CN109711242A - Modification method, device and the storage medium of lane line - Google Patents
Modification method, device and the storage medium of lane line Download PDFInfo
- Publication number
- CN109711242A CN109711242A CN201811291589.XA CN201811291589A CN109711242A CN 109711242 A CN109711242 A CN 109711242A CN 201811291589 A CN201811291589 A CN 201811291589A CN 109711242 A CN109711242 A CN 109711242A
- Authority
- CN
- China
- Prior art keywords
- lane line
- target
- breakpoint
- image
- line
- 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
Abstract
The present invention provides the modification method, device and storage medium of a kind of lane line, this method comprises: obtaining the image on current driving road surface；According to image, first lane line and target lane line on current driving road surface are obtained, first lane line is the lane line that lane line missing occurs in image, target lane line are as follows: is greater than the lane line of similarity threshold in image with the similarity of first lane line；Using target lane line, the first lane line in image is modified.The present invention can be modified the lane line of missing according to the target lane line in image, so that unmanned vehicle obtains correct, continuous lane line, improve the safety of automatic Pilot.
Description
Technical field
The present invention relates to automatic Pilot technical fields more particularly to a kind of modification method of lane line, device and storage to be situated between
Matter.
Background technique
With the continuous development of automatic Pilot technology, more and more automatic Pilot services initially enter people's lives
Circle.Unmanned vehicle can acquire the point cloud data of ambient enviroment in the process of moving, determine current unmanned vehicle institute according to point cloud data
Position；It wherein, include the point cloud data of lane line in the point cloud data, unmanned vehicle is according to the point cloud number of the lane line of acquisition
According to correctly being travelled in lane.
In the prior art, lane line is due to the problems such as blocking, wearing, and will lead to that lane line is unclear, broken string, alternatively,
There is the problems such as the phenomenon that missing of lane line in crossing region；And then cause unmanned vehicle that can not obtain correct, continuous lane line,
Further affected to the accurate traveling of unmanned vehicle.
Summary of the invention
The present invention provides the modification method, device and storage medium of a kind of lane line, can be according to the target carriage in image
Diatom is modified the lane line of missing, so that unmanned vehicle obtains correct, continuous lane line, improves the peace of automatic Pilot
Quan Xing.
The first aspect of the present invention provides a kind of modification method of lane line, comprising:
Obtain the image on current driving road surface；
According to described image, the first lane line and target lane line on the current driving road surface are obtained, described first
Lane line is the lane line that lane line missing occurs in described image, the target lane line are as follows: with described the in described image
The similarity of one lane line is greater than the lane line of similarity threshold；
Using the target lane line, the first lane line in described image is modified.
Optionally, the image for obtaining unmanned vehicle current driving road surface, comprising:
Obtain the point cloud data on the current driving road surface；
The point cloud data is projected into two-dimensional surface, the corresponding reflected value base map of the point cloud data is obtained；
Using the reflected value base map as described image.
Optionally, the first lane line is the lane line lacked in the middle part of lane line, obtains the current driving road
Target lane line on face, comprising:
Obtain the first breakpoint and the second breakpoint of the first lane line；
Multiple candidate lane lines are obtained within the scope of the pre-determined distance of first breakpoint；
According to the similarity of each candidate lane line and the first lane line, maximum similarity is greater than the phase if it exists
Like degree threshold value, then using the corresponding candidate lane line of maximum similarity as the target lane line.
Optionally, the similarity according to each candidate lane line and the first lane line, and by maximum similarity
Before corresponding candidate lane line is as the target lane line, comprising:
Each candidate lane line is translated, until first breakpoint and first breakpoint are described each
Subpoint on the candidate lane line is overlapped；
Obtain the distance of second breakpoint to each candidate lane line；
According to second breakpoint to the distance of each candidate lane line, each candidate lane line and described the are obtained
The similarity of one lane line.
Optionally, described to use the target lane line, the first lane line in described image is modified, comprising:
According to the target lane line, first breakpoint and second breakpoint are attached.
Optionally, the first lane line is the lane line that lane line one end lacks, and the first lane line is more
It is a, obtain the target lane line on the current driving road surface, comprising:
According to the ascending order of the length of the lane line lacked in each first lane line, to multiple first lane lines
It is ranked up；
For sequence after each of the first lane line, using the previous first lane line of the first lane line as
The target lane line of the first lane line.
Optionally, described to use the target lane line, the first lane line in described image is modified, comprising:
The target lane line is moved to and the first lane line overlap.
The second aspect of the present invention provides a kind of correcting device of lane line, comprising:
Image collection module, for obtaining the image on current driving road surface；
Lane line obtains module, for according to described image, obtain first lane line on the current driving road surface and
Target lane line, the first lane line are the lane line that lane line missing occurs in described image, the target lane line are as follows:
It is greater than the lane line of similarity threshold in described image with the similarity of the first lane line；
Correction module is modified the first lane line in described image for using the target lane line.
Optionally, described image obtains module, specifically for obtaining the point cloud data on the current driving road surface；It will be described
Point cloud data is projected into two-dimensional surface, obtains the corresponding reflected value base map of the point cloud data；The reflected value base map is made
For described image.
Optionally, the lane line obtains module, specifically for obtaining the first breakpoint and second of the first lane line
Breakpoint；Multiple candidate lane lines are obtained within the scope of the pre-determined distance of first breakpoint；According to each candidate lane line and institute
The similarity of first lane line is stated, maximum similarity is greater than the similarity threshold if it exists, then maximum similarity is corresponding
Candidate lane line is as the target lane line.
Optionally, described device further include: similarity obtains module；
The similarity obtains module, for translating each candidate lane line, until first breakpoint
It is overlapped with subpoint of first breakpoint on each candidate lane line；Second breakpoint is obtained to each institute
State the distance of candidate lane line；According to second breakpoint to the distance of each candidate lane line, each candidate vehicle is obtained
The similarity of diatom and the first lane line.
Optionally, the correction module, is specifically used for according to the target lane line, to first breakpoint and described the
Two breakpoints are attached.
Optionally, the lane line obtains module, is specifically also used to according to the vehicle lacked in each first lane line
The ascending order of the length of diatom is ranked up multiple first lane lines；For sequence after each of the first lane line,
Using the previous first lane line of the first lane line as the target lane line of the first lane line.
Optionally, the correction module is specifically also used to for the target lane line being moved to and the first lane line
Overlapping.
The third aspect of the present invention provides a kind of correcting device of lane line, comprising: at least one processor and memory；
The memory stores computer executed instructions；
At least one described processor executes the computer executed instructions of the memory storage, so that the lane line
Correcting device executes the modification method of above-mentioned lane line.
The fourth aspect of the present invention provides a kind of computer readable storage medium, deposits on the computer readable storage medium
Computer executed instructions are contained, when the computer executed instructions are executed by processor, realize the amendment side of above-mentioned lane line
Method.
The present invention provides the modification method, device and storage medium of a kind of lane line, this method comprises: obtaining current driving
The image on road surface；According to image, first lane line and target lane line on current driving road surface are obtained, first lane line is figure
The lane line of lane line missing, target lane line occur as in are as follows: be greater than similarity with the similarity of first lane line in image
The lane line of threshold value；Using target lane line, the first lane line in image is modified.The present invention can be according in image
Target lane line the lane line of missing is modified so that unmanned vehicle obtains correct, continuous lane line, improve automatic
The safety of driving.
Detailed description of the invention
Fig. 1 is the flow diagram one of the modification method of lane line provided by the invention；
Fig. 2 is the schematic diagram one for the lane line that the present invention obtains；
Fig. 3 is the flow diagram two of the modification method of lane line provided by the invention；
Fig. 4 is the lane line schematic diagram after being modified to the lane line A in Fig. 2；
Fig. 5 is the flow diagram three of the modification method of lane line provided by the invention；
Fig. 6 is the schematic diagram two for the lane line that the present invention obtains；
Fig. 7 is the lane line schematic diagram after being modified to the lane line in Fig. 6；
Fig. 8 is the structural schematic diagram one of the correcting device of lane line provided by the invention；
Fig. 9 is the structural schematic diagram two of the correcting device of lane line provided by the invention；
Figure 10 is the structural schematic diagram three of the correcting device of lane line provided by the invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with the embodiment of the present invention, to this
Technical solution in inventive embodiments is clearly and completely described, it is clear that described embodiment is that a part of the invention is real
Example is applied, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creation
Property labour under the premise of every other embodiment obtained, shall fall within the protection scope of the present invention.
Collecting vehicle acquires the point cloud data of all positions in unmanned region in advance, and position and point cloud data are carried out pair
High-precision map should be obtained；Unmanned vehicle can acquire the point cloud data of ambient enviroment in the process of moving, according to what is currently acquired
Point cloud data is matched with the point cloud data in high-precision map, obtains the position where unmanned vehicle.Further, at nobody
It include the point cloud data of lane line in the point cloud data of vehicle acquisition, unmanned vehicle is known according to the point cloud data of the lane line of acquisition
Not Chu lane line, travelled according to the lane line of identification.
But lane line is unclear, broken string will lead to due to wearing, block etc. for lane line；Alternatively, in road surface inlet
The problems such as lane line lacks causes unmanned vehicle that can not obtain correct, continuous lane line, and then can not normally travel.
Technical term paraphrase of the present invention:
Point cloud data: the prior art mainly adopts the modes of laser scanning, obtain the point cloud data of environment；Work as beam of laser
When being irradiated to body surface, the laser reflected can carry the information such as orientation, distance.If laser beam is carried out according to certain track
The laser point information of reflection will be recorded in scanning in scanning, extremely fine due to scanning, then can obtain a large amount of laser
Point, thus the laser point cloud data of object can be formed.Point cloud data is the set in the massive point cloud of target surface characteristic.
Reflected value base map: the point cloud obtained according to laser measurement principle, including three-dimensional coordinate (XYZ) and laser reflection letter
Breath；The point cloud obtained according to photogrammetry principles, including three-dimensional coordinate (XYZ)；It is obtained in conjunction with laser measurement and photogrammetry principles
To a cloud, including three-dimensional coordinate (XYZ) and laser reflection information.Point cloud data is indicated according to the reflective information in cloud
Obtain the corresponding reflected value base map of point cloud data.
Semantic segmentation: doing each pixel in the image of acquisition and classify, i.e., each pixel belongs to assorted in acquisition image
The pixel of object.The prior art mainly adopts the modes of convolutional neural networks to classify to pixel.
Fig. 1 is the flow diagram one of the modification method of lane line provided by the invention, the execution of method flow shown in Fig. 1
Main body can be lane line correcting device, wherein the correcting device of lane line can be set on unmanned vehicle, may also set up with
Server set is integrally formed setting, can be set in following embodiments with the correcting device of lane line and says in unmanned vehicle progress example
Bright, the correcting device of the lane line can be by arbitrary software and or hardware realization.As shown in Figure 1, lane provided in this embodiment
The modification method of line may include:
S101 obtains the image on current driving road surface.
In the present embodiment, the correcting device for the lane line being arranged on unmanned vehicle available unmanned vehicle current driving road surface
Image.
Wherein, a method of obtaining the image on current driving road surface are as follows: filming apparatus is provided on unmanned vehicle, the shooting
The road surface that device can in real time travel unmanned vehicle is shot；Wherein, which can be with the correcting device of lane line
It is provided separately, after the image on filming apparatus shooting current driving road surface, the correcting device of lane line can be sent an image to；
Filming apparatus can also become one setting with the correcting device of lane line, directly acquire nobody by the correcting device of lane line
The image on vehicle current driving road surface.
Another kind obtains the mode of the image on current driving road surface are as follows: the correcting device of lane line can acquire unmanned vehicle and work as
The point cloud data of preceding track, due to including the Reflection intensity information of acquisition target in point cloud data, according to the prior art
In point cloud data conversion method, the point cloud data on current driving road surface is converted into reflected value base map, by the reflected value base map
Image as current driving road surface.
S102 obtains first lane line and target lane line on current driving road surface according to image, and first lane line is
The lane line of lane line missing, target lane line occur in image are as follows: to the similarity of first lane line greater than similar in image
Spend the lane line of threshold value.
Specifically, what the image that the correcting device of lane line obtains was made of multiple block of pixels, each block of pixels is corresponding
Object may be different.It is previously stored with disaggregated model in the correcting device of lane line in the present embodiment, which uses
Classify in the block of pixels in image, i.e. progress semantic segmentation, obtains the lane line in image；The disaggregated model can be
Training in advance obtains by the way of convolutional neural networks in the prior art, to the acquisition side of disaggregated model in the present embodiment
Formula is with no restrictions.
In the present embodiment, after the lane line in image is obtained according to disaggregated model, first on current driving road surface is obtained
Lane line and target lane line.It wherein, include multiple lane lines, multiple vehicle in the image that the correcting device of lane line obtains
Diatom can be identical or different lane line.Fig. 2 is the schematic diagram one for the lane line that the present invention obtains, as shown in Fig. 2,
It include that there are three lane line, respectively lane line A, lane line B and lane line C in the image that the correcting device of lane line obtains.
Specifically, first lane line is the lane line that lane line missing occurs in image.The correcting device of lane line according to
Lane line in the available image of the mode of above-mentioned semantic segmentation, and, the corresponding block of pixels of lane line；According to block of pixels
Continuity obtains the first lane line in image, i.e. discontinuous situation occurs for the corresponding block of pixels of lane line, in Fig. 2
Shown in lane line A.
In the present embodiment, after the correcting device of lane line obtains first lane line, in multiple lane lines in the picture,
Target lane line is obtained in lane line other than first lane line；Specifically, the target lane line are as follows: with first in image
The similarity of lane line is greater than the lane line of similarity threshold.
Wherein it is possible to obtain the similarity of lane line and first lane line according to the radian of lane line, direction, type etc.；
Illustratively, if the difference of the radian of the radian of lane line and first lane line is less than difference threshold, determine that the lane line is the
The target lane line of one lane line；Further, when obtaining the similarity of lane line and first lane line, it is also necessary to so that obtaining
The target lane line taken is identical as the direction of first lane line and type.The direction of lane line refers to the circular arc that lane line is formed
The direction in the corresponding center of circle, it is solid line or dotted line etc. that the type of lane line, which can be lane line,.According to above-mentioned acquisition similarity
Mode obtain the similarities of multiple lane lines Yu first lane line, will be greater than the lane line of similarity threshold as target lane
Line.It is envisioned that when the lane line for being greater than similarity threshold is multiple, using the corresponding lane line of maximum similarity as mesh
Mark lane line.
It is envisioned that when first lane line is straightway, it can be according to whether parallel with first lane line obtain
Target lane line.In the present embodiment, for different lane line types, target lane line can be obtained using corresponding mode,
As long as the similarity of the target lane line and first lane line that obtain is greater than similarity threshold.
Illustratively, as shown in Fig. 2, by the lane line B and lane line C calculating with the similarity of lane line A respectively, really
Determine the target lane line that lane line B is lane line A.
S103 is modified the first lane line in image using target lane line.
In the present embodiment, since the target lane line of acquisition and the similarity of first lane line are greater than similarity threshold.It is right
A kind of mode that first lane line in image is modified may is that the corresponding block of pixels of target lane line is replicated,
It is moved to the position of first lane line, is made up using the part that target lane line lacks first lane line, is corrected, so that
First lane line forms complete lane line.It is envisioned that translated using target lane line top divided lane line, it is right
The part of first lane line missing is modified.
Another way may is that the breakpoint for obtaining first lane line, which is projected on target lane line,
Obtain subpoint corresponding with the breakpoint on target lane line；Illustratively, as shown in Fig. 2, there are two breakpoint on lane line A,
Respectively breakpoint a and breakpoint b；According to projection pattern in the prior art, it is right respectively that breakpoint a and breakpoint b is obtained on lane line B
The subpoint a ' and b ' answered, is attached, so that lane line A forms complete vehicle using formula a pair of lane line A and lane line B
Diatom.
Wherein, target lane line and first lane line are attached according to pre-set connection formula, so that first
Lane line forms complete lane line, specifically, the connection formula can be as follows shown in formula one:
Wherein, XaFor the abscissa of the first breakpoint a on first lane line (lane line A), YaFor first on lane line A
The ordinate of breakpoint a, XbFor the abscissa of the second breakpoint b on lane line A, YbFor the vertical seat of the second breakpoint b on lane line A
Mark, Ya' the ordinate for being subpoint a ' of the first breakpoint a on target lane line (lane line B), Yb' it is the second breakpoint b in vehicle
The ordinate of subpoint b ' on diatom B, X ' are the abscissa at any point on the Projection Line Segment a ' b ' on lane line B, Y '
For the ordinate at any point on the Projection Line Segment a ' b ' on lane line B, Y is on the Projection Line Segment a ' b ' on lane line B
Any point is connected to the ordinate on lane line A.
It refers to using on target lane line specifically, target lane line and first lane line are attached, with first
The corresponding lane line of lane line of lane line missing, is attached the lane line of the missing, makes up.The connection type avoids
Lane line is replicated, is translated, is connected, can directly according to subpoint to target lane line and first lane line into
Row connection.
The modification method of lane line provided in this embodiment includes: the image for obtaining current driving road surface；According to image, obtain
The first lane line and target lane line on current driving road surface are taken, first lane line is the vehicle that lane line missing occurs in image
Diatom, target lane line are as follows: be greater than the lane line of similarity threshold in image with the similarity of first lane line；Using target carriage
Diatom is modified the first lane line in image.The modification method of lane line provided in this embodiment can be according to image
In target lane line the lane line of missing is modified so that unmanned vehicle obtains correct, continuous lane line, improve from
The dynamic safety driven.
The missing of lane line can be divided into: missing, lane line A as shown in Figure 2 in the middle part of lane line；It is also possible to lane
Line one end missing, such case are suitable for having multiple lanes at link entry, and each vehicle lane line has equal length, if but
When one end missing of lane line, so that the lane line end and normal lane line are uneven, image unmanned vehicle normally takes part in Taoism and goes
It sails.
On the basis of the above embodiments, below with reference to Fig. 3 to the vehicle in the modification method of lane line provided by the invention
The case where missing, is described in detail in the middle part of diatom, and Fig. 3 is the flow diagram of the modification method of lane line provided by the invention
Two, as shown in figure 3, the modification method of lane line provided in this embodiment may include:
S301 obtains the point cloud data on current driving road surface.
Wherein, point cloud data acquisition device is provided on unmanned vehicle, which can be in real time to nobody
The road surface of vehicle traveling is scanned, and obtains the point cloud data on current driving road surface；Point cloud data acquisition device in the present embodiment
It can be laser radar system in the prior art.
Specifically, the point cloud data acquisition device can be provided separately with the correcting device of lane line, when point cloud data obtains
After taking device to obtain the point cloud data on current driving road surface, the point cloud data that can will acquire current driving road surface is sent to lane
The correcting device of line；Point cloud data acquisition device can also become one setting with the correcting device of lane line, by lane line
Correcting device directly acquire the point cloud data on current driving road surface.
S302 projects point cloud data into two-dimensional surface, obtains the corresponding reflected value base map of point cloud data；By reflected value
Image of the base map as current driving road surface.
Point cloud data is three-dimensional state, projects the point cloud data of the three-dimensional to two according to projection pattern in the prior art
In dimensional plane, the corresponding reflected value base map of point cloud data is obtained, specifically, including the anti-of each block of pixels on reflected value base map
Penetrate value strength information.In the present embodiment, using reflected value base map as the image on current driving road surface.
S303 obtains the first lane line on current driving road surface according to image, and first lane line is hair in the middle part of lane line
The lane line of raw missing.
Wherein, which is the reflected value base map on current driving road surface；It is previously stored in the correcting device of lane line point
Class model, specifically, the disaggregated model is to be obtained according to the feature of the block of pixels in reflected value base map as training parameter, it should
Disaggregated model can classify each block of pixels in reflected value base map, obtain the corresponding object of each block of pixels.
Specifically, according to multiple lane lines in the available reflected value base map of the disaggregated model；Wherein, according to lane line
Whether corresponding block of pixels is continuous, obtains first lane line, and first lane line is the lane line that lane line missing occurs；This implementation
In example, the first lane line lacked in the middle part of lane line is obtained according to the type of lane line missing.
S304 obtains the first breakpoint and the second breakpoint of first lane line.
First lane line in the present embodiment is the lane line of middle part missing, i.e. first lane line is two broken string lane lines
Composition, middle part have two adjacent breakpoints, are the first breakpoint and the second breakpoint.It can belong to according to by judgement in the present embodiment
First lane line whether is continuously acquired in the block of pixels of same lane line, in the block of pixels that the first lane line lacks
Obtain the first breakpoint and the second breakpoint.
Illustratively, as shown in Fig. 2, there are two breakpoints, respectively breakpoint a and breakpoint b on lane line A.
S305 obtains multiple candidate lane lines within the scope of the pre-determined distance of the first breakpoint.
Illustratively, as shown in Fig. 2, the first breakpoint can be breakpoint a or breakpoint b；In the preset range of the first breakpoint,
Obtain multiple candidate lane lines.Wherein, the preset range of first breakpoint can be pre-determined distance or predetermined angle；It is corresponding
, candidate lane line can be the lane line in the range of pre-determined distance of the first breakpoint of distance；Alternatively, candidate lane line can
Think the lane line in the range of predetermined angle of the first breakpoint of distance.
S306 translates each candidate lane line, until the first breakpoint and the first breakpoint are in each candidate lane line
On subpoint be overlapped, obtain the second breakpoint to each candidate lane line distance.
In the present embodiment, the first breakpoint is projected on each candidate lane line, is obtained on each candidate lane line
Take the corresponding subpoint of the first breakpoint；Specifically, when candidate lane line is linear vehicle diatom, by the first breakpoint in candidate lane
Make vertical line section on line, the focus of the vertical line section and candidate lane line is subpoint of first breakpoint on candidate lane line；Or
Person, when candidate lane line is camber line lane line, and first lane line is linear vehicle diatom, by the first breakpoint on first lane line
Make vertical line section, the focus of the vertical line section and candidate lane line is subpoint of first breakpoint on candidate lane line；Alternatively, working as
Candidate lane line is that camber line lane line can be according to first lane line and candidate lane when first lane line is camber line lane line
The radian of line determines subpoint of first breakpoint on candidate lane line.
The first breakpoint is being obtained after the subpoint on each candidate lane line, each candidate lane line is being translated,
Until the first breakpoint is overlapped with subpoint of first breakpoint on each candidate lane line, the second breakpoint is obtained to each candidate vehicle
The distance of diatom.
In the present embodiment, by the way of the subpoint on candidate lane line, it can be obtained using with above-mentioned the first breakpoint of acquisition
Take subpoint of second breakpoint on candidate lane line, the distance of the second breakpoint to each candidate lane line be the second breakpoint extremely
The distance of subpoint on corresponding each candidate lane line.
Illustratively, as shown in Fig. 2, first lane line is lane line A, candidate lane line is lane line B and lane line C；
The subpoint a ' and a " of the first breakpoint a are obtained on lane line B and lane line C respectively, and, the subpoint b ' of the second breakpoint b
With b "；Lane line B and lane line C is translated respectively；Until the subpoint a ' on lane line B is overlapped with the first breakpoint a, and, vehicle
Subpoint a " on diatom C is overlapped with the first breakpoint a；Wherein, (a ') shown in Fig. 2, (b ') indicate the lane line B after translation
On a ' and b ', (a "), (b ") indicate a " and b " on the lane line B after translation；Since lane line B and lane line A are complete in Fig. 2
Exactly the same, therefore, when the subpoint a ' on lane line B is overlapped with the first breakpoint a, lane line B is completely coincident with lane line A,
The distance b (b ') for obtaining (b ') of the first breakpoint b on the lane line B to after translating at this time is S1, wherein S1 is equal to 0；Similarly
Ground obtains the second breakpoint b on the lane line C to after translating until the subpoint a " on lane line C is overlapped with the first breakpoint a
The distance b (b ") of (b ") is S2, and one section of lane line thereon after lane line C is translated illustratively is depicted in Fig. 2.
S307 obtains each candidate lane line and first lane according to the second breakpoint to the distance of each candidate lane line
The similarity of line.
In the present embodiment, the distance of the second breakpoint to each candidate lane line is smaller, then each candidate lane line and first
The similarity of lane line is bigger.Obtained by above-mentioned, the second breakpoint b on lane line A respectively to the distance of lane line B sum be S1 and
S2, wherein S2 is greater than S1；That is the similarity between lane line B and lane line A is greater than, the phase between lane line C and lane line A
Like degree.
It is and similar specifically, be stored with the distance of the second breakpoint to each candidate lane line in the correcting device of lane line
The corresponding relationship of degree.The correcting device of lane line is right according to this after obtaining distance of second breakpoint to each candidate lane line
It should be related to, obtain the similarity of each candidate lane line and first lane line.
S308, according to the similarity of each candidate lane line and first lane line, maximum similarity is greater than similar if it exists
Threshold value is spent, then using the corresponding candidate lane line of maximum similarity as target lane line.
The correcting device of lane line after obtaining the similarity of each candidate lane line and first lane line, to similarity by
Arrive greatly it is small be ranked up, if it exists maximum similarity be greater than similarity threshold, then by the corresponding candidate lane line of maximum similarity
As target lane line.
It is envisioned that if maximum similarity is still less than similarity threshold, it is determined that the lane in the reflected value base map
The target lane line of first lane line is not present in line, it is corresponding, processing is not modified to first lane line.
S309 is attached the first breakpoint and the second breakpoint according to target lane line.
It can be carried out using the formula twin target lane line in above-described embodiment with first lane line in the present embodiment
Connection.Specifically, be using on target lane line, lane line corresponding with the lane line lacked on first lane line, to the
One lane line is modified.
Fig. 4 is the lane line schematic diagram after being modified to the lane line A in Fig. 2, as shown in figure 4, in above-mentioned formula one
Connection type be specifically: using lane line a ' corresponding with the lane line ab lacked on first lane line A on target lane line B
B ' is attached first lane line A.Specifically, according to the connection method in above-mentioned connection formula, so that first lane line A
Form complete lane line.
In the present embodiment, first lane line is the lane line that lacks in the middle part of lane line, the of acquisition first lane line
One breakpoint and the second breakpoint；Multiple candidate lane lines are obtained within the scope of the pre-determined distance of the first breakpoint；According to each candidate vehicle
The similarity of diatom and first lane line, specifically, obtaining the mode of similarity are as follows: each candidate lane line is translated,
Until the first breakpoint is overlapped with subpoint of first breakpoint on each candidate lane line；The second breakpoint is obtained to each candidate vehicle
The distance of diatom；According to the second breakpoint to the distance of each candidate lane line, each candidate lane line and first lane line are obtained
Similarity；Further, according to target lane line, the first breakpoint and the second breakpoint are attached.Make it possible to according to anti-
The target lane line penetrated in value base map is modified the lane line of missing, so that unmanned vehicle obtains correct, continuous lane line,
Improve the safety of automatic Pilot.
On the basis of the above embodiments, below with reference to Fig. 5 to the vehicle in the modification method of lane line provided by the invention
The case where missing, is described in detail in the middle part of diatom, and Fig. 5 is the flow diagram of the modification method of lane line provided by the invention
Three, as shown in figure 5, the modification method of lane line provided in this embodiment may include:
S501 obtains the point cloud data on current driving road surface.
S502 projects point cloud data into two-dimensional surface, obtains the corresponding reflected value base map of point cloud data；By reflected value
Image of the base map as current driving road surface.
S503 obtains the first lane line on current driving road surface according to image, and first lane line is lane line one end hair
The lane line of raw missing.
First lane line is lane lines multiple, that first lane line lacks for lane line one end, and such case is applicable in
There are multiple lanes at link entry, each lane line has equal length, if but when one end missing of lane line, so that vehicle
The diatom end and normal lane line are uneven, and image unmanned vehicle is normally taken part in Taoism and travelled.
Fig. 6 is the schematic diagram two for the lane line that the present invention obtains, as shown in fig. 6, the lane line that unmanned vehicle obtains includes vehicle
Diatom A ', lane line B ', lane line C ' and lane line D '；Under normal circumstances, lane line A ', lane line B ', lane line C ' and vehicle
Diatom D ' also length having the same, but due to abrasion or pixel display lamp, so that lane line A ', lane line B ', vehicle
Diatom C's ' lacks once lane line occurs.So that the arrival end of lane line A ', lane line B ', lane line C ' and lane line D ' are not
Together, the normal of image unmanned vehicle is taken part in Taoism and is travelled.
S504, according to the ascending order of the length of the lane line lacked in each first lane line, to multiple first lane lines into
Row sequence.
In the present embodiment, the sequence of the lane line lacked in each first lane line is obtained；As shown in fig. 6, lane line A ',
The length of lane line B ', lane line C ' missing are as follows: lane line A ' is greater than lane line B ', and lane line B ' is greater than lane line C '；I.e.
The ascending order of the length of the lane line lacked in one lane line are as follows: lane line C ', lane line B ', lane line A '.The first final vehicle
The sequence of diatom are as follows: lane line C ', lane line B ', lane line A '.
S505, for each first lane line after sequence, using the previous first lane line of first lane line as the
The target lane line of one lane line.
It in the present embodiment, after obtaining first lane line, needs to obtain the target lane line of each first lane line, uses
Target lane line is modified first lane line；Specifically, for each first lane line after sequence, by first lane line
Target lane line of the previous first lane line as first lane line.As shown in fig. 6, the first lane line in the present embodiment
Sequence are as follows: lane line C ', lane line B ', lane line A ', then the target lane line of lane line A ' be lane line B ', lane line B '
Target lane line be lane line C '.
Wherein, since lane line C ', lane line B ' are to use lane line there are the lane line of lane line missing
The mode that C ', lane line B ' are respectively modified lane line B ', lane line A ', can make lane line B ', lane line A ' with
Lane line C ' is identical.Still fall within the lane line that do not complete.
At this point, for the smallest lane line of length of missing, can be used for lane line C ' and obtain vehicle in above-described embodiment
The mode of the target lane line of diatom C ', is first modified lane line C ', so that the complete lane line of lane line C ' formation, such as
Shown in Fig. 6, the target lane line for obtaining lane line C ' can be lane line D '.Further, using in revised lane
Line C ' is modified lane line B ', so that the complete lane line of lane line B ' formation, correspondingly, using in revised lane
Line B ' is modified lane line A ', so that the complete lane line of lane line A ' formation.
Target lane line is moved to and first lane line overlap by S506.
In the present embodiment, due to each lane line be it is identical, in the target lane for obtaining each first lane line
After line, target lane line can be translated, until target lane line can be realized completely with first lane line overlap to the
The amendment of one lane line.
It is worth noting that, in the present embodiment, it can be first to the shortest first lane of lane line lacked in first lane line
Line is modified, and according still further to the ascending order of the lane line length lacked in a lane line, is successively modified, is made to first lane line
Complete lane line must be obtained.Fig. 7 is the lane line schematic diagram after being modified to the lane line in Fig. 6, as shown in fig. 7, repairing
Lane line C ', lane line B ', lane line A ' after just are respectively formed complete lane line, wherein the dotted portion in Fig. 7 is lane
The lane line of polishing is carried out in line.
In the present embodiment, the specific embodiment in S501-S503 is referred in the S301-S303 in above-described embodiment
Associated description, this will not be repeated here.
In the present embodiment, first lane line is the lane line that lane line one end lacks, and first lane line is multiple, tool
Body, according to the ascending order of the length of the lane line lacked in each first lane line, multiple first lane lines are ranked up；It is right
Each first lane line after sequence, using the previous first lane line of first lane line as the target carriage of first lane line
When being modified, target lane line is moved to and first lane line overlap for diatom.The present embodiment can be according to the mesh in image
Mark lane line is modified the lane line of missing, so that unmanned vehicle obtains correct, continuous lane line, improves automatic Pilot
Safety.
Fig. 8 is the structural schematic diagram one of the correcting device of lane line provided by the invention, as shown in figure 8, the lane line
Correcting device 800 includes: image collection module 801, lane line acquisition module 802, correction module 803.
Image collection module 801, for obtaining the image on current driving road surface.
Lane line obtains module 802, for obtaining the first lane line and target carriage on current driving road surface according to image
Diatom, first lane line be image in occur lane line missing lane line, target lane line are as follows: in image with first lane line
Similarity be greater than similarity threshold lane line.
Correction module 803 is modified the first lane line in image for using target lane line.
The principle and technology that the modification method of the correcting device of lane line provided in this embodiment and above-mentioned lane line is realized
Effect is similar, and therefore not to repeat here.
Optionally, Fig. 9 is the structural schematic diagram two of the correcting device of lane line provided by the invention, as shown in figure 9, the vehicle
The correcting device 800 of diatom further include: similarity obtains module 804.
Optionally, image collection module 801, specifically for obtaining the point cloud data on current driving road surface；By point cloud data
Projection obtains the corresponding reflected value base map of point cloud data into two-dimensional surface；Using reflected value base map as image.
Optionally, lane line obtains module 802, specifically for obtaining the first breakpoint and the second breakpoint of first lane line；
Multiple candidate lane lines are obtained within the scope of the pre-determined distance of the first breakpoint；According to each candidate lane line and first lane line
Similarity, maximum similarity is greater than similarity threshold if it exists, then using the corresponding candidate lane line of maximum similarity as target
Lane line.
Similarity obtains module 804, for translating each candidate lane line, until the first breakpoint and the first breakpoint
Subpoint on each candidate lane line is overlapped；Obtain the distance of the second breakpoint to each candidate lane line；It is disconnected according to second
Point obtains the similarity of each candidate lane line and first lane line to the distance of each candidate lane line.
Optionally, correction module 803 are specifically used for connecting the first breakpoint and the second breakpoint according to target lane line
It connects.
Optionally, lane line obtains module 802, is specifically also used to according to the lane line lacked in each first lane line
The ascending order of length is ranked up multiple first lane lines；For each first lane line after sequence, by first lane line
Target lane line of the previous first lane line as first lane line.
Optionally, correction module 803 are specifically also used to for target lane line being moved to and first lane line overlap.
Figure 10 is the structural schematic diagram three of the correcting device of lane line provided by the invention, the correcting device example of the lane line
It such as can be terminal device, such as smart phone, tablet computer, computer etc..As shown in Figure 10, the correcting device of the lane line
1000 include: memory 1001 and at least one processor 1002.
Memory 1001, for storing program instruction.
Processor 1002, for being performed the modification method for realizing the lane line in the present embodiment, tool in program instruction
Body realization principle can be found in above-described embodiment, and details are not described herein again for the present embodiment.
The correcting device 1000 of the lane line can also include and input/output interface 1003.
Input/output interface 1003 may include independent output interface and input interface, or integrated input and
The integrated interface of output.Wherein, output interface is used for output data, and input interface is used to obtain the data of input, above-mentioned output
Data be the general designation that exports in above method embodiment, the data of input are the general designation inputted in above method embodiment.
The present invention also provides a kind of readable storage medium storing program for executing, it is stored with and executes instruction in readable storage medium storing program for executing, when lane line
When at least one processor of correcting device executes this and executes instruction, when computer executed instructions are executed by processor, realize
The modification method of lane line in above-described embodiment.
The present invention also provides a kind of program product, the program product include execute instruction, this execute instruction be stored in it is readable
In storage medium.At least one processor of the correcting device of lane line can read this from readable storage medium storing program for executing and execute instruction,
At least one processor executes this and executes instruction so that the correcting device of lane line implements what above-mentioned various embodiments provided
The modification method of lane line.
In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit
Letter connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one
In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) or processor (English: processor) execute this hair
The part steps of bright each embodiment the method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory
(English: Read-Only Memory, abbreviation: ROM), random access memory (English: Random Access Memory, letter
Claim: RAM), the various media that can store program code such as magnetic or disk.
In the embodiment of the above-mentioned network equipment or terminal device, it should be appreciated that processor can be central processing unit
(English: Central Processing Unit, referred to as: CPU), it can also be other general processors, digital signal processor
(English: Digital Signal Processor, abbreviation: DSP), specific integrated circuit (English: Application
Specific Integrated Circuit, referred to as: ASIC) etc..General processor can be microprocessor or the processor
It is also possible to any conventional processor etc..Hardware handles can be embodied directly in conjunction with the step of method disclosed in the present application
Device executes completion, or in processor hardware and software module combination execute completion.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to
So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into
Row equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (10)
1. a kind of modification method of lane line characterized by comprising
Obtain the image on current driving road surface；
According to described image, the first lane line and target lane line on the current driving road surface, the first lane are obtained
Line be described image in occur lane line missing lane line, the target lane line are as follows: in described image with first vehicle
The similarity of diatom is greater than the lane line of similarity threshold；
Using the target lane line, the first lane line in described image is modified.
2. the method according to claim 1, wherein the image for obtaining current driving road surface, comprising:
Obtain the point cloud data on the current driving road surface；
The point cloud data is projected into two-dimensional surface, the corresponding reflected value base map of the point cloud data is obtained；
Using the reflected value base map as described image.
3. method according to claim 1 or 2, which is characterized in that the first lane line is to occur to lack in the middle part of lane line
The lane line of mistake obtains the target lane line on the current driving road surface, comprising:
Obtain the first breakpoint and the second breakpoint of the first lane line；
Multiple candidate lane lines are obtained within the scope of the pre-determined distance of first breakpoint；
According to the similarity of each candidate lane line and the first lane line, maximum similarity is greater than the similarity if it exists
Threshold value, then using the corresponding candidate lane line of maximum similarity as the target lane line.
4. according to the method described in claim 3, it is characterized in that, described according to each candidate lane line and the first lane
The similarity of line, and using the corresponding candidate lane line of maximum similarity as the target lane line before, comprising:
Each candidate lane line is translated, until first breakpoint and first breakpoint are described each described
Subpoint on candidate lane line is overlapped；
Obtain the distance of second breakpoint to each candidate lane line；
According to second breakpoint to the distance of each candidate lane line, each candidate lane line and first vehicle are obtained
The similarity of diatom.
5. according to the method described in claim 4, it is characterized in that, described use the target lane line, in described image
First lane line be modified, comprising:
According to the target lane line, first breakpoint and second breakpoint are attached.
6. method according to claim 1 or 2, which is characterized in that the first lane line is that lane line one end occurs to lack
The lane line of mistake, the first lane line be it is multiple, obtain the target lane line on the current driving road surface, comprising:
According to the ascending order of the length of the lane line lacked in each first lane line, multiple first lane lines are carried out
Sequence；
For the first lane line each of after sequence, using the previous first lane line of the first lane line as described in
The target lane line of first lane line.
7. according to the method described in claim 6, it is characterized in that, described use the target lane line, in described image
First lane line be modified, comprising:
The target lane line is moved to and the first lane line overlap.
8. a kind of correcting device of lane line characterized by comprising
Image collection module, for obtaining the image on current driving road surface；
Lane line obtains module, for obtaining the first lane line and target on the current driving road surface according to described image
Lane line, the first lane line are the lane line that lane line missing occurs in described image, the target lane line are as follows: described
It is greater than the lane line of similarity threshold in image with the similarity of the first lane line；
Correction module is modified the first lane line in described image for using the target lane line.
9. a kind of correcting device of lane line characterized by comprising at least one processor and memory；
The memory stores computer executed instructions；
At least one described processor executes the computer executed instructions of the memory storage, so that the amendment of the lane line
Device perform claim requires the described in any item methods of 1-7.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
It executes instruction, when the computer executed instructions are executed by processor, realizes the described in any item methods of claim 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811291589.XA CN109711242B (en) | 2018-10-31 | 2018-10-31 | Lane line correction method, lane line correction device, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811291589.XA CN109711242B (en) | 2018-10-31 | 2018-10-31 | Lane line correction method, lane line correction device, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109711242A true CN109711242A (en) | 2019-05-03 |
| CN109711242B CN109711242B (en) | 2021-04-20 |
Family
ID=66254134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811291589.XA Active CN109711242B (en) | 2018-10-31 | 2018-10-31 | Lane line correction method, lane line correction device, and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109711242B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110866449A (en) * | 2019-10-21 | 2020-03-06 | 北京京东尚科信息技术有限公司 | Method and device for identifying target object in road |
| CN111114544A (en) * | 2019-11-25 | 2020-05-08 | 福瑞泰克智能系统有限公司 | Method and device for determining compensation lane line, movable equipment and storage medium |
| CN111460986A (en) * | 2020-03-30 | 2020-07-28 | 深圳市凯立德科技股份有限公司 | Lane line processing method and device |
| CN112560558A (en) * | 2019-09-26 | 2021-03-26 | 阿里巴巴集团控股有限公司 | Lane line correction method and device and electronic equipment |
| CN114889650A (en) * | 2021-01-11 | 2022-08-12 | 广东科学技术职业学院 | Method for driving vehicle and unmanned vehicle |
| CN115049997A (en) * | 2022-06-07 | 2022-09-13 | 北京百度网讯科技有限公司 | Method and device for generating edge lane line, electronic device and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160269704A1 (en) * | 2015-03-12 | 2016-09-15 | Omron Corporation | Image processing apparatus and image processing method |
| CN108052869A (en) * | 2017-11-23 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Lane detection method, apparatus and computer readable storage medium |
| CN108229406A (en) * | 2018-01-11 | 2018-06-29 | 海信集团有限公司 | A kind of method for detecting lane lines, device and terminal |
| CN108297866A (en) * | 2018-01-03 | 2018-07-20 | 西安交通大学 | A kind of track holding control method of vehicle |
-
2018
- 2018-10-31 CN CN201811291589.XA patent/CN109711242B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160269704A1 (en) * | 2015-03-12 | 2016-09-15 | Omron Corporation | Image processing apparatus and image processing method |
| CN108052869A (en) * | 2017-11-23 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Lane detection method, apparatus and computer readable storage medium |
| CN108297866A (en) * | 2018-01-03 | 2018-07-20 | 西安交通大学 | A kind of track holding control method of vehicle |
| CN108229406A (en) * | 2018-01-11 | 2018-06-29 | 海信集团有限公司 | A kind of method for detecting lane lines, device and terminal |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112560558A (en) * | 2019-09-26 | 2021-03-26 | 阿里巴巴集团控股有限公司 | Lane line correction method and device and electronic equipment |
| CN112560558B (en) * | 2019-09-26 | 2023-04-18 | 阿里巴巴集团控股有限公司 | Lane line correction method and device and electronic equipment |
| CN110866449A (en) * | 2019-10-21 | 2020-03-06 | 北京京东尚科信息技术有限公司 | Method and device for identifying target object in road |
| CN111114544A (en) * | 2019-11-25 | 2020-05-08 | 福瑞泰克智能系统有限公司 | Method and device for determining compensation lane line, movable equipment and storage medium |
| CN111460986A (en) * | 2020-03-30 | 2020-07-28 | 深圳市凯立德科技股份有限公司 | Lane line processing method and device |
| CN114889650A (en) * | 2021-01-11 | 2022-08-12 | 广东科学技术职业学院 | Method for driving vehicle and unmanned vehicle |
| CN115049997A (en) * | 2022-06-07 | 2022-09-13 | 北京百度网讯科技有限公司 | Method and device for generating edge lane line, electronic device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109711242B (en) | 2021-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109711242A (en) | Modification method, device and the storage medium of lane line | |
| Banerjee et al. | Online camera lidar fusion and object detection on hybrid data for autonomous driving | |
| Jörgensen et al. | Monocular 3d object detection and box fitting trained end-to-end using intersection-over-union loss | |
| CN102521884B (en) | Three-dimensional roof reconstruction method based on LiDAR data and ortho images | |
| CN109509230A (en) | A kind of SLAM method applied to more camera lens combined type panorama cameras | |
| US8442307B1 (en) | Appearance augmented 3-D point clouds for trajectory and camera localization | |
| CN108764187A (en) | Extract method, apparatus, equipment, storage medium and the acquisition entity of lane line | |
| CN102750537B (en) | Automatic registering method of high accuracy images | |
| CN110223351B (en) | Depth camera positioning method based on convolutional neural network | |
| CN105303615A (en) | Combination method of two-dimensional stitching and three-dimensional surface reconstruction of image | |
| CN110136182A (en) | Method for registering, device, equipment and the medium of laser point cloud and 2D image | |
| CN104299260A (en) | Contact network three-dimensional reconstruction method based on SIFT and LBP point cloud registration | |
| CN107679537A (en) | A kind of texture-free spatial target posture algorithm for estimating based on profile point ORB characteristic matchings | |
| CN110135323A (en) | Image labeling method, device, system and storage medium | |
| Bastani et al. | Machine-assisted map editing | |
| WO2020237516A1 (en) | Point cloud processing method, device, and computer readable storage medium | |
| CN104408772A (en) | Grid projection-based three-dimensional reconstructing method for free-form surface | |
| CN113409459A (en) | Method, device and equipment for producing high-precision map and computer storage medium | |
| CN105513083A (en) | PTAM camera tracking method and device | |
| CN110084743A (en) | Image mosaic and localization method based on more air strips starting track constraint | |
| CN114969221A (en) | Method for updating map and related equipment | |
| Huang et al. | Measuring the absolute distance of a front vehicle from an in-car camera based on monocular vision and instance segmentation | |
| US11948344B2 (en) | Method, system, medium, equipment and terminal for inland vessel identification and depth estimation for smart maritime | |
| CN111023994B (en) | Grating three-dimensional scanning method and system based on multiple measurement | |
| US11348261B2 (en) | Method for processing three-dimensional point cloud data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |