CN101303803A - Method and system for discriminating license plate - Google Patents
Method and system for discriminating license plate Download PDFInfo
- Publication number
- CN101303803A CN101303803A CNA200810114753XA CN200810114753A CN101303803A CN 101303803 A CN101303803 A CN 101303803A CN A200810114753X A CNA200810114753X A CN A200810114753XA CN 200810114753 A CN200810114753 A CN 200810114753A CN 101303803 A CN101303803 A CN 101303803A
- Authority
- CN
- China
- Prior art keywords
- character
- image
- car plate
- license plate
- black
- 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
- 238000000034 method Methods 0.000 title claims abstract description 57
- 230000011218 segmentation Effects 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 18
- 238000012937 correction Methods 0.000 claims abstract description 10
- 230000000007 visual effect Effects 0.000 claims description 35
- 230000001915 proofreading effect Effects 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 11
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000009434 installation Methods 0.000 abstract description 8
- 238000012423 maintenance Methods 0.000 abstract description 6
- 230000001960 triggered effect Effects 0.000 abstract 2
- 238000007689 inspection Methods 0.000 abstract 1
- 230000006698 induction Effects 0.000 description 15
- 230000000875 corresponding effect Effects 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 230000000903 blocking effect Effects 0.000 description 3
- 230000005484 gravity Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012856 packing Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Landscapes
- Traffic Control Systems (AREA)
- Character Input (AREA)
Abstract
The invention discloses a license plate recognition method, which comprises collecting video stream; whether the license plate recognition is to be triggered or not is judged according to the video stream and a vehicle deceleration enabling signal; when the license plate recognition is judged to be carried out, the video stream is taken as an input image; according to the input image, the edge information of the license plate is calculated and the position of a license plate region is determined, furthermore, a license plate image after correction is obtained according to the edge information; character segmentation is carried out on the license plate image after correction so as to obtain the image of each character; a classifier obtained by pre-training is utilized for recognizing the image of each character, and a final license plate recognition result is obtained according to the recognized character. Meanwhile, the invention also discloses a license plate recognition system. The license plate recognition method and the system disclosed by the invention determines whether the license plate recognition is to be triggered or not according to the video stream and the vehicle deceleration enabling signal, reduces erroneous inspection caused by interferent, carries out installation and debugging without damaging roads, and has lower installation and maintenance cost.
Description
Technical field
The present invention relates to license plate recognition technology, be specifically related to a kind of method and system of car plate identification.
Background technology
Current, along with computer technology rapid development, the information processing capability of computing machine improves constantly.Image recognition technology along with the raising of computer process ability has also obtained fast development, wherein, has worldwide obtained using widely based on the intelligent transportation and the electronic police system of multimedia and artificial intelligence technology.During these are used, have 96% automated system to use automatic Recognition of License Plate, the system more than 75% is the application that is identified as core with car plate.
Car plate identification is meant the license plate area that exists in the certain zone that obtains image from static state or dynamic video image, and further identifies the character in the license plate area.Car plate identification has important practical value as the most basic collecting vehicle information technology in each fields such as video monitoring and intelligent traffic administration systems.Such as fields such as expressway tol lcollection, vehicle safety checks, the flow process of the car plate of widespread use identification as shown in Figure 1, comprising:
Step 101: bury ground induction coil underground or set up infrared induction light curtain in detection site in advance, by described ground induction coil or induction light curtain, the perception vehicle arrives information.Simultaneously, set up video camera in advance, with the camera points appointed area, the focus of described video camera and focal length all are set in this regional center.When vehicle arrived appointment regional, described ground induction coil or induction light curtain perception vehicle sent trigger pip after arriving, and corresponding trigger pip place frame constantly is set to the target frame of car plate identification in the notice car plate recognition device video flowing.
Step 102: the car plate recognition device receives trigger pip, from camera acquisition to video flowing select target frame.
Step 103: adopt feature location and the extractive technique set, from target frame, select license plate area, utilize selected recognizer (for example recognizer of block letter font), identify the character content in the car plate.
Adopt said method, can from video flowing, identify required license board information, but, existing method need be buried ground induction coil in advance underground or set up the infrared ray curtain, need spend more manpower and materials when debugging ground induction coil or light curtain susceptibility, in case and system operation goes wrong or ageing equipment need upgrade the time, also need to damage again road and lay ground induction coil again or set up the induction light curtain, so maintenance cost is higher.
Simultaneously, flase drop appears in described ground induction coil and infrared induction light curtain easily.For example: as non-vehicle object process ground induction coil, perhaps when barriers such as human or animal enter the induction region of infrared induction light curtain, might trigger the car plate recognition device equally and detect, thereby flase drop occur, cause frequent system triggers, cause the waste of device resource.
As fully visible, current scheme of carrying out car plate identification is higher at the cost of actual installation and maintenance and false drop rate when using is higher.
Summary of the invention
The embodiment of the invention provides a kind of method and system of car plate identification, can discern automatically with lower installation and maintenance cost realization car plate, reduces flase drop generation probability simultaneously.
For achieving the above object, technical scheme of the present invention specifically is achieved in that
A kind of method of car plate identification, this method comprises:
Gather video flowing, judge whether to trigger car plate identification according to described video flowing and vehicle deceleration enable signal;
When judging that not carrying out car plate discerns, continue to gather video flowing and also carry out described judgement; Otherwise, use described video flowing as input picture;
Calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
License plate image after the described correction is carried out the image that Character segmentation obtains each character;
The sorter that utilizes training in advance to obtain is discerned the image of each character, obtains final license plate recognition result according to the character that identifies.
Judge whether that according to described video flowing and vehicle deceleration enable signal the method that triggers car plate identification comprises:
Judged whether that according to described video flowing vehicle enters the visual field; When judgement has vehicle to enter the visual field, judge further whether the vehicle deceleration enable signal is in effective status, if described enable signal is effective, then trigger car plate identification; Otherwise, do not trigger car plate identification.
Judge whether that according to described video flowing the method that vehicle enters the visual field comprises:
Set up background image in advance, utilize each present frame in the described video flowing poor with background image respectively, obtain the multiframe frame difference image;
The object size parameter that obtains entering the visual field according to described frame difference image, when described object size parameter during greater than the value set, thinking has vehicle to enter the visual field; Otherwise thinking does not have vehicle to enter the visual field.
The described method of setting up background image in advance comprises:
Calculate the average brightness of Same Scene each pixel in the background image modeling time, according to described mean value generation background image, the length of described background image modeling time preestablishes;
Every predefined update time, initiate to carry out the step of the new described average brightness of calculating, the background image after upgrading according to the new average brightness generation that calculates.
Calculates the marginal information of car plate according to described input picture, determine the position of license plate area, and the method for the license plate image after obtaining proofreading and correct according to described marginal information comprises:
Utilize estimation to select the slowest pairing two field picture of the moment of car speed from described input picture, the detection template of the horizontal direction of employing sobel operator is calculated the horizontal edge of edge pixel point in this two field picture;
Utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtain the license plate area of black and whiteization;
According to described horizontal edge, calculate the angle of the horizontal direction of edge pixel point in the described license plate area again, the license plate image after obtaining proofreading and correct.
Calculates the marginal information of car plate according to described input picture, determine the position of license plate area, and the method for the license plate image after obtaining proofreading and correct according to described marginal information comprises:
The detection template of the horizontal direction of employing sobel operator is calculated the horizontal edge of edge pixel point in the described input picture respectively;
Utilize setting threshold that the image-region between the horizontal edge in the described input picture is carried out the black and white processing, obtain the image of the license plate area of black and whiteization;
Again according to described horizontal edge, calculate the angle of the horizontal direction of edge pixel point in the image of described license plate area respectively, obtain the license plate image after multiframe is proofreaied and correct.
The described sorter that utilizes training in advance to obtain is discerned the image of each character, and the method that obtains final license plate recognition result according to the character that identifies comprises:
The sorter that utilizes training in advance to obtain is discerned the image of each character;
Merge the recognition result that multiframe comprises the image of each character,, generate final recognition result when described recognition result satisfies predefined playing a card during condition; The described predefined condition of playing a card is: the recognition sample of this car plate reaches default quantity, and the degree of confidence of each character identification result in the car plate reaches default confidence weighting value.
License plate image after the described correction is carried out the method that Character segmentation obtains the image of each character to be comprised:
License plate image after proofreading and correct is carried out projection on perpendicular to the direction on car plate surface;
Vertically, to the other end, the image after the projection is lined by line scan until running into first black pixel point, determine the reference position of whole character zone since an end; Begin in the other direction image to be lined by line scan until running into first black pixel point from the other end, determine the final position of whole character zone; The altitude range of determining whole character zone according to the reference position and the final position of described whole character zone;
In the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
In the width range of each character of determining, respectively vertically, line by line scan until running into first black pixel point since an end, determine the reference position of each character; Begin to line by line scan until running into first black pixel point in the other direction from the other end, determine the final position of each character; The altitude range of determining each character according to the reference position and the final position of described this character;
The image that obtains each character by the width range and the altitude range of each character of determining.
Further comprise in the method for described car plate identification:
After car plate identification was finished, it is invalid that the vehicle deceleration enable signal is changed to, and lets pass and finished the vehicle pass-through of car plate identification, through the vehicle deceleration enable signal being changed to effectively after the time delay of setting again.
A kind of system of car plate identification, this system comprises: video acquisition device, vehicle deceleration device, car plate identification flip flop equipment and car plate APU;
Described video acquisition device is used to gather video flowing and sends to car plate identification flip flop equipment and car plate APU;
Described vehicle deceleration device is used for the state of vehicle deceleration enable signal is sent to car plate identification flip flop equipment;
Described car plate identification flip flop equipment is used for judging whether to trigger car plate identification according to video flowing that receives and vehicle deceleration enable signal; When judging that not triggering car plate discerns, continue the video flowing of receiver, video harvester transmission and also carry out described judgement; Otherwise, trigger the car plate APU and carry out car plate identification;
Described car plate APU, be used for trigger pip according to described car plate identification flip flop equipment, the video flowing that the receiver, video harvester sends is as input picture, calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information; License plate image after the described correction is carried out the image that Character segmentation obtains each character; The sorter that utilizes training in advance to obtain is discerned the image of each character, obtains license plate recognition result according to the character that identifies.
Comprise in the described car plate identification flip flop equipment: target discrimination module and trigger module;
Described target discrimination module is used for having judged whether that according to described video flowing vehicle enters the visual field, and with judged result notification triggers module;
Described trigger module, the judged result that is used for the receiving target determination module, when described target discrimination module judgement has vehicle to enter the visual field, further the state of the vehicle deceleration enable signal of sending according to the vehicle deceleration device is judged, when described vehicle deceleration enable signal when being effective, trigger described car plate APU and carry out car plate identification; Otherwise, do not trigger described car plate APU.
Comprise in the described target discrimination module: background image modeling unit, target sizes computing unit and target sizes comparing unit;
Described background image modeling unit is used to calculate the average brightness of Same Scene each pixel in the background image modeling time, and according to described mean value generation background image, the length of described background image modeling time preestablishes; Every predefined update time, initiate to carry out the step of the new described average brightness of calculating, the background image after upgrading according to the new average brightness generation that calculates;
Described target sizes computing unit, it is poor with background image respectively to be used for each present frame of described video flowing, obtains the multiframe frame difference image, and obtains entering the object size parameter of visual field according to described frame difference image;
Described target sizes comparing unit is used for described object size parameter and predefined value are compared, and when described object size parameter during greater than predefined value, judging has vehicle to enter the visual field; Otherwise judging does not have vehicle to enter the visual field; With judged result notification triggers module.
Described car plate APU comprises: license plate image locating module, Character segmentation module and character recognition module;
Described license plate image locating module, be used for trigger pip according to described car plate identification flip flop equipment, the video flowing that the receiver, video harvester sends is as input picture, calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
Described Character segmentation module is used for the license plate image after proofreading and correct is carried out the image that Character segmentation obtains each character;
Described character recognition module, the sorter that is used to utilize training in advance to obtain is discerned the image of each character, obtains license plate recognition result according to the character that identifies.
Comprise in the described license plate image locating module: edge calculations unit, black and white processing unit and car plate correcting unit;
Described edge calculations unit is used for utilizing the detection template of the horizontal direction of sobel operator to calculate the horizontal edge of described input picture edge pixel point;
Described black and white processing unit is used to utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtains the license plate area of black and whiteization;
Described car plate correcting unit is used for according to described horizontal edge, calculates the angle of the horizontal direction of edge pixel point in the license plate area of described black and whiteization, the license plate image after obtaining proofreading and correct;
Further comprise the warm module of multiframe in the described car plate APU;
The warm module of described multiframe is used to merge the recognition result that multiframe comprises the image of each character, when described recognition result satisfies predefined playing a card during condition, generates final recognition result; The described predefined condition of playing a card is: the recognition sample of this car plate reaches default quantity, and the degree of confidence of each character identification result in the car plate reaches default confidence weighting value.
Comprise the input picture determining unit in the described license plate image locating module, edge calculations unit, black and white processing unit and car plate correcting unit;
Described input picture determining unit is used for utilizing estimation to select a pairing two field picture of the slowest moment of car speed to carry out car plate identification as final input picture from described video flowing;
Described edge calculations unit is used for utilizing the detection template of the horizontal direction of sobel operator to calculate the horizontal edge of described input picture edge pixel point;
Described black and white processing unit is used to utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtains the license plate area of black and whiteization;
Described car plate correcting unit is used for calculating the angle of the horizontal direction of edge pixel point in the described license plate area according to described horizontal edge the license plate image after obtaining proofreading and correct.
Described Character segmentation module comprises: projecting cell, vertical sweep unit, horizontal scanning unit and character picture generation unit;
Described projecting cell is used for the license plate image after proofreading and correct is carried out projection on perpendicular to the direction on car plate surface;
Described vertical sweep unit is used for vertically,, to the other end image after the projection is lined by line scan until running into first black pixel point since an end, determines the reference position of whole character zone; Begin in the other direction image to be lined by line scan until running into first black pixel point from the other end, determine the final position of whole character zone; The altitude range of determining whole character zone according to the reference position and the final position of described whole character zone; Also be used in the width range of each character of determining, respectively vertically, line by line scan until running into first black pixel point, determine the reference position of each character since an end; Begin to line by line scan until running into first black pixel point in the other direction from the other end, determine the final position of each character; The altitude range of determining each character according to the reference position and the final position of described this character;
Described horizontal scanning unit, in the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
Described character picture generation unit is used for the image that width range and altitude range by each character of determining obtain each character.
Further comprise the rear end control device in the system of described car plate identification;
Described car plate APU is further used for after car plate identification is finished license plate recognition result being sent to the rear end control device by Ethernet or wireless network;
Described rear end control device is used for after receiving license plate recognition result, and it is invalid that the vehicle deceleration enable signal in the vehicle deceleration device is changed to, through it being reverted to effectively after the time delay of setting again; Also be used for license plate recognition result that will receive and the permission car plate catalogue of preserving in advance and compare, when the car plate that identifies is not included in the described permission car plate catalogue, described license board information is preserved, and carry out unusual car plate in the suitable time and report.
As seen from the above technical solutions, the method and system of this car plate identification of the embodiment of the invention, judge whether to trigger car plate identification according to video flowing and vehicle deceleration enable signal, can reduce because the flase drop that chaff interference causes, and the car plate identification trigger mechanism of the method and system of this car plate identification needn't destroy the road surface and carry out Installation and Debugging, so the cost of installation and maintenance is lower.
Description of drawings
Fig. 1 is the schematic flow sheet of the method for car plate identification in the prior art.
Fig. 2 is the schematic flow sheet of the method for car plate identification in the embodiment of the invention.
Fig. 3 is a kind of composition structural representation of the system of car plate identification in the embodiment of the invention.
Fig. 4 is that the another kind of the system of car plate identification in the embodiment of the invention is formed structural representation.
Embodiment
For making purpose of the present invention, technical scheme and advantage clearer, below with reference to the accompanying drawing embodiment that develops simultaneously, the present invention is described in more detail.
The embodiment of the invention provides the method for a kind of car plate identification, its flow process as shown in Figure 2, comprising:
Step 201: gather video flowing, judge whether to carry out car plate identification according to described video flowing and vehicle deceleration enable signal;
Step 202: when judging that not carrying out car plate discerns, continue to gather video flowing and also carry out described judgement; Otherwise, use described video flowing as input picture;
Step 203: calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
Step 204: the license plate image after the described correction is carried out the image that Character segmentation obtains each character;
Step 205: the sorter that utilizes training in advance to obtain is discerned the image of each character, obtains final license plate recognition result according to the character that identifies.
Below above-mentioned steps 201～205 is elaborated:
Judge whether that according to described video flowing and vehicle deceleration enable signal the method for carrying out car plate identification is in the step 201:
Judged whether that according to described video flowing vehicle enters the visual field; When judgement has vehicle to enter the visual field, judge whether the vehicle deceleration enable signal is in effective status, if described enable signal is effective, then carry out car plate identification; If described enable signal is invalid, or judge when not having vehicle to enter the visual field, then do not carry out car plate identification.
The method of the described car plate identification of the embodiment of the invention, can be applied under the various vehicle safety check scenes, be provided with the vehicle deceleration device this moment slows down to vehicle, and vehicle decelerates to lower-speed state or static fully after entering the visual field, and can obtain the license plate image more clearly of vehicle this moment.Such as pay parking commonly used, be typically provided with at the Entrance place and block bar, can sensor be set on the bar blocking, to put down when blocking bar, expression is changed to effective status with the vehicle deceleration enable signal; When blocking bar and packing up, then the vehicle deceleration enable signal is changed to disarmed state.
Wherein, judge whether that according to described video flowing the method that vehicle enters the visual field comprises:
A, set up background image in advance, utilize each present frame in the described video flowing poor with background image respectively, obtain the multiframe frame difference image;
B, the object size parameter that obtains entering the visual field according to described frame difference image, when described object size parameter during greater than the value set, thinking has vehicle to enter the visual field; Otherwise thinking does not have vehicle to enter the visual field.
Because the application scenarios relative fixed of car plate identification, therefore the method for embodiment of the invention employing is for gathering Same Scene (hereinafter is called the background image modeling time with this certain hour section) in the certain hour section image, the average brightness of each pixel in the computed image is according to described mean value generation background image.Background image is in case generate, (it also is the predefined time at set intervals, but be different from the background image modeling time in the preamble, be the expression difference, update time hereinafter referred to as), background image is upgraded, promptly re-execute the average brightness of each pixel in the described computed image, step according to described mean value generation background image, like this, just can reduce the influence of uncertain factors such as irradiate light round the clock effectively, thereby can when guaranteeing to detect accuracy rate, reduce the complexity of system greatly.
Further comprise in the step 203:
Step 203a: the horizontal edge of edge pixel point in the detection template calculating input image of the horizontal direction of employing sobel operator;
Step 203b: utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtain the license plate area of black and whiteization; The black and white processing of this moment is meant the black and white binary conversion treatment, is about to image and changes into the image that two kinds of colors of black and white constitute, and is different from the black white image that common described multi-stage grey scale constitutes.
Step 203c: again according to described horizontal edge, calculate the angle of the horizontal direction of edge pixel point in the described license plate area, the license plate image after obtaining proofreading and correct.
In the input picture, the intersection of car plate profile and background image has strong edge, and the edge of car plate profile and background intersection is the important information that carries out car plate identification, car plate profile and background image can be made a distinction according to marginal information; Simultaneously, the edge not only has intensity, but also has certain direction, because license plate image does not always face video capture device but may have certain angle, therefore before carrying out character recognition,, need proofread and correct this license plate image in order to guarantee good discrimination.
The acquiring method at pixel edge has a variety of, and is commonly used as sobel or prewitt operator, adopts the sobel operator in the embodiment of the invention, and concrete use-pattern is same as the prior art, so repeat no more.But it is pointed out that and equally also can adopt other ripe edge calculations methods herein.
Among the step 203c, the method that a kind of feasible license plate image is proofreaied and correct is: ask the distribution covariance matrix of the matrix of original car plate picture signal earlier, further obtain the eigenwert and the corresponding proper vector (each proper vector is called principal component again) of this covariance matrix.Each eigenwert is arranged by descending, and each eigenwert and whole ratio of eigenwert summations are called the contribution rate of the corresponding principal component of this eigenwert, characterize the number percent that this component is represented the signal energy in the original license plate image; The number of principal component is chosen in decision as required, and the angle that calculates the principal component respectively chosen and coordinate axis is respectively determined license plate image to be rotated the license plate image after obtaining proofreading and correct according to the angle of inclination in the angle of inclination of license plate image.
Further comprise in the step 204:
Step 204a: the license plate image after will proofreading and correct carries out projection on perpendicular to the direction on car plate surface;
Step 204b: vertically, to the other end, the image after the projection is lined by line scan until running into first black pixel point, determine the reference position of whole character zone since an end; Begin in the other direction image to be lined by line scan until running into first black pixel point from the other end, determine the final position of whole character zone; The altitude range of determining whole character zone according to the reference position and the final position of described whole character zone;
Step 204c: in the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
Step 204d: in the width range of each character of determining, respectively vertically, line by line scan until running into first black pixel point, determine the reference position of each character since an end; Begin to line by line scan until running into first black pixel point in the other direction from the other end, determine the final position of each character; The altitude range of determining each character according to the reference position and the final position of described this character;
So far, the image that just can obtain each character according to the width range and the altitude range of each character of determining.
In step 205, because the characteristic information relative fixed that comprises in the car plate can be trained sorter targetedly according to various dissimilar car plate templates when therefore described sorter being trained in advance.
For example: the common car plate of domestic application comprises Chinese character, English alphabet and digital three classes usually, Chinese character generally is called for short beginning by the each province, as capital, Shanghai, Guangdong etc., English alphabet be among A～Z any one or a plurality of, also comprise different colouring informations in the car plate simultaneously, as white, blueness and yellow car plate etc.; And special car plate also all comprises special separately characteristic information usually as People's Armed Police, army's licence plate etc.For numeral and English alphabet, utilizable characteristic information comprises:
1) number of edges, the edge of a character of expression is communicated with the number of profile, and for example the number of edges of character 6 is 2, and the number of edges of letter e is 1 etc.
2) edge gravity center is primarily aimed at number of edges and is 2 character, calculates the coordinate of top, two edges and bottom respectively, is designated as y
Edge1 Top, y
Edge1 Bottom, y
Edge2 Top, y
Edge2 Bottom, the difference at two edges of this that ask respectively again
So work as D
Top/ D
Bottom＜0.5, edge gravity center is at the first half; Work as D
Top/ D
Bottom＞2, edge gravity center is in Lower Half.
3) contour feature value uses the single order differential variation tendency definition of profile to constitute the elementary cell of character outline, be divided into a left side tiltedly, right tiltedly, straight line, circular arc, sudden change and uncertain etc.
4) stroke number, a certain position of character up and down or about draw a straight line arbitrarily, according to its through white (the character track is a black) zone number of times character is divided.
5) projection is carried out vertically and the projection of horizontal direction character, the projection amount of statistics diverse location.
6) nose calculates vertically or the white line length of horizontal direction connected region, finds out wherein the longest one then, can look in entire image also and can look in the zone of certain setting.
7) area ratio, calculate character area occupied in a certain setting regions, be mainly used in differentiation such as 0, character such as D, O or Q, need when quadraturing white connected region is filled the area that the black region after filling is calculated in the back, and then calculate the area of white connected region, obtain described character area occupied after doing difference.
In actual applications, also have many further feature information can be used for training classifier, enumerate no longer one by one as space is limited herein, particular content can list of references " design of Number-Plate Character Recognition algorithm ", " video technique is used and engineering ", 1002-8692 (2007) 05-0088-03, the author opens and thanks to China.Understand easily, by the combination of using various features information sorter is trained in advance, when carrying out character recognition, the character picture after will cutting apart is again sent in the sorter after the training, and sorter just can extract feature and discerns according to the sorting technique of setting from character to be identified.Sorter after described training classifier and the utilization training carries out the known technology that character recognition belongs to those skilled in the art, so no longer describe in detail.
It is pointed out that above description only is for several alternative characteristic informations being described, should not be construed as limiting.Simultaneously, selected characteristic information when the Chinese character that comprises in the car plate is discerned can repeat no more with reference to pertinent literature herein.
The system that the embodiment of the invention also provides a kind of car plate identification forms structure as shown in Figure 3, comprising: video acquisition device 310, vehicle deceleration device 340, car plate identification flip flop equipment 320 and car plate APU 330;
Car plate identification flip flop equipment 320 is used for judging whether to trigger car plate identification according to video flowing that receives and vehicle deceleration enable signal; When judging that not triggering car plate discerns, continue the video flowing of receiver, video harvester 310 transmissions and also carry out described judgement; Otherwise, trigger car plate APU 330 and carry out car plate identification;
Vehicle deceleration device 340 is used for the state of vehicle deceleration enable signal is sent to car plate identification flip flop equipment 320;
In actual applications, video acquisition device 310 can adopt special-purpose monitoring camera or traditional camera, can also add corresponding lighting compensating device in case of necessity can be applied to dark surrounds.
Wherein, comprise in the car plate identification flip flop equipment 320: target discrimination module 321 and trigger module 322;
Further comprise again in the target discrimination module 321: background image modeling unit 321a, target sizes computing unit 321b and target sizes comparing unit 321c;
Background image modeling unit 321a is used to calculate the average brightness of Same Scene each pixel in the background image modeling time, and according to described mean value generation background image, the length of described background image modeling time preestablishes; Every predefined update time, initiate to carry out the step of the new described average brightness of calculating, the background image after upgrading according to the new average brightness generation that calculates;
Target sizes computing unit 321b, it is poor with background image respectively to be used for each present frame of described video flowing, obtains the multiframe frame difference image, and obtains entering the object size parameter of visual field according to described frame difference image;
Target sizes comparing unit 321c is used for described object size parameter and predefined value are compared, and when described object size parameter during greater than predefined value, judging has vehicle to enter the visual field; Otherwise judging does not have vehicle to enter the visual field; With judged result notification triggers module 322.
License plate image locating module 331, be used for trigger pip according to described car plate identification flip flop equipment 320, the video flowing that receiver, video harvester 310 sends is as input picture, calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
Character recognition module 333, the sorter that is used to utilize training in advance to obtain is discerned the image of each character, obtains license plate recognition result according to the character that identifies.
Need to prove, above-mentioned when calculating the marginal information of car plate according to described input picture, because what video acquisition device 310 collected is multiple image, can take this moment two kinds of methods to carry out car plate identification with described multiple image, correspondingly, 330 pairs of car plate APUs should have two kinds of different composition structures:
One, as shown in Figure 3, the multiple image that 330 pairs of video acquisition devices of car plate APU this moment 310 send carries out car plate identification:
At this moment, license plate image locating module 331 comprises: edge calculations unit 331a, black and white processing unit 331b and car plate correcting unit 331c;
Edge calculations unit 331a is used for utilizing the detection template of the horizontal direction of sobel operator to calculate the horizontal edge of described input picture edge pixel point;
Black and white processing unit 331b is used to utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtains the license plate area of black and whiteization;
Car plate correcting unit 331c is used for according to described horizontal edge, calculates the angle of the horizontal direction of edge pixel point in the license plate area of described black and whiteization, the license plate image after obtaining proofreading and correct;
Correspondingly, further comprise the warm module 334 of multiframe in the car plate APU 330 at this moment;
The warm module 334 of described multiframe is used to merge the recognition result that multiframe comprises the image of each character, when described recognition result satisfies predefined playing a card during condition, generates final recognition result; The described predefined condition of playing a card is: the recognition sample of this car plate reaches default quantity, and the degree of confidence of each character identification result in the car plate reaches default confidence weighting value.
Its two, as shown in Figure 4, this moment, car plate APU 330 selected a frame to carry out car plate identification as input picture from the multiple image that video acquisition device 310 sends:
At this moment, also further comprise input picture determining unit 331d in the license plate image locating module 331;
Input picture determining unit 331d is used for utilizing estimation to select a pairing two field picture of the slowest moment of car speed to carry out car plate identification as final input picture from described video flowing;
Correspondingly, no longer comprise the warm module 334 of multiframe in the car plate APU 330 at this moment.
Understand easily, the difference of above-mentioned two kinds of embodiments only is the difference of car plate APU 330 structures, and other parts of the system of described vehicle identification are identical, therefore, will continue to introduce other ingredients of this system below.
Projecting cell 332a is used for the license plate image after proofreading and correct is carried out projection on perpendicular to the direction on car plate surface;
Horizontal scanning unit 332c, in the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
Character picture generation unit 332d is used for the image that width range and altitude range by each character of determining obtain each character.
Preferably, can further include rear end control device 350 in the system of the identification of the car plate described in the embodiment of the invention;
At this moment, described car plate APU 330 is further used for after car plate identification is finished license plate recognition result being sent to rear end control device 350 by Ethernet or wireless network;
Rear end control device 350 is used for after receiving license plate recognition result, and it is invalid that the vehicle deceleration enable signal in the described vehicle deceleration device 340 is changed to, through it being reverted to effectively after the time delay of setting again; Also be used for license plate recognition result that will receive and the permission car plate catalogue of preserving in advance and compare, when the car plate that identifies is not included in the described permission car plate catalogue, described license board information is preserved, and carry out unusual car plate in the suitable time and report.
In actual applications, fax mechanism can be set in vehicle deceleration device 340, the default state of remaining valid of described vehicle deceleration enable signal, this moment, fax mechanism was in corresponding effective status (for example block bar be in down state or block door be in closed condition).When car plate identification is finished, rear end control device 350 sends asserts signal (for example negate signal) to described vehicle deceleration device 340, it is invalid that described vehicle deceleration enable signal is changed to, then fax mechanism carries out corresponding actions (for example block bar and lift or block a unlatching) according to this enable signal after invalid, through (to guarantee being arranged normally by described vehicle deceleration device 340 the enough time) after the time delay of setting by the clearance vehicle, described rear end control device 350 sends reset signal described vehicle deceleration enable signal is changed to default setting again, then the state (for example block bar relay down or block door close once more) before the car plate identification recovers to carry out in fax mechanism, prepares to begin next time car plate and discerns.
Rear end control device 350 herein is an example just, difference according to application scenarios, can also further in the system of described car plate identification, add various functional modules, for example in pay parking, can be in the vehicle deceleration device of the gateway in garage the integrated display screen that is provided with, be used for showing respectively that vehicle enters and leave the time in parking lot, simultaneously in the control device of rear end accounting module is set, will entering and leave toll amount that the Time Calculation in parking lot goes out according to vehicle, to be presented at the display screen in exit when vehicle leaves EXIT first-class.
By as seen above-mentioned, the method and system of the car plate identification that the embodiment of the invention provides, judge whether to trigger car plate identification according to video flowing and vehicle deceleration enable signal, thereby can reduce because the flase drop that chaff interference causes, and the car plate identification trigger mechanism of the method and system of this car plate identification needn't destroy the road surface and carry out Installation and Debugging, so the cost of installation and maintenance is lower.
Therefore; understand easily, the above is preferred embodiment of the present invention only, is not to be used to limit spirit of the present invention and protection domain; equivalent variations that any those of ordinary skill in the art made or replacement all should be considered as being encompassed within protection scope of the present invention.
Claims (16)
1, a kind of method of car plate identification is characterized in that this method comprises:
Gather video flowing, judge whether to trigger car plate identification according to described video flowing and vehicle deceleration enable signal;
When judging that not carrying out car plate discerns, continue to gather video flowing and also carry out described judgement; Otherwise, use described video flowing as input picture;
Calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
License plate image after the described correction is carried out the image that Character segmentation obtains each character;
The sorter that utilizes training in advance to obtain is discerned the image of each character, obtains final license plate recognition result according to the character that identifies.
2, method according to claim 1 is characterized in that, judges whether that according to described video flowing and vehicle deceleration enable signal the method that triggers car plate identification comprises:
Judged whether that according to described video flowing vehicle enters the visual field; When judgement has vehicle to enter the visual field, judge further whether the vehicle deceleration enable signal is in effective status, if described enable signal is effective, then trigger car plate identification; Otherwise, do not trigger car plate identification.
3, method according to claim 2 is characterized in that, judges whether that according to described video flowing the method that vehicle enters the visual field comprises:
Set up background image in advance, utilize each present frame in the described video flowing poor with background image respectively, obtain the multiframe frame difference image;
The object size parameter that obtains entering the visual field according to described frame difference image, when described object size parameter during greater than the value set, thinking has vehicle to enter the visual field; Otherwise thinking does not have vehicle to enter the visual field.
4, method according to claim 3 is characterized in that, the described method of setting up background image in advance comprises:
Calculate the average brightness of Same Scene each pixel in the background image modeling time, according to described mean value generation background image, the length of described background image modeling time preestablishes;
Every predefined update time, initiate to carry out the step of the new described average brightness of calculating, the background image after upgrading according to the new average brightness generation that calculates.
5, method according to claim 1 is characterized in that, calculates the marginal information of car plate according to described input picture, determines the position of license plate area, and the method for the license plate image after obtaining proofreading and correct according to described marginal information comprises:
Utilize estimation to select the slowest pairing two field picture of the moment of car speed from described input picture, the detection template of the horizontal direction of employing sobel operator is calculated the horizontal edge of edge pixel point in this two field picture;
Utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtain the license plate area of black and whiteization;
According to described horizontal edge, calculate the angle of the horizontal direction of edge pixel point in the described license plate area again, the license plate image after obtaining proofreading and correct.
6, method according to claim 1 is characterized in that, calculates the marginal information of car plate according to described input picture, determines the position of license plate area, and the method for the license plate image after obtaining proofreading and correct according to described marginal information comprises:
The detection template of the horizontal direction of employing sobel operator is calculated the horizontal edge of edge pixel point in the described input picture respectively;
Utilize setting threshold that the image-region between the horizontal edge in the described input picture is carried out the black and white processing, obtain the image of the license plate area of black and whiteization;
Again according to described horizontal edge, calculate the angle of the horizontal direction of edge pixel point in the image of described license plate area respectively, obtain the license plate image after multiframe is proofreaied and correct.
The described sorter that utilizes training in advance to obtain is discerned the image of each character, and the method that obtains final license plate recognition result according to the character that identifies comprises:
The sorter that utilizes training in advance to obtain is discerned the image of each character;
Merge the recognition result that multiframe comprises the image of each character,, generate final recognition result when described recognition result satisfies predefined playing a card during condition; The described predefined condition of playing a card is: the recognition sample of this car plate reaches default quantity, and the degree of confidence of each character identification result in the car plate reaches default confidence weighting value.
7, method according to claim 1 is characterized in that, the license plate image after the described correction is carried out the method that Character segmentation obtains the image of each character comprise:
License plate image after proofreading and correct is carried out projection on perpendicular to the direction on car plate surface;
Vertically, to the other end, the image after the projection is lined by line scan until running into first black pixel point, determine the reference position of whole character zone since an end; Begin in the other direction image to be lined by line scan until running into first black pixel point from the other end, determine the final position of whole character zone; The altitude range of determining whole character zone according to the reference position and the final position of described whole character zone;
In the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
In the width range of each character of determining, respectively vertically, line by line scan until running into first black pixel point since an end, determine the reference position of each character; Begin to line by line scan until running into first black pixel point in the other direction from the other end, determine the final position of each character; The altitude range of determining each character according to the reference position and the final position of described this character;
The image that obtains each character by the width range and the altitude range of each character of determining.
8, according to each described method in the claim 1 to 7, it is characterized in that this method further comprises:
After car plate identification was finished, it is invalid that the vehicle deceleration enable signal is changed to, and lets pass and finished the vehicle pass-through of car plate identification, through the vehicle deceleration enable signal being changed to effectively after the time delay of setting again.
9, a kind of system of car plate identification is characterized in that this system comprises: video acquisition device, vehicle deceleration device, car plate identification flip flop equipment and car plate APU;
Described video acquisition device is used to gather video flowing and sends to car plate identification flip flop equipment and car plate APU;
Described vehicle deceleration device is used for the state of vehicle deceleration enable signal is sent to car plate identification flip flop equipment;
Described car plate identification flip flop equipment is used for judging whether to trigger car plate identification according to video flowing that receives and vehicle deceleration enable signal; When judging that not triggering car plate discerns, continue the video flowing of receiver, video harvester transmission and also carry out described judgement; Otherwise, trigger the car plate APU and carry out car plate identification;
Described car plate APU, be used for trigger pip according to described car plate identification flip flop equipment, the video flowing that the receiver, video harvester sends is as input picture, calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information; License plate image after the described correction is carried out the image that Character segmentation obtains each character; The sorter that utilizes training in advance to obtain is discerned the image of each character, obtains license plate recognition result according to the character that identifies.
10, system according to claim 9 is characterized in that, comprises in the described car plate identification flip flop equipment: target discrimination module and trigger module;
Described target discrimination module is used for having judged whether that according to described video flowing vehicle enters the visual field, and with judged result notification triggers module;
Described trigger module, the judged result that is used for the receiving target determination module, when described target discrimination module judgement has vehicle to enter the visual field, further the state of the vehicle deceleration enable signal of sending according to the vehicle deceleration device is judged, when described vehicle deceleration enable signal when being effective, trigger described car plate APU and carry out car plate identification; Otherwise, do not trigger described car plate APU.
11, system according to claim 10 is characterized in that, comprises in the described target discrimination module: background image modeling unit, target sizes computing unit and target sizes comparing unit;
Described background image modeling unit is used to calculate the average brightness of Same Scene each pixel in the background image modeling time, and according to described mean value generation background image, the length of described background image modeling time preestablishes; Every predefined update time, initiate to carry out the step of the new described average brightness of calculating, the background image after upgrading according to the new average brightness generation that calculates;
Described target sizes computing unit, it is poor with background image respectively to be used for each present frame of described video flowing, obtains the multiframe frame difference image, and obtains entering the object size parameter of visual field according to described frame difference image;
Described target sizes comparing unit is used for described object size parameter and predefined value are compared, and when described object size parameter during greater than predefined value, judging has vehicle to enter the visual field; Otherwise judging does not have vehicle to enter the visual field; With judged result notification triggers module.
12, system according to claim 9 is characterized in that, described car plate APU comprises: license plate image locating module, Character segmentation module and character recognition module;
Described license plate image locating module, be used for trigger pip according to described car plate identification flip flop equipment, the video flowing that the receiver, video harvester sends is as input picture, calculate the marginal information of car plate according to described input picture, determine the position of license plate area, and the license plate image after obtaining proofreading and correct according to described marginal information;
Described Character segmentation module is used for the license plate image after proofreading and correct is carried out the image that Character segmentation obtains each character;
Described character recognition module, the sorter that is used to utilize training in advance to obtain is discerned the image of each character, obtains license plate recognition result according to the character that identifies.
13, system according to claim 12, its feature with, comprise in the described license plate image locating module: edge calculations unit, black and white processing unit and car plate correcting unit;
Described edge calculations unit is used for utilizing the detection template of the horizontal direction of sobel operator to calculate the horizontal edge of described input picture edge pixel point;
Described black and white processing unit is used to utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtains the license plate area of black and whiteization;
Described car plate correcting unit is used for according to described horizontal edge, calculates the angle of the horizontal direction of edge pixel point in the license plate area of described black and whiteization, the license plate image after obtaining proofreading and correct;
Further comprise the warm module of multiframe in the described car plate APU;
The warm module of described multiframe is used to merge the recognition result that multiframe comprises the image of each character, when described recognition result satisfies predefined playing a card during condition, generates final recognition result; The described predefined condition of playing a card is: the recognition sample of this car plate reaches default quantity, and the degree of confidence of each character identification result in the car plate reaches default confidence weighting value.
14, system according to claim 12, its feature with, comprise the input picture determining unit in the described license plate image locating module, edge calculations unit, black and white processing unit and car plate correcting unit;
Described input picture determining unit is used for utilizing estimation to select a pairing two field picture of the slowest moment of car speed to carry out car plate identification as final input picture from described video flowing;
Described edge calculations unit is used for utilizing the detection template of the horizontal direction of sobel operator to calculate the horizontal edge of described input picture edge pixel point;
Described black and white processing unit is used to utilize setting threshold that the image-region between the horizontal edge is carried out the black and white processing, obtains the license plate area of black and whiteization;
Described car plate correcting unit is used for calculating the angle of the horizontal direction of edge pixel point in the described license plate area according to described horizontal edge the license plate image after obtaining proofreading and correct.
15, system according to claim 12 is characterized in that, described Character segmentation module comprises: projecting cell, vertical sweep unit, horizontal scanning unit and character picture generation unit;
Described projecting cell is used for the license plate image after proofreading and correct is carried out projection on perpendicular to the direction on car plate surface;
Described vertical sweep unit is used for vertically,, to the other end image after the projection is lined by line scan until running into first black pixel point since an end, determines the reference position of whole character zone; Begin in the other direction image to be lined by line scan until running into first black pixel point from the other end, determine the final position of whole character zone; The altitude range of determining whole character zone according to the reference position and the final position of described whole character zone; Also be used in the width range of each character of determining, respectively vertically, line by line scan until running into first black pixel point, determine the reference position of each character since an end; Begin to line by line scan until running into first black pixel point in the other direction from the other end, determine the final position of each character; The altitude range of determining each character according to the reference position and the final position of described this character;
Described horizontal scanning unit, in the altitude range of the whole character zone of determining, along continuous straight runs begins to pursue column scan until running into first black pixel point to opposite side from a side, determines the reference position of cutting apart of first character; Continuation is pursued column scan until running into the position that first row do not comprise black pixel point to opposite side, determines the final position of cutting apart of first character; Continuation is pursued column scan to opposite side, repeats described the cutting apart reference position and cutting apart the step of final position of character of determining, and up to being scanned up to the opposite side end points, determines the width range of each character in the whole character zone;
Described character picture generation unit is used for the image that width range and altitude range by each character of determining obtain each character.
16, according to each described system in the claim 9 to 15, it is characterized in that, further comprise the rear end control device in this system;
Described car plate APU is further used for after car plate identification is finished license plate recognition result being sent to the rear end control device by Ethernet or wireless network;
Described rear end control device is used for after receiving license plate recognition result, and it is invalid that the vehicle deceleration enable signal in the vehicle deceleration device is changed to, through it being reverted to effectively after the time delay of setting again; Also be used for license plate recognition result that will receive and the permission car plate catalogue of preserving in advance and compare, when the car plate that identifies is not included in the described permission car plate catalogue, described license board information is preserved, and carry out unusual car plate in the suitable time and report.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200810114753XA CN101303803A (en) | 2008-06-11 | 2008-06-11 | Method and system for discriminating license plate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200810114753XA CN101303803A (en) | 2008-06-11 | 2008-06-11 | Method and system for discriminating license plate |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101303803A true CN101303803A (en) | 2008-11-12 |
Family
ID=40113695
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA200810114753XA Pending CN101303803A (en) | 2008-06-11 | 2008-06-11 | Method and system for discriminating license plate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101303803A (en) |
Cited By (42)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833859A (en) * | 2010-05-14 | 2010-09-15 | 山东大学 | Self-triggering license plate recognition method based on virtual coil |
| CN101859382A (en) * | 2010-06-03 | 2010-10-13 | 复旦大学 | A license plate detection and recognition method based on the maximum stable extremum region |
| CN102096821A (en) * | 2010-04-15 | 2011-06-15 | 西安理工大学 | Number plate identification method under strong interference environment on basis of complex network theory |
| CN102243705A (en) * | 2011-05-09 | 2011-11-16 | 东南大学 | Method for positioning license plate based on edge detection |
| CN102685364A (en) * | 2011-02-11 | 2012-09-19 | 微软公司 | Updating a low frame rate image by using a high frame rate image stream |
| CN102915574A (en) * | 2012-08-30 | 2013-02-06 | 甘肃紫光智能交通与控制技术有限公司 | Embedded intelligent traffic control toll terminal |
| CN103000029A (en) * | 2012-11-20 | 2013-03-27 | 河南亚视软件技术有限公司 | Vehicle video recognition method and application thereof |
| CN103136528A (en) * | 2011-11-24 | 2013-06-05 | 同济大学 | Double-edge detection based vehicle license plate identification method |
| CN103236167A (en) * | 2013-03-27 | 2013-08-07 | 河南亚视软件技术有限公司 | Vehicular video monitoring method for vehicular access entries, and application thereof |
| CN103258431A (en) * | 2012-02-20 | 2013-08-21 | 三星泰科威株式会社 | Automatic number plate recognition system and method for transmitting vehicle information |
| CN103366577A (en) * | 2013-05-16 | 2013-10-23 | 王斌 | Vehicle license authenticity recognition system |
| CN103903005A (en) * | 2012-12-25 | 2014-07-02 | 财团法人交大思源基金会 | License plate image identification system and method |
| CN104036241A (en) * | 2014-05-30 | 2014-09-10 | 宁波海视智能系统有限公司 | License plate recognition method |
| CN104268596A (en) * | 2014-09-25 | 2015-01-07 | 深圳市捷顺科技实业股份有限公司 | License plate recognizer and license plate detection method and system thereof |
| WO2015066984A1 (en) * | 2013-11-08 | 2015-05-14 | 广州中智融通金融科技有限公司 | Complex background-oriented optical character recognition method and device |
| CN105156010A (en) * | 2015-08-28 | 2015-12-16 | 苏州市博得立电源科技有限公司 | Method for controlling garage roller shutter door on basis of image analysis |
| CN105391917A (en) * | 2015-11-05 | 2016-03-09 | 清华大学 | Interferent-detection-based video optimization method |
| CN105608446A (en) * | 2016-02-02 | 2016-05-25 | 北京大学深圳研究生院 | Video stream abnormal event detection method and apparatus |
| CN106203416A (en) * | 2016-07-08 | 2016-12-07 | 武汉工程大学 | A kind of express delivery document information input method based on scanogram and input device |
| CN106408950A (en) * | 2016-11-18 | 2017-02-15 | 北京停简单信息技术有限公司 | Parking lot entrance and exit license plate recognition system and method |
| CN103679191B (en) * | 2013-09-04 | 2017-02-22 | 西交利物浦大学 | An automatic fake-licensed vehicle detection method based on static state pictures |
| CN106683073A (en) * | 2015-11-11 | 2017-05-17 | 杭州海康威视数字技术股份有限公司 | License plate detection method, camera and server |
| CN108182808A (en) * | 2017-12-28 | 2018-06-19 | 广东傲智创新科技有限公司 | A kind of movable type car plate data collecting system |
| CN108229483A (en) * | 2018-01-11 | 2018-06-29 | 中国计量大学 | Based on the doorplate pressed characters identification device under caffe and soft triggering |
| CN108985137A (en) * | 2017-06-02 | 2018-12-11 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method, apparatus and system |
| CN110728278A (en) * | 2018-07-16 | 2020-01-24 | 杭州海康威视数字技术股份有限公司 | License plate recognition method and device |
| TWI690857B (en) * | 2018-03-14 | 2020-04-11 | 台達電子工業股份有限公司 | License plate recognition methods and systems thereof |
| CN111047723A (en) * | 2019-12-12 | 2020-04-21 | 杭州昊恒科技有限公司 | An urban smart behavior analysis system based on image processing |
| CN111340966A (en) * | 2019-12-12 | 2020-06-26 | 山东中创软件工程股份有限公司 | Vehicle detection method, device and system |
| CN111401360A (en) * | 2020-03-02 | 2020-07-10 | 杭州雄迈集成电路技术股份有限公司 | Method and system for optimizing license plate detection model and license plate detection method and system |
| CN111582272A (en) * | 2020-04-30 | 2020-08-25 | 平安科技（深圳）有限公司 | Double-row license plate recognition method, device and equipment and computer readable storage medium |
| CN111610925A (en) * | 2016-12-19 | 2020-09-01 | 西安艾润物联网技术服务有限责任公司 | Method and device for auxiliary input of vehicle license plate number |
| CN111856611A (en) * | 2019-04-17 | 2020-10-30 | 阿里巴巴集团控股有限公司 | Trigger detection method, device and system |
| CN112071083A (en) * | 2020-09-15 | 2020-12-11 | 台州市远行客网络技术有限公司 | Motor vehicle license plate relay identification system and license plate relay identification method |
| CN112101274A (en) * | 2020-09-24 | 2020-12-18 | 苏州科达科技股份有限公司 | License plate recognition method, device, equipment and readable storage medium |
| CN112418221A (en) * | 2020-12-02 | 2021-02-26 | 福建亿安智能技术有限公司 | Method for realizing recognition and positioning of non-motor vehicle license plate |
| CN112651309A (en) * | 2020-12-15 | 2021-04-13 | 广州小鹏自动驾驶科技有限公司 | Parking space number acquisition method, device, equipment and storage medium |
| CN112712692A (en) * | 2019-10-24 | 2021-04-27 | 上海宗保科技有限公司 | Vehicle monitoring and recognizing system |
| CN114677682A (en) * | 2022-03-22 | 2022-06-28 | 深圳市平方科技股份有限公司 | Method and system for identifying number of truck top |
| US11443535B2 (en) | 2018-03-14 | 2022-09-13 | Delta Electronics, Inc. | License plate identification method and system thereof |
| CN115052189A (en) * | 2022-06-06 | 2022-09-13 | 重庆法链科技有限责任公司 | Video stream identification system and method |
| CN117152732A (en) * | 2023-10-18 | 2023-12-01 | 北京精英智通科技股份有限公司 | Multi-feature-assisted license plate recognition method and system |
-
2008
- 2008-06-11 CN CNA200810114753XA patent/CN101303803A/en active Pending
Cited By (61)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102096821A (en) * | 2010-04-15 | 2011-06-15 | 西安理工大学 | Number plate identification method under strong interference environment on basis of complex network theory |
| CN102096821B (en) * | 2010-04-15 | 2014-05-07 | 西安理工大学 | Number plate identification method under strong interference environment on basis of complex network theory |
| CN101833859A (en) * | 2010-05-14 | 2010-09-15 | 山东大学 | Self-triggering license plate recognition method based on virtual coil |
| CN101833859B (en) * | 2010-05-14 | 2012-05-23 | 山东大学 | Self-triggering license plate identification method based on virtual coil |
| CN101859382A (en) * | 2010-06-03 | 2010-10-13 | 复旦大学 | A license plate detection and recognition method based on the maximum stable extremum region |
| CN102685364A (en) * | 2011-02-11 | 2012-09-19 | 微软公司 | Updating a low frame rate image by using a high frame rate image stream |
| CN102243705A (en) * | 2011-05-09 | 2011-11-16 | 东南大学 | Method for positioning license plate based on edge detection |
| CN103136528B (en) * | 2011-11-24 | 2016-06-15 | 同济大学 | A kind of licence plate recognition method based on dual edge detection |
| CN103136528A (en) * | 2011-11-24 | 2013-06-05 | 同济大学 | Double-edge detection based vehicle license plate identification method |
| CN103258431A (en) * | 2012-02-20 | 2013-08-21 | 三星泰科威株式会社 | Automatic number plate recognition system and method for transmitting vehicle information |
| CN103258431B (en) * | 2012-02-20 | 2017-04-26 | 韩华泰科株式会社 | Automatic number plate recognition system and method for transmitting vehicle information |
| CN102915574A (en) * | 2012-08-30 | 2013-02-06 | 甘肃紫光智能交通与控制技术有限公司 | Embedded intelligent traffic control toll terminal |
| CN103000029A (en) * | 2012-11-20 | 2013-03-27 | 河南亚视软件技术有限公司 | Vehicle video recognition method and application thereof |
| CN103903005A (en) * | 2012-12-25 | 2014-07-02 | 财团法人交大思源基金会 | License plate image identification system and method |
| CN103903005B (en) * | 2012-12-25 | 2016-12-28 | 财团法人交大思源基金会 | License plate image identification system and method |
| CN103236167B (en) * | 2013-03-27 | 2015-08-05 | 河南亚视软件技术有限公司 | Porte-cochere entrance automobile video frequency method for supervising and application thereof |
| CN103236167A (en) * | 2013-03-27 | 2013-08-07 | 河南亚视软件技术有限公司 | Vehicular video monitoring method for vehicular access entries, and application thereof |
| CN103366577B (en) * | 2013-05-16 | 2015-12-23 | 王斌 | Vehicle license authenticity recognition system |
| CN103366577A (en) * | 2013-05-16 | 2013-10-23 | 王斌 | Vehicle license authenticity recognition system |
| CN103679191B (en) * | 2013-09-04 | 2017-02-22 | 西交利物浦大学 | An automatic fake-licensed vehicle detection method based on static state pictures |
| WO2015066984A1 (en) * | 2013-11-08 | 2015-05-14 | 广州中智融通金融科技有限公司 | Complex background-oriented optical character recognition method and device |
| US9613266B2 (en) | 2013-11-08 | 2017-04-04 | Grg Banking Equipment Co., Ltd. | Complex background-oriented optical character recognition method and device |
| AU2014346263B2 (en) * | 2013-11-08 | 2016-11-17 | Grg Banking Equipment Co., Ltd. | Complex background-oriented optical character recognition method and device |
| CN104036241B (en) * | 2014-05-30 | 2018-11-09 | 宁波海视智能系统有限公司 | A kind of licence plate recognition method |
| CN104036241A (en) * | 2014-05-30 | 2014-09-10 | 宁波海视智能系统有限公司 | License plate recognition method |
| CN104268596B (en) * | 2014-09-25 | 2017-11-10 | 深圳市捷顺科技实业股份有限公司 | A kind of Car license recognition device and its detection method of license plate and system |
| CN104268596A (en) * | 2014-09-25 | 2015-01-07 | 深圳市捷顺科技实业股份有限公司 | License plate recognizer and license plate detection method and system thereof |
| CN105156010A (en) * | 2015-08-28 | 2015-12-16 | 苏州市博得立电源科技有限公司 | Method for controlling garage roller shutter door on basis of image analysis |
| CN105391917B (en) * | 2015-11-05 | 2018-10-16 | 清华大学 | Method for optimizing video based on interference analyte detection |
| CN105391917A (en) * | 2015-11-05 | 2016-03-09 | 清华大学 | Interferent-detection-based video optimization method |
| CN106683073A (en) * | 2015-11-11 | 2017-05-17 | 杭州海康威视数字技术股份有限公司 | License plate detection method, camera and server |
| CN106683073B (en) * | 2015-11-11 | 2020-02-18 | 杭州海康威视数字技术股份有限公司 | License plate detection method, camera and server |
| CN105608446B (en) * | 2016-02-02 | 2019-02-12 | 北京大学深圳研究生院 | A method and device for detecting abnormal events in a video stream |
| CN105608446A (en) * | 2016-02-02 | 2016-05-25 | 北京大学深圳研究生院 | Video stream abnormal event detection method and apparatus |
| CN106203416A (en) * | 2016-07-08 | 2016-12-07 | 武汉工程大学 | A kind of express delivery document information input method based on scanogram and input device |
| CN106408950A (en) * | 2016-11-18 | 2017-02-15 | 北京停简单信息技术有限公司 | Parking lot entrance and exit license plate recognition system and method |
| CN111610925A (en) * | 2016-12-19 | 2020-09-01 | 西安艾润物联网技术服务有限责任公司 | Method and device for auxiliary input of vehicle license plate number |
| CN108985137A (en) * | 2017-06-02 | 2018-12-11 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method, apparatus and system |
| CN108182808A (en) * | 2017-12-28 | 2018-06-19 | 广东傲智创新科技有限公司 | A kind of movable type car plate data collecting system |
| CN108229483A (en) * | 2018-01-11 | 2018-06-29 | 中国计量大学 | Based on the doorplate pressed characters identification device under caffe and soft triggering |
| TWI690857B (en) * | 2018-03-14 | 2020-04-11 | 台達電子工業股份有限公司 | License plate recognition methods and systems thereof |
| US11443535B2 (en) | 2018-03-14 | 2022-09-13 | Delta Electronics, Inc. | License plate identification method and system thereof |
| CN110728278A (en) * | 2018-07-16 | 2020-01-24 | 杭州海康威视数字技术股份有限公司 | License plate recognition method and device |
| CN111856611A (en) * | 2019-04-17 | 2020-10-30 | 阿里巴巴集团控股有限公司 | Trigger detection method, device and system |
| CN112712692A (en) * | 2019-10-24 | 2021-04-27 | 上海宗保科技有限公司 | Vehicle monitoring and recognizing system |
| CN111340966A (en) * | 2019-12-12 | 2020-06-26 | 山东中创软件工程股份有限公司 | Vehicle detection method, device and system |
| CN111047723B (en) * | 2019-12-12 | 2021-01-05 | 杭州昊恒科技有限公司 | City wisdom behavior analysis system based on image processing |
| CN111047723A (en) * | 2019-12-12 | 2020-04-21 | 杭州昊恒科技有限公司 | An urban smart behavior analysis system based on image processing |
| CN111340966B (en) * | 2019-12-12 | 2022-03-08 | 山东中创软件工程股份有限公司 | Vehicle detection method, device and system |
| CN111401360A (en) * | 2020-03-02 | 2020-07-10 | 杭州雄迈集成电路技术股份有限公司 | Method and system for optimizing license plate detection model and license plate detection method and system |
| CN111401360B (en) * | 2020-03-02 | 2023-06-20 | 杭州雄迈集成电路技术股份有限公司 | Method and system for optimizing license plate detection model, license plate detection method and system |
| CN111582272A (en) * | 2020-04-30 | 2020-08-25 | 平安科技（深圳）有限公司 | Double-row license plate recognition method, device and equipment and computer readable storage medium |
| CN112071083B (en) * | 2020-09-15 | 2022-03-01 | 深圳市领航城市科技有限公司 | Motor vehicle license plate relay identification system and license plate relay identification method |
| CN112071083A (en) * | 2020-09-15 | 2020-12-11 | 台州市远行客网络技术有限公司 | Motor vehicle license plate relay identification system and license plate relay identification method |
| CN112101274A (en) * | 2020-09-24 | 2020-12-18 | 苏州科达科技股份有限公司 | License plate recognition method, device, equipment and readable storage medium |
| CN112418221A (en) * | 2020-12-02 | 2021-02-26 | 福建亿安智能技术有限公司 | Method for realizing recognition and positioning of non-motor vehicle license plate |
| CN112651309A (en) * | 2020-12-15 | 2021-04-13 | 广州小鹏自动驾驶科技有限公司 | Parking space number acquisition method, device, equipment and storage medium |
| CN114677682A (en) * | 2022-03-22 | 2022-06-28 | 深圳市平方科技股份有限公司 | Method and system for identifying number of truck top |
| CN115052189A (en) * | 2022-06-06 | 2022-09-13 | 重庆法链科技有限责任公司 | Video stream identification system and method |
| CN117152732A (en) * | 2023-10-18 | 2023-12-01 | 北京精英智通科技股份有限公司 | Multi-feature-assisted license plate recognition method and system |
| CN117152732B (en) * | 2023-10-18 | 2024-02-27 | 北京精英智通科技股份有限公司 | Multi-feature-assisted license plate recognition method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101303803A (en) | Method and system for discriminating license plate | |
| CN201159903Y (en) | License plate recognition device | |
| CN110197589B (en) | Deep learning-based red light violation detection method | |
| CN101030256B (en) | Method and apparatus for cutting vehicle image | |
| CN109284674B (en) | Method and device for determining lane line | |
| CN106373426B (en) | Parking stall based on computer vision and violation road occupation for parking monitoring method | |
| CN101751744B (en) | Detection and early warning method of smoke | |
| CN101373517B (en) | Method and system for recognizing license plate | |
| CN101388145B (en) | Auto alarming method and device for traffic safety | |
| CN101877058B (en) | People flow rate statistical method and system | |
| US20030190058A1 (en) | Apparatus and method for measuring queue length of vehicles | |
| EP2535843A1 (en) | Method and system for population flow statistics | |
| CN104805784B (en) | Gate fare evasion detection system and gate fare evasion detection method | |
| CN106778648A (en) | Vehicle tracing and Vehicle License Plate Recognition System and recognition methods | |
| WO2008064621A1 (en) | Detection and categorization of light spots using a camera in a vehicle environment | |
| CN102867417A (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| KR102434154B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN110956104A (en) | Method, device and system for detecting overflow of garbage can | |
| CN101187985A (en) | Method and device for classification boundary of identifying object classifier | |
| WO2018008461A1 (en) | Image processing device | |
| CN109447090B (en) | Shield door obstacle detection method and system | |
| CN106339657A (en) | Straw incineration monitoring method and device based on monitoring video | |
| CN112307840A (en) | Indicator light detection method, device, equipment and computer readable storage medium | |
| CN112085018A (en) | License plate recognition system based on neural network | |
| CN110458093A (en) | A kind of Safe belt detection method and corresponding equipment based on driver's monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20081112 |