CN102074113B - License tag recognizing and vehicle speed measuring method based on videos - Google Patents
License tag recognizing and vehicle speed measuring method based on videos Download PDFInfo
- Publication number
- CN102074113B CN102074113B CN 201010285620 CN201010285620A CN102074113B CN 102074113 B CN102074113 B CN 102074113B CN 201010285620 CN201010285620 CN 201010285620 CN 201010285620 A CN201010285620 A CN 201010285620A CN 102074113 B CN102074113 B CN 102074113B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- car light
- car
- line
- headstock
- 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.)
- Active
Links
Images
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a license tag recognizing and vehicle speed measuring method based on videos. With the method, license tags from every frame of a video image are recognized, and the license tag recognizing results are calculated out according to the historical information, a higher recognition rate is realized. Meanwhile, the central position of characters can be positioned and tracked with the method to realize the accurate measurement of the vehicle speed. A vehicle head lower edge line detecting and tracking module positions a vehicle head lower edge line according to prospective motion and prospective edge detecting results, efficiently avoids the interference of shadows, captures the position of the vehicle head lower edge line, and tracks. After a trigger line is reached, the vehicle head lower edge line detecting and tracking module carries out space mapping on the historical tracking position to realize the accurate measurement of the vehicle speed. A vehicle lamp pair detecting and tracking module eliminates the interference of the circumference and headlamps of the vehicle in a night mode, accurately detects the vehicle lamp pair, and tracks. After the trigger line is reached, the vehicle lamp pair detecting and tracking module carries out space mapping on the historical tracking position to realize the accurate measurement of the vehicle speed.
Description
Technical field
The invention belongs to the intelligent traffic monitoring management domain, relate in particular to a kind of license plate identification and Vehicle Velocity Measurement Method of pure video.
Background technology
At present, the mode that the plate recognition system on the market mainly relies on hardware trigger or video to trigger is obtained the video image that a frame has vehicle, then by the related algorithm of image processing with pattern-recognition, identifies corresponding license plate information.
The mode normal operation magnetic induction loop of hardware trigger, toggle rate is higher, but the installation of coil need to destroy the road surface, and maintenance cost is higher; The mode that video triggers generally adopts licence plate location or virtual coil, installs simply, and is easy to maintenance, but because the complicacy of scene and the interference of light easily cause too many false declaration.
License plate identification based on the video triggering mode, only identify for triggering a frame vehicle image that comes, often owing to the travel direction of vehicle and instantaneous external interference, poor image quality or license plate area are blocked, be unfavorable for license plate identification, cause final identification error.
In the measurement of car speed, the modes that adopt detections of radar on the market more, accuracy of detection is high, but input cost is large; Large at vehicle flow, and the lower road surface of car speed, the phenomenon low with accuracy of detection also can appear reporting by mistake.
Summary of the invention
The object of the invention is to for the deficiencies in the prior art, a kind of license plate identification based on video and Vehicle Velocity Measurement Method are provided.
The objective of the invention is to be achieved through the following technical solutions: a kind of license plate identification and Vehicle Velocity Measurement Method based on video comprise the steps:
(1) live video stream obtains and transmits: obtain the real-time video flow data by the high definition camera that is installed on portal frame or the L bar, then by Internet Transmission to industrial computer.
(2) detect along the line under car plate identification, the headstock, the car light center line detecting: detect the video data identification license plate number that obtains, detect along the line under the headstock or car light center line.
(3) vehicle triggers: follow the tracks of along the line under car plate, the headstock or car light center line, when along the line under car plate, the headstock or car light center arrive the triggering line that sets in advance, send trigger pip, and by network delivery to high definition camera and LED light compensating lamp.
(4) triggering picture preserves: when the LED light compensating lamp obtains trigger pip, become quick-fried sudden strain of a muscle by stroboscopic, then with this frame picture by Internet Transmission to industrial computer and preserve.
(5) vehicle speed measurement: after vehicle triggers, calculate car speed according to different triggering modes.
(6) night on daytime, detecting pattern switched: along the line under the detection headstock under the monitoring pattern by day, detect the car light center line at night under the monitoring pattern.Relatively based on the vehicle toggle rate under the vehicle toggle rate of car plate identification and daytime, the night monitoring pattern, if the vehicle toggle rate under the daytime monitoring pattern can't meet the demands, then automatically switch to the detecting pattern at night.If the vehicle toggle rate under the night monitoring pattern can't meet the demands, then automatically switch to the detecting pattern on daytime.
Compared with prior art, the present invention has following technique effect:
1, the present invention is comprised of a high definition camera, many LED light compensating lamps, do not need other vehicle equipment, every two field picture is carried out license plate identification, headstock detection/car light detects, export license plate information after arriving trigger position, car speed and trigger flashing lamp, the present invention is simple in structure, and the interface arranges conveniently.Capture one high-definition image clearly can clearly be debated people's face and body color.
2, in every two field picture, carry out license plate identification, after vehicle arrives the triggering line, the a series of recognition result of statistical study, according to corresponding confidence level size, obtain the license plate identification result of a high confidence level, avoided because the discrimination Problem-Error that a frame image quality problem causes has improved the license plate identification rate greatly.
3, in every two field picture, detect the center position of licence plate, edge/car light position of center line under the headstock, and by the geometrical correspondence that the depth of field is calibrated, calculate world coordinates corresponding to image coordinate; After arriving the triggering line, according to historical position information, the instantaneous velocity of analyzing multiframe finally provides car speed, and accuracy of detection is high.
4, larger at vehicle flowrate, the perhaps lower environment of the speed of a motor vehicle, because the present invention carries out every frame to car plate, headstock or car light to detect and follow the tracks of, the different different vehicles of pursuit path representative still can carry out the detection of speed accurately.
5, according to the historical information of license plate identification and the historical information of headstock detection/car light detection, can identify fast and accurately current is daytime or evening, select the algorithm that headstock detects or car light detects, reach high detection and location rate, and can feed back to camera.
6, install simply, easy to maintenance, do not need brokenly the road to install; Accurate trigger flashing lamp when guaranteeing to have board and unlicensed vehicle to pass through obtains clearly high definition licence plate.
Description of drawings
Fig. 1 is based on the license plate identification of video and Vehicle Velocity Measurement Method figure;
Fig. 2 is the output logic process flow diagram of this method;
Fig. 3 is car plate identification and speed of a motor vehicle overhaul flow chart;
Fig. 4 is day mode headstock and lower overhaul flow chart along the line;
Fig. 5 is pattern car light and center line detecting process flow diagram in night.
Embodiment
Vehicle license plate information, vehicle flowrate and car speed information are data messages important among the ITS, and it provides effective parameter foundation for obtaining Traffic Information.
The present invention has adopted that every frame carries out license plate identification in video flowing, when vehicle arrive trigger line after, the mode with statistical history license plate information output licence plate result solves the mistake that single image identification brings, and greatly improves the license plate identification rate; By detection and the real-time follow-up to along the line under licence plate center, the headstock and car light center, pass through again new depth of field calibration algorithm, realize the measurement of car speed, guaranteed measuring accuracy, solved again the measurement problem that vehicle flow is large, travel speed is slow.
The license plate identification and the Vehicle Velocity Measurement Method that the present invention is based on video comprise the steps:
1, live video stream obtains and transmits: obtain the real-time video flow data by the high definition camera that is installed on portal frame or the L bar, then by Internet Transmission to industrial computer;
2, detect along the line under car plate identification, the headstock, the car light center line detecting: detect the video data identification license plate number that obtains, detect along the line under the headstock or car light center line;
3, vehicle triggers: follow the tracks of along the line under car plate, the headstock or car light center line, when along the line under car plate, the headstock or car light center arrive the triggering line that sets in advance, send trigger pip, and by network delivery to high definition camera and LED light compensating lamp;
4, triggering picture preserves: when the LED light compensating lamp obtains trigger pip, become quick-fried sudden strain of a muscle by stroboscopic, then with this frame picture by Internet Transmission to industrial computer and preserve;
5, vehicle speed measurement: after vehicle triggers, calculate car speed according to different triggering modes.A) licence plate triggers: by depth of field calibration algorithm, mapping licence plate center obtains instantaneous velocity to the displacement relation of world coordinate system according to the frame of being separated by, and exports accurately car speed information according to statistical algorithms again; B) trigger along the line under the headstock: by depth of field calibration algorithm, center along the line is to world coordinate system under the mapping headstock, and the displacement relation according to the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again; C) the car light center line triggers: by depth of field calibration algorithm, mapping car light center line center is to world coordinate system, and the displacement relation according to the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again;
6, night on daytime, detecting pattern switched: this method detects under the headstock under the pattern along the line by day, detects the car light center line under night mode.Relatively trigger vehicle toggle rate under toggle rate and daytime (night) pattern based on the vehicle of car plate identification, if the vehicle toggle rate under daytime (night) pattern can't meet the demands, then automatically switch to detecting pattern at night (daytime).
The present invention will be further described below in conjunction with the drawings and specific embodiments:
Fig. 1 be this method installation, schematic diagram is set, mainly comprise: high-definition camera 1, LED light compensating lamp 2 and 3, netting twine 7, switch 8, industrial computer 9; 4 is the ground lane line; 5 for needing the lane line of setting, and 6 is the triggering line that need to arrange.Wherein, high-definition camera is aimed at many tracks, selects suitable resolution of video camera and camera lens to guarantee the size of licence plate; Every the track needs a LED light compensating lamp, possesses stroboscopic and expose to the sun to dodge two kinds of functions; Lane line 5 needs manual the setting, corresponding output car Taoist monastic name when output licence plate and velocity information; Triggering 6 needs manual the setting, after vehicle arrival contacts this line, begins statistics and generates license plate information and corresponding velocity information.
This method only needs to set up at L bar or portal frame the LED light compensating lamp of a high-definition camera and corresponding number of track-lines in actual applications, does not need other vehicle equipment; Lane line, a parameter that triggers line and the calibration of several depth of field on arranging, the interface only need to be set.Then utilize the related algorithm of Video processing and pattern-recognition, in live video stream, finish the identification of licence plate, headstock detects and car light detects, and carry out the license plate information of high confidence level and the output of speed, the present invention is simple in structure, the interface arranges portable, and the license plate identification rate is high, and the speed accuracy of detection is high.
Fig. 2 is this method output logic process flow diagram.After the vehicle that licence plate is arranged entered scene, the license plate identification algorithm by frame by frame can obtain corresponding license plate information, and licence plate center information; Detect the headstock positional information by motion detection and track algorithm daytime simultaneously; Detect with track algorithm by car light and detect the car light positional information night.Like this, two kinds of algorithms of this method parallel running, a kind of is for the license plate identification algorithm that the board car is arranged, a kind of is to detect and the car light detection algorithm for the headstock of unlicensed vehicle.Fig. 2 process flow diagram is in order to solve the decision problem of net result output.
Step S21 after vehicle arrives the triggering line, need to add up historical information, exports final license plate information and car speed information.
Step S22, if the existence of license plate information is arranged in historical information, preferential output, directly statistical history license plate information, and licence plate center information is according to the highest output license plate information of confidence level, according to statistics output car speed information.
Step S23, output license plate information and car speed information.
Step S24, according to license plate identification and number of times that daytime, headstock car light in detection/night detected, judgement should be day mode, or night pattern.
Step S25, day mode and night pattern switching
Step S26, the historical information by car light center in night calculates running velocity
Step S27, the headstock by day mode detect and headstock under the historical information of position along the line, calculate running velocity.
Fig. 3 is license plate identification output information process flow diagram.This flow process is main flow of the present invention, the vehicle overwhelming majority of travelling on the road all is to hang car plate, and car plate is apparent, in this case, result's output all is to trust licence plate to detect and the result who identifies, and the detection of car speed also is the position of character center line after the trust license plate identification.
Step S31, real-time video flowing
Step S32 by the training to a large amount of car plate samples, extracts the aspect of model of car plate, travel through license plate area at video image, finally obtain accurately position, in this process, false zone might be introduced, this zone can be filtered out in the process of later stage segmentation and recognition.
Step S33, by the connected domain analysis to license plate area, and the analysis of horizontal projective histogram, be partitioned into each character in the license plate area.
Step S34 by the template matches in the pattern-recognition and support vector machine (SVM) algorithm, identifies each character, and provides corresponding confidence level.And generate the confidence level of overall identification.
Step S35 according to the result of segmentation of the characters and their identification, obtains the rectangle frame of being close to character, and adjusts the license plate area of obtaining among the S32, obtains accurately licence plate position
Step S36 preserves license number, recognition credibility, licence plate position
Step S37, whether license plate area arrives is triggered the line differentiation.
Step S38 according to historical license number and reliability information, exports accurately license plate information
Step S39, by depth of field calibration algorithm, mapping licence plate center is to world coordinate system.
Step S310, the displacement relation by the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again.
Fig. 4 is the process flow diagram that detects along the line under headstock under the day mode and the headstock.Headstock under the day mode detects with motion detection and is tracked as Main Basis, take marginal information as supplementary.The vehicle foreground area is followed the tracks of the detection that realizes vehicle, the filtration of noise; Confirm vehicle and to positioning under the headstock, trigger video behind the arrival triggering line along the line by marginal information again, calculate simultaneously car speed.As shown in Figure 4, headstock under the day mode detect with headstock under positioning flow figure along the line, mainly comprise:
Step S41, real-time video flowing
Step S42 utilizes the algorithm of background subtraction to detect sport foreground.At first set up a background model by the self study of initial multi-frame video image, then by the difference of current frame image and background model, obtain the foreground area of moving target.
Step S43, according to the result of motion detection in the S42 background subtraction, selectively update background module, and preservation background model.
Step S44 according to the result of motion detection in the S42 background subtraction, identifies connected region, and tentatively filters the processing of making an uproar according to a series of priori.
Step S45 processes according to the connected domain of S44, and each connected region is considered as a vehicle candidate region, obtains the information such as the position of vehicle candidate region, wide, high, area and perimeter by the connected region sign.
Step S46 obtains the edge of current frame image.
Step S47, edge background difference, its principle is similar to S42, and the source images that difference is to carry out difference is the present frame outline map that S46 obtains.
Step S48, the updating edge background model, principle and S43 are similar, and preserve the edge background model.
Step S49, along the line under the headstock of location.Because vehicle shadow is along with vehicle moves together, the foreground area that step S42 obtains comprises vehicle region and shadow region, and the texture of shade is consistent with the texture of background, and according to this characteristic, associating S45 and S47 obtain under the headstock along the line.
Step S410, by information along the line and vehicle candidate region information under the comparison headstock, filtered noise, and obtain accurately vehicle location.
Step S411 follows the tracks of the testing result of S410.Use space time information, coupling and prediction scheduling algorithm, vehicle is followed the tracks of accurately, obtain the running orbit of Vehicle Object, and preserve the trace information of Vehicle Object.
Step S412 judges whether vehicle arrives the triggering line position, does not arrive in this way, then carries out the detection of next frame, if arrive then execution in step S413 and S414.
Step S413, by depth of field calibration algorithm, center along the line is to world coordinate system under the mapping headstock.
Step S414, the displacement relation by the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again.
Fig. 5 is that car light detects and car light center positioning flow figure under the night mode.Because car light has caused very large impact to motion detection in the night environment, and the car light of vehicle is to very outstanding when the night, therefore detects with car light to substitute motion detection and carry out the detection of vehicle and the measurement of speed.As shown in Figure 5, the car light detection mainly comprises under the night mode:
Step S51, live video stream.
Step S52 utilizes the algorithm of background subtraction to detect sport foreground.At first set up a background model by the self study of initial multi-frame video image, then by the difference of current frame image and background model, obtain the foreground area of moving target.
Step S53, according to the result of motion detection in the S52 background subtraction, selectively update background module, and preservation background model.
Step S54, when having automobile front lamp in the video, the brightness of the image-region at car light place is apparently higher than other zones, as the car light candidate region.
Step S55 according to the result that step S52 and S54 obtain, obtains the prospect of doubtful car light.
Step S56 according to S55 foreground detection result, identifies connected region, and tentatively filters the processing of making an uproar according to a series of priori,
Step S57 processes according to the connected domain of S56, and each connected region is considered as a car light candidate region, obtains the information such as the position of car light candidate region, wide, high, area and perimeter by the connected region sign.
Step S58 according to the symmetry of two car lights of vehicle, by the similarity of the area of two car light candidate regions relatively, judges that whether the car light candidate region is a car light pair.
Step S59 according to the symmetry of two car lights of vehicle, by the position relationship of two car light candidate regions relatively, judges that whether two car light candidate regions are a car light pair.
Step S510 according to the symmetry of two car lights of vehicle, by the class circularity of two car light candidate regions relatively, judges that whether two car light candidate regions are a car light pair.
Step S511 according to the symmetry of two car lights of vehicle, by the ratio of width to height of two car light candidate regions relatively, judges that whether two car light candidate regions are a car light pair.
Step S512 considers S58, S59, and the result of S510 and S511 is mated one by one the car light candidate region and to be returned car light to formation.
Step S513 calculates the right center point coordinate of each car light.
Step S514 follows the tracks of the testing result of S512.Use space time information, coupling and prediction scheduling algorithm, vehicle is followed the tracks of accurately, obtain the running orbit of Vehicle Object, and preserve the trace information of Vehicle Object.
Step S515 judges whether vehicle arrives the triggering line position, does not arrive in this way, then carries out the detection of next frame, if arrive then execution in step S516 and S517.
Step S516, by depth of field calibration algorithm, the mapping car light arrives world coordinate system to the center.
Step S517, the displacement relation by the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again.
Claims (4)
1. license plate identification and Vehicle Velocity Measurement Method based on a video is characterized in that, comprise the steps:
(1) live video stream obtains and transmits: obtain the real-time video flow data by the high definition camera that is installed on portal frame or the L bar, then by Internet Transmission to industrial computer;
(2) detect along the line under car plate identification, the headstock, the car light center line detecting: detect the video data identification license plate number that obtains, detect along the line under the headstock or car light center line;
(3) vehicle triggers: follow the tracks of along the line under car plate, the headstock or car light center line, when along the line under car plate, the headstock or car light center arrive the triggering line that sets in advance, send trigger pip, and by network delivery to high definition camera and LED light compensating lamp;
(4) triggering picture preserves: when the LED light compensating lamp obtains trigger pip, become quick-fried sudden strain of a muscle by stroboscopic, then with this frame picture by Internet Transmission to industrial computer and preserve;
(5) vehicle speed measurement: after vehicle triggers, calculate car speed according to different triggering modes;
(6) night on daytime, detecting pattern switched: along the line under the detection headstock under the monitoring pattern by day, detect the car light center line at night under the monitoring pattern; Relatively based on the vehicle toggle rate under the vehicle toggle rate of car plate identification and daytime, the night monitoring pattern, if the vehicle toggle rate under the daytime monitoring pattern can't meet the demands, then automatically switch to the detecting pattern at night; If the vehicle toggle rate under the night monitoring pattern can't meet the demands, then automatically switch to the detecting pattern on daytime.
2. according to claim 1 described license plate identification and Vehicle Velocity Measurement Method based on video is characterized in that, described step (5) is specially:
(A) licence plate triggers: by depth of field calibration algorithm, mapping licence plate center obtains instantaneous velocity to the displacement relation of world coordinate system according to the frame of being separated by, and exports accurately car speed information according to statistical algorithms again;
(B) trigger along the line under the headstock: by depth of field calibration algorithm, center along the line is to world coordinate system under the mapping headstock, and the displacement relation according to the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again;
(C) the car light center line triggers: by depth of field calibration algorithm, mapping car light center line center is to world coordinate system, and the displacement relation according to the frame of being separated by obtains instantaneous velocity, exports accurately car speed information according to statistical algorithms again.
3. according to claim 1 described license plate identification and Vehicle Velocity Measurement Method based on video is characterized in that, in the described step (6), described daytime, detecting pattern comprised following substep:
(a) obtain real-time video flowing;
(b) utilize the algorithm of background subtraction to detect sport foreground: at first to set up a background model by the self study of initial multi-frame video image, then by the difference of current frame image and background model, obtain the foreground area of moving target;
(c) according to the result of motion detection in step (b) background subtraction, update background module selectively, and preserve background model;
(d) according to the result of motion detection in step (b) background subtraction, identify connected region, and tentatively filter the processing of making an uproar according to a series of priori;
(e) process according to the connected domain of step (d), each connected region is considered as a vehicle candidate region, obtains the position of vehicle candidate region, wide, high, area and perimeter information by the connected region sign;
(f) obtain the outline map of current frame image;
(g) edge background difference, its principle is similar to step (b), and the source images that difference is to carry out difference is the outline map of the current frame image that obtains of step (f);
(h) updating edge background model, (c) is similar for principle and step, and preserves the edge background model;
(i) along the line under the headstock of location; Because vehicle shadow is along with vehicle moves together, the foreground area that step (b) obtains comprises vehicle region and shadow region, and the texture of shade is consistent with the texture of background, and according to this characteristic, joint step (e) and step (g) obtain under the headstock along the line;
(j) pass through to compare information along the line and vehicle candidate region information under the headstock, filtered noise, and obtain accurately vehicle location;
(k) testing result of step (j) is followed the tracks of: use space time information, coupling and prediction algorithm, vehicle is followed the tracks of accurately, obtain the running orbit of Vehicle Object, and preserve the trace information of Vehicle Object;
(l) judge whether vehicle arrives the triggering line position, do not arrive in this way, then carry out the detection of next frame, if arrive then execution in step (m) and step (n);
(m) by depth of field calibration algorithm, center along the line is to world coordinate system under the mapping headstock;
(n) by the displacement relation of the frame of being separated by, obtain instantaneous velocity, export accurately car speed information according to statistical algorithms again.
4. according to claim 1 described license plate identification and Vehicle Velocity Measurement Method based on video is characterized in that, in the described step (6), described night, detecting pattern comprised following substep:
(a) obtain live video stream;
(b) utilize the algorithm of background subtraction to detect sport foreground: at first to set up a background model by the self study of initial multi-frame video image, then by the difference of current frame image and background model, obtain the foreground area of moving target;
(c) according to the result of motion detection in step (b) background subtraction, update background module selectively, and preserve background model;
(d) when having automobile front lamp in the video, the brightness of the image-region at car light place is apparently higher than other zones, as the car light candidate region;
(e) result who obtains according to step (b) and step (d) obtains the prospect of doubtful car light;
(f) according to step (e) foreground detection result, identify connected region, and tentatively filter the processing of making an uproar according to a series of priori,
(g) process according to the connected domain of step (f), each connected region is considered as a car light candidate region, obtains the position of car light candidate region, wide, high, area and perimeter information by the connected region sign;
(h) according to the symmetry of two car lights of vehicle, by the similarity of the area of two car light candidate regions relatively, judge that whether the car light candidate region is a car light pair;
(i) according to the symmetry of two car lights of vehicle, by the position relationship of two car light candidate regions relatively, judge that whether two car light candidate regions are a car light pair;
(j) according to the symmetry of two car lights of vehicle, by the class circularity of two car light candidate regions relatively, judge that whether two car light candidate regions are a car light pair;
(k) according to the symmetry of two car lights of vehicle, by the ratio of width to height of two car light candidate regions relatively, judge that whether two car light candidate regions are a car light pair;
(l) according to step (h), (i), (j) and result (k), the car light candidate region mated one by one return car light to formation;
(m) calculate the right center point coordinate of each car light;
(n) testing result of step (l) is followed the tracks of: use space time information, coupling and prediction algorithm, vehicle is followed the tracks of accurately, obtain the running orbit of Vehicle Object, and preserve the trace information of Vehicle Object;
(o) judge whether vehicle arrives the triggering line position, do not arrive in this way, then carry out the detection of next frame, if arrive then execution in step (p) and (q);
(p) by depth of field calibration algorithm, the mapping car light arrives world coordinate system to the center;
(q) by the displacement relation of the frame of being separated by, obtain instantaneous velocity, export accurately car speed information according to statistical algorithms again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010285620 CN102074113B (en) | 2010-09-17 | 2010-09-17 | License tag recognizing and vehicle speed measuring method based on videos |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010285620 CN102074113B (en) | 2010-09-17 | 2010-09-17 | License tag recognizing and vehicle speed measuring method based on videos |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102074113A CN102074113A (en) | 2011-05-25 |
| CN102074113B true CN102074113B (en) | 2013-04-10 |
Family
ID=44032637
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010285620 Active CN102074113B (en) | 2010-09-17 | 2010-09-17 | License tag recognizing and vehicle speed measuring method based on videos |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102074113B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102509457B (en) * | 2011-10-09 | 2014-03-26 | 青岛海信网络科技股份有限公司 | Vehicle tracking method and device |
| CN102592454A (en) * | 2012-02-29 | 2012-07-18 | 北京航空航天大学 | Intersection vehicle movement parameter measuring method based on detection of vehicle side face and road intersection line |
| CN102610104B (en) * | 2012-03-16 | 2014-11-05 | 江苏苏科畅联科技有限公司 | Onboard front vehicle detection method |
| CN102867416B (en) * | 2012-09-13 | 2014-08-06 | 中国科学院自动化研究所 | Vehicle part feature-based vehicle detection and tracking method |
| CN103679688B (en) * | 2012-09-18 | 2016-11-09 | 浙江大华技术股份有限公司 | A kind of method and device extracting target image |
| CN103197090B (en) * | 2013-03-07 | 2015-03-04 | 上海电科智能系统股份有限公司 | Video speed measuring method based on feature point change |
| CN104063878B (en) * | 2013-03-20 | 2017-08-08 | 富士通株式会社 | Moving Objects detection means, Moving Objects detection method and electronic equipment |
| CN104240512A (en) * | 2013-06-09 | 2014-12-24 | 北京智通视讯科技发展有限公司 | Control technology achieving plate number video detecting and flash shooting for vehicle recognition system |
| US10015394B2 (en) | 2015-10-06 | 2018-07-03 | Genetec Inc. | Camera-based speed estimation and system calibration therefor |
| CN106331656B (en) * | 2016-10-26 | 2023-08-15 | 桂林理工大学 | License plate-assisted visible light communication car lamp positioning and tracking device |
| CN109916485B (en) * | 2017-12-13 | 2021-09-17 | 北京万集科技股份有限公司 | Dynamic vehicle weighing method and device |
| CN108154687B (en) * | 2017-12-28 | 2020-11-10 | 江苏航天大为科技股份有限公司 | Night traffic flow detection method, device and system based on car lamp |
| CN109979206B (en) * | 2017-12-28 | 2020-11-03 | 杭州海康威视系统技术有限公司 | Vehicle speed measuring method, device and system, electronic equipment and storage medium |
| CN108280412A (en) * | 2018-01-12 | 2018-07-13 | 西安电子科技大学 | High Resolution SAR image object detection method based on structure changes CNN |
| CN108621924A (en) * | 2018-02-08 | 2018-10-09 | 常州星宇车灯股份有限公司 | A kind of control system for rear light and its control method with prompt facility |
| CN110276789B (en) * | 2018-03-15 | 2021-10-29 | 杭州海康威视系统技术有限公司 | Target tracking method and device |
| CN110444026B (en) * | 2019-08-06 | 2021-07-09 | 北京万集科技股份有限公司 | Triggering snapshot method and system for vehicle |
| CN110824188B (en) * | 2019-10-17 | 2022-10-04 | 浙江大华技术股份有限公司 | Speed measuring method and device for highway vehicles, coder-decoder and storage device |
| CN111198561B (en) * | 2019-12-05 | 2021-10-22 | 浙江大华技术股份有限公司 | Motion control method and device for target tracking, computer equipment and storage medium |
| CN113990101B (en) * | 2021-11-19 | 2023-04-07 | 深圳市捷顺科技实业股份有限公司 | Method, system and processing device for detecting vehicles in no-parking area |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002074368A (en) * | 2000-08-25 | 2002-03-15 | Matsushita Electric Ind Co Ltd | Moving object recognizing and tracking device |
| JP2006235921A (en) * | 2005-02-24 | 2006-09-07 | Hitachi Ltd | Vehicle number recognition apparatus, and method for adjusting electric lens of vehicle number recognition apparatus |
| CN101533557A (en) * | 2009-03-02 | 2009-09-16 | 上海高德威智能交通系统有限公司 | Traffic information detection method and traffic information detection device |
| CN101807346A (en) * | 2009-02-16 | 2010-08-18 | 汉王科技股份有限公司 | Automatic license plate identification system at urban checkpoint |
-
2010
- 2010-09-17 CN CN 201010285620 patent/CN102074113B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002074368A (en) * | 2000-08-25 | 2002-03-15 | Matsushita Electric Ind Co Ltd | Moving object recognizing and tracking device |
| JP2006235921A (en) * | 2005-02-24 | 2006-09-07 | Hitachi Ltd | Vehicle number recognition apparatus, and method for adjusting electric lens of vehicle number recognition apparatus |
| CN101807346A (en) * | 2009-02-16 | 2010-08-18 | 汉王科技股份有限公司 | Automatic license plate identification system at urban checkpoint |
| CN101533557A (en) * | 2009-03-02 | 2009-09-16 | 上海高德威智能交通系统有限公司 | Traffic information detection method and traffic information detection device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102074113A (en) | 2011-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102074113B (en) | License tag recognizing and vehicle speed measuring method based on videos | |
| US20160034778A1 (en) | Method for detecting traffic violation | |
| CN101656023B (en) | Management method of indoor car park in video monitor mode | |
| Buch et al. | Detection and classification of vehicles for urban traffic scenes | |
| KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN102867417B (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| CN102496281B (en) | Vehicle red-light violation detection method based on combination of tracking and virtual loop | |
| CN103871079A (en) | Vehicle tracking method based on machine learning and optical flow | |
| CN1725266A (en) | Vehicle intelligent monitoring recording syste mand method based on video trigger and speed measuring | |
| CN105611244A (en) | Method for detecting airport foreign object debris based on monitoring video of dome camera | |
| CN102819764A (en) | Method for counting pedestrian flow from multiple views under complex scene of traffic junction | |
| CN103051887A (en) | Eagle eye-imitated intelligent visual sensing node and work method thereof | |
| CN102915433A (en) | Character combination-based license plate positioning and identifying method | |
| CN102136196A (en) | Vehicle velocity measurement method based on image characteristics | |
| CN111354221A (en) | Intelligent road parking equipment and system based on millimeter wave radar detection | |
| CN104573811A (en) | Pedestrian flow counting method based on infrared image and color image fusion | |
| CN106296677A (en) | A kind of remnant object detection method of double mask context updates based on double-background model | |
| CN209225052U (en) | A kind of pilotless automobile drive-control system with variable headlamp | |
| CN110334666A (en) | A kind of intersection divided lane detailed traffic parameter collecting method based on twin camera | |
| CN211928753U (en) | Intelligent road parking equipment and system based on millimeter wave radar detection | |
| CN114781479A (en) | Traffic incident detection method and device | |
| CN110516600A (en) | A kind of bus passenger flow detection method based on Face datection | |
| Huang et al. | A real-time and color-based computer vision for traffic monitoring system | |
| Taha et al. | Day/night detector for vehicle tracking in traffic monitoring systems | |
| Nguyen et al. | An adaptive vision-based outdoor car parking lot 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 | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |