CN102646274B - Lane boundary detecting device and lane boundary detecting method - Google Patents
Lane boundary detecting device and lane boundary detecting method Download PDFInfo
- Publication number
- CN102646274B CN102646274B CN201210031690.8A CN201210031690A CN102646274B CN 102646274 B CN102646274 B CN 102646274B CN 201210031690 A CN201210031690 A CN 201210031690A CN 102646274 B CN102646274 B CN 102646274B
- Authority
- CN
- China
- Prior art keywords
- marking
- speed reduction
- graticule
- road traffic
- live width
- 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
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a lane boundary detecting device and a lane boundary detecting method, which can reduce processing loads and stably detect lane boundaries. The lane boundary detecting method comprises the following steps of detecting the road traffic markings on the road according to shot images about a vehicle travlling road; judging whether the road traffic markings are a triplet according to the detected road traffic markings; detecting pairs of decelerating markings based on detected marking widths and marking spaces; and determining the lane boundary positions according to the positions of the detected decelerating marking pairs if the road traffic markings are judged to be a triplet.
Description
Technical field
The present invention relates to a kind of lane boundary detection technique that the lane boundary indicating is detected on travel.
Background technology
As the device that detects lane boundary, for example, there is the device described in patent documentation 1.In this device, by template matches, judge complex line (for example 3 heavy lines).In addition, according to the results of comparison judgement track of the template matches with each scounting line, distinguish line.And, track is distinguished to the characteristic quantity of line road traffic marking in addition and is removed as interference.By above processing, from complex line, remove track and distinguish line road traffic marking in addition.
No. 3748811 communique of patent documentation 1 Jap.P.
Summary of the invention
In above-mentioned prior art, utilize template matches to judge the structure of complex line, and utilize this result of determination to judge to distinguish line position, therefore must be corresponding to because of the different various speed reduction markings in region.Therefore, must in navigational system, store fully the speed reduction marking of each region.
Its result, in above-mentioned prior art, in order to increase versatility, must increase the memory capacity that template file is used.In addition,, in above-mentioned prior art, there is the problem of the calculation process load increase of matching treatment.
Therefore, the object of the invention is, reduce to process load, meanwhile, stably detect lane boundary.
In order to address the above problem, the present invention is according to the road traffic marking on the image detection road surface of this Vehicle Driving Cycle of shooting road, according to detected road traffic marking, judge whether road traffic marking is 3 heavy lines, and the graticule live width of the road traffic marking based on above-mentioned detection and graticule interval, be detected as right speed reduction marking group.And, when being judged to be above-mentioned 3 heavy line, according to each position deduction track of the paired speed reduction marking of above-mentioned detection, distinguish the position of line.
Paired speed reduction marking group, symmetrically describes with respect to being positioned at the differentiation line in deceleration strip interval.And, according to the present invention, by being conceived to the simple testing conditions of the feature of above-mentioned speed reduction marking, be detected as right speed reduction marking group.And, according to the speed reduction marking detecting, to inferring track, distinguish the position of line.Its result can reduce processes load.
In addition, based on compare with distinguish line than track the speed reduction marking that the frequency that clearly presents is high in photographic images, infer the position that line is distinguished in track, therefore can stably detect lane boundary.
Accompanying drawing explanation
Fig. 1 means the figure of the vehicle relating to based on embodiments of the present invention.
Fig. 2 is the key diagram that the structure of the deviation warning system relating to based on the 1st embodiment of the present invention is described.
Fig. 3 means the figure of testing result example of the candidate point of the road traffic marking relating to based on embodiments of the present invention.
Fig. 4 is the figure that the structure of the lane recognition device relating to based on the 1st embodiment of the present invention is described.
Fig. 5 means the figure of the processing image of road traffic marking characteristic quantity test section 31.
Fig. 6 describes the key diagram of speed reduction marking to the processing of test section 32.
Fig. 7 is the figure that the presuming method of line position is distinguished in explanation track.
Fig. 8 means the figure of image of the output information of road traffic marking candidate point test section.
Fig. 9 means the figure of the treatment scheme example relating to based on the 1st embodiment of the present invention.
Figure 10 is the figure that the structure of the deviation warning system relating to based on the 2nd embodiment of the present invention is described.
Figure 11 means the figure of the treatment scheme example relating to based on the 2nd embodiment of the present invention.
Figure 12 means the schematic diagram of the kind example of the speed reduction marking that can detect.
Embodiment
Embodiments of the present invention are described with reference to the accompanying drawings.
(the 1st embodiment)
In the present embodiment, take lane recognition device 1 has as the situation of the lane boundary pick-up unit of the first embodiment of the present invention and describes as example.And in the present embodiment, the deviation warning system with above line fare recognition device 1 of take describes as example.
In addition, the user of the output information of lane recognition device 1 is not limited to the control part of deviation.For example, user as the output information of lane recognition device 1, can be the output based on from lane recognition device 1 and drive the drive assistance device of steering actuator and vehicle sensors etc., or for display control unit of the display part of navigational system etc. etc.
(structure)
Fig. 1 means the schematic diagram of the vehicle MM that has carried the deviation warning system with above line fare recognition device 1.Fig. 2 means the schematic block diagram of the deviation warning system of the lane boundary pick-up unit having according to the embodiment of the present invention.
As shown in Figure 1, by vehicle, MM carries in above-mentioned deviation warning system.As shown in Figure 2, this deviation warning system comprises lane recognition device 1, deviation circuit for alarming 5, vehicle speed sensor 6, hummer 7.The lane identification control part 1A that lane recognition device carries out lane identifying processing by camera 2 and the photographic images based on camera 2 forms.In addition, above line fare identification control part 1A and above-mentioned deviation circuit for alarming 5, carry in vehicle MM in the mode of 1 or 2 or 2 above control module.And above-mentioned deviation warning system, utilizes lane recognition device 1 to infer road shape, vehicle state quantity, utilizes the output of this lane recognition device 1, the vehicle that to carry out take deviation be object is controlled.
Lane recognition device 1 is this Vehicle Driving Cycle of identification road, detects with respect to the vehicle state quantity in track and the device of road shape.
The camera 2 of present embodiment is taken the track of vehicle front, particularly vehicle front.Camera 2 is for example arranged on the front center part of vehicle MM indoor canopy, thereby can take vehicle front, and via front windshield, takes the travel of vehicle front.But the installation method of this camera 2 is the general installation examples of carrying out the camera 2 of lane identification.The installation method of camera 2 is not limited to this, so long as can take the installation method of this Vehicle Driving Cycle road.For example, also can be arranged on vehicle front end.In addition, the direction of camera 2 is also not limited to vehicle front, for example, also can, as shown in rear view cameras, at rear view of vehicle, camera be installed.As long as corresponding to the application target that uses vehicle identification result, suitably determine the kind of camera and the setting party of camera to.
Above line fare identification control part 1A as shown in Figure 2, is comprised of road traffic marking candidate point test section 3 and lane identification part 4.The photographic images of road traffic marking candidate point test section 3 based on above-mentioned camera 2, detects road traffic marking candidate point.Lane identification, according to the testing result of the road traffic marking candidate being obtained by this road traffic marking candidate point test section 3, is carried out in lane identification part 4.About the details of road traffic marking candidate point test section 3 and lane identification part 4 as described later.
In addition, as shown in Figure 2, deviation circuit for alarming 5 utilizes by the detected vehicle location with respect to track of lane recognition device 1 and vehicle attitude, and from this vehicle speed of vehicle speed sensor 6, detects this vehicle MM and have or not deviation.And deviation circuit for alarming 5 is implemented following processing: if be judged to be deviation, use hummer 7 to notify alarm to driver.
Above-mentioned road traffic marking candidate point test section 3, the image of this route or travel by vehicle that acquisition camera 2 is taken, as shown in Figure 3, by the image of this acquisition is carried out to image processing, detects road traffic marking edge.And, corresponding to running environment, according to road traffic marking edge, generate road traffic marking characteristic quantity.In Fig. 3, dotted line frame presentation video processing block, thick line represents track characteristic quantity.In addition detect, the characteristic quantity of the road traffic marking position corresponding with each road traffic marking.As shown in Figure 4, this road traffic marking candidate point test section 3 comprises that road traffic marking characteristic quantity test section 31, speed reduction marking distinguish line position to test section 32, track and infer the interval test section 34 of the heavy line of portion 33,3 and road traffic marking characteristic quantity efferent 35.
Above-mentioned road traffic marking characteristic quantity test section 31, by the image obtaining from camera 2 is carried out to image processing as shown in Figure 5, detects the road traffic marking edge as road traffic marking characteristic quantity.In Fig. 5, dotted line frame presentation video processing block, thick line represents track characteristic quantity, dotted arrow represents the direction (searching method) of sweep trace.The concrete processing of above-mentioned road traffic marking characteristic quantity test section 31 is described with reference to Fig. 3 and Fig. 5 below.Road traffic marking characteristic quantity test section 3 first as shown in Figure 3, based on pre-stored road parameters in lane recognition device 1 inside etc., is distinguished line (candidate) with respect to the track that is positioned at the left and right on image, sets the position of image processing block.Then, from extracting the edge of road traffic marking in the image processing block of setting.For example, inner for the image processing block of setting, for example, by primary space differential filter (Sobel wave filter), implement primary space differential, and behind the edge on border of emphasizing road traffic marking and road surface, in the enterprising line scanning of the direction with road traffic marking quadrature, extract road traffic marking edge.And record extracts result.Extraction and recording method as above-mentioned road traffic marking edge, as shown in Fig. 5 (b), detect the edge of road surface-road traffic marking, the edge on road traffic marking-road surface and by these two edges pair that edge forms, right by two edges, as the edge (characteristic quantity) corresponding with road traffic marking and record.In addition, by measuring the distance between above-mentioned detected paired edge, detect lane width (graticule live width), for the edge pair of each road traffic marking, using detected lane width as additional information, carry out record.In addition, the storage of the data of storing as characteristic quantity example, with respect to the coordinate points at predefined true origin, detected each edge and the information of graticule live width.
Above-mentioned speed reduction marking, to test section 32, as shown in Figure 6, based on by by the road traffic marking characteristic quantity test section 31 definite road traffic markings in detected road traffic marking edge, is detected as right speed reduction marking to (paired speed reduction marking group).Above-mentioned speed reduction marking is to test section 32, in the situation that detect, is more than or equal to the road traffic marking row of two and processes.
As concrete method, the pairing of road traffic marking edge is determined after each road traffic marking, graticule live width is more than or equal to predefined setting and graticule live width each other probably to be equated, and graticule interval is more than or equal to the group of predefined value, as paired speed reduction marking, to detecting, above-mentioned predefined value is the value being more than or equal to by the graticule live width of road law regulation.For detected speed reduction marking pair, each edge of two speed reduction markings will be represented, as a pair of record that carries out of speed reduction marking.
For example, in the situation that road traffic marking characteristic quantity test section 31 extracts two road traffic marking (edge), as shown in Figure 6, based on following condition, detect speed reduction marking to (speed reduction marking group).; as following (1)～(3) formula; these two graticule live width W1, W2 are all more than or equal to predefined live width threshold value d0; and the difference of two graticule live width W1, W2 is within predefined range of allowable error d1; and these two graticule interval L1 are more than or equal in the situation apart from d2; as paired speed reduction marking, (speed reduction marking group) detected, above-mentionedly apart from d2, be more than or equal to the graticule live width by road law regulation.
W1、W2≥d0…(1)
|W1-W2|＜d1…(2)
L1＞d2…(3)
Here, as shown in the sweep trace B of Fig. 6, in the situation that detect the road traffic marking row that are more than or equal to three, judge that whether every two road traffic markings meet respectively the condition of above-mentioned speed reduction marking to (group of speed reduction marking), detect (group of speed reduction marking) using qualified group as speed reduction marking.
Here, above-mentioned speed reduction marking is to test section 32, when detecting the many groups of speed reduction markings that meet above-mentioned condition to (group of speed reduction marking), selects the speed reduction marking of graticule live width maximum to (group of speed reduction marking).
And, in the situation that exist speed reduction marking that a plurality of graticule live widths are each other less than or equal to predefined allowable value to (group of speed reduction marking), the group of selecting graticule interval maximum as speed reduction marking to (group of speed reduction marking).
In addition, track is distinguished line position and is inferred portion 33 as shown in Figure 7, according to by speed reduction marking to the right marginal position of the detected speed reduction marking of test section 32, infer the marginal position that line is distinguished in track.Specifically, first, with respect to the Width central side of this route or travel by vehicle, according to the right outermost marginal position of speed reduction marking and the marginal position of inner side, calculate track and distinguish the center of line (with reference to (4) formula).
Then, with respect to the center of width of roadway direction, line is distinguished in track from calculating, half position of line width is distinguished in skew by the track of road law regulation, be estimated as track and distinguish the marginal position of line (with reference to (4) formula).
Half of line standard width distinguished in road traffic marking characteristic quantity coordinate=(the outermost X of outside road traffic marking coordinate+inner side road traffic marking inner side X coordinate) ÷ 2+ track ... (4)
Take with the direction (width of roadway direction) of track quadrature here, is X-axis.
In addition, when calculating the center of track differentiation line, also can obtain according to two edges determining the inner side that speed reduction marking is right.
In addition, track differentiation line position is inferred portion 33 and also can be configured to, and only, in the situation that the interval test sections 34 of 3 heavy line are judged to be 3 heavy line interval, implements the processing of inferring that line marginal position is distinguished in track.
The interval test section 34 of 3 heavy line, according to the testing result of road traffic marking characteristic quantity test section 31, judges that whether this route or travel by vehicle is interval as 3 heavy lines.For example, if 3 road traffic markings that are more than or equal to specified length detected, be judged to be 3 heavy lines interval.
As shown in Figure 8, the Output rusults corresponding to speed reduction marking to test section 32 and the interval test section 34 of 3 heavy line, changes the output to lane identification part 4 to road traffic marking characteristic quantity efferent 35.Specifically, judgement based on the interval test section 34 of 3 heavy line, the road traffic marking of current identification is judged to be to 3 heavy lines, and in the situation that speed reduction marking detects the group of paired speed reduction marking to test section 32, using distinguished line position by track, infer track that portion 33 infers and distinguish the edge of line and as track, distinguish the road traffic marking of line and export (with reference to Fig. 8 (b)).On the other hand, in the situation that not meeting above-mentioned condition, road traffic marking characteristic quantity efferent 35, only by according to by road traffic marking characteristic quantity test section 31 detected edges to calculate, with respect to the edge of the inner side at center, track, as track, distinguish the road traffic marking characteristic quantity of line and export (with reference to Fig. 8 (a)).
The road traffic marking characteristic quantity (road traffic marking edge) of lane identification part 4 based on input calculates the near linear of road shape, based on its result of calculation, by predefined condition estimating device, infers road parameters.As shown in Figure 4, this lane identification part 4 is inferred portion 43 by road shape calculating part 41, road parameters and is formed.
Road shape calculating part 41, according to the detected road traffic marking characteristic quantity of road traffic marking candidate point test section 3, carries out straight line to road shape and is similar to.Specifically, road shape calculating part 41, by being carried out to Hough transformation, following straight line detects, thereby obtain the near linear of road shape, this straight line is to make pixel corresponding to road traffic marking characteristic quantity detecting with road traffic marking candidate point test section 3 preset setting by being more than or equal to, and 1 of the top of surveyed area is linked with following 1.
Road parameters is inferred portion 43, and the near linear of the road shape calculating according to road shape calculating part 41, and following (5) formula based on as road model formula, infer road parameters (with respect to the vehicle posture of road shape and this road).But formula (5) is the general example of the formula of performance road model, the road model formula as inferring travel, is not limited to this formula.
[several 1]
?
Here, parameter A, B, C, D, the H in formula (5) is that road parameters is inferred road parameters, the vehicle state quantity of inferring in portion 43.
Shown in specific as follows.
A: the transversal displacement with respect to track of this vehicle MM
B: road curvature
C: the yaw angle with respect to track of this vehicle MM
D: the pitch angle of this vehicle MM
H: apart from the height of the described image acquisition portion on road surface
In addition, W is for representing the constant of lane width (distance between the inner side in track, left and right), and f is camera perspective transform constant, and j is the parameter in difference track, left and right.Wherein, left road traffic marking is that j=0, right road traffic marking are j=1.In addition, (x, y) is arbitrfary point on the left or right road traffic marking the inner coordinate on road image, and getting road image upper left is initial point, and right is x axle positive dirction, and lower direction is y axle positive dirction.
Below, with reference to the processing example of the treatment scheme explanation above line fare identification control part 1A shown in Fig. 9.
Following processing, during 1 action of lane recognition device, obtains the execution repeatedly of cycle synchronisation ground with the image of camera 2.
First, in step S101, obtain the road image on this Vehicle Driving Cycle road surface of camera 2 shootings, jump to step S102.
In step S102, utilize road traffic marking characteristic quantity test section 31, the road image obtaining in step S101 is carried out to image processing, detect road traffic marking edge.Afterwards, jump to step S103.
In step S103, utilize the interval test section 34 of 3 heavy line, whether according to the information at road traffic marking edge, detect is that 3 heavy lines are interval.Jump to afterwards step S104.
In step S104, utilize speed reduction marking to test section 32, the information according to the road traffic marking edge extracting in step S102, is detected as right speed reduction marking pair.
Then, in step S105, the testing result based on the heavy line of 3 in step S103 interval, and the right amount detection of the speed reduction marking in step S104, judge that whether travel route is interval in 3 heavy lines, and whether speed reduction marking is to being detected.While meeting two conditions, jump to step S106.On the other hand, while not satisfying condition, jump to step S107.
In step S106, by road traffic marking characteristic quantity efferent 35, line edge is distinguished in the track of inferring according to the right testing result of the speed reduction marking in step S104 and as road traffic marking characteristic quantity, export, jump to step S108.
In step S107, from by road traffic marking characteristic quantity test section 31 according to the edge of selecting inner side in detected road traffic marking edge in step S102, as road traffic marking characteristic quantity, export, jump to step S108.
In step S108, road shape calculating part 41 is used the road traffic marking characteristic quantity of input, road shape is carried out to straight line and be similar to.In addition, this processing is divided into region, left and right centered by vehicle MM, regional is calculated to the near linear formula of road shape.Afterwards, jump to step S109.
In step S109, utilize road parameters to infer portion 43, the near linear of the road shape that use step S108 calculates, infers road parameters.Return afterwards, at next control cycle, jump to step S101.
(other actions)
Camera 2, with predefined shooting cycle (shooting time), is taken the travel route of this vehicle.
Road traffic marking candidate point test section 3, carries out image processing by the image that camera 2 is taken, and detects road traffic marking edge, according to this detected road traffic marking edge, generates the road traffic marking characteristic quantity that line is distinguished in track.
Now, the in the situation that of there is a plurality of road traffic marking in an image processing block, implement following processing.That is, the interval test section 34 of 3 heavy line judges whether this route or travel by vehicle is that 3 heavy lines are interval.Now, speed reduction marking carries out the detection of paired speed reduction marking to (group of speed reduction marking) to test section 32.And, the in the situation that of 3 heavy line interval, track differentiation line position is inferred portion 33 speed reduction marking is estimated as to the center that line is distinguished in track to the right center of the detected speed reduction marking of test section 32, infers result obtain the road traffic marking characteristic quantity that line is distinguished in track based on it.
And, lane identification part 4, the road traffic marking characteristic quantity of obtaining based on road traffic marking candidate point test section 3 (distinguishing the road traffic marking edge of line in present embodiment for track), road shape is calculated as near linear, and infer road parameters based on its result of calculation.
By above processing, interval for deceleration strip, in view of the interval distinctive feature of deceleration strip, the fact of symmetrically drawing with respect to distinguishing line, road traffic marking candidate point test section 3 is with specific and simple testing conditions the detection for speed reduction marking, detection speed reduction marking pair.And, according to the speed reduction marking detecting, to inferring track, distinguish the position of line.Its result, can alleviate processing load.
In addition, deceleration road traffic marking is distinguished line with track and is compared, and it is high that deceleration road traffic marking clearly presents frequency on camera 2 images.Therefore, by generating and distinguish line characteristic quantity by deceleration road traffic marking, a large amount of road traffic marking characteristic quantities can be obtained, therefore lane boundary can be stably detected.
Here, camera 2 forms shoot part.Road traffic marking characteristic quantity test section 31 forms graticule candidate detecting unit.The interval test section 34 of 3 heavy line forms 3 heavy line identifying units.Speed reduction marking forms speed reduction marking to detecting unit to test section 32.Track differentiation line position infers portion 33 and road traffic marking characteristic quantity efferent 35 formation differentiation line positions are inferred unit.
(effect of present embodiment)
(1) road traffic marking characteristic quantity test section 31, according to the candidate of the road traffic marking on the image detection road surface of shooting unit photographs.The interval test section 34 of 3 heavy line, based on the detected road traffic marking of graticule candidate detecting unit, judges whether road traffic marking is 3 heavy lines.The graticule live width of the road traffic marking that speed reduction marking detects based on road traffic marking characteristic quantity test section 31 test section 32 and graticule interval, be detected as the group of right speed reduction marking.And, road traffic marking candidate point test section 3, in the situation that interval test section 34 judgements of 3 heavy line are 3 heavy line, the position of paired speed reduction marking test section 32 being detected according to speed reduction marking, infers the position that line is distinguished in track.
Speed reduction marking pair, interval at deceleration strip, with respect to track, distinguish line and symmetrically draw.And, according to the present invention, to be conceived to the simple testing conditions of the feature of such speed reduction marking, detect speed reduction marking pair, according to the speed reduction marking detecting, to inferring track, distinguish the position of line.Its result, compared with prior art can reduce processing load.
And deceleration road traffic marking is distinguished line with track and compared, the frequency that deceleration road traffic marking clearly presents on camera 2 images is high.Therefore, by being generated and distinguished line characteristic quantity by deceleration road traffic marking, can obtain a large amount of road traffic marking characteristic quantities, its result, can stably detect lane boundary.
(2) above-mentioned speed reduction marking, to test section 32, is less than or equal to predefined allowable value two graticule linewidth difference, and these two graticule intervals are while being more than or equal to predefined value, as the group of paired speed reduction marking, detects.
Predefined value is set as being more than or equal to the normal width that line is distinguished in track.
Take and be positioned at track to distinguish the graticule of the same width of left and right be condition, detect speed reduction marking pair.Therefore, can reduce processing load, and speed reduction marking shape that can corresponding various shapes.
(3) above-mentioned speed reduction marking, to test section 32, in the situation that detect the candidates that form the group of right speed reduction marking, selects the group of graticule live width maximum as the group of speed reduction marking more.
Even be road traffic marking in a plurality of interference by error-detecting, in the situation that the left and right of speed reduction marking to beyond detect right, by select wide right of width, can reduce to be caused by interference to error-detecting.
(4) above-mentioned speed reduction marking, to test section 32, in the situation that detect the candidates that form the group of right speed reduction marking, selects the group of graticule interval maximum as speed reduction marking pair more.
Therefore, even in the situation that the live width of track differentiation line and speed reduction marking is identical, also can suitably select the right of speed reduction marking.
(the 2nd embodiment)
Below, with reference to the accompanying drawings of the 2nd embodiment.And, to the structure identical with above-mentioned the 1st embodiment, adopt identical label to describe.
The basic structure of present embodiment is identical with above-mentioned the 1st embodiment.But, as shown in figure 10, in the present embodiment, comprise speed reduction marking live width recording unit 36.And speed reduction marking has the processing of the storage resume based on this speed reduction marking live width recording unit 36 to test section 32.Other processing structure is identical with above-mentioned the 1st embodiment.
Above-mentioned speed reduction marking live width recording unit 36, is recorded in speed reduction marking in recording unit to the right graticule live width of the detected speed reduction marking of test section 32.But, be only the situation of 3 heavy lines that is identified as by the interval test section 34 of 3 heavy line.By the interval test sections 34 of 3 heavy lines, be judged to be not 3 heavy line in the situation that, the past resume of the graticule live width recording in this recording unit are cleared.That is, above-mentioned speed reduction marking live width recording unit 36, the result of determination based on the interval test section 34 of 3 heavy line, if be judged to be 3 heavy lines, the right graticule live width of speed reduction marking speed reduction marking being detected detecting unit is recorded in recording unit.In addition, if be judged to be, be not 3 heavy lines, above-mentioned speed reduction marking live width recording unit 36 is removed the past resume of the graticule live width of storing in recording unit.
In addition, speed reduction marking is to test section 32, in the situation that the resume of the graticule live width of speed reduction marking live width recording unit 36 records are more than or equal to predefined quantity, obtains the mean value of a plurality of graticule live widths of its storage as graticule live width reference value.And, speed reduction marking is to test section 32, the graticule live width of the graticule of more detected speed reduction marking candidate and above-mentioned graticule live width reference value, in the situation that the difference of the graticule live width of the graticule of speed reduction marking candidate and above-mentioned graticule live width reference value is more than or equal to predefined value, from speed reduction marking candidate, get rid of the graticule of this speed reduction marking candidate.That is, using the graticule that is only less than predefined value with the difference of above-mentioned graticule live width reference value as speed reduction marking candidate.
Below, the processing example about the speed reduction marking in embodiments of the present invention to test section 32, describes with reference to Figure 11.
First, in step S201, road traffic marking characteristic quantity test section 31 detected road traffic marking edges are matched, jump to step S202.
In step S202, road traffic marking edge based on pairing, select graticule live width be more than or equal to predefined setting and graticule live width each other about equally, graticule interval is more than or equal to normal width right that line is distinguished in track, as speed reduction marking pair.
Concrete, determine whether and meet described (1)～(3) formula.
Now, in step S202, by speed reduction marking live width recording unit 36, stored in the situation of the graticule live width that is more than or equal to predetermined amount, according to the graticule live width of storage, calculated the mean value of speed reduction marking live width, as graticule reference value.And, in graticule live width, compare short predefined value with graticule live width reference value above in the situation that, or in situation more than long predefined value, do not select it as speed reduction marking pair.And after processing finishes, jump to step S203.
In step S203, judge whether the speed reduction marking of being selected by step S202 is to being more than or equal to two.Existence is more than or equal in the situation of two and jumps to step S204.Exist processing in the situation of to finish.
In step S204, relatively speed reduction marking between graticule live width, select graticule live width maximum to as speed reduction marking pair.But, use predefined allowable error, in error range, exist a plurality of to time, select these a plurality of to as speed reduction marking.And, after processing, jump to step S205.
In step S205, judge whether the speed reduction marking of being selected by step S204 is to being more than or equal to two.Existence is more than or equal in the situation of two and jumps to step S206.Exist processing in the situation of to finish
In step S206, the right graticule interval of each speed reduction marking relatively.Select graticule interval maximum to as speed reduction marking pair.Afterwards, processing finishes.
Other structure is identical with above-mentioned the 1st embodiment with processing.
(other action)
On actual road surface, the situation that exists tire trace, seam etc. to disturb.And existing the above-mentioned Interference Detection beyond road traffic marking is the problem at road traffic marking edge.
The width of the speed reduction marking of setting in intercity high speed in addition, is mostly wide than the width of track differentiation line.In expressway around city is waited at a high speed in capital, the width of the width of speed reduction marking and track differentiation line is roughly the same.State described above, speed reduction marking to the processing of test section 32 in, by interference error-detecting described above, be likely speed reduction marking pair.
To this, in present embodiment, even if the situation that exists tire trace, seam etc. to disturb also can reduce the generation of error-detecting or it is not occurred.For example, as shown in figure 12, even each various speed reduction markings that all city or community exists also can detect speed reduction marking pair, even if it is interval at the deceleration strip consisting of speed reduction marking mode as Figure 12 therefore to travel, also can continues to detect original track and distinguish line.Here, in Figure 12, label A represents speed reduction marking.
Here, speed reduction marking live width recording unit 36 forms graticule live width storage unit.The correspondence of other structure is identical with above-mentioned the 1st embodiment.
(effect of present embodiment)
Present embodiment, not only has by the effect described in above-mentioned the 1st embodiment, and has following effect.
(1) speed reduction marking live width recording unit 36, stores the graticule live width of above-mentioned speed reduction marking to the speed reduction marking of detecting unit detection.Above-mentioned speed reduction marking is to test section 32, in the stored situation of graticule live width of preset value, graticule live width based on above-mentioned graticule live width cell stores is determined graticule benchmark live width, only will form the road traffic marking of graticule live width that difference with above-mentioned graticule benchmark live width is less than or equal to preset value as the candidate of speed reduction marking.
Therefore, for example, according to the mean value of past resume, infer the width of speed reduction marking, can remove is not obviously the speed reduction marking pair of speed reduction marking.Its result, can reduce error-detecting.
(2) above-mentioned speed reduction marking live width recording unit 36, the result of determination based on the interval test section 34 of 3 heavy line judges that current track is not 3 heavy lines, removes the record of graticule live width, defines the past resume of graticule live width.
Therefore,, by only detecting definitely speed reduction marking to time record resume in the past, can prevent detect and the wrong interference that can not detect speed reduction marking equity is removed.
Claims (5)
1. a lane boundary pick-up unit, is characterized in that, has:
Take unit, it takes this Vehicle Driving Cycle road;
Graticule candidate detecting unit, it detects the candidate of the road traffic marking on road surface according to the image of taking unit photographs;
3 heavy line identifying units, it judges based on the detected road traffic marking of graticule candidate detecting unit whether road traffic marking is 3 heavy lines;
Speed reduction marking is to detecting unit, and it,, based on by the detected graticule live width of graticule candidate detecting unit and graticule interval, is detected as the group of right speed reduction marking; And
Distinguish line position and infer unit, it is in the situation that 3 heavy line identifying units are judged to be 3 heavy line, and the position according to speed reduction marking to the detected paired speed reduction marking of detecting unit, infers the position that line is distinguished in track,
Above-mentioned speed reduction marking is to detecting unit, in the difference of two graticule live widths, is less than or equal to predefined allowable value, and these two graticule intervals are more than or equal in the situation of predefined value, as the group of paired speed reduction marking, detect,
Above-mentioned speed reduction marking, to detecting unit, in the situation that detect the candidates that form the group of right speed reduction marking, selects the group of graticule live width maximum as the group of speed reduction marking more.
2. lane boundary pick-up unit according to claim 1, is characterized in that,
Above-mentioned speed reduction marking, to detecting unit, in the situation that detect the candidates that form the group of right speed reduction marking, selects the group of graticule interval maximum as speed reduction marking pair more.
3. lane boundary pick-up unit according to claim 1 and 2, is characterized in that,
Have graticule live width storage unit, it stores the graticule live width of above-mentioned speed reduction marking to the detected speed reduction marking of detecting unit,
Above-mentioned speed reduction marking is to detecting unit, in the stored situation of graticule live width of predetermined amount, graticule live width based on above-mentioned graticule live width cell stores is determined graticule benchmark live width, by the difference with above-mentioned graticule benchmark live width, be only that predefined value is with the road traffic marking of interior graticule live width, as the candidate of speed reduction marking.
4. lane boundary pick-up unit according to claim 3, is characterized in that,
Above-mentioned graticule live width storage unit, in the situation that the result of determination based on 3 heavy line identifying units judges that current track is not 3 heavy lines, removes the record of graticule live width.
5. a lane boundary detection method, is characterized in that,
According to the road traffic marking on the image detection road surface of this Vehicle Driving Cycle of shooting road, according to detected road traffic marking, judge whether road traffic marking is 3 heavy lines, and the graticule live width of the road traffic marking based on above-mentioned detection and graticule interval, be detected as the group of right speed reduction marking
In the situation that being judged to be above-mentioned 3 heavy line, according to the position deduction track of above-mentioned detected paired speed reduction marking, distinguish the position of line,
Difference two graticule live widths is less than or equal to predefined allowable value, and these two graticule intervals are more than or equal in the situation of predefined value, as the group of paired speed reduction marking, detect,
In the situation that detect the candidates that form the group of right speed reduction marking, select the group of graticule live width maximum as the group of speed reduction marking more.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2011031121A JP5716443B2 (en) | 2011-02-16 | 2011-02-16 | Lane boundary detection device and lane boundary detection method |
| JP2011-031121 | 2011-02-16 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102646274A CN102646274A (en) | 2012-08-22 |
| CN102646274B true CN102646274B (en) | 2014-11-05 |
Family
ID=46659086
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210031690.8A Active CN102646274B (en) | 2011-02-16 | 2012-02-13 | Lane boundary detecting device and lane boundary detecting method |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP5716443B2 (en) |
| CN (1) | CN102646274B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3019508B1 (en) * | 2014-04-08 | 2017-12-08 | Alstom Transp Tech | METHOD FOR DETECTING RAILS IN WHICH CIRCULATES A RAILWAY VEHICLE |
| CN104477103B (en) * | 2014-12-25 | 2016-08-24 | 刘志平 | A kind of Automobile broadside circle comes back and shows drafting, use and implementation |
| JP6483446B2 (en) * | 2015-01-15 | 2019-03-13 | トヨタ自動車株式会社 | Composite line determination apparatus and composite line determination method |
| JP6628189B2 (en) * | 2016-05-19 | 2020-01-08 | パナソニックＩｐマネジメント株式会社 | Detection device and detection method |
| CN108304749A (en) * | 2017-01-13 | 2018-07-20 | 比亚迪股份有限公司 | The recognition methods of road speed line, device and vehicle |
| JP6776213B2 (en) * | 2017-11-01 | 2020-10-28 | 株式会社東芝 | Orbit identification device |
| JP6364565B1 (en) * | 2018-02-15 | 2018-07-25 | 阪神高速技術株式会社 | Analysis device, crack detection processing device, and analysis program |
| CN109740469B (en) * | 2018-12-24 | 2021-01-22 | 百度在线网络技术（北京）有限公司 | Lane line detection method, lane line detection device, computer device, and storage medium |
| CN109657641B (en) * | 2018-12-29 | 2021-02-02 | 北京经纬恒润科技股份有限公司 | Method and device for judging main road and auxiliary road of vehicle |
| CN111002990B (en) * | 2019-12-05 | 2021-06-08 | 华南理工大学 | Lane departure early warning method and system based on dynamic departure threshold |
| US11579250B2 (en) | 2020-12-23 | 2023-02-14 | Robert Bosch Gmbh | Lane boundary detection using radar signature trace data |
| CN113532425B (en) * | 2021-09-16 | 2021-11-30 | 西南交通大学 | Tunnel base station-free vehicle-road cooperative positioning method based on oscillation marked lines and mobile phone sensing |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1836267A (en) * | 2003-09-24 | 2006-09-20 | 爱信精机株式会社 | Device for detecting road traveling lane |
| CN101264755B (en) * | 2008-03-06 | 2010-07-28 | 上海交通大学 | Vehicle running safety intelligent monitoring and controlling device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3748811B2 (en) * | 2001-11-30 | 2006-02-22 | 株式会社豊田中央研究所 | Lane boundary judgment device |
| JP3930366B2 (en) * | 2002-04-18 | 2007-06-13 | 株式会社デンソー | White line recognition device |
| JP4506790B2 (en) * | 2007-07-05 | 2010-07-21 | アイシン・エィ・ダブリュ株式会社 | Road information generation apparatus, road information generation method, and road information generation program |
-
2011
- 2011-02-16 JP JP2011031121A patent/JP5716443B2/en active Active
-
2012
- 2012-02-13 CN CN201210031690.8A patent/CN102646274B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1836267A (en) * | 2003-09-24 | 2006-09-20 | 爱信精机株式会社 | Device for detecting road traveling lane |
| CN101264755B (en) * | 2008-03-06 | 2010-07-28 | 上海交通大学 | Vehicle running safety intelligent monitoring and controlling device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102646274A (en) | 2012-08-22 |
| JP5716443B2 (en) | 2015-05-13 |
| JP2012168884A (en) | 2012-09-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102646274B (en) | Lane boundary detecting device and lane boundary detecting method | |
| US10977504B2 (en) | Vehicle-mounted image target objection recognition device | |
| CN103630122B (en) | Monocular vision lane line detection method and distance measurement method thereof | |
| JP4962581B2 (en) | Lane marking detector | |
| JP5281664B2 (en) | Lane departure warning device and lane departure warning system | |
| US20110320163A1 (en) | Method and system for determining road data | |
| JP4607193B2 (en) | Vehicle and lane mark detection device | |
| US6122597A (en) | Vehicle monitoring apparatus | |
| US10885358B2 (en) | Method for detecting traffic signs | |
| CN109435942A (en) | A kind of parking stall line parking stall recognition methods and device based on information fusion | |
| CN109871732B (en) | Parking grid identification system and method thereof | |
| CN109791598A (en) | The image processing method of land mark and land mark detection system for identification | |
| US20050278112A1 (en) | Process for predicting the course of a lane of a vehicle | |
| JP2013530435A (en) | How to find roadway transitions for automobiles | |
| US20080266139A1 (en) | Method for detecting parking area by using range sensor | |
| CN102915532A (en) | Method of determining extrinsic parameters of a vehicle vision system and vehicle vision system | |
| US9792509B2 (en) | Method and device for determining a lane course of a vehicle | |
| CN104937648A (en) | Road environment recognition system | |
| CN107097794A (en) | The detecting system and its method of terrain vehicle diatom | |
| CN112124304B (en) | Library position positioning method and device and vehicle-mounted equipment | |
| CN113874914A (en) | Method for determining an operating angle between a tractor and a trailer of a tractor | |
| CN105320928A (en) | Object recognition apparatus | |
| JP2018055222A (en) | Runway detection method and runway detection device | |
| KR20160055413A (en) | Automatic scoring system for driving license test using around-view monitering video, and method for the same | |
| JP2005173899A (en) | Surrounding situation display unit |
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 |