CN1581212A - Method for plucking license-plate region from vehicle image - Google Patents
Method for plucking license-plate region from vehicle image Download PDFInfo
- Publication number
- CN1581212A CN1581212A CN 200410044488 CN200410044488A CN1581212A CN 1581212 A CN1581212 A CN 1581212A CN 200410044488 CN200410044488 CN 200410044488 CN 200410044488 A CN200410044488 A CN 200410044488A CN 1581212 A CN1581212 A CN 1581212A
- Authority
- CN
- China
- Prior art keywords
- value
- image
- gradient
- license plate
- binaryzation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Character Input (AREA)
Abstract
The present invention relates to a method for picking license plate region from car image. Said method mainly includes the following steps: calculating horizontal gradient and vertical gradient of car image to produce horizontal gradient image and vertical gradient image; binarizating the horizontal gradient image and vertical gradient image to obtain horizontal gradient binarizated image and verticalgradient binarizated image so as to make the numerical value of every pixel in image only be high value or low value; then combining horizontal gradient binarizated image and vertical gradient binarizated image into integrated gradient binarizated image; collecting adjacent high value pixel group in the integrated gradient binarizated image into license plate candidate region, then defining every license plate candidate region to produce defined license plate region image.
Description
Technical field
The present invention relates to a kind of from automobile image the method for picking license plate area, particularly can be applicable to the acquisition of car plate automatic identification system and traffic monitoring, vehicle gate inhibition's license plate area image.
Background technology
The location of the license plate area in the automobile image and acquisition are the pre-process programs of the automatic identification of car plate, so its accuracy influences the overall performance of the automatic identification of car plate; In addition, and even in vehicle entrance guard management on the traffic monitoring, managerial personnel need see through the number-plate number that video camera and display are watched car, if can be automatically with the image capture of license plate area and show, can make things convenient for managerial personnel to watch.Therefore, how automatic and effective place, position, the acquisition of finding out car plate from automobile image come out, and is the important topic that car plate automatic identification system and traffic monitoring, vehicle gate inhibition etc. use.
Prior art is for the location and the acquisition of license plate area in the automobile image, generally be to use pixel shading value (Intensity) to come calculation process and the possible case of the operation result of all pixels is drawn histogram (Histogram), from histogram, calculate the desire critical value that the operation result binaryzation of all pixels is required (Threshold) again, so be easy to follow the accuracy that can't find suitable critical value so that binaryzation poor effect to influence license plate area location and acquisition because of the histogram no mark that rises and falls up and down; In addition, because the many distribution scenario of possible license plate area not being analyzed its high value pixel of prior art, so that the chance of the license plate area that locates errors is bigger.
Summary of the invention
The object of the present invention is to provide a kind of from automobile image the method for picking license plate area, can be applicable to automatic identification of car plate and traffic monitoring, vehicle gate inhibition.
The present invention mainly is to use the shading value (Intensity) of pixel or logarithm shading value (Logarithmic intensity) to come calculation process and the possible case of the operation result of all pixels is drawn stacked bar graph (Cumulative histogram), and calculates the critical value that the operation result binaryzation of all pixels is required (Threshold) from stacked bar graph; Because stacked bar graph system is and increases progressively propradation and do not have the general histogram governed phenomenon of no mark of rising and falling up and down, more easily calculate the desire critical value that the operation result binaryzation of all pixels is required, the present invention obtains stacked bar graph with approximatioss and increases progressively rise turnover or the most tangible place of easing up as the critical value of binaryzation, can obtain preferable binaryzation result.
In addition, the present invention is cut into several block of cells respectively to each license plate candidate area and is analyzed the mean number and the standard deviation (Standarddeviation) of the high value pixel that all block of cells comprise, the too high person of the block of license plate candidate area high value number standard deviation can be disallowable and do not regard as license plate area.
Provided by the present invention from automobile image the method for picking license plate area, compared with prior art, have the following advantages:
1. obtain stacked bar graph with approximatioss and increase progressively rise turnover or the most tangible place of easing up, can obtain preferable image binaryzation result as the critical value of binaryzation.
2. license plate candidate area is analyzed the distribution scenario of its high value pixel, the chance that obtains wrong license plate area is less.
Description of drawings
Below in conjunction with preferred embodiment of the present invention and accompanying drawing thereof, further specify technology contents of the present invention and purpose effect thereof.
Fig. 1 is the embodiment process flow diagram of the method for the present invention's picking license plate area from automobile image;
Fig. 2 draws the embodiment synoptic diagram of the binaryzation critical value of horizontal gradient value (or VG (vertical gradient) value) from the stacked bar graph of horizontal gradient value (or VG (vertical gradient) value) for the present invention;
Fig. 3 is cut into several block of cells for the present invention respectively to each license plate candidate area and analyzes the mean number of the high value pixel that all block of cells comprise and the embodiment synoptic diagram of standard deviation;
Fig. 4 comes the embodiment process flow diagram of the method for picking license plate area for the present invention from the horizontal gradient of automobile image;
Fig. 5 comes the embodiment process flow diagram of the method for picking license plate area for the present invention from the VG (vertical gradient) of automobile image.
Embodiment
Embodiment one:
As shown in Figure 1, of the present invention from automobile image the method for picking license plate area, mainly may further comprise the steps:
(1) the calculated level gradient 1: the horizontal gradient (Horizontalgradient) that calculates automobile image produces the horizontal gradient image; The mode of its calculated level gradient be the shading value of adjacent pixels is subtracted each other or shading value taken the logarithm after become the logarithm shading value and subtract each other again.
(2) the horizontal gradient binaryzation 2: then the horizontal gradient image is given two-value and change into horizontal gradient binaryzation image and make that the numerical value of each pixel only is high value (High value) or low value (Low value) in the image, way as shown in Figure 2, the occurrence number of all possible horizontal gradient value that is calculated level gradient image is to draw the stacked bar graph of horizontal gradient value, the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all horizontal gradient values, the longitudinal axis is the accumulation occurrence number that each horizontal gradient value is less than or equal to this horizontal gradient value, between floor level Grad and highest level Grad, look for a horizontal gradient value, feasible occurrence number with the floor level Grad, the accumulation occurrence number that is less than or equal to this horizontal gradient value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum horizontal Grad constituted and the gap minimum (promptly approaching most stacked bar graph) of horizontal gradient value stacked bar graph, the binaryzation critical value that this horizontal gradient value is decided to be the horizontal gradient value, and will be decided to be high value greater than the horizontal gradient value of critical value, the horizontal gradient value that is less than or equal to critical value is decided to be low value.
(3) calculate VG (vertical gradient) 3: the VG (vertical gradient) (Vertical gradient) of calculating automobile image produces the VG (vertical gradient) image; Its mode of calculating VG (vertical gradient) be the shading value of adjacent pixels is subtracted each other or shading value taken the logarithm after become the logarithm shading value and subtract each other again.
(4) the VG (vertical gradient) binaryzation 4: and the VG (vertical gradient) image is given two-value change into VG (vertical gradient) binaryzation image, make that the numerical value of each pixel only is high value or low value in the image; Way is to calculate the occurrence number of all possible VG (vertical gradient) value of VG (vertical gradient) image, to draw the stacked bar graph of VG (vertical gradient) value, the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all VG (vertical gradient) values, the longitudinal axis is the accumulation occurrence number that each VG (vertical gradient) value is less than or equal to this VG (vertical gradient) value, between minimum VG (vertical gradient) value and the highest VG (vertical gradient) value, look for a VG (vertical gradient) value, feasible occurrence number with minimum VG (vertical gradient) value, the accumulation occurrence number that is less than or equal to this VG (vertical gradient) value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum perpendicular Grad constituted and the gap minimum (promptly approaching most stacked bar graph) of VG (vertical gradient) value stacked bar graph, the binaryzation critical value that this VG (vertical gradient) value is decided to be the VG (vertical gradient) value, and will be decided to be high value greater than the VG (vertical gradient) value of critical value, the VG (vertical gradient) value that is less than or equal to critical value is decided to be low value.
(5) merge horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image 5: horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image are merged into integration gradient binaryzation image, way is for to do " OR " computing with horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image, as long as two pixels that are same coordinate position in horizontal gradient binaryzation image and the VG (vertical gradient) binaryzation image have one to be to present high value, the pixel of then integrating same coordinate position in the gradient binaryzation image also presents high value; If two pixels of same coordinate position all present low value in horizontal gradient binaryzation image and the VG (vertical gradient) binaryzation image, the pixel of then integrating same coordinate position in the gradient binaryzation image also presents low value.
The high value bunching of picture element that (6) will be close to becomes license plate candidate area 6: will integrate high value bunching of picture element contiguous in the gradient binaryzation image and become license plate candidate area (Candidate region): way is for integrating the distance between the high value pixel in the gradient binaryzation image, if less than a particular value, then these high value pixels are indicated identical label, the region with all pixels of same volume target is decided to be license plate candidate area again.
(7) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area 7: each license plate candidate area is determined whether to be license plate area according to the distribution scenario of high value pixel in its zone: as shown in Figure 3, each license plate candidate area is cut into several block of cells respectively and analyzes the mean number and the standard deviation of the high value pixel that all block of cells comprise, the high value of the block of license plate candidate area number standard deviation is too high, can be disallowable and do not regard as license plate area, and the license plate area image of output after determining.
Embodiment two:
As shown in Figure 4, of the present invention from automobile image the method for picking license plate area, can be only come the binaryzation image at the horizontal gradient of automobile image, mainly may further comprise the steps:
(1) the calculated level gradient 1: the horizontal gradient (Horizontalgradient) that calculates automobile image produces the horizontal gradient image; The mode of its calculated level gradient be the shading value of adjacent pixels is subtracted each other or shading value taken the logarithm after become the logarithm shading value and subtract each other again.
(2) the horizontal gradient binaryzation 2: then the horizontal gradient image is given two-value and change into horizontal gradient binaryzation image and make that the numerical value of each pixel only is high value (High value) or low value (Low value) in the image, way is that the occurrence number of all possible horizontal gradient value of calculated level gradient image is to draw the stacked bar graph of horizontal gradient value, the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all horizontal gradient values, the longitudinal axis is the accumulation occurrence number that each horizontal gradient value is less than or equal to this horizontal gradient value, between floor level Grad and highest level Grad, look for a horizontal gradient value, feasible occurrence number with the floor level Grad, the accumulation occurrence number that is less than or equal to this horizontal gradient value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum horizontal Grad constituted and the gap minimum (promptly approaching most stacked bar graph) of horizontal gradient value stacked bar graph, the binaryzation critical value that this horizontal gradient value is decided to be the horizontal gradient value, and will be decided to be high value greater than the horizontal gradient value of critical value, the horizontal gradient value that is less than or equal to critical value is decided to be low value.
The high value bunching of picture element that (3) will be close to becomes license plate candidate area 6: will integrate high value bunching of picture element contiguous in the gradient binaryzation image and become license plate candidate area (Candidate region); If way less than particular value, then indicates identical label with these high value pixels for integrating the distance between the high value pixel in the gradient binaryzation image, and the region with all pixels of same volume target is decided to be license plate candidate area again.
(4) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area 7: each license plate candidate area is determined whether to be license plate area according to the distribution scenario of high value pixel in its zone: as shown in Figure 3, each license plate candidate area is cut into several block of cells respectively and analyzes the mean number and the standard deviation of the high value pixel that all block of cells comprise, the high value of the block of license plate candidate area number standard deviation is too high, can be disallowable and do not regard as license plate area, and the license plate area image of output after determining.
Embodiment three:
As shown in Figure 5, of the present invention from automobile image the method for picking license plate area, can only come the binaryzation image at the VG (vertical gradient) of automobile image, mainly comprise following steps:
(1) calculate VG (vertical gradient) 3: the VG (vertical gradient) (Vertical gradient) of calculating automobile image produces the VG (vertical gradient) image; Its mode of calculating VG (vertical gradient) be the shading value of adjacent pixels is subtracted each other or shading value taken the logarithm after become the logarithm shading value and subtract each other again.
(2) the VG (vertical gradient) binaryzation 4: the VG (vertical gradient) image is given two-value change into VG (vertical gradient) binaryzation image and make that the numerical value of each pixel only is high value or low value in the image; Way is to calculate the occurrence number of all possible VG (vertical gradient) value of VG (vertical gradient) image to draw the stacked bar graph of VG (vertical gradient) value, the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all VG (vertical gradient) values, the longitudinal axis is the accumulation occurrence number that each VG (vertical gradient) value is less than or equal to this VG (vertical gradient) value, between minimum VG (vertical gradient) value and the highest VG (vertical gradient) value, look for a VG (vertical gradient) value, feasible occurrence number with minimum VG (vertical gradient) value, the accumulation occurrence number that is less than or equal to this VG (vertical gradient) value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum perpendicular Grad constituted and the gap minimum (promptly approaching most stacked bar graph) of VG (vertical gradient) value stacked bar graph, the binaryzation critical value that this VG (vertical gradient) value is decided to be the VG (vertical gradient) value, and will be decided to be high value greater than the VG (vertical gradient) value of critical value, the VG (vertical gradient) value that is less than or equal to critical value is decided to be low value.
The high value bunching of picture element that (3) will be close to becomes license plate candidate area 6: will integrate high value bunching of picture element contiguous in the gradient binaryzation image and become license plate candidate area (Candidate region); If way then indicates identical label with these high value pixels for integrating distance between the high value pixel in the gradient binaryzation image less than particular value, the region with all pixels of same volume target is decided to be license plate candidate area again.
(4) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area 7: each license plate candidate area is determined whether to be license plate area according to the distribution scenario of high value pixel in its zone: as shown in Figure 3, each license plate candidate area is cut into several block of cells respectively and analyzes the mean number and the standard deviation of the high value pixel that all block of cells comprise, the high value of the block of license plate candidate area number standard deviation is too high, can be disallowable and do not regard as license plate area, and the license plate area image of output after determining.
Claims (12)
1. the method for a picking license plate area from automobile image is characterized in that, mainly comprises following steps:
(1) calculated level gradient: the horizontal gradient that calculates automobile image produces the horizontal gradient image;
(2) horizontal gradient binaryzation: then the horizontal gradient image is given two-value changes into horizontal gradient binaryzation image, makes that the numerical value of each pixel only is high value or low value in the image;
(3) calculate VG (vertical gradient): the VG (vertical gradient) of calculating automobile image produces the VG (vertical gradient) image;
(4) VG (vertical gradient) binaryzation: and the VG (vertical gradient) image is given two-value change into VG (vertical gradient) binaryzation image, make that the numerical value of each pixel only is high value or low value in the image;
(5) merge horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image: horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image are merged into integration gradient binaryzation image;
The high value bunching of picture element that (6) will be close to becomes license plate candidate area: will integrate high value bunching of picture element contiguous in the gradient binaryzation image then and become license plate candidate area;
(7) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area: at last each license plate candidate area is determined whether to be license plate area according to the distribution scenario of high value pixel in its zone, and the license plate area image of output after determining.
2. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: the mode of calculating horizontal gradient in the described step (1) is that the shading value of adjacent pixels is subtracted each other.
3. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: the mode of calculating horizontal gradient in the described step (1) is to become the logarithm shading value after the shading value of adjacent pixels is taken the logarithm to subtract each other again.
4. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: the occurrence number of all possible horizontal gradient value of calculating horizontal gradient image is to draw the stacked bar graph of horizontal gradient value in the described step (2), the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all horizontal gradient values, the longitudinal axis is the accumulation occurrence number that each horizontal gradient value is less than or equal to this horizontal gradient value, between floor level Grad and highest level Grad, look for a horizontal gradient value, feasible occurrence number with the floor level Grad, the accumulation occurrence number that is less than or equal to this horizontal gradient value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum horizontal Grad constituted and the gap minimum of horizontal gradient value stacked bar graph, promptly approach most stacked bar graph, this horizontal gradient value is decided to be the binaryzation critical value of horizontal gradient value and will be decided to be high value greater than the horizontal gradient value of critical value, and the horizontal gradient value that is less than or equal to critical value is decided to be low value.
By claim 1 described from automobile image the method for picking license plate area, it is characterized in that: the mode of calculating horizontal gradient in the described step (3) is that the shading value of adjacent pixels is subtracted each other.
6. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: the mode of calculating VG (vertical gradient) in the described step (3) is to become the logarithm shading value after the shading value of neighbor is taken the logarithm to subtract each other again.
7. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: the occurrence number of all possible VG (vertical gradient) value of calculating VG (vertical gradient) image is to draw the stacked bar graph of VG (vertical gradient) value in the described step (4), the transverse axis of stacked bar graph is the ascending arrangements in regular turn of all VG (vertical gradient) values, the longitudinal axis is the accumulation occurrence number that each VG (vertical gradient) value is less than or equal to this VG (vertical gradient) value, between minimum VG (vertical gradient) value and the highest VG (vertical gradient) value, look for a VG (vertical gradient) value, feasible occurrence number with minimum VG (vertical gradient) value, the accumulation occurrence number that is less than or equal to this VG (vertical gradient) value, be less than or equal to the linear fold lines that the accumulation occurrence number three of maximum perpendicular Grad constituted and the gap minimum of VG (vertical gradient) value stacked bar graph, promptly approach most stacked bar graph, this VG (vertical gradient) value is decided to be the binaryzation critical value of VG (vertical gradient) value and will be decided to be high value greater than the VG (vertical gradient) value of critical value, and the VG (vertical gradient) value that is less than or equal to critical value is decided to be low value.
8. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: in the described step (5) horizontal gradient binaryzation image and VG (vertical gradient) binaryzation image are done " OR " computing, as long as two pixels that are same coordinate position in horizontal gradient binaryzation image and the VG (vertical gradient) binaryzation image have one to be to present high value, the pixel of then integrating same coordinate position in the gradient binaryzation image also presents high value; If two pixels of same coordinate position all present low value in horizontal gradient binaryzation image and the VG (vertical gradient) binaryzation image, the pixel of then integrating same coordinate position in the gradient binaryzation image also presents low value.
9. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: integrate the distance between the high value pixel in the gradient binaryzation image in the described step (6), if less than a particular value, then these high value pixels are indicated identical label, the region with all pixels of same volume target is decided to be license plate candidate area again.
10. according to claim 1 from automobile image the method for picking license plate area, it is characterized in that: in the described step (7) each license plate candidate area is cut into several block of cells respectively and calculates the mean number and the standard deviation of the high value pixel that all block of cells comprise, the high value of the block of license plate candidate area number standard deviation is too high, disallowable and do not regard as license plate area, and the license plate area image of output after determining.
11. the method for a picking license plate area from automobile image is characterized in that, mainly comprises following steps:
(1) calculated level gradient: the horizontal gradient that calculates automobile image produces the horizontal gradient image;
(2) horizontal gradient binaryzation: then the horizontal gradient image is given two-value and change into horizontal gradient binaryzation image and make that the numerical value of each pixel only is high value or low value in the image;
The high value bunching of picture element that (3) will be close to becomes license plate candidate area: then high value bunching of picture element contiguous in the horizontal gradient binaryzation image is become license plate candidate area;
(4) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area: at last each license plate candidate area is determined whether to be license plate area and the output license plate area image after determining according to the distribution scenario of high value pixel in its zone.
12. the method for a picking license plate area from automobile image is characterized in that, mainly comprises following steps:
(1) calculate VG (vertical gradient): the VG (vertical gradient) of calculating automobile image produces the VG (vertical gradient) image;
(2) VG (vertical gradient) binaryzation: then the VG (vertical gradient) image is given two-value and change into VG (vertical gradient) binaryzation image and make that the numerical value of each pixel only is high value or low value in the image;
The high value bunching of picture element that (3) will be close to becomes license plate candidate area: then high value bunching of picture element contiguous in the VG (vertical gradient) binaryzation image is become license plate candidate area;
(4) determine with the distribution scenario of high value pixel whether license plate candidate area is license plate area: at last each license plate candidate area is determined whether to be license plate area and the output license plate area image after determining according to the distribution scenario of high value pixel in its zone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100444884A CN100349175C (en) | 2004-05-14 | 2004-05-14 | Method for plucking license-plate region from vehicle image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100444884A CN100349175C (en) | 2004-05-14 | 2004-05-14 | Method for plucking license-plate region from vehicle image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1581212A true CN1581212A (en) | 2005-02-16 |
| CN100349175C CN100349175C (en) | 2007-11-14 |
Family
ID=34582202
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004100444884A Expired - Fee Related CN100349175C (en) | 2004-05-14 | 2004-05-14 | Method for plucking license-plate region from vehicle image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100349175C (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017091060A1 (en) * | 2015-11-27 | 2017-06-01 | Mimos Berhad | A system and method for detecting objects from image |
| CN106845484A (en) * | 2017-02-28 | 2017-06-13 | 浙江华睿科技有限公司 | A kind of localization method and device in one-dimension code region |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW544643B (en) * | 2001-05-08 | 2003-08-01 | Chunghwa Telecom Co Ltd | Method for cropping area of license plate from vehicle image |
| FR2825173B1 (en) * | 2001-05-23 | 2003-10-31 | France Telecom | METHOD FOR DETECTING TEXT AREAS IN A VIDEO IMAGE |
| CN1198235C (en) * | 2002-03-28 | 2005-04-20 | 中华电信股份有限公司 | Method and device for picking license plate area and correcting skew license plate from automobile image |
-
2004
- 2004-05-14 CN CNB2004100444884A patent/CN100349175C/en not_active Expired - Fee Related
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017091060A1 (en) * | 2015-11-27 | 2017-06-01 | Mimos Berhad | A system and method for detecting objects from image |
| CN106845484A (en) * | 2017-02-28 | 2017-06-13 | 浙江华睿科技有限公司 | A kind of localization method and device in one-dimension code region |
| CN106845484B (en) * | 2017-02-28 | 2019-09-17 | 浙江华睿科技有限公司 | A kind of localization method and device in one-dimension code region |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100349175C (en) | 2007-11-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8902053B2 (en) | Method and system for lane departure warning | |
| EP2919189B1 (en) | Pedestrian tracking and counting method and device for near-front top-view monitoring video | |
| CN106373426B (en) | Parking stall based on computer vision and violation road occupation for parking monitoring method | |
| US7787656B2 (en) | Method for counting people passing through a gate | |
| US8290265B2 (en) | Method and apparatus for segmenting an object region of interest from an image | |
| US8019122B2 (en) | Vehicle side image recognition method and apparatus | |
| US20070127778A1 (en) | Object detecting system and object detecting method | |
| KR20150102546A (en) | Apparatus and method for recognizing lane | |
| US10726277B2 (en) | Lane line detection method | |
| KR20100034409A (en) | Method for recognizing lane and lane departure with single lane extraction | |
| CN110335467B (en) | Method for realizing highway vehicle behavior detection by using computer vision | |
| Janda et al. | Road boundary detection for run-off road prevention based on the fusion of video and radar | |
| KR101264282B1 (en) | detection method vehicle in road using Region of Interest | |
| CN114092903A (en) | Lane line marking method, lane line detection model determining method, lane line detection method and related equipment | |
| CN111523385B (en) | Stationary vehicle detection method and system based on frame difference method | |
| KR101236223B1 (en) | Method for detecting traffic lane | |
| Albiol et al. | Video-based traffic queue length estimation | |
| Zong et al. | Traffic light detection based on multi-feature segmentation and online selecting scheme | |
| CN102610103A (en) | Intelligent traffic line analysis and recognition method | |
| CN1581212A (en) | Method for plucking license-plate region from vehicle image | |
| JP2861638B2 (en) | Travel lane detection device | |
| Li et al. | A fog level detection method based on grayscale features | |
| Wang et al. | Detection of lane lines on both sides of road based on monocular camera | |
| CN109886120B (en) | Zebra crossing detection method and system | |
| CN114882709A (en) | Vehicle congestion detection method and device and computer storage medium |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20071114 Termination date: 20190514 |