CN103366578A - Image-based vehicle detection method - Google Patents
Image-based vehicle detection method Download PDFInfo
- Publication number
- CN103366578A CN103366578A CN2013102594349A CN201310259434A CN103366578A CN 103366578 A CN103366578 A CN 103366578A CN 2013102594349 A CN2013102594349 A CN 2013102594349A CN 201310259434 A CN201310259434 A CN 201310259434A CN 103366578 A CN103366578 A CN 103366578A
- Authority
- CN
- China
- Prior art keywords
- image
- vehicle
- surveyed area
- imaging device
- information
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses an image-based vehicle detection method, which comprises the steps: setting a detection area of an imaging device, wherein the detection area comprises a fixation object; collecting an image in the detection area, analyzing the collected image, and storing the image information into the imaging device as prior structural information; shooting the detection area at a high speed with the imaging device; processing the shot image in the detection area, extracting image information, and comparing the image information with the prior structural information, if the image information and the prior structural information are totally or partly matched, no vehicle occurring is judged, and if edge information extraction in the image is failed or the image information and the prior structural information are not matched, vehicle occurring is judged. With the adoption of the technical scheme disclosed by the invention, by changing detection objects, the detection rate of vehicles is obviously improved and the false detection rate is reduced.
Description
Technical field
The present invention relates to the vehicle detection field, particularly a kind of image-based vehicle checking method.
Background technology
The common purpose of vehicle detection is captured vehicle or vehicle is counted, and vehicle detection technology commonly used comprises the various ways such as buried coil detection, infrared detection, radar detection technique, video detection.Adopt video to detect can to avoid destroying the road surface, needn't the additional external checkout equipment, need not correct trigger position, reduce expenses, and be more suitable for the requirement of movable type, portable use.
System carries out video frequency vehicle and detects, and need to possess very high processing speed and adopt outstanding algorithm, realizes image acquisition, processing in the situation of substantially not frame losing.If processing speed is slow, then cause frame losing, make system can't detect faster vehicle of travel speed, be difficult to also guarantee that the position being conducive to identify begins identifying processing simultaneously, affect system recognition rate.The Key Performance Indicator that video frequency vehicle detects be included under the various photoenvironments and the vehicle flowrate condition under the accuracy of vehicle verification and measurement ratio, false detection rate, detection position.The video frequency vehicle detection technique has become focus at intelligent transportation field, but low with respect to its verification and measurement ratio of buried coil, false detection rate is high and the poor accuracy of detection position, so range of application is very restricted.
One of existing technical scheme, dummy line or virtual coil technology are the most frequently used detection meanss during video frequency vehicle detects, the know-why of dummy line and virtual coil is identical, equal limited pixels of the extracted region of vehicle process in image, form a set, add up certain statistical characteristics (such as average gray, gradient distribution etc.) of the interior pixel of set in every two field picture.When the eigenwert gap of eigenwert and background characteristics value or consecutive frame during greater than a certain threshold value, think vehicle to occur in the zone that the pixel set covers.Based on the motion detection of the object detection of background subtraction point-score, time differencing method and optical flow-based all in this category.
The defective of virtual coil technology has following 2 points:
1 because the principle of virtual coil is based on the statistical characteristics variation detection vehicle of pixel set, not merely occurred causing by vehicle but change in the practical application, cause in some situation false detection rate high, the shade of vehicle for example, the factors such as the lights of vehicle at night all can cause error detection.
2 judge whether to exist the threshold value of vehicle to be difficult to determine, this threshold value and photoenvironment, and the reflection coefficient on road surface, camera properties parameter have substantial connection, cause the adaptability of algorithm relatively poor.
Using vehicle characteristics to carry out vehicle detection is another kind of technical scheme, and the vehicle characteristics of often being selected comprises shape, color, the texture of vehicle, the car light of vehicle at night.Use vehicle characteristics to detect the vehicle detection that is known as based on model, common step is before carrying out vehicle detection, set up a vehicle characteristics model, for example night car light model may comprise the gray feature of car light, the shape of car light, the symmetric form of car light.After setting up the vehicle characteristics model, in each two field picture, be partitioned into area-of-interest, and extract eigenwert and characteristic model mates, when eigenwert meets characteristic model, think that the area-of-interest that splits comprises vehicle.Often adopt various features to merge in the actual algorithm, could obtain reasonable robustness.
Vehicle detection based on feature has lower false detection rate with respect to virtual coil, but it is existent defect also: because the diversity of detected object and the uncertainty of the factors such as photoenvironment, shooting angle, the difficulty that causes model to set up increases the bad adaptability of algorithm.
Summary of the invention
The purpose of this invention is to provide a kind of image-based vehicle checking method, can obviously improve the vehicle detection rate and reduce false detection rate.
For achieving the above object, the present invention adopts following technical scheme:
A kind of image-based vehicle checking method comprises the steps:
The surveyed area of imaging device is set, and described surveyed area comprises fixation object;
Gather the image in the described surveyed area, analyze the image that gathers, described image information is stored in imaging device as the priori structural information;
Use described imaging device at a high speed described surveyed area to be taken;
Captured surveyed area image is processed, extracted image information, and compare with described priori structural information, if both mate or the part coupling fully, then judge without vehicle to occur, if the edge extraction in the image is failed or both do not mate, then is determined with vehicle and occurs.
As the preferred embodiments of the present invention, described image information comprises the marginal information of described fixation object.
As the preferred embodiments of the present invention, described method also comprises the steps:
When judging that vehicle occurs, described imaging device is switched to the full resolution imaging pattern take;
Analyze the fixation object that the other place's surveyed area in the captured image comprises and whether be blocked, if be blocked, judge that then the vehicle that passes through is large car, if be not blocked, judge that then the vehicle that passes through is compact car.
As the preferred embodiments of the present invention, also comprise the step that fixation object is set.
As the preferred embodiments of the present invention, by arranging reflecting coating on the road surface or utilize light filling equipment that projection arranges described fixation object on the road surface.
As the preferred embodiments of the present invention, the fixation object that comprises in the described surveyed area is the identifier marking of vehicle deceleration strip or road.
As the preferred embodiments of the present invention, described imaging device is set as the mode of operation of windowing, use described imaging device at a high speed surveyed area to be taken.
As the preferred embodiments of the present invention, the fixation object in the described surveyed area is fixedly installed.
Image-based vehicle checking method provided by the present invention by the change detection target, has obviously improved the verification and measurement ratio of vehicle and has reduced false detection rate, also detects the positional precision that the frame per second raising detects by improving.
Description of drawings
Fig. 1 is the synoptic diagram of a specific embodiment of the present invention;
Fig. 2 is synoptic diagram during without the car state in a specific embodiment of the present invention;
Fig. 3 is synoptic diagram when vehicle occurs in a specific embodiment of the present invention;
The synoptic diagram of Fig. 4 another specific embodiment of the present invention.
Reference numeral: 1, fixation object; 2, imaging device; 3, surveyed area one; 4, vehicle; 5, capture the zone; 6, the surveyed area image; 7, the priori structural information; 8, the structural information of extraction; 9, surveyed area two.
Embodiment
In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, is not intended to limit the present invention.
This image-based vehicle checking method comprises the steps:
The surveyed area of imaging device is set, and this surveyed area comprises fixation object.
Can be by arranging reflecting coating on the road surface or utilize light filling equipment that projection arranges fixation object on the road surface; Fixation object also can be the identifier marking of vehicle deceleration strip or road.
Fixation object in the surveyed area is fixedly installed.
Image in the acquisition testing zone is analyzed the image that gathers, and image information is stored in imaging device as the priori structural information, and this image information comprises the marginal information of this fixation object.
Use imaging device at a high speed described surveyed area to be taken, when judging that vehicle occurs, imaging device is switched to the full resolution imaging pattern take.
Also imaging device can be set as the mode of operation of windowing, use this imaging device at a high speed surveyed area to be taken.
Captured surveyed area image is processed, extracted image information, and compare with described priori structural information, if both mate or the part coupling fully, then judge without vehicle to occur, if the edge extraction in the image is failed or both do not mate, then is determined with vehicle and occurs.
Analyze the fixation object that the other place's surveyed area in the captured image comprises and whether be blocked, if be blocked, judge that then the vehicle that passes through is large car, if be not blocked, judge that then the vehicle that passes through is compact car.
As shown in Figure 1, arrange some fixation objects 1 in the zone that vehicle must pass through, the shape of target comprises abundant structural information, for example triangle or capitalize large English character.The material of fixation object 1 should be reflectorized material and wear-resisting, to improve the visibility at night, for example uses the hot melt method of scoring that extensively adopts at present to lay on the road surface.
Fixedly imaging device 2, and imaging device 2 works in the full resolution pattern, make fixation object 1 in the visual field of video camera, and surveyed area 1 is set comprises fixation object 1.
Image processing algorithm is analyzed the image in the surveyed area 1, extracts the edge of fixation object 1, is kept at imaging device 2 inside as priori structural information 7.Priori structural information 7 can be utilized the existing facility in application scenario, the identifier marking of vehicle deceleration strip or road for example, also can use reflecting coating to arrange on the road surface according to the shape of design, can also utilize light filling equipment (for example laser instrument) to throw on the road surface.Reflecting coating and laser instrument are not limited to use visible-range, also can use infrared light reflecting coating and laser instrument.
Image processing algorithm is processed surveyed area image 6, extract the marginal information in the image, and compare with priori structural information 7, if the structural information 8 of extracting mate fully or part is mated then think that fixation object 1 is not blocked, illustrate current without the vehicle appearance, as shown in Figure 2; If structural information extraction failure or the structural information 8 of extracting are not mated with priori structural information 7 then thought that fixation object is blocked, there is vehicle to occur, as shown in Figure 3.Wherein, fixation object is not limited to be blocked fully, and fixation object shape and production method and detection method are irrelevant, all should be considered as same detection method.
When vehicle occurred, imaging device 2 switched to the full resolution imaging pattern and takes capturing zone 5.
Whether the fixation object of analyzing in the surveyed area 29 in the full resolution image is blocked, judge the length of vehicle, compact car and large car are classified, if be blocked, judge that then the vehicle that passes through is large car, if be not blocked, judge that then the vehicle that passes through is compact car, as shown in Figure 4.Can also use more fixation objects to be used for more vehicle dimension classification.
By the structural information that realizes determining uncertain object detection is converted to definite object detection, the Detection accuracy of structural information can reach more than 99%, can significantly improve under the various photoenvironments verification and measurement ratio and reduce false detection rate.
Owing to use high frame per second to carry out vehicle detection, can obviously improve the positional precision of detection, can capture in place vehicle, can adapt to the vehicle that passes through at a high speed simultaneously.
The relative position of imaging device and vehicle changes the work that does not affect algorithm, so low to the installation requirement of video camera, it is convenient to implement.
Because fixation object is fixed, can not impact detection algorithm so prolong the time shutter of imaging device, after the time shutter of prolongation imaging device, night, the power of light filling equipment can correspondingly reduce, and reduced the purpose that energy consumption reduces light pollution to reach.
The above is preferred embodiment of the present invention only, is not to limit practical range of the present invention; If do not break away from the spirit and scope of the present invention, the present invention is made amendment or is equal to replacement, all should be encompassed in the middle of the protection domain of claim of the present invention.
Claims (8)
1. an image-based vehicle checking method is characterized in that, comprises the steps:
The surveyed area of imaging device is set, and described surveyed area comprises fixation object;
Gather the image in the described surveyed area, analyze the image that gathers, described image information is stored in imaging device as the priori structural information;
Use described imaging device at a high speed described surveyed area to be taken;
Captured surveyed area image is processed, extracted image information, and compare with described priori structural information, if both mate or the part coupling fully, then judge without vehicle to occur, if the edge extraction in the image is failed or both do not mate, then is determined with vehicle and occurs.
2. image-based vehicle checking method according to claim 1 is characterized in that, described image information comprises the marginal information of described fixation object.
3. according to claim 1 or 2 described image-based vehicle checking methods, it is characterized in that described method also comprises the steps:
When judging that vehicle occurs, described imaging device is switched to the full resolution imaging pattern take;
Analyze the fixation object that the other place's surveyed area in the captured image comprises and whether be blocked, if be blocked, judge that then the vehicle that passes through is large car, if be not blocked, judge that then the vehicle that passes through is compact car.
4. according to claim 1 or 2 described image-based vehicle checking methods, it is characterized in that, also comprise the step that fixation object is set.
5. image-based vehicle checking method according to claim 4 is characterized in that, by arranging reflecting coating on the road surface or utilize light filling equipment that projection arranges described fixation object on the road surface.
6. image-based vehicle checking method according to claim 1 and 2 is characterized in that, the fixation object that comprises in the described surveyed area is the identifier marking of vehicle deceleration strip or road.
7. image-based vehicle checking method according to claim 1 and 2 is characterized in that, described imaging device is set as the mode of operation of windowing, and uses described imaging device at a high speed surveyed area to be taken.
8. image-based vehicle checking method according to claim 1 and 2 is characterized in that, the fixation object in the described surveyed area is fixedly installed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013102594349A CN103366578A (en) | 2013-06-27 | 2013-06-27 | Image-based vehicle detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013102594349A CN103366578A (en) | 2013-06-27 | 2013-06-27 | Image-based vehicle detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103366578A true CN103366578A (en) | 2013-10-23 |
Family
ID=49367822
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2013102594349A Pending CN103366578A (en) | 2013-06-27 | 2013-06-27 | Image-based vehicle detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103366578A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103839415A (en) * | 2014-03-19 | 2014-06-04 | 重庆攸亮科技有限公司 | Traffic flow and occupation ratio information acquisition method based on road surface image feature identification |
CN105117693A (en) * | 2015-08-12 | 2015-12-02 | 杜宪利 | Video detection method based on optical identification |
CN105575125A (en) * | 2015-12-15 | 2016-05-11 | 上海微桥电子科技有限公司 | Vehicle flow video detection and analysis system |
CN110896449A (en) * | 2019-11-10 | 2020-03-20 | 张晓东 | Real-time camera resolution adjustment system and method |
CN112602127A (en) * | 2019-03-22 | 2021-04-02 | 深圳市大疆创新科技有限公司 | System and method for lane monitoring and providing lane departure warning |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2003023716A2 (en) * | 2001-09-11 | 2003-03-20 | Pips Technology Limited | Dtermining occupancy of a region monitored by a video apparatus |
EP1744292A2 (en) * | 2005-07-08 | 2007-01-17 | Van de Weijdeven, Everhardus Franciscus | Method for determining data of vehicles |
CN101765527A (en) * | 2007-07-31 | 2010-06-30 | 株式会社丰田自动织机 | Parking assistance device, vehicle-side device for parking assistance device, parking assistance method, and parking assistance program |
CN102610103A (en) * | 2012-01-13 | 2012-07-25 | 深圳市黄河数字技术有限公司 | Intelligent traffic line analysis and recognition method |
CN102682602A (en) * | 2012-05-15 | 2012-09-19 | 华南理工大学 | Collecting method for road traffic parameters based on video technique |
-
2013
- 2013-06-27 CN CN2013102594349A patent/CN103366578A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2003023716A2 (en) * | 2001-09-11 | 2003-03-20 | Pips Technology Limited | Dtermining occupancy of a region monitored by a video apparatus |
EP1744292A2 (en) * | 2005-07-08 | 2007-01-17 | Van de Weijdeven, Everhardus Franciscus | Method for determining data of vehicles |
CN101765527A (en) * | 2007-07-31 | 2010-06-30 | 株式会社丰田自动织机 | Parking assistance device, vehicle-side device for parking assistance device, parking assistance method, and parking assistance program |
CN102610103A (en) * | 2012-01-13 | 2012-07-25 | 深圳市黄河数字技术有限公司 | Intelligent traffic line analysis and recognition method |
CN102682602A (en) * | 2012-05-15 | 2012-09-19 | 华南理工大学 | Collecting method for road traffic parameters based on video technique |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103839415A (en) * | 2014-03-19 | 2014-06-04 | 重庆攸亮科技有限公司 | Traffic flow and occupation ratio information acquisition method based on road surface image feature identification |
CN103839415B (en) * | 2014-03-19 | 2016-08-24 | 重庆攸亮科技股份有限公司 | Traffic flow based on pavement image feature identification and occupation rate information getting method |
CN105117693A (en) * | 2015-08-12 | 2015-12-02 | 杜宪利 | Video detection method based on optical identification |
CN105575125A (en) * | 2015-12-15 | 2016-05-11 | 上海微桥电子科技有限公司 | Vehicle flow video detection and analysis system |
CN105575125B (en) * | 2015-12-15 | 2018-05-29 | 上海微桥电子科技有限公司 | A kind of wagon flow video detecting analysis system |
CN112602127A (en) * | 2019-03-22 | 2021-04-02 | 深圳市大疆创新科技有限公司 | System and method for lane monitoring and providing lane departure warning |
CN110896449A (en) * | 2019-11-10 | 2020-03-20 | 张晓东 | Real-time camera resolution adjustment system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105809679B (en) | Mountain railway side slope rockfall detection method based on visual analysis | |
CN104392224B (en) | A kind of highway pavement crack detecting method | |
CN105373135B (en) | A kind of method and system of aircraft docking guidance and plane type recognition based on machine vision | |
CN100545867C (en) | Aerial shooting traffic video frequency vehicle rapid checking method | |
US8798314B2 (en) | Detection of vehicles in images of a night time scene | |
CN105203552A (en) | 360-degree tread image detecting system and method | |
CN109489724B (en) | Tunnel train safe operation environment comprehensive detection device and detection method | |
CN109871728B (en) | Vehicle type recognition method and device | |
CN109242035B (en) | Vehicle bottom fault detection device and method | |
CN103366578A (en) | Image-based vehicle detection method | |
CN103983426B (en) | The detection of a kind of defect of optical fiber based on machine vision and sorting technique | |
JP6139088B2 (en) | Vehicle detection device | |
CN102445453B (en) | Automatic detection device and identification method for integrality of guardrail of high-speed railway line | |
CN108197523A (en) | Vehicle detection at night method and system based on image conversion and profile neighborhood difference | |
CN102496281A (en) | Vehicle red-light violation detection method based on combination of tracking and virtual loop | |
CN103010258A (en) | System and method for detecting cracks of fasteners of high-speed rails and subways | |
CN104978746A (en) | Running vehicle body color identification method | |
CN202058221U (en) | Passenger flow statistic device based on binocular vision | |
CN102779348A (en) | Method for tracking and measuring moving targets without marks | |
KR101089029B1 (en) | Crime Preventing Car Detection System using Optical Flow | |
Ryu et al. | Feature-based pothole detection in two-dimensional images | |
Shafie et al. | Smart video surveillance system for vehicle detection and traffic flow control | |
Saha et al. | Development of an automated red light violation detection system (RLVDS) for Indian vehicles | |
CN113221603A (en) | Method and device for detecting shielding of monitoring equipment by foreign matters | |
CN105117693A (en) | Video detection method based on optical identification |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20131023 |
|
RJ01 | Rejection of invention patent application after publication |