CN109711242B - Lane line correction method, lane line correction device, and storage medium - Google Patents
Lane line correction method, lane line correction device, and storage medium Download PDFInfo
- Publication number
- CN109711242B CN109711242B CN201811291589.XA CN201811291589A CN109711242B CN 109711242 B CN109711242 B CN 109711242B CN 201811291589 A CN201811291589 A CN 201811291589A CN 109711242 B CN109711242 B CN 109711242B
- Authority
- CN
- China
- Prior art keywords
- lane line
- lane
- image
- breakpoint
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a method, a device and a storage medium for correcting a lane line, wherein the method comprises the following steps: acquiring an image of a current driving road surface; according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the missing lane line in the image, and the target lane line is as follows: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value; and correcting the first lane line in the image by adopting the target lane line. The method and the device can correct the missing lane lines according to the target lane lines in the image, so that the unmanned vehicle can obtain correct and continuous lane lines, and the safety of automatic driving is improved.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to a lane line correction method, a lane line correction device and a storage medium.
Background
With the continuous development of automatic driving technology, more and more automatic driving services are beginning to enter people's life circle. The unmanned vehicle can acquire point cloud data of the surrounding environment in the driving process, and the current position of the unmanned vehicle is determined according to the point cloud data; the point cloud data comprises point cloud data of a lane line, and the unmanned vehicle can drive in the lane correctly according to the acquired point cloud data of the lane line.
In the prior art, the problems of unclear lane lines, broken lane lines, or the phenomenon of missing lane lines in an intersection area and the like can be caused due to the problems of shielding, abrasion and the like of the lane lines; and then the unmanned vehicle can not obtain correct and continuous lane lines, and further the accurate driving of the unmanned vehicle is influenced.
Disclosure of Invention
The invention provides a method and a device for correcting a lane line and a storage medium, which can correct the missing lane line according to a target lane line in an image, so that an unmanned vehicle can acquire a correct and continuous lane line, and the safety of automatic driving is improved.
A first aspect of the present invention provides a lane line correction method, including:
acquiring an image of a current driving road surface;
according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the lane line missing in the image, and the target lane line is as follows: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value;
and correcting the first lane line in the image by adopting the target lane line.
Optionally, the acquiring the image of the current driving road surface of the unmanned vehicle includes:
acquiring point cloud data of the current driving road surface;
projecting the point cloud data to a two-dimensional plane to obtain a reflection value base map corresponding to the point cloud data;
and taking the reflection value base map as the image.
Optionally, the first lane line is a lane line in which a missing occurs in the middle of the lane line, and the target lane line on the current driving road is obtained, including:
acquiring a first breakpoint and a second breakpoint of the first lane line;
obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint;
and according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than the similarity threshold, taking the candidate lane line corresponding to the maximum similarity as the target lane line.
Optionally, before the step of taking the candidate lane line corresponding to the maximum similarity as the target lane line according to the similarity between each candidate lane line and the first lane line, the method includes:
translating each candidate lane line until the first breakpoint is overlapped with a projection point of the first breakpoint on each candidate lane line;
acquiring the distance from the second breakpoint to each candidate lane line;
and according to the distance from the second breakpoint to each candidate lane line, obtaining the similarity between each candidate lane line and the first lane line.
Optionally, the correcting the first lane line in the image by using the target lane line includes:
and connecting the first breakpoint and the second breakpoint according to the target lane line.
Optionally, the first lane line is a lane line with one end of the lane line missing, the first lane line is multiple, and the target lane line on the current driving road surface is obtained, including:
sequencing the plurality of first lane lines according to the ascending order of the length of the missing lane line in each first lane line;
for each sorted first lane line, taking a first lane line before the first lane line as a target lane line of the first lane line.
Optionally, the correcting the first lane line in the image by using the target lane line includes:
translating the target lane line to overlap the first lane line.
A second aspect of the present invention provides a lane line correction device, including:
the image acquisition module is used for acquiring an image of a current driving road surface;
a lane line obtaining module, configured to obtain, according to the image, a first lane line and a target lane line on the current driving road surface, where the first lane line is a lane line in the image where a lane line is missing, and the target lane line is: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value;
and the correction module is used for correcting the first lane line in the image by adopting the target lane line.
Optionally, the image obtaining module is specifically configured to obtain point cloud data of the current driving road surface; projecting the point cloud data to a two-dimensional plane to obtain a reflection value base map corresponding to the point cloud data; and taking the reflection value base map as the image.
Optionally, the lane line obtaining module is specifically configured to obtain a first breakpoint and a second breakpoint of the first lane line; obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint; and according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than the similarity threshold, taking the candidate lane line corresponding to the maximum similarity as the target lane line.
Optionally, the apparatus further comprises: a similarity obtaining module;
the similarity obtaining module is configured to translate each candidate lane line until the first breakpoint coincides with a projection point of the first breakpoint on each candidate lane line; acquiring the distance from the second breakpoint to each candidate lane line; and according to the distance from the second breakpoint to each candidate lane line, obtaining the similarity between each candidate lane line and the first lane line.
Optionally, the correction module is specifically configured to connect the first breakpoint and the second breakpoint according to the target lane line.
Optionally, the lane line obtaining module is further specifically configured to sort the plurality of first lane lines according to an ascending order of lengths of missing lane lines in each of the first lane lines; for each sorted first lane line, taking a first lane line before the first lane line as a target lane line of the first lane line.
Optionally, the correction module is specifically further configured to translate the target lane line to overlap with the first lane line.
A third aspect of the present invention provides a lane line correction device, including: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executes computer-executable instructions stored in the memory, so that the lane line correction device executes the lane line correction method.
A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer-executable instructions, which, when executed by a processor, implement the lane line correction method described above.
The invention provides a method, a device and a storage medium for correcting a lane line, wherein the method comprises the following steps: acquiring an image of a current driving road surface; according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the missing lane line in the image, and the target lane line is as follows: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value; and correcting the first lane line in the image by adopting the target lane line. The method and the device can correct the missing lane lines according to the target lane lines in the image, so that the unmanned vehicle can obtain correct and continuous lane lines, and the safety of automatic driving is improved.
Drawings
Fig. 1 is a first schematic flow chart of a lane line correction method provided by the present invention;
FIG. 2 is a first schematic view of a lane line obtained by the present invention;
FIG. 3 is a second flowchart illustrating a lane line correction method according to the present invention;
FIG. 4 is a schematic view of the lane line A in FIG. 2 after being corrected;
fig. 5 is a third schematic flow chart of the lane line correction method provided by the present invention;
FIG. 6 is a second schematic view of a lane line obtained by the present invention;
FIG. 7 is a lane marking diagram after a lane marking correction of FIG. 6;
FIG. 8 is a first schematic structural diagram of a lane line correction apparatus according to the present invention;
fig. 9 is a second schematic structural view of the lane line correction device according to the present invention;
fig. 10 is a third schematic structural view of the lane line correction device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method comprises the steps that a collection vehicle collects point cloud data of all positions of an unmanned area in advance, and the positions and the point cloud data are corresponding to obtain a high-precision map; the unmanned vehicle can acquire point cloud data of the surrounding environment in the driving process, and the position of the unmanned vehicle is acquired by matching the currently acquired point cloud data with the point cloud data in the high-precision map. Further, the point cloud data collected by the unmanned vehicle comprises point cloud data of the lane line, the unmanned vehicle identifies the lane line according to the acquired point cloud data of the lane line, and the unmanned vehicle drives according to the identified lane line.
However, the lane line is not clear and broken due to abrasion, shielding and the like; or, the lane line at the road entrance is lost, and the like, so that the unmanned vehicle cannot acquire the correct and continuous lane line, and further cannot normally run.
The term of the present invention is defined as:
point cloud data: in the prior art, a laser scanning mode is mostly adopted to obtain point cloud data of an environment; when a laser beam irradiates the surface of an object, the reflected laser beam carries information such as direction, distance and the like. When the laser beam is scanned along a certain trajectory, the reflected laser spot information is recorded while scanning, and since the scanning is extremely fine, a large number of laser spots can be obtained, and thus, laser point cloud data of an object can be formed. The point cloud data is a collection of a large number of point clouds at the target surface features.
Bottom graph of reflection value: the point cloud obtained according to the laser measurement principle comprises three-dimensional coordinates (XYZ) and laser reflection information; a point cloud obtained according to photogrammetry principles, comprising three-dimensional coordinates (XYZ); and combining laser measurement and photogrammetry principles to obtain a point cloud comprising three-dimensional coordinates (XYZ) and laser reflection information. And representing the point cloud data according to the reflection information in the point cloud, and acquiring a reflection value base map corresponding to the point cloud data.
Semantic segmentation: each pixel in the acquired image is classified, i.e. the pixel of what object each pixel in the image belongs to is acquired. In the prior art, a convolutional neural network mode is mostly adopted to classify pixels.
Fig. 1 is a first schematic flow chart of a lane line correction method provided by the present invention, and an execution main body of the method flow shown in fig. 1 may be a lane line correction device, where the lane line correction device may be disposed on an unmanned vehicle, or may be disposed integrally with a server. As shown in fig. 1, the method for correcting a lane line provided in this embodiment may include:
and S101, acquiring an image of the current running road surface.
In this embodiment, the correction device for the lane line provided on the unmanned vehicle may acquire an image of the current driving road surface of the unmanned vehicle.
One way to obtain the image of the current driving road surface is as follows: the unmanned vehicle is provided with a shooting device, and the shooting device can shoot the road surface where the unmanned vehicle runs in real time; the shooting device can be arranged separately from the lane line correcting device, and can send the image to the lane line correcting device after the shooting device shoots the image of the current driving road surface; the shooting device can also be integrated with a lane line correcting device, and the lane line correcting device directly acquires the image of the current running road surface of the unmanned vehicle.
Another way to obtain the image of the current driving road surface is as follows: the correction device of the lane line can collect the point cloud data of the current driving road surface of the unmanned vehicle, and because the point cloud data comprises the reflection intensity information of a collection object, the point cloud data of the current driving road surface is converted into a reflection value base map according to a point cloud data conversion method in the prior art, and the reflection value base map is used as an image of the current driving road surface.
S102, according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the lane line missing in the image, and the target lane line is as follows: and the similarity between the image and the first lane line is greater than the lane line with the similarity threshold value.
Specifically, the image obtained by the lane line correction device is composed of a plurality of pixel blocks, and the object corresponding to each pixel block may be different. In the lane line correction device in this embodiment, a classification model is stored in advance, and the classification model is used to classify pixel blocks in an image, that is, perform semantic segmentation to obtain a lane line in the image; the classification model may be obtained by pre-training in a convolutional neural network manner in the prior art, and the obtaining manner of the classification model is not limited in this embodiment.
In this embodiment, after the lane line in the image is obtained according to the classification model, a first lane line and a target lane line on the current driving road surface are obtained. The image obtained by the lane line correction device includes a plurality of lane lines, and the lane lines may be the same or different lane lines. Fig. 2 is a schematic view of a lane line obtained by the present invention, and as shown in fig. 2, an image obtained by the lane line correction device includes three lane lines, which are a lane line a, a lane line B, and a lane line C.
Specifically, the first lane line is a lane line in which a lane line is missing in the image. The lane line correcting device can acquire the lane lines in the image and pixel blocks corresponding to the lane lines according to the semantic segmentation mode; according to the continuity of the pixel blocks, the first lane line in the image, that is, the pixel blocks corresponding to the lane line are discontinuous, is obtained, as shown by the lane line a in fig. 2.
In this embodiment, after the correction device of the lane line acquires the first lane line, a target lane line is acquired from lane lines other than the first lane line among the plurality of lane lines in the image; specifically, the target lane line is: and the similarity between the image and the first lane line is greater than the lane line with the similarity threshold value.
The similarity between the lane line and the first lane line can be obtained according to the radian, the orientation, the type and the like of the lane line; for example, if the difference between the radian of the lane line and the radian of the first lane line is smaller than the difference threshold, the lane line is determined to be the target lane line of the first lane line; further, when the similarity between the lane line and the first lane line is obtained, the orientation and the type of the obtained target lane line and the first lane line also need to be the same. The orientation of the lane line refers to the direction of the circle center corresponding to the arc formed by the lane line, and the type of the lane line may be a solid line or a dotted line. And acquiring the similarity between the plurality of lane lines and the first lane line according to the similarity acquiring mode, and taking the lane line larger than the similarity threshold value as the target lane line. When there are a plurality of lane lines greater than the similarity threshold, the lane line corresponding to the maximum similarity is considered as the target lane line.
It is conceivable that, when the first lane line is a straight line segment, the target lane line may be acquired according to whether or not it is parallel to the first lane line. In this embodiment, for different lane line types, the target lane line may be obtained in a corresponding manner as long as the similarity between the obtained target lane line and the first lane line is greater than the similarity threshold.
For example, as shown in fig. 2, it is determined that the lane line B is the target lane line of the lane line a by calculating the similarity between the lane line B and the lane line C and the lane line a, respectively.
S103, correcting the first lane line in the image by adopting the target lane line.
In this embodiment, the similarity between the acquired target lane line and the first lane line is greater than the similarity threshold. One way to correct the first lane line in the image may be: and copying and translating the pixel blocks corresponding to the target lane line to the position of the first lane line, and adopting the target lane line to make up for the missing part of the first lane line and correcting so that the first lane line forms a complete lane line. It is conceivable that the portion where the first lane line is missing is corrected by performing translation using a portion of the lane line on the target lane line.
Another way may be: acquiring a breakpoint of a first lane line, projecting the breakpoint on a target lane line, and acquiring a projection point corresponding to the breakpoint on the target lane line; exemplarily, as shown in fig. 2, there are two breakpoints on the lane line a, which are a breakpoint a and a breakpoint b respectively; according to a projection mode in the prior art, projection points a 'and B' corresponding to the break point a and the break point B are obtained on the lane line B, and the lane line A and the lane line B are connected by adopting a formula, so that the lane line A forms a complete lane line.
The target lane line is connected with the first lane line according to a preset connection formula, so that the first lane line forms a complete lane line, and specifically, the connection formula can be shown as the following formula I:
wherein, XaIs the abscissa, Y, of the first breakpoint a on the first lane line (lane line A)aIs the ordinate, X, of the first breakpoint a on the lane line AbIs the abscissa, Y, of the second breakpoint b on the lane line AbIs the ordinate, Y, of the second breakpoint b on the lane line Aa'is the ordinate, Y, of the projection point a' of the first breakpoint a on the target lane line (lane line B)b' is the second breakpoint B on the lane line BX 'is an abscissa of any one point on a projected line segment a' B 'on the lane line B, Y' is an ordinate of any one point on a projected line segment a 'B' on the lane line B, and Y is a ordinate of any one point on a projected line segment a 'B' on the lane line B connected to the lane line a.
Specifically, connecting the target lane line with the first lane line means connecting and compensating the missing lane line by using the lane line corresponding to the lane line missing from the first lane line on the target lane line. The connection mode avoids the steps of copying, translating, connecting and the like of the lane line, and can directly connect the target lane line with the first lane line according to the projection point.
The method for correcting the lane line provided by the embodiment comprises the following steps: acquiring an image of a current driving road surface; according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the missing lane line in the image, and the target lane line is as follows: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value; and correcting the first lane line in the image by adopting the target lane line. The lane line correction method provided by the embodiment can correct the missing lane line according to the target lane line in the image, so that the unmanned vehicle can acquire the correct and continuous lane line, and the safety of automatic driving is improved.
The absence of lane lines can be classified as: lane line mid-miss, lane line a shown in fig. 2; or one end of the lane line is missing, the condition is suitable for the situation that a plurality of lanes are arranged at the road entrance, each lane line has the same length, if one end of the lane line is missing, the end of the lane line is not aligned with the normal lane line, and the image is used for normal lane entering and running of the unmanned vehicle.
On the basis of the foregoing embodiment, the following describes in detail a situation that the middle portion of the lane line is missing in the lane line correction method provided by the present invention with reference to fig. 3, where fig. 3 is a schematic flow chart of the lane line correction method provided by the present invention, and as shown in fig. 3, the lane line correction method provided by this embodiment may include:
s301, point cloud data of the current driving road surface are obtained.
The unmanned vehicle is provided with a point cloud data acquisition device, and the point cloud data acquisition device can scan the road surface on which the unmanned vehicle runs in real time to acquire the point cloud data of the current running road surface; the point cloud data acquiring device in this embodiment may be a laser radar system in the prior art.
Specifically, the point cloud data acquisition device may be provided separately from the lane line correction device, and when the point cloud data acquisition device acquires the point cloud data of the current driving road surface, the point cloud data of the current driving road surface may be sent to the lane line correction device; the point cloud data acquisition device can also be integrated with a lane line correction device, and the lane line correction device directly acquires the point cloud data of the current driving road surface.
S302, projecting the point cloud data to a two-dimensional plane to obtain a reflection value base map corresponding to the point cloud data; and taking the reflection value base map as an image of the current running road surface.
The point cloud data is in a three-dimensional state, the three-dimensional point cloud data is projected to a two-dimensional plane according to a projection mode in the prior art, and a reflection value base map corresponding to the point cloud data is obtained, and specifically, the reflection value base map comprises reflection value intensity information of each pixel block. In the present embodiment, the reflection value base map is used as an image of the current traveling road surface.
And S303, acquiring a first lane line on the current driving road surface according to the image, wherein the first lane line is a lane line with a missing middle part of the lane line.
Wherein, the image is a reflection value base map of the current driving road surface; the lane line correction device stores a classification model in advance, specifically, the classification model is obtained by using the characteristics of the pixel blocks in the reflection value base map as training parameters, and the classification model can classify each pixel block in the reflection value base map to obtain an object corresponding to each pixel block.
Specifically, a plurality of lane lines in the reflection value base map can be obtained according to the classification model; acquiring a first lane line according to whether pixel blocks corresponding to the lane line are continuous or not, wherein the first lane line is a lane line with a lane line loss; in this embodiment, the first lane line missing in the middle of the lane line is obtained according to the type of the lane line missing.
S304, acquiring a first breakpoint and a second breakpoint of the first lane line.
The first lane line in this embodiment is the lane line that the middle part lacks, and first lane line constitutes for two broken string lane lines promptly, and the middle part has two adjacent breakpoints, is first breakpoint and second breakpoint. In this embodiment, the first breakpoint and the second breakpoint may be obtained in the pixel block where the first lane line is missing, according to whether the first lane line is continuously obtained by determining whether the pixel blocks belonging to the same lane line are consecutive.
Illustratively, as shown in fig. 2, there are two breakpoints on the lane line a, i.e., a breakpoint a and a breakpoint b.
S305, a plurality of candidate lane lines are obtained within the preset distance range of the first breakpoint.
For example, as shown in fig. 2, the first breakpoint may be breakpoint a or breakpoint b; and obtaining a plurality of candidate lane lines within the preset range of the first breakpoint. The preset range of the first breakpoint can be a preset distance or a preset angle; correspondingly, the candidate lane line may be a lane line within a preset distance range from the first breakpoint; alternatively, the candidate lane line may be a lane line within a range of a preset angle from the first breakpoint.
S306, translating each candidate lane line until the first break point is overlapped with the projection point of the first break point on each candidate lane line, and obtaining the distance from the second break point to each candidate lane line.
In this embodiment, the first breakpoint is projected on each candidate lane line, and a projection point corresponding to the first breakpoint is obtained on each candidate lane line; specifically, when the candidate lane line is a straight lane line, the first breakpoint is made into a vertical line segment on the candidate lane line, and the focus of the vertical line segment and the candidate lane line is the projection point of the first breakpoint on the candidate lane line; or when the candidate lane line is an arc lane line and the first lane line is a straight lane line, making a vertical line segment of the first breakpoint on the first lane line, wherein the focus of the vertical line segment and the candidate lane line is the projection point of the first breakpoint on the candidate lane line; or, when the candidate lane line is an arc lane line and the first lane line is an arc lane line, the projection point of the first breakpoint on the candidate lane line may be determined according to the radians of the first lane line and the candidate lane line.
And after the projection point of the first breakpoint on each candidate lane line is obtained, translating each candidate lane line until the first breakpoint is overlapped with the projection point of the first breakpoint on each candidate lane line, and obtaining the distance from the second breakpoint to each candidate lane line.
In this embodiment, the projection point of the second breakpoint on the candidate lane line may be obtained in a manner similar to the above-described manner of obtaining the projection point of the first breakpoint on the candidate lane line, where a distance from the second breakpoint to each candidate lane line is a distance from the second breakpoint to the corresponding projection point on each candidate lane line.
Illustratively, as shown in fig. 2, the first lane line is lane line a, and the candidate lane lines are lane line B and lane line C; respectively acquiring projection points a 'and a' of a first breakpoint a and projection points B 'and B' of a second breakpoint B on a lane line B and a lane line C; respectively translating the lane line B and the lane line C; until the projection point a 'on the lane line B coincides with the first breakpoint a, and the projection point a' on the lane line C coincides with the first breakpoint a; wherein (a '), (B') shown in fig. 2 denote a 'and B' on the lane line B after the translation, and (a "), (B") denote a "and B" on the lane line B after the translation; in fig. 2, since the lane line B is completely the same as the lane line a, when the projection point a ' on the lane line B is overlapped with the first breakpoint a, the lane line B is completely overlapped with the lane line a, and at this time, the distance B (B ') from the first breakpoint B to (B ') on the lane line B after translation is obtained is S1, where S1 is equal to 0; similarly, until the projected point a ″ on the lane line C coincides with the first breakpoint a, the distance b (b ″) from the second breakpoint b to (b ″) on the translated lane line C is obtained as S2, and a segment of the lane line on the translated lane line C is exemplarily depicted in fig. 2.
S307, according to the distance from the second breakpoint to each candidate lane line, the similarity between each candidate lane line and the first lane line is obtained.
In this embodiment, the smaller the distance from the second breakpoint to each candidate lane line, the greater the similarity between each candidate lane line and the first lane line. From the above, the distances from the second break point B on the lane line a to the lane lines B and B are S1 and S2, respectively, wherein S2 is greater than S1; that is, the similarity between the lane line B and the lane line a is greater than the similarity between the lane line C and the lane line a.
Specifically, the device for correcting the lane line stores a correspondence between the distance from the second breakpoint to each candidate lane line and the similarity. And after the correction device of the lane line obtains the distance from the second breakpoint to each candidate lane line, obtaining the similarity between each candidate lane line and the first lane line according to the corresponding relation.
And S308, according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than the similarity threshold, taking the candidate lane line corresponding to the maximum similarity as the target lane line.
And after the similarity between each candidate lane line and the first lane line is obtained, the lane line correcting device sorts the similarities from large to small, and if the maximum similarity is larger than a similarity threshold, the candidate lane line corresponding to the maximum similarity is taken as the target lane line.
It is conceivable that, if the maximum similarity is still smaller than the similarity threshold, it is determined that the target lane line of the first lane line does not exist in the lane lines in the reflection value base map, and accordingly, the first lane line is not corrected.
S309, connecting the first breakpoint and the second breakpoint according to the target lane line.
In this embodiment, the formula in the above embodiment may be used to connect the target lane line and the first lane line. Specifically, the first lane line is corrected by using a lane line corresponding to the missing lane line on the target lane line.
Fig. 4 is a schematic diagram of the lane line a in fig. 2 after being corrected, and as shown in fig. 4, the connection manner in the first formula is specifically: and connecting the first lane line A by adopting a lane line a 'B' corresponding to the missing lane line ab on the target lane line B. Specifically, according to the connection method in the connection formula, the first lane line a forms a complete lane line.
In this embodiment, the first lane line is a lane line in which the middle of the lane line is missing, and a first breakpoint and a second breakpoint of the first lane line are obtained; obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint; according to the similarity between each candidate lane line and the first lane line, specifically, the method for obtaining the similarity is as follows: translating each candidate lane line until the first break point is overlapped with the projection point of the first break point on each candidate lane line; acquiring the distance from the second breakpoint to each candidate lane line; according to the distance from the second breakpoint to each candidate lane line, the similarity between each candidate lane line and the first lane line is obtained; and further, connecting the first breakpoint and the second breakpoint according to the target lane line. The missing lane lines can be corrected according to the target lane lines in the reflection value base map, so that the unmanned vehicle can obtain correct and continuous lane lines, and the safety of automatic driving is improved.
On the basis of the foregoing embodiment, the following describes in detail a situation that the middle portion of the lane line is missing in the lane line correction method provided by the present invention with reference to fig. 5, where fig. 5 is a schematic flow chart of the lane line correction method provided by the present invention, and as shown in fig. 5, the lane line correction method provided by this embodiment may include:
s501, point cloud data of the current driving road surface are obtained.
S502, projecting the point cloud data to a two-dimensional plane to obtain a reflection value base map corresponding to the point cloud data; and taking the reflection value base map as an image of the current running road surface.
S503, according to the image, a first lane line on the current driving road surface is obtained, wherein the first lane line is a lane line with one end of the lane line missing.
The first lane lines are multiple, the first lane lines are lane lines with one ends of the lane lines missing, the condition is suitable for the situation that a plurality of lanes exist at a road entrance, each lane line has the same length, if one ends of the lane lines are missing, the ends of the lane lines are not aligned with the normal lane lines, and the image of normal lane entering and driving of the unmanned vehicle is obtained.
Fig. 6 is a schematic view of a lane line obtained by the present invention, and as shown in fig. 6, the lane line obtained by the unmanned vehicle includes a lane line a ', a lane line B', a lane line C 'and a lane line D'; normally, the lane line a ', the lane line B ', the lane line C ' and the lane line D ' have the same length, but due to abrasion or pixel display lamps, the lane line a ', the lane line B ', and the lane line C ' are once lost. So that the entrance ends of the lane line A ', the lane line B', the lane line C 'and the lane line D' are not uniform, and the image unmanned vehicle normally enters the lane and runs.
S504, sorting the plurality of first lane lines according to the ascending order of the length of the missing lane line in each first lane line.
In this embodiment, the order of missing lane lines in each first lane line is obtained; as shown in fig. 6, the missing length of lane line a ', lane line B ', and lane line C ' is: the lane line A 'is greater than the lane line B', and the lane line B 'is greater than the lane line C'; namely, the ascending order of the length of the missing lane line in the first lane line is as follows: lane line C ', lane line B ', lane line a '. The final first lane line sequence is: lane line C ', lane line B ', lane line a '.
And S505, regarding each sorted first lane line, taking the previous first lane line of the first lane line as a target lane line of the first lane line.
In this embodiment, after the first lane line is obtained, a target lane line of each first lane line needs to be obtained, and the first lane line is corrected by using the target lane line; specifically, for each first lane line after the sorting, a first lane line preceding the first lane line is used as a target lane line of the first lane line. As shown in fig. 6, the first lane line in this embodiment is ordered as: the lane line C ', the lane line B ', and the lane line a ', the target lane line of the lane line a ' is the lane line B ', and the target lane line of the lane line B ' is the lane line C '.
The lane line C ' and the lane line B ' are lane lines with missing lane lines, so that the lane line B ' and the lane line a ' are completely the same as the lane line C ' by respectively correcting the lane line B ' and the lane line a '. Still belonging to an incomplete lane line.
At this time, for the lane line C ', which is the lane line with the minimum missing length, the lane line C' may be corrected first by adopting the manner of obtaining the target lane line of the lane line C 'in the above embodiment, so that the lane line C' forms a complete lane line, as shown in fig. 6, the target lane line obtained from the lane line C 'may be the lane line D'. Further, the lane line B 'is corrected by the corrected lane line C' so that the lane line B 'forms a complete lane line, and correspondingly, the lane line a' is corrected by the corrected lane line B 'so that the lane line a' forms a complete lane line.
S506, translating the target lane line to be overlapped with the first lane line.
In this embodiment, since each lane line is identical, after the target lane line of each first lane line is obtained, the target lane line may be translated until the target lane line is completely overlapped with the first lane line, and the correction of the first lane line may be implemented.
It should be noted that, in this embodiment, the first lane line with the shortest missing lane line in the first lane line may be corrected first, and then the first lane line may be corrected sequentially according to the ascending order of the lengths of the missing lane lines in the first lane line, so as to obtain a complete lane line. Fig. 7 is a schematic view of the lane lines in fig. 6 after the lane lines are corrected, and as shown in fig. 7, the corrected lane lines C ', B ', a ' all form complete lane lines, wherein the dotted line portion in fig. 7 is the lane line in which the lane lines are filled.
In this embodiment, the specific implementation in S501-S503 may refer to the related description in S301-S303 in the above embodiment, which is not described herein again.
In this embodiment, the first lane lines are lane lines with one missing end, and the first lane lines are multiple, specifically, the multiple first lane lines are sorted according to the ascending order of the length of the missing lane line in each first lane line; and for each sorted first lane line, taking the first lane line which is one first lane line ahead of the first lane line as a target lane line of the first lane line, and translating the target lane line to be overlapped with the first lane line when correcting. According to the embodiment, the missing lane lines can be corrected according to the target lane lines in the image, so that the unmanned vehicle can obtain correct and continuous lane lines, and the safety of automatic driving is improved.
Fig. 8 is a schematic structural diagram of a lane line correction device according to the present invention, and as shown in fig. 8, the lane line correction device 800 includes: an image acquisition module 801, a lane line acquisition module 802, and a correction module 803.
An image obtaining module 801, configured to obtain an image of a current driving road surface.
The lane line acquiring module 802 is configured to acquire a first lane line and a target lane line on a current driving road according to the image, where the first lane line is a lane line in the image where the lane line is missing, and the target lane line is: and the similarity between the image and the first lane line is greater than the lane line with the similarity threshold value.
And a correcting module 803, configured to correct the first lane line in the image by using the target lane line.
The principle and technical effect of the lane line correction device provided in this embodiment are similar to those of the lane line correction method, and are not described herein again.
Optionally, fig. 9 is a schematic structural diagram of a lane line correction device provided in the present invention, and as shown in fig. 9, the lane line correction device 800 further includes: a similarity obtaining module 804.
Optionally, the image obtaining module 801 is specifically configured to obtain point cloud data of a current driving road surface; projecting the point cloud data to a two-dimensional plane to obtain a reflection value base map corresponding to the point cloud data; and taking the reflection value base map as an image.
Optionally, the lane line obtaining module 802 is specifically configured to obtain a first breakpoint and a second breakpoint of the first lane line; obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint; and according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than the similarity threshold, taking the candidate lane line corresponding to the maximum similarity as the target lane line.
A similarity obtaining module 804, configured to translate each candidate lane line until the first break point coincides with a projection point of the first break point on each candidate lane line; acquiring the distance from the second breakpoint to each candidate lane line; and according to the distance from the second breakpoint to each candidate lane line, obtaining the similarity between each candidate lane line and the first lane line.
Optionally, the correcting module 803 is specifically configured to connect the first breakpoint and the second breakpoint according to the target lane line.
Optionally, the lane line obtaining module 802 is further configured to sort the plurality of first lane lines according to an ascending order of the length of the missing lane line in each first lane line; and for each first lane line after sequencing, taking the first lane line which is one first lane line before the first lane line as a target lane line of the first lane line.
Optionally, the correcting module 803 is further configured to translate the target lane line to overlap with the first lane line.
Fig. 10 is a schematic structural diagram of a lane line correction device provided in the present invention, where the lane line correction device may be, for example, a terminal device, such as a smart phone, a tablet computer, a computer, and the like. As shown in fig. 10, the lane line correction device 1000 includes: a memory 1001 and at least one processor 1002.
The processor 1002 is configured to implement the lane line correction method in this embodiment when the program instruction is executed, and the specific implementation principle may refer to the foregoing embodiment, which is not described herein again.
The lane line correction apparatus 1000 may further include an input/output interface 1003.
The input/output interface 1003 may include a separate output interface and input interface, or may be an integrated interface that integrates input and output. The output interface is used for outputting data, the input interface is used for acquiring input data, the output data is a general name output in the method embodiment, and the input data is a general name input in the method embodiment.
The invention further provides a readable storage medium, wherein the readable storage medium stores an execution instruction, and when the execution instruction is executed by at least one processor of the lane line correction device, the computer execution instruction is executed by the processor to realize the lane line correction method in the above embodiment.
The present invention also provides a program product comprising execution instructions stored in a readable storage medium. The at least one processor of the lane line correction apparatus may read the execution instruction from the readable storage medium, and the execution of the execution instruction by the at least one processor causes the lane line correction apparatus to implement the lane line correction method provided in the various embodiments described above.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In the foregoing embodiments of the network device or the terminal device, it should be understood that the Processor may be a Central Processing Unit (CPU), or may be other general-purpose processors, Digital Signal Processors (DSP), Application Specific Integrated Circuits (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present application may be embodied directly in a hardware processor, or in a combination of the hardware and software modules in the processor.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. A lane line correction method is characterized by comprising the following steps:
acquiring point cloud data of a current driving road surface;
projecting the point cloud data to a two-dimensional plane, and acquiring a reflection value base map corresponding to the point cloud data, wherein the reflection value base map is an image of a current driving road surface;
according to the image, a first lane line and a target lane line on the current driving road surface are obtained, the first lane line is the lane line with the lane line missing in the image, and the target lane line is as follows: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value;
correcting a first lane line in the image by adopting the target lane line; if the first lane line is a lane line with a missing lane line in the middle, acquiring a target lane line on the current driving road, including:
acquiring a first breakpoint and a second breakpoint of the first lane line;
obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint;
according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than the similarity threshold, taking the candidate lane line corresponding to the maximum similarity as the target lane line;
alternatively, the first and second electrodes may be,
if the first lane line is a lane line with one end of the lane line missing, the first lane lines are multiple, and the target lane line on the current driving road surface is obtained, including:
sequencing the plurality of first lane lines according to the ascending order of the length of the missing lane line in each first lane line;
for each sorted first lane line, taking a first lane line before the first lane line as a target lane line of the first lane line.
2. The method according to claim 1, wherein before the step of taking the candidate lane line corresponding to the maximum similarity as the target lane line according to the similarity between each candidate lane line and the first lane line, the method comprises:
translating each candidate lane line until the first breakpoint is overlapped with a projection point of the first breakpoint on each candidate lane line;
acquiring the distance from the second breakpoint to each candidate lane line;
and according to the distance from the second breakpoint to each candidate lane line, obtaining the similarity between each candidate lane line and the first lane line.
3. The method of claim 2, wherein the correcting the first lane line in the image using the target lane line comprises:
and connecting the first breakpoint and the second breakpoint according to the target lane line.
4. The method of claim 1, wherein the correcting the first lane line in the image using the target lane line comprises:
translating the target lane line to overlap the first lane line.
5. A lane line correction apparatus, comprising:
the image acquisition module is used for acquiring point cloud data of a current driving road surface; projecting the point cloud data to a two-dimensional plane, and acquiring a reflection value base map corresponding to the point cloud data, wherein the reflection value base map is an image of a current driving road surface;
a lane line obtaining module, configured to obtain, according to the image, a first lane line and a target lane line on the current driving road surface, where the first lane line is a lane line in the image where a lane line is missing, and the target lane line is: the similarity between the image and the first lane line is larger than the lane line with the similarity threshold value;
the correction module is used for correcting the first lane line in the image by adopting the target lane line;
wherein the content of the first and second substances,
the lane line acquisition module is specifically used for acquiring a first breakpoint and a second breakpoint of a first lane line; obtaining a plurality of candidate lane lines within a preset distance range of the first breakpoint; according to the similarity between each candidate lane line and the first lane line, if the maximum similarity is larger than a similarity threshold value, taking the candidate lane line corresponding to the maximum similarity as a target lane line;
the lane line obtaining module is specifically configured to sort the plurality of first lane lines according to an ascending order of the length of the missing lane line in each of the first lane lines; for each sorted first lane line, taking a first lane line before the first lane line as a target lane line of the first lane line.
6. A lane line correction apparatus, comprising: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executing the computer-executable instructions stored by the memory causes the lane line modification apparatus to perform the method of any of claims 1-4.
7. A computer-readable storage medium having computer-executable instructions stored thereon which, when executed by a processor, implement the method of any one of claims 1-4.
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 CN109711242A (en) | 2019-05-03 |
| CN109711242B true 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) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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***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 |
| CN112507852A (en) * | 2020-12-02 | 2021-03-16 | 上海眼控科技股份有限公司 | Lane line identification method, device, equipment and storage medium |
| CN112874524B (en) * | 2021-01-11 | 2022-08-23 | 广东科学技术职业学院 | Method and device for driving vehicle and unmanned vehicle |
| CN115049997B (en) * | 2022-06-07 | 2023-03-17 | 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 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6432399B2 (en) * | 2015-03-12 | 2018-12-05 | オムロン株式会社 | Image processing apparatus and image processing method |
| CN108052869B (en) * | 2017-11-23 | 2020-09-29 | 深圳市易成自动驾驶技术有限公司 | Lane line recognition method, lane line recognition device and computer-readable storage medium |
| CN108297866B (en) * | 2018-01-03 | 2019-10-15 | 西安交通大学 | A kind of lane holding control method of vehicle |
| CN108229406B (en) * | 2018-01-11 | 2022-03-04 | 海信集团有限公司 | Lane line detection method, device and terminal |
-
2018
- 2018-10-31 CN CN201811291589.XA patent/CN109711242B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109711242A (en) | 2019-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109711242B (en) | Lane line correction method, lane line correction device, and storage medium | |
| CN110147706B (en) | Obstacle recognition method and device, storage medium, and electronic device | |
| CN110148196B (en) | Image processing method and device and related equipment | |
| CN110705405B (en) | Target labeling method and device | |
| CN110869974B (en) | Point cloud processing method, equipment and storage medium | |
| EP3620964A1 (en) | Lane line processing method and device | |
| CN111222395A (en) | Target detection method and device and electronic equipment | |
| CN109583345B (en) | Road recognition method, device, computer device and computer readable storage medium | |
| JP6457278B2 (en) | Object detection apparatus and object detection method | |
| US20150363668A1 (en) | Traffic lane boundary line extraction apparatus and method of extracting traffic lane boundary line | |
| CN110400338B (en) | Depth map processing method and device and electronic equipment | |
| CN111753757B (en) | Image recognition processing method and device | |
| CN109583313B (en) | Lane line extraction method, device and storage medium | |
| CN112819895A (en) | Camera calibration method and device | |
| CN114863380B (en) | Lane line identification method and device and electronic equipment | |
| JP2020038617A (en) | Lane line processing method and device | |
| US20210326613A1 (en) | Vehicle detection method and device | |
| WO2023116113A1 (en) | Traffic light state detection method and apparatus, vehicle and storage medium | |
| JP7389729B2 (en) | Obstacle detection device, obstacle detection system and obstacle detection method | |
| CN114037966A (en) | High-precision map feature extraction method, device, medium and electronic equipment | |
| CN114708426A (en) | Target detection method, model training method, device, equipment and storage medium | |
| US11256929B2 (en) | Image-based road cone recognition method and apparatus, storage medium, and vehicle | |
| CN111238365A (en) | Subway train distance measurement and positioning method and system based on stereoscopic vision | |
| CN111862208B (en) | Vehicle positioning method, device and server based on screen optical communication | |
| US11538260B2 (en) | Object identification apparatus, object identification method, and nontransitory computer readable medium storing control program |
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 |