CN109711372A - A kind of recognition methods of lane line and system, storage medium, server - Google Patents
A kind of recognition methods of lane line and system, storage medium, server Download PDFInfo
- Publication number
- CN109711372A CN109711372A CN201811640595.1A CN201811640595A CN109711372A CN 109711372 A CN109711372 A CN 109711372A CN 201811640595 A CN201811640595 A CN 201811640595A CN 109711372 A CN109711372 A CN 109711372A
- Authority
- CN
- China
- Prior art keywords
- key point
- prediction
- current
- lane line
- ordinate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Image Analysis (AREA)
Abstract
The present embodiments relate to a kind of recognition methods of lane line and system, storage medium, server.Wherein, this method comprises: obtaining the image including lane line, key point corresponding with lane line is chosen from image, set of keypoints is obtained, obtains the slope of key point, polymerize based on key point and slope, the lane line after being polymerize.The technical solution provided through this embodiment, when avoiding in the prior art through image, semantic segmentation or the segmentation of image sample, the technology drawback that the lane line of various radian bends can not be precisely identified, it realizes even there are the bend of various radians, it still can be to the technical effect that lane line is accurately identified.I.e. the present embodiment realizes accurately Lane detection, and then realize the technical effect of safety traffic by the technical solution of " key point identification+angle recognition ".
Description
Technical field
The present embodiments relate to the recognition methods of field of intelligent transportation technology more particularly to a kind of lane line and system,
Storage medium, server.
Background technique
Lane detection based on camera is an important technology in current automatic Pilot and robot field, is based on
Line walking movement may be implemented in correct identification to lane line in environment, automatic driving vehicle or robot or auxiliary determines itself
Position.
Current existing Lane detection is often based on image, semantic segmentation or the segmentation of image sample.Specifically, first
Lane line in image is distinguished with image segmentation algorithm and background, then different lane lines is distinguished.It was distinguishing
Clustering algorithm can be used in Cheng Zhong.It is of course also possible to the method by the way that different lane lines are given with different labels, or
Increase the knowledge method for distinguishing to lane line end point in model.
However, these existing methods are for the lane line in real scene, recognition effect is all not ideal enough, because true
In scene, the angle multiplicity of lane line in image, there are the bends of various radians, and there are roads to import remittance abroad mouth, and are disappearing
It loses at point, the distance of lane line from each other is all very small or even adhesion, these can all influence have algorithm to different at present
The differentiation of lane line, and the fitting of subsequent lane line equation is further influenced, obstruction is brought to the practical application of technology.
Summary of the invention
The technical problem to be solved by the present invention is to provide a kind of lane for the drawbacks described above in the presence of the prior art
The recognition methods of line and system, storage medium, server are low to solve the problems, such as accuracy of identification exists in the prior art.
According to an aspect of an embodiment of the present invention, the embodiment of the invention provides a kind of recognition methods of lane line, institutes
The method of stating includes:
Obtain the image including lane line;
Key point corresponding with the lane line is chosen from described image, obtains set of keypoints;
Obtain the slope of the key point;
It is polymerize based on the key point and the slope, the lane line after being polymerize.
Further, the key point is the corresponding pixel of center line of the lane line;Alternatively,
The key point is the corresponding pixel of edge line of the lane line.
Further, the key point includes corresponding coordinate, which comprises
Obtain the corresponding changing coordinates of current key point and current slope, wherein the changing coordinates are sat corresponding to image
Mark system;
Based on the changing coordinates and the current slope, prediction coordinate is determined;
The coordinate of other key points in the set of keypoints in addition to the current key point and the prediction are sat
Mark is compared, and next key point of the current key point is determined according to comparison result.
Further, the changing coordinates include current abscissa and current ordinate, described to be based on the changing coordinates
With the current slope, determines prediction coordinate, specifically includes:
Prediction ordinate is determined based on the current ordinate;
Prediction abscissa is determined based on the current abscissa and the current slope;
The prediction coordinate is determined based on the prediction ordinate and the prediction abscissa.
Further, described that prediction ordinate is determined based on the current ordinate, it specifically includes:
By the preset first distance of the current ordinate Forward, the prediction ordinate is obtained;
It is described that prediction abscissa is determined based on the current abscissa and the current slope, it specifically includes:
The current abscissa is based on the current slope and moves to right preset second distance, obtains the horizontal seat of prediction
Mark, wherein the second distance is determined according to the first distance and the current slope.
Further, described by the coordinate of other key points in set of keypoints in addition to the current key point and institute
It states prediction coordinate to be compared, and determines next key point of the current key point according to comparison result, specifically include:
The key point to match with the prediction ordinate is screened from the set of keypoints, wherein with the prediction
The key point that ordinate matches is: the absolute value of the difference of the ordinate of key point and the prediction ordinate is preset the
In one threshold value;
The key point to match with the prediction abscissa is chosen from the key point to match with the prediction ordinate,
And the key point is determined as to next key point of the current key point, wherein match with the prediction abscissa
Key point is: the absolute value of the difference of key point abscissa and the prediction abscissa is in preset second threshold.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of server, feature exists
In the server includes:
One or more processors;
Memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes method as described above.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of computer-readable storage mediums
Matter, including instruction make the computer execute method as described above when described instruction is run on computers.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of identification systems of lane line
System, the system comprises: it obtains module, choose module and aggregation module, wherein
The module that obtains is used for: obtaining the image including lane line;
The selection module is used for: being chosen key point corresponding with the lane line from described image, is obtained key point
Set;
The acquisition module is also used to: obtaining the slope of the key point;
The aggregation module is used for: being polymerize based on the key point and the slope, the lane line after being polymerize.
Further, the key point is the corresponding pixel of center line of the lane line;Alternatively,
The key point is the corresponding pixel of edge line of the lane line.
Further, the key point includes corresponding coordinate, and the aggregation module is specifically used for:
Obtain the corresponding changing coordinates of current key point and current slope, wherein the changing coordinates are sat corresponding to image
Mark system;
Based on the changing coordinates and the current slope, prediction coordinate is determined;
The coordinate of other key points in the set of keypoints in addition to the current key point and the prediction are sat
Mark is compared, and next key point of the current key point is determined according to comparison result.
Further, the changing coordinates include current abscissa and current ordinate, and the aggregation module is specifically used for:
Prediction ordinate is determined based on the current ordinate;
Prediction abscissa is determined based on the current abscissa and the current slope;
The prediction coordinate is determined based on the prediction ordinate and the prediction abscissa.
Further, the aggregation module is specifically used for:
By the preset first distance of the current ordinate Forward, the prediction ordinate is obtained;
The current abscissa is based on the current slope and moves to right preset second distance, obtains the horizontal seat of prediction
Mark, wherein the second distance is determined according to the first distance and the current slope.
Further, the aggregation module is specifically used for:
The key point to match with the prediction ordinate is screened from the set of keypoints, wherein with the prediction
The key point that ordinate matches is: the absolute value of the difference of the ordinate of key point and the prediction ordinate is preset the
In one threshold value;
The key point to match with the prediction abscissa is chosen from the key point to match with the prediction ordinate,
And the key point is determined as to next key point of the current key point, wherein match with the prediction abscissa
Key point is: the absolute value of the difference of key point abscissa and the prediction abscissa is in preset second threshold.
The beneficial effect of the embodiment of the present invention is, obtains the image including lane line due to using, and from image
Key point corresponding with lane line is chosen, set of keypoints is obtained, obtains the slope of key point, is carried out based on key point and slope
Polymerization, the technical solution of the lane line after being polymerize avoid in the prior art through image, semantic segmentation or image sample
When segmentation, the technology drawback that can not be precisely identified to the lane line of various radian bends realizes and even exists respectively
, still can be to the technical effect that lane line is accurately identified in the case of the bend of kind of radian, and then realize and ensure to pacify
The technical effect travelled entirely.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the recognition methods of lane line provided in an embodiment of the present invention;
Fig. 2 is a kind of module diagram of the identifying system of lane line provided in an embodiment of the present invention;
Appended drawing reference:
1, module is obtained;2, module is chosen;3, aggregation module.
Specific embodiment
In being described below, for illustration and not for limitation, propose such as specific system structure, interface, technology it
The detail of class, to understand thoroughly the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system and method
Detailed description, in order to avoid unnecessary details interfere description of the invention.
The embodiment of the invention provides a kind of recognition methods of lane line and system, storage medium, server.
According to an aspect of an embodiment of the present invention, the embodiment of the invention provides a kind of recognition methods of lane line.
Referring to Fig. 1, Fig. 1 is a kind of flow diagram of the recognition methods of lane line provided in an embodiment of the present invention.
As shown in Figure 1, this method comprises:
S1: the image including lane line is obtained.
S2: key point corresponding with lane line in image is chosen, set of keypoints is obtained.
Preferably, key point is the corresponding pixel of center line of lane line;Alternatively,
Key point is the corresponding pixel of edge line of lane line.
In this step, described image is identified and detected, the key point relevant to lane line in described image is obtained.?
In some embodiments, the key point can be selected from the center line of lane line or the edge line of lane line.In some implementations
In example, the key point can be the center line of lane line or the edge line corresponding pixel in the picture of lane line.
It specifically, can be by generating one centered on the position of each key point when being chosen to key point
Gaussian Profile, the lable label extracted using based on the Gaussian Profile as key point.And by using KL divergence as loss
Function is learnt and is predicted to the position of key point, to obtain the position (such as coordinate information) of key point.
It is of course also possible to be chosen otherwise to key point.Such as: the corresponding image of lane line is corroded
After processing, then by Hough transform, (quantity of such as center line or edge line, edge line can be chosen on lane line for extraction
One or two) all amount point pixels as key point.
It is, of course, also possible to by first obtaining all foreground points on lane line, then according to (or being arrived from top to bottom by down
On), the higher point of probability value is successively chosen from these foreground points as key point.(or being arrived from top to bottom by down herein
On) be with relative to lane line when driving for.
S3: the slope of key point is obtained.
Before executing the present embodiment, i.e., before S1, further include the steps that initialization.Specifically, it including models or sets
The step of setting database.That is, the slope of the point on lane line is trained and is stored in advance.
In the present embodiment, when extracting key point, the slope of the key point of extraction can be obtained from database.
The slope of key point, that is, key point position lines angle.
S4: being polymerize based on key point and slope, the lane line after being polymerize.
In this step, when obtaining key point and the corresponding slope of key point, can be polymerize based on key point and slope.
Mode due in the prior art, being based only on image, semantic segmentation or the segmentation of image sample carries out lane line
Identification.And during actual travel, the angle of lane line is diversification.Such as there is the bend of various radians, there are roads
Remittance abroad mouth is imported, and at end point, all very small very adhesion of the distance of lane line from each other.If being based only on existing skill
The mode of image, semantic segmentation or the segmentation of image sample in art identifies lane line, then can not accurately identify lane
Line, so as to cause events such as traveling deviations.
And in the present embodiment, after obtaining key point and slope, it is polymerize based on key point and slope, is polymerize
Accurately lane line then can be obtained in lane line afterwards, to realize the technical effect precisely travelled.That is, passing through this reality
The technical solution for applying " the key point identification+angle recognition " of example offer, may be implemented even there are the bends of various radians
In the case of, still lane line can accurately be identified.
Preferably, key point includes corresponding coordinate, and S4 is specifically included:
S4-1: the corresponding changing coordinates of current key point and current slope are obtained, wherein the corresponding coordinate system of coordinate refers to
Image coordinate system, the image coordinate system refer to using the upper left corner of image as origin, are to the right X-axis, are downwards Y-axis.
S4-2: being based on changing coordinates and current slope, determines prediction coordinate.
S4-3: the coordinate of other key points in set of keypoints in addition to current key point and prediction coordinate are compared
Compared with, and determine according to comparison result next key point of current key point.
Now the scheme of S4 is explained in detail in the form of specific example.It should be noted that the data in example
Only exemplary explanation to deepen the understanding to the present embodiment, and should not be understood as the limit to the present embodiment protection scope
It is fixed.
M key point is shared in set of keypoints.Current key point is A, and A includes coordinate A1 and slope A2.Obtain the seat of A
Mark A1 and slope A2.Next key point of A is B (key point is ideal key point).
It is then calculated based on A1 and A2, obtains the corresponding coordinate B1 of B.By (m-1) a key point (i.e. n key point, and
Do not include key point A in n key point) coordinate be compared respectively with B1, obtain comparison result, according to comparison result from
(m-1) a next key point C (key point is the corresponding practical key point of B) is determined in a key point, A and C is connected, to realize
Identification to lane line.
Coordinate B1 can be determined namely based on A1 and A2, (do not include its of key point A in set of keypoints by n key point
Its key point) coordinate be compared respectively with coordinate B1, obtain comparison result, according to comparison result from n key point really
Determine next key point of key point A.
Preferably, changing coordinates include current abscissa and current ordinate, and S4-2 is specifically included:
S4-2-1: prediction ordinate is determined based on current ordinate.
S4-2-2: prediction abscissa is determined based on current abscissa and current slope.
S4-2-3: prediction coordinate is determined based on prediction ordinate and prediction abscissa.
In the present embodiment, respectively according to the current abscissa and the determining prediction abscissa of current ordinate in changing coordinates
With prediction ordinate, then prediction coordinate can be determined based on prediction abscissa and prediction ordinate.
Specifically, S4-2-1 is specifically included: by the preset first distance of current ordinate Forward, obtaining prediction ordinate.
Specifically, S4-2-2 is specifically included: current abscissa being based on current slope and moves to right preset second distance, is obtained
Predict abscissa, wherein second distance is determined according to first distance and current slope.
Prediction coordinate determining in the present embodiment is explained in detail in the form of specific example.Such as:
Prediction ordinate Y is determined based on formula 11, formula 1:
Y1=Y0+a
Prediction abscissa X is determined based on formula 22, formula 2:
X1=(Y1-Y0)/k+X0
Wherein, Y0For the corresponding ordinate of current key point, a is first distance, and k is current slope, X0For second distance.
It is understood that the corresponding coordinate system of coordinate refers to image coordinate system when identifying to lane line.Also
It is to say, is successively to recruit key point down along Y-coordinate (ordinate).Therefore, a is first distance, that is, can be regarded as Δ Y.Tool
The a of body, that is, Δ Y can be set according to the actual situation.Such as 1 unit or 0.5 unit (length unit in practice).
Preferably, S4-3 is specifically included:
S4-3-1: the key point to match with prediction ordinate is screened from set of keypoints, wherein with prediction ordinate
The key point to match is: the absolute value of the difference of the ordinate and prediction ordinate of key point is in preset first threshold.
S4-3-2: choosing the key point to match with prediction abscissa from the key point to match with prediction ordinate,
And the key point is determined as to next key point of current key point, wherein the key point to match with prediction abscissa is:
The absolute value of the difference of key point abscissa and prediction abscissa is in preset second threshold.
Now the scheme of S4-3 is explained in detail based on above-mentioned S4 specific example.Such as:
It is Y that ordinate is chosen from n key point2Key point, ordinate Y2Key point share z.Wherein, |
Y1-Y2| it is less than or equal to first threshold.Then judge the abscissa and X of each key point in z key point1Difference and second
The size of threshold value.If key point C abscissa and X1Difference be less than second threshold, then connect A and C, illustrate A and C is lane
Two adjacent key points on line.
Certainly, there is also another situation, i.e. the abscissa and X of any key point in z key point1Difference not
Less than threshold value.The D point then chosen on lane line is current key point, executes the specific steps of S4.And so on, until obtaining essence
Quasi- identification line.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of server, server packets
It includes:
One or more processors;
Memory, for storing one or more programs;
When one or more programs are executed by one or more processors, so that one or more processors realize institute as above
The method stated.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of computer-readable storage mediums
Matter, including instruction make computer execute method as described above when instruction is run on computers.
Other side according to an embodiment of the present invention, the embodiment of the invention also provides a kind of identification systems of lane line
System.
Referring to Fig. 2, Fig. 2 is a kind of module diagram of the identifying system of lane line provided in an embodiment of the present invention.
As shown in Fig. 2, the system includes: to obtain module 1, module 2 and aggregation module 3 are chosen, wherein
Obtain module with 1 in: obtain including lane line image.
It chooses module 2 to be used for: choosing key point corresponding with lane line from image, obtain set of keypoints.
It obtains module 1 to be also used to: obtaining the slope of key point.
Aggregation module 3 is used for: being polymerize based on key point and slope, the lane line after being polymerize.
In a kind of technical solution in the cards,
Key point is the corresponding pixel of center line of lane line;Alternatively,
Key point is the corresponding pixel of edge line of lane line.
In a kind of technical solution in the cards, key point includes corresponding coordinate, and aggregation module 3 is specifically used
In:
Obtain the corresponding changing coordinates of current key point and current slope, wherein changing coordinates correspond to image coordinate system.
Based on changing coordinates and current slope, prediction coordinate is determined.
The coordinate of other key points in set of keypoints in addition to current key point is compared with prediction coordinate, and
Next key point of current key point is determined according to comparison result.
In a kind of technical solution in the cards, changing coordinates include current abscissa and current ordinate, polymerize mould
Block is specifically used for:
Prediction ordinate is determined based on current ordinate;
Prediction abscissa is determined based on current abscissa and current slope;
Prediction coordinate is determined based on prediction ordinate and prediction abscissa.
In a kind of technical solution in the cards, aggregation module is specifically used for:
By the preset first distance of current ordinate Forward, prediction ordinate is obtained;
By current abscissa be based on current slope move to right preset second distance, obtain prediction abscissa, wherein second away from
From being determined according to first distance and current slope.
In a kind of technical solution in the cards, aggregation module 3 is specifically used for:
Prediction ordinate Y is determined based on formula 11, formula 1:
Y1=Y0+a
Prediction abscissa X is determined based on formula 22, formula 2:
X1=(Y1-Y0)/k+X0
Wherein, Y0For the corresponding ordinate of current key point, a is first distance, and k is current slope, X0For second distance.
In a kind of technical solution in the cards, aggregation module 3 is specifically used for:
The key point to match with prediction ordinate is screened from set of keypoints, wherein match with prediction ordinate
Key point be: the absolute value of the difference of the ordinate of key point and prediction ordinate is in preset first threshold;
Choose the key point to match with prediction abscissa from the key point that matches of prediction ordinate, and by the pass
Key point is determined as next key point of current key point, wherein the key point to match with prediction abscissa is: key point is horizontal
The absolute value of the difference of coordinate and prediction abscissa is in preset second threshold.
The embodiment of the present invention chooses key corresponding with lane line by obtaining the image including lane line from image
Point, obtains set of keypoints, obtains the slope of key point, is polymerize based on key point and slope, the lane after being polymerize
The technical solution of line can not be to various radians when avoiding in the prior art through image, semantic segmentation or the segmentation of image sample
The technology drawback that the lane line of bend is precisely identified realizes even there are the bend of various radians, according to
So can be to the technical effect that lane line is accurately identified, and then realize the technical effect for ensuring to drive safely.
Reader should be understood that in the description of this specification reference term " one embodiment ", " is shown " some embodiments "
The description of example ", " specific example " or " some examples " etc. means specific features described in conjunction with this embodiment or example, structure
Or feature is included at least one embodiment or example of the invention.In the present specification, to the schematic of above-mentioned term
Statement need not be directed to identical embodiment or example.Moreover, specific features, structure or the feature of description can be any
It can be combined in any suitable manner in a or multiple embodiment or examples.In addition, without conflicting with each other, the technology of this field
The feature of different embodiments or examples described in this specification and different embodiments or examples can be combined by personnel
And combination.
It is apparent to those skilled in the art that for convenience of description and succinctly, the dress of foregoing description
The specific work process with unit is set, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
In several embodiments provided herein, 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 unit, only
A kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or components can combine or
Person is desirably integrated into another system, or some features can be ignored or not executed.
Unit may or may not be physically separated as illustrated by the separation member, shown as a unit
Component may or may not be physical unit, it can and it is in one place, or may be distributed over multiple networks
On unit.It can select some or all of unit therein according to the actual needs to realize the mesh of the embodiment of the present invention
'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, is also possible to two or more units and is integrated in one unit.It is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.
It, can if integrated unit is realized in the form of SFU software functional unit and when sold or used as an independent product
To be stored in a computer readable storage medium.Based on this understanding, technical solution of the present invention substantially or
Say that all or part of the part that contributes to existing technology or the technical solution can embody in the form of software products
Out, which 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.) executes all or part of each embodiment method of the present invention
Step.And storage medium above-mentioned include: USB flash disk, it is mobile hard disk, read-only memory (ROM, Read-Only Memory), random
Access various Jie that can store program code such as memory (RAM, Random Access Memory), magnetic or disk
Matter.
It should also be understood that magnitude of the sequence numbers of the above procedures are not meant to execute sequence in various embodiments of the present invention
It is successive, the execution of each process sequence should be determined by its function and internal logic, the implementation without coping with the embodiment of the present invention
Journey constitutes any restriction.
More than, only a specific embodiment of the invention, but scope of protection of the present invention is not limited thereto, and it is any to be familiar with
Those skilled in the art in the technical scope disclosed by the present invention, can readily occur in various equivalent modifications or substitutions,
These modifications or substitutions should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be wanted with right
Subject to the protection scope asked.
Claims (14)
1. a kind of recognition methods of lane line, which is characterized in that the described method includes:
Obtain the image including lane line;
Key point corresponding with the lane line is chosen from described image, obtains set of keypoints;
Obtain the slope of the key point;
It is polymerize based on the key point and the slope, the lane line after being polymerize.
2. the recognition methods of lane line according to claim 1, which is characterized in that
The key point is the corresponding pixel of center line of the lane line;Alternatively,
The key point is the corresponding pixel of edge line of the lane line.
3. the recognition methods of lane line according to claim 1 or 2, which is characterized in that the key point includes right with it
The coordinate answered, which comprises
Obtain the corresponding changing coordinates of current key point and current slope, wherein the changing coordinates correspond to image coordinate system;
Based on the changing coordinates and the current slope, prediction coordinate is determined;
By the coordinate of other key points in the set of keypoints in addition to the current key point and the prediction coordinate into
Row compares, and next key point of the current key point is determined according to comparison result.
4. the recognition methods of lane line according to claim 3, which is characterized in that the changing coordinates include current horizontal seat
Mark and current ordinate, it is described to be based on the changing coordinates and the current slope, it determines prediction coordinate, specifically includes:
Prediction ordinate is determined based on the current ordinate;
Prediction abscissa is determined based on the current abscissa and the current slope;
The prediction coordinate is determined based on the prediction ordinate and the prediction abscissa.
5. the recognition methods of lane line according to claim 4, which is characterized in that
It is described that prediction ordinate is determined based on the current ordinate, it specifically includes:
By the preset first distance of the current ordinate Forward, the prediction ordinate is obtained;
It is described that prediction abscissa is determined based on the current abscissa and the current slope, it specifically includes:
The current abscissa is based on the current slope and moves to right preset second distance, obtains the prediction abscissa,
In, the second distance is determined according to the first distance and the current slope.
6. the recognition methods of lane line according to claim 4, which is characterized in that it is described will be in set of keypoints except described
The coordinate of other key points except current key point is compared with the prediction coordinate, and according to comparison result determination
Next key point of current key point, specifically includes:
The key point to match with the prediction ordinate is screened from the set of keypoints, wherein indulge and sit with the prediction
Marking the key point to match is: the absolute value of the difference of the ordinate of key point and the prediction ordinate is in preset first threshold
In value;
The key point to match with the prediction abscissa is chosen from the key point to match with the prediction ordinate, and will
The key point is determined as next key point of the current key point, wherein the key to match with the prediction abscissa
Point is: the absolute value of the difference of key point abscissa and the prediction abscissa is in preset second threshold.
7. a kind of server, which is characterized in that the server includes:
One or more processors;
Memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as method described in any one of claims 1 to 6.
8. a kind of computer readable storage medium, including instruction, when described instruction is run on computers, make the computer
Execute method according to any one of claim 1 to 6.
9. a kind of identifying system of lane line, which is characterized in that the system comprises: it obtains module, choose module and polymerization mould
Block, wherein
The module that obtains is used for: obtaining the image including lane line;
The selection module is used for: being chosen key point corresponding with the lane line from described image, is obtained set of keypoints;
The acquisition module is also used to: obtaining the slope of the key point;
The aggregation module is used for: being polymerize based on the key point and the slope, the lane line after being polymerize.
10. the identifying system of lane line according to claim 9, which is characterized in that
The key point is the corresponding pixel of center line of the lane line;Alternatively,
The key point is the corresponding pixel of edge line of the lane line.
11. the identifying system of lane line according to claim 9 or 10, which is characterized in that the key point include and its
Corresponding coordinate, the aggregation module are specifically used for:
Obtain the corresponding changing coordinates of current key point and current slope, wherein the changing coordinates correspond to image coordinate system;
Based on the changing coordinates and the current slope, prediction coordinate is determined;
By the coordinate of other key points in the set of keypoints in addition to the current key point and the prediction coordinate into
Row compares, and next key point of the current key point is determined according to comparison result.
12. the identifying system of lane line according to claim 11, which is characterized in that the changing coordinates include current horizontal
Coordinate and current ordinate, the aggregation module are specifically used for:
Prediction ordinate is determined based on the current ordinate;
Prediction abscissa is determined based on the current abscissa and the current slope;
The prediction coordinate is determined based on the prediction ordinate and the prediction abscissa.
13. the identifying system of lane line described in 12 according to claim, which is characterized in that the aggregation module is specifically used
In:
By the preset first distance of the current ordinate Forward, the prediction ordinate is obtained;
The current abscissa is based on the current slope and moves to right preset second distance, obtains the prediction abscissa,
In, the second distance is determined according to the first distance and the current slope.
14. the identifying system of lane line described in 12 according to claim, which is characterized in that the aggregation module is specifically used
In:
The key point to match with the prediction ordinate is screened from the set of keypoints, wherein indulge and sit with the prediction
Marking the key point to match is: the absolute value of the difference of the ordinate of key point and the prediction ordinate is in preset first threshold
In value;
The key point to match with the prediction abscissa is chosen from the key point to match with the prediction ordinate, and will
The key point is determined as next key point of the current key point, wherein the key to match with the prediction abscissa
Point is: the absolute value of the difference of key point abscissa and the prediction abscissa is in preset second threshold.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811640595.1A CN109711372A (en) | 2018-12-29 | 2018-12-29 | A kind of recognition methods of lane line and system, storage medium, server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811640595.1A CN109711372A (en) | 2018-12-29 | 2018-12-29 | A kind of recognition methods of lane line and system, storage medium, server |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109711372A true CN109711372A (en) | 2019-05-03 |
Family
ID=66260390
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811640595.1A Pending CN109711372A (en) | 2018-12-29 | 2018-12-29 | A kind of recognition methods of lane line and system, storage medium, server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109711372A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110472580A (en) * | 2019-08-16 | 2019-11-19 | 北京茵沃汽车科技有限公司 | The method, apparatus of detection parking stall based on full-view image, storage medium |
CN112800902A (en) * | 2021-01-15 | 2021-05-14 | 北京航空航天大学 | Method for fitting track line in complex scene based on image segmentation |
CN114445597A (en) * | 2022-01-28 | 2022-05-06 | 禾多科技(北京)有限公司 | Three-dimensional lane line generation method and device, electronic device and computer readable medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105069415A (en) * | 2015-07-24 | 2015-11-18 | 深圳市佳信捷技术股份有限公司 | Lane line detection method and device |
CN105760812A (en) * | 2016-01-15 | 2016-07-13 | 北京工业大学 | Hough transform-based lane line detection method |
CN106092121A (en) * | 2016-05-27 | 2016-11-09 | 百度在线网络技术(北京)有限公司 | Automobile navigation method and device |
CN107229908A (en) * | 2017-05-16 | 2017-10-03 | 浙江理工大学 | A kind of method for detecting lane lines |
US20180285659A1 (en) * | 2017-03-31 | 2018-10-04 | Here Global B.V. | Method, apparatus, and system for a parametric representation of lane lines |
CN108932472A (en) * | 2018-05-23 | 2018-12-04 | 中国汽车技术研究中心有限公司 | A kind of automatic Pilot running region method of discrimination based on lane detection |
-
2018
- 2018-12-29 CN CN201811640595.1A patent/CN109711372A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105069415A (en) * | 2015-07-24 | 2015-11-18 | 深圳市佳信捷技术股份有限公司 | Lane line detection method and device |
CN105760812A (en) * | 2016-01-15 | 2016-07-13 | 北京工业大学 | Hough transform-based lane line detection method |
CN106092121A (en) * | 2016-05-27 | 2016-11-09 | 百度在线网络技术(北京)有限公司 | Automobile navigation method and device |
US20180285659A1 (en) * | 2017-03-31 | 2018-10-04 | Here Global B.V. | Method, apparatus, and system for a parametric representation of lane lines |
CN107229908A (en) * | 2017-05-16 | 2017-10-03 | 浙江理工大学 | A kind of method for detecting lane lines |
CN108932472A (en) * | 2018-05-23 | 2018-12-04 | 中国汽车技术研究中心有限公司 | A kind of automatic Pilot running region method of discrimination based on lane detection |
Non-Patent Citations (2)
Title |
---|
王晓锦 等: "基于消失点检测与分段直线模型的车道线识别", 《机电一体化》 * |
王珂娜 等: "一种弯道标志线启发式分段搜索算法", 《电子测量与仪器学报》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110472580A (en) * | 2019-08-16 | 2019-11-19 | 北京茵沃汽车科技有限公司 | The method, apparatus of detection parking stall based on full-view image, storage medium |
CN112800902A (en) * | 2021-01-15 | 2021-05-14 | 北京航空航天大学 | Method for fitting track line in complex scene based on image segmentation |
CN112800902B (en) * | 2021-01-15 | 2023-09-05 | 北京航空航天大学 | Image segmentation-based track line fitting method under complex scene |
CN114445597A (en) * | 2022-01-28 | 2022-05-06 | 禾多科技(北京)有限公司 | Three-dimensional lane line generation method and device, electronic device and computer readable medium |
CN114445597B (en) * | 2022-01-28 | 2022-11-11 | 禾多科技(北京)有限公司 | Three-dimensional lane line generation method and device, electronic device and computer readable medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109426801B (en) | Lane line instance detection method and device | |
KR102421855B1 (en) | Method and apparatus of identifying driving lane | |
US10133941B2 (en) | Method, apparatus and device for detecting lane boundary | |
Foucher et al. | Detection and recognition of urban road markings using images | |
Satzoda et al. | Multipart vehicle detection using symmetry-derived analysis and active learning | |
Gupte et al. | Detection and classification of vehicles | |
EP3686779B1 (en) | Method and device for attention-based lane detection without post-processing by using lane mask and testing method and testing device using the same | |
Rauter | Reliable human detection and tracking in top-view depth images | |
US20170344855A1 (en) | Method of predicting traffic collisions and system thereof | |
CN109791603B (en) | Method for capturing objects in an environment region of a motor vehicle by predicting the movement of the objects, camera system and motor vehicle | |
US9501703B2 (en) | Apparatus and method for recognizing traffic sign board | |
KR101992398B1 (en) | Method and Apparatus for Recognizing Road Symbols and Lanes | |
Chougule et al. | Reliable multilane detection and classification by utilizing cnn as a regression network | |
Gomez et al. | Traffic lights detection and state estimation using hidden markov models | |
CN109711372A (en) | A kind of recognition methods of lane line and system, storage medium, server | |
Jang et al. | Multiple exposure images based traffic light recognition | |
KR101508310B1 (en) | Apparatus and method for tracking multiple moving objects in video surveillance system | |
CN109101932B (en) | Multi-task and proximity information fusion deep learning method based on target detection | |
CN110543807A (en) | method for verifying obstacle candidate | |
WO2016002900A1 (en) | Method and device for detecting objects, and computer program product | |
CN111723724B (en) | Road surface obstacle recognition method and related device | |
CN111191611A (en) | Deep learning-based traffic sign label identification method | |
Suhr et al. | Fast symbolic road marking and stop-line detection for vehicle localization | |
Mistry et al. | Survey: Vision based road detection techniques | |
KR20170137273A (en) | Apparatus and Method for Pedestrian Detection using Deformable Part Model |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190503 |