ABCD, A (x1, Y1), B (x1, y2), C (x2, y1), D (x2, y2)；1.2) picture pre-processes：Gray processing processing and binary conversion treatment, both the above are carried out to the rectangular area picture got Image processing method is the standard preprocess method before image procossing, after above preprocess method, at this time in image Pixel only has 0 and 255 two value, and the wherein corresponding pixel value of characters on license plate is 255, and the corresponding pixel value of car plate background is 0；1.3) Character segmentation：Character quantity is 8 in car plate, according to the wide principle of character in car plate, etc. by car plate from left to right It is wide to be arranged to 8 parts, it can ensure each two character at this time using character midpoint as cut-off rule, car plate can be separated into according to the method 8 orderly characters；1.4) character match：The length and width of the single character handled at this time are all consistent, and the pixel value of length and width is both configured to fix, i.e. m* n。
- A kind of 4. officer's car monitoring method based on road camera as claimed in claim 3, it is characterised in that：The step 1.4) in, the process of the character match is as follows：A) will likely appear in advance all characters in car plate be put into pixel be m*n sizes image in, respectively to picture into The vertical scanning of row from left to right, horizontal sweep from up to down, the vertical scanning of right-to-left and bottom-up level are swept Retouch, record respectively in the case of above-mentioned four kinds the transverse axis coordinate of the black pixel point of first appearance, ordinate of orthogonal axes, transverse axis coordinate and Ordinate of orthogonal axes value, is denoted as T (p1, p2, p3, p4), and each character of empirical tests corresponds to a different characteristic value T, therefore car plate In all characters being likely to occur can form a feature database T { }；B) during character match, the image where treating glyphomancy symbol carries out vertical scanning from left to right, level from up to down is swept Retouch, the vertical scanning of right-to-left and bottom-up horizontal sweep, record in the case of above-mentioned four kinds first appearance respectively Transverse axis coordinate, ordinate of orthogonal axes, transverse axis coordinate and the ordinate of orthogonal axes value of black pixel point, are denoted as T2 (p5, p6, p7, p8)；C) the T2 coordinates measured and each characteristic value in feature database are compared, and deviation range is set to each characteristic value, i.e., Each coordinate value is allowed to have certain deviation range during 4 coordinate values for comparing T and T2；When matching, individual features value is immediate During characteristic character, can determine current character to be measured why character, this step is circulated successively, when all characters have all matched Finish, you can obtain the required number-plate number.
- 5. the ROI acquisition algorithms as described in claim 3 or 4 obtain car plate rectangular area, it is characterised in that：The step 1.1) In, the region of driver's face is determined according to face recognition algorithms first, and then license plate area is determined according to human face region, its Implementation method is：1.1.1) recognition of face, recognition of face is carried out using Adaboost algorithm, determines the human face region histogram of driver Width is width, a height of height, and it is p (x, y) that the rectangle bottom right angular coordinate, which is manually set,；1.1.2 car plate up-and-down boundary region) is obtained, 6 times of height are moved down according to the ordinate value of face coordinate, obtain y1 =y-6*height, the value using y1 as the coboundary ordinate of orthogonal axes of car plate rectangle；Point y1 is moved down 2/3 times again Height, obtains point y2=y-20/3*height, the value using y2 as the lower boundary ordinate of orthogonal axes of car plate rectangle；1.1.3 car plate right boundary region) is obtained, according to the textural characteristics of car plate, the vertical edge of car plate is compared with horizontal edge collection In, so making upright projection to the license plate image for having oriented up-and-down boundary, seven characters on car plate are corresponding in perspective view Be distribution is continuous and concentrates wave crest, using upright projection change rate can positioning licence plate right boundary, it judges Process is as follows：A) scanned from left to right in vertical projection diagram first, if projection change rate is more than maximum rate of change, this row is labeled as The left margin x1 of car plate；So obtain the coordinate of car plate rectangle top left corner apex A and lower-left angular vertex B, i.e. A (x1, y1), B (x1, y2)；B) continued to scan in vertical projection diagram, when the row of projection value change rate maximum are arrived in scanning again just labeled as car plate Right margin x2；C) it is 3 according to the ratio of width to height of car plate：1, if x2-x1 is more than or equal to 3 times that height is surveyed in car plate scanning, x2 to x1 is car The right boundary of board, otherwise returns and b) continues to scan to the right from x2, scanning is just labeled as car when arriving the row of car plate change rate maximum The right margin of board；So obtain the coordinate of car plate rectangle bottom right angular vertex C and upper right angular vertex D, i.e. C (x2, y1), D (x2, y2)；Therefore it is ABCD regions to obtain car plate rectangle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711361709.4A CN107977643A (en) | 2017-12-18 | 2017-12-18 | A kind of officer's car monitoring method based on road camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711361709.4A CN107977643A (en) | 2017-12-18 | 2017-12-18 | A kind of officer's car monitoring method based on road camera |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107977643A true CN107977643A (en) | 2018-05-01 |
Family
ID=62006632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711361709.4A Pending CN107977643A (en) | 2017-12-18 | 2017-12-18 | A kind of officer's car monitoring method based on road camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107977643A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108417047A (en) * | 2018-05-10 | 2018-08-17 | 杭州盈蝶科技有限公司 | A kind of vehicle location method for tracing and its system |
| CN110070706A (en) * | 2019-04-28 | 2019-07-30 | 西安艾润物联网技术服务有限责任公司 | Officer's car monitoring method, apparatus and system |
| CN112068566A (en) * | 2020-09-15 | 2020-12-11 | 北京百度网讯科技有限公司 | Guiding path determining method, and vehicle driving control method, device and equipment |
| CN112818954A (en) * | 2021-03-17 | 2021-05-18 | 重庆紫光华山智安科技有限公司 | Vehicle state identification method, device, equipment and medium |
| CN113223190A (en) * | 2020-01-21 | 2021-08-06 | 武汉德云祥科技发展有限公司 | Monitoring equipment and method for non-inductive payment |
| CN113239929A (en) * | 2021-04-30 | 2021-08-10 | 南京钢铁股份有限公司 | Method for identifying numbers in computer display screen |
| CN113255395A (en) * | 2020-02-07 | 2021-08-13 | 深圳云天励飞技术有限公司 | Driver region positioning method and device, electronic equipment and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751785A (en) * | 2010-01-12 | 2010-06-23 | 杭州电子科技大学 | Automatic license plate recognition method based on image processing |
| CN102411800A (en) * | 2011-08-25 | 2012-04-11 | 甘肃万华金慧信息科技有限公司 | System and method for realizing digital service management of official vehicle |
| CN102880863A (en) * | 2012-09-20 | 2013-01-16 | 北京理工大学 | Method for positioning license number and face of driver on basis of deformable part model |
| CN102999749A (en) * | 2012-12-21 | 2013-03-27 | 广东万安科技股份有限公司 | Intelligent safety belt regulation violation event detecting method based on face detection |
| CN104167095A (en) * | 2014-08-05 | 2014-11-26 | 江苏省邮电规划设计院有限责任公司 | Method for checking vehicle behavior modes on basis of smart cities |
| CN104966066A (en) * | 2015-06-26 | 2015-10-07 | 武汉大学 | Traffic block port monitoring oriented in-car human face detection method and system |
| CN105825184A (en) * | 2016-03-15 | 2016-08-03 | 深圳市佳信捷技术股份有限公司 | Vehicle data processing method and device |
| CN106295630A (en) * | 2016-07-21 | 2017-01-04 | 北京小米移动软件有限公司 | Character identifying method and device |
-
2017
- 2017-12-18 CN CN201711361709.4A patent/CN107977643A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751785A (en) * | 2010-01-12 | 2010-06-23 | 杭州电子科技大学 | Automatic license plate recognition method based on image processing |
| CN102411800A (en) * | 2011-08-25 | 2012-04-11 | 甘肃万华金慧信息科技有限公司 | System and method for realizing digital service management of official vehicle |
| CN102880863A (en) * | 2012-09-20 | 2013-01-16 | 北京理工大学 | Method for positioning license number and face of driver on basis of deformable part model |
| CN102999749A (en) * | 2012-12-21 | 2013-03-27 | 广东万安科技股份有限公司 | Intelligent safety belt regulation violation event detecting method based on face detection |
| CN104167095A (en) * | 2014-08-05 | 2014-11-26 | 江苏省邮电规划设计院有限责任公司 | Method for checking vehicle behavior modes on basis of smart cities |
| CN104966066A (en) * | 2015-06-26 | 2015-10-07 | 武汉大学 | Traffic block port monitoring oriented in-car human face detection method and system |
| CN105825184A (en) * | 2016-03-15 | 2016-08-03 | 深圳市佳信捷技术股份有限公司 | Vehicle data processing method and device |
| CN106295630A (en) * | 2016-07-21 | 2017-01-04 | 北京小米移动软件有限公司 | Character identifying method and device |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108417047A (en) * | 2018-05-10 | 2018-08-17 | 杭州盈蝶科技有限公司 | A kind of vehicle location method for tracing and its system |
| CN110070706A (en) * | 2019-04-28 | 2019-07-30 | 西安艾润物联网技术服务有限责任公司 | Officer's car monitoring method, apparatus and system |
| CN113223190A (en) * | 2020-01-21 | 2021-08-06 | 武汉德云祥科技发展有限公司 | Monitoring equipment and method for non-inductive payment |
| CN113255395A (en) * | 2020-02-07 | 2021-08-13 | 深圳云天励飞技术有限公司 | Driver region positioning method and device, electronic equipment and storage medium |
| CN113255395B (en) * | 2020-02-07 | 2024-06-11 | 深圳云天励飞技术有限公司 | Driver region positioning method and device, electronic equipment and storage medium |
| CN112068566A (en) * | 2020-09-15 | 2020-12-11 | 北京百度网讯科技有限公司 | Guiding path determining method, and vehicle driving control method, device and equipment |
| CN112818954A (en) * | 2021-03-17 | 2021-05-18 | 重庆紫光华山智安科技有限公司 | Vehicle state identification method, device, equipment and medium |
| CN113239929A (en) * | 2021-04-30 | 2021-08-10 | 南京钢铁股份有限公司 | Method for identifying numbers in computer display screen |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107977643A (en) | A kind of officer's car monitoring method based on road camera | |
| CN106529424B (en) | A kind of logo detection recognition method and system based on selective search algorithm | |
| CN104217427B (en) | Lane line localization method in a kind of Traffic Surveillance Video | |
| CN107657259A (en) | Distorted image detection method, electronic installation and readable storage medium storing program for executing | |
| CN104809461B (en) | The licence plate recognition method and system of binding sequence image super-resolution rebuilding | |
| CN107194398A (en) | Car damages recognition methods and the system at position | |
| CN105354568A (en) | Convolutional neural network based vehicle logo identification method | |
| CN105976392B (en) | Vehicle tyre detection method and device based on maximum output probability | |
| CN106022232A (en) | License plate detection method based on deep learning | |
| CN107491753A (en) | A kind of parking offense detection method based on background modeling | |
| CN109726717A (en) | A kind of vehicle comprehensive information detection system | |
| CN104573707B (en) | A kind of license plate Chinese character recognition method based on multi-feature fusion | |
| CN103530608A (en) | Vehicle type judgment method and vehicle type judgment device | |
| CN106558051A (en) | A kind of improved method for detecting road from single image | |
| Xia et al. | The study of license plate character segmentation algorithm based on vetical projection | |
| CN104766078A (en) | Slant angle detecting method and slant angle detecting device of target image | |
| CN106991821A (en) | Vehicles peccancy hand-held mobile terminal data collecting system | |
| CN105224937A (en) | Based on the semantic color pedestrian of the fine granularity heavily recognition methods of human part position constraint | |
| CN106407959A (en) | Low-illumination complicated background license plate positioning method based on wavelet transform and SVM | |
| CN109190483A (en) | A kind of method for detecting lane lines of view-based access control model | |
| CN109961013A (en) | Recognition methods, device, equipment and the computer readable storage medium of lane line | |
| CN108960079A (en) | A kind of image-recognizing method and device | |
| CN108717789A (en) | A kind of the acquisition mask method and device of vehicle sample | |
| CN103679205A (en) | Preceding car detection method based on shadow hypothesis and layered HOG (histogram of oriented gradient) symmetric characteristic verification | |
| CN109460722A (en) | A kind of license plate intelligent identification 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180501 |