CN109325388A - Recognition methods, system and the automobile of lane line - Google Patents
Recognition methods, system and the automobile of lane line Download PDFInfo
- Publication number
- CN109325388A CN109325388A CN201710642211.9A CN201710642211A CN109325388A CN 109325388 A CN109325388 A CN 109325388A CN 201710642211 A CN201710642211 A CN 201710642211A CN 109325388 A CN109325388 A CN 109325388A
- Authority
- CN
- China
- Prior art keywords
- lane line
- lane
- frame
- image
- line image
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of recognition methods of lane line, system and automobiles, wherein, method is the following steps are included: acquire the video information of vehicle front lane line, and carry out frame decoding processing to video information to obtain M frame lane line image, wherein, M is the integer greater than 1；Processing is split to obtain multiple ROI regions interested of the dynamic comprising lane line in the line image of every frame lane to every frame lane line image；First localization process is carried out to the lane line in the line image of every frame lane according to multiple ROI regions to obtain corresponding just localization region；The lane line in each ROI region in the frame lane line image is scanned for handling to obtain the lane line center of gravity in each ROI region based on first localization region；Processing is fitted to obtain the lane line of vehicle front in the line image of every frame lane to multiple lane line centers of gravity.This method is not only able to achieve the identification of straight way, moreover it is possible to which the identification for realizing bend helps to improve the active safety of automobile.
Description
Technical field
The present invention relates to automotive field more particularly to a kind of recognition methods of lane line, a kind of identifying system of lane line
With a kind of automobile.
Background technique
With popularizing for automobile, the identification capability of road pavement object such as lane line is improved using sensor, to give driver
More security warnings and auxiliary are provided, to improve the active safety of driving, it has also become intelligent transport system field development
Important directions.
Currently, mainly describing lane line in Lane detection with straight line model, such as carried out by Hough transform straight
Line identification, the method can effectively identify straight way lane line, but poor for the recognition effect of bend lane line.In other words, it is
The drive assistance function for the higher orders such as keeping, needs to accurately identify lane line, and describes lane line with straight line model
It is impossible to meet the demands.
Summary of the invention
The present invention is directed to solve one of the technical problem in above-mentioned technology at least to a certain extent.For this purpose, of the invention
One purpose is to propose a kind of recognition methods of lane line, and this method is not only able to achieve the identification of straight way, moreover it is possible to realize bend
Identification, help to improve the active safety of automobile.
Second object of the present invention is to propose a kind of identifying system of lane line.
Third object of the present invention is to propose a kind of automobile.
In order to achieve the above objectives, first aspect present invention embodiment proposes a kind of recognition methods of lane line, including with
Lower step: the video information of acquisition vehicle front lane line, and frame decoding processing is carried out to obtain M frame lane to the video information
Line image, wherein M is the integer greater than 1；Processing is split to obtain in the line image of every frame lane to every frame lane line image
Multiple ROI regions interested of the dynamic comprising lane line；According to multiple ROI regions to the lane line in the line image of every frame lane into
The first localization process of row is to obtain corresponding just localization region；Based on first localization region to each ROI in the frame lane line image
Lane line in region scans for handling to obtain the lane line center of gravity in each ROI region；To multiple lane line centers of gravity into
Row process of fitting treatment is to obtain the lane line of vehicle front described in the line image of every frame lane.
The recognition methods of lane line according to an embodiment of the present invention is split processing to every frame lane line image to obtain
Multiple ROI regions interested of the dynamic comprising lane line, and then according to multiple ROI regions to the lane in the line image of every frame lane
Line carries out first localization process to obtain corresponding just localization region, and based on first localization region to every in the frame lane line image
Lane line in a ROI region scans for handling to obtain the lane line center of gravity in each ROI region, and then to multiple lanes
Line center of gravity is fitted processing to obtain the lane line of vehicle front in the line image of every frame lane, is not only able to achieve straight way as a result,
Identification, moreover it is possible to the identification for realizing bend helps to improve the active safety of automobile.
In order to achieve the above objectives, second aspect of the present invention embodiment proposes a kind of identifying system of lane line, comprising: solution
Frame module carries out frame decoding processing for acquiring the video information of vehicle front lane line, and to the video information to obtain M frame
Lane line image, wherein M is the integer greater than 1；Divide module, for being split processing to every frame lane line image to obtain
Obtain multiple ROI regions interested of the dynamic comprising lane line in the line image of every frame lane；First locating module, for according to multiple
ROI region carries out first localization process to the lane line in the line image of every frame lane to obtain corresponding just localization region；Search for mould
Block, for based on first localization region to the lane line in each ROI region in the frame lane line image scan for processing with
Obtain the lane line center of gravity in each ROI region；Fitting module, for being fitted processing to multiple lane line centers of gravity to obtain
The lane line of vehicle front described in the line image of every frame lane.
The identifying system of lane line according to an embodiment of the present invention divides every frame lane line image by dividing module
Processing is cut to obtain multiple ROI regions interested of the dynamic comprising lane line, and then by first locating module according to multiple areas ROI
Domain carries out first localization process to the lane line in the line image of every frame lane to obtain corresponding just localization region, and passes through search mould
Block scans for processing to the lane line in each ROI region in the frame lane line image based on first localization region to obtain often
Lane line center of gravity in a ROI region, and then processing is fitted to multiple lane line centers of gravity to obtain often by fitting module
The lane line of vehicle front, is not only able to achieve the identification of straight way as a result, in the line image of frame lane, moreover it is possible to realize the identification of bend,
Help to improve the active safety of automobile.
Further, the invention proposes a kind of automobiles comprising the identification system of the lane line of the above embodiment of the present invention
System.
The automobile of the embodiment of the present invention is not only able to achieve straight way using the identifying system of the lane line of above-described embodiment
Identification, moreover it is possible to which the identification for realizing bend helps to improve the active safety of automobile.
Detailed description of the invention
Fig. 1 is the flow chart of the recognition methods of lane line according to an embodiment of the invention；
Fig. 2 is the schematic diagram of the lane line image after an exemplary dividing processing according to the present invention；
Fig. 3 is the schematic diagram of the ROI region in an exemplary lane line image comprising left and right lane line according to the present invention；
Fig. 4 is the block diagram of the identifying system of lane line according to an embodiment of the invention；
Fig. 5 is the block diagram of the identifying system of lane line in accordance with another embodiment of the present invention；
Fig. 6 is the block diagram of the identifying system of the lane line of another embodiment according to the present invention；
Fig. 7 is the block diagram of automobile according to an embodiment of the present invention.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
Below with reference to the accompanying drawings recognition methods, system and the automobile of lane line according to an embodiment of the present invention are described.
Fig. 1 is the flow chart of the recognition methods of lane line according to an embodiment of the invention.As shown in Figure 1, the lane
The recognition methods of line the following steps are included:
S101 acquires the video information of vehicle front lane line, and carries out frame decoding processing to video information to obtain M frame vehicle
Road line image, wherein M is the integer greater than 1.
Specifically, vehicle front vehicle can be included by the visual sensor such as forward sight camera shooting being arranged on automobile
The video information of diatom, so video information can be input to processor such as DSP (Digital Signal Processing,
Digital Signal Processing) chip, dsp chip can by video input interface thereon to video information carry out frame decoding processing, with
M frame lane line image is collected, and collected lane line image is stored in corresponding memory.
Further, after obtaining M frame lane line image, each image can also be pre-processed, such as utilizes Gauss
Filter to every frame lane line image carry out denoising and smoothing processing, to improve raising picture quality, and carry out gray proces with
Obtain the corresponding gray level image of every frame lane line image.
S102, it is multiple comprising lane line in the line image of every frame lane to obtain to be split processing to every frame lane line image
Dynamic ROI region interested.
Specifically, the maximum entropy of each gray value in the line image of every frame lane is calculated, and is obtained in the line image of every frame lane
Maximum value in all maximum entropies；Using the corresponding gray value of maximum value in the line image of every frame lane in all maximum entropies as this
The segmentation threshold of frame lane line image, and processing is split to the frame lane line image to obtain the frame vehicle according to segmentation threshold
Multiple ROI regions comprising lane line in road line image.
In information theory, entropy is used to indicate average information, is defined as follows formula (1):
Wherein, p (x) is the probability density function of stochastic variable x.
In digital picture, stochastic variable x can be one of gray-scale value, area grayscale and Gradient Features.One
It ties up in maximum entropy dividing method, calculating separately target corresponding with each gray level T of image according to formula (1), (i.e. gray value is lower than T
Pixel) maximum entropy and background (i.e. gray value be higher than T pixel) maximum entropy, and find out they maximum entropy and,
So that maximum entropy and maximum gray value T are the segmentation threshold of image.
In an embodiment of the present invention, due to that in collected lane line video information, may include sky, lane
Information, these information such as the driving vehicle in the woods shrub on both sides, lane can have an impact the calculating of maximum entropy.
In this example, in order to which above-mentioned influence is reduced or avoided, can first according to include in image object (such as sky,
Trees, vehicle, lane line etc.) attribute carries out initial segmentation to lane line image, it such as obtains including the region of lane line and not wrapping
The region of lane line is included, and then the corresponding maximum entropy of each gray value in two regions can be calculated, and obtain all maximum entropies
Maximum value, lane line image is split with the maximum value corresponding gray value, the image after segmentation can be as shown in Figure 2.
Thereby, it is possible to improve to the segmentation effect for having the stronger lane line image of vehicle driving, light in lane.It should be noted that
After being split to lane line image, ROI region division can be carried out according to the feature of image after segmentation, Fig. 2 can be obtained
Shown in include lane line rectangle frame (i.e. ROI region).
It, in this embodiment, can also be respectively in the line image of lane in order to adapt to the identification of different shape lane line
Left-lane line and right-lane line dynamically set up corresponding ROI region.
Specifically, can first determine first in the line image of the every frame lane ROI region comprising left-lane line and comprising
The separation of the ROI region of right-lane line, and above one comprising left-lane line ROI region center of gravity and include right-lane line
ROI region center of gravity midpoint as next ROI region comprising left-lane line and the ROI region comprising right-lane line
Separation, to dynamically set up multiple ROI regions comprising left-lane line and comprising the ROI region of right-lane line.
Wherein, the separation of first ROI region comprising left-lane line and the ROI region comprising right-lane line can be
The geometry midpoint of lane line image.
For example, referring to Fig. 3, for a frame lane line image, continuity and direction characteristic are had according to lane line,
The geometry midpoint O of near-sighted field in the frame lane line image is determined first, and point O is that first of the frame lane line image includes
The separation of the ROI region A1 of left-lane line, ROI region A2 comprising right-lane line；Then can with the center of gravity of ROI region A1 and
Separation of the midpoint of the center of gravity of ROI region A2 as ROI region B1, ROI region B2, and become according to the offset of separation O2
Gesture determines the search range for working as front left and right lane line, and so on.Thus the ROI region comprising left-lane line is dynamically set up
A1, B1, C1 ..., and ROI region A2, B2 comprising right-lane line, C2 ....
Need to illustrate, in the line image of lane right simultaneously include comprising left-lane line ROI region and comprising a left side
The ROI region of lane line can be obtained according to specific lane line image.
S103, according to multiple ROI regions to the lane line in the line image of every frame lane carry out just localization process to obtain pair
The first localization region of the lane line answered.
Specifically, when carrying out just localization process to the lane line in the line image of every frame lane, every frame lane can be calculated
The lane line probability density of each ROI region in line image, and the maximum ROI region of lane line probability density is obtained, by lane
The maximum ROI region of line probability density is as localization region at the beginning of the lane line of the frame lane line image.
For example, for a frame lane line image, if n is that lane line is corresponding in each ROI region in the frame image
The pixel number of gray value, N*N are the total pixel number in each ROI region, then the lane line probability density in each ROI region isWherein, n is variable, and N is quantitative.Thus, it is possible to the determination maximum ROI region of lane line probability density, and then can incite somebody to action
Localization region at the beginning of its lane line as the frame lane line image.Similarly, the first positioning area of available every frame lane line image
Domain.
S104 scans for the lane line in each ROI region in the frame lane line image based on first localization region
Processing is to obtain the lane line center of gravity in each ROI region.
In this embodiment it is possible to scan for locating to the lane line in each ROI region using Meanshift algorithm
Reason.Specifically, first localization region is the initial position of search window, for each ROI region, calculates search window first
Lane line center of gravity, and (2) calculate mean-shift vector according to the following formula, then adjust search window according to mean-shift vector
The position of geometric center opposite lane line center of gravity, until the distance between the center of search window and lane line center of gravity are set less than one
Definite value, at this point it is possible to think lane line center of gravity at the center of search window, i.e., the lane line center of gravity in the ROI region has been detected
It surveys.
Wherein, x is the vector value at search window center, xiIt is i-th point in search window of vector value, h is search window
The width of mouth, g (x)=- k ' (x), k (x) are sample probability density function.
Optionally, sample probability density function can choose kernel density function shown in following formula (3):
Wherein, c is normalization coefficient.
S105 is fitted processing to multiple lane line centers of gravity to obtain the lane of vehicle front in the line image of every frame lane
Line.
Specifically, processing can be fitted to M lane line center of gravity by default lane line model, wherein default vehicle
Road line model can be indicated by formula (4):
Y=0.5*kx2+mx+b (4)
Wherein, y is the lane line of vehicle front, and x is lane line center of gravity, and k, m, b are default lane line model parameter.
It should be noted that scanning for processing to the lane line in the ROI region in the line image of every frame lane to obtain
When obtaining the lane line center of gravity in each ROI region, it is possible to error detection occurs.In this regard, in this embodiment it is possible to further
Screening Treatment is carried out to the lane line center of gravity that detection obtains, to remove the lane line center of gravity of error detection.
For example, the lane line center of gravity of error detection can be screened out, using RANSAC algorithm to obtain in process of fitting treatment
To optimal lane line model parameter, i.e., so that matched curve more meets practical lane line.Specifically, using RANSAC to lane
The step of line center of gravity is screened is as follows:
Step 1: minimum frequency in sampling is calculated according to fiducial probability P and data error probability ε；
Step 2: randomly selecting multiple such as 3 lane line centers of gravity and carry out least square method curve matchings, obtain lane line mould
Shape parameter examines lane line model parameter with all lane line focus points, obtains the interior quantity of every group of lane line model parameter；
Step 3: optimal lane line model parameter is selected according to interior quantity and error variance；
Step 4: utilizing the corresponding interior point estimation final mask parameter of optimal model parameters.
It is appreciated that lane line is straight line when k=0；When k ≠ 0, lane line is curve.
Further, after obtaining the lane line of the vehicle front in the line image of present frame lane, it can be determined that present frame
With previous frame lane line image recognition to lane line transition whether occurs and whether smooth curve can be obtained according to changing value；Such as
Fruit according to present frame and previous frame lane line image recognition to lane line cannot obtain there are transition and according to changing value it is smooth
Curve then judges that the Lane detection of present frame lane line image is invalid, abandons the Lane detection of present frame lane line image
As a result, and carrying out the Lane detection of next frame lane line image.
If it is determined that the lane line of continuous m frame such as 3 frame lane line images there is no transition or can obtain light according to changing value
Sliding curve, then can control automobile and enter lane line tracing mode, wherein when automobile enters lane line tracing mode, present frame
Multiple ROI regions in the line image of lane are determined according to multiple search windows in the line image of previous frame lane, thus, it is possible to subtract
The calculation amount of few Lane detection, improves the recognition speed of lane line.
In other words, in order to guarantee identification lane line accuracy, can use the successional feature of lane line, to before and after frames
The lane line parameter of lane line image recognition is compared, if lane line parameter has transition, and continuous several frames such as 3 frame lane lines
The lane line variation of image illustrates that Lane detection is wrong, abandons the recognition result of the frame, into the knowledge of next frame without rule
Not.In the effective situation of identification of continuous 3 frame lane line image, it can control automobile and enter lane line tracing mode, herein
Under mode, each ROI region in the line image of present frame lane can be according to each search window in the line image of previous frame lane
Mouth determines, to reduce the calculation amount of Lane detection.
It should be noted that after automobile enters lane line tracing mode, still can to the lane line center of gravity recognized into
Row Screening Treatment, to guarantee the validity of the lane line recognized.
In addition, it should also be noted that, the partitioning algorithm of the embodiment of the present invention can be not limited to above-mentioned maximum entropy partitioning algorithm,
Search process algorithm can be not limited to above-mentioned Meanshift algorithm, and filtering algorithm can be not limited to above-mentioned RANSAC.
To sum up, the recognition methods of lane line according to an embodiment of the present invention is split processing to every frame lane line image
To obtain multiple ROI regions interested of the dynamic comprising lane line, and then according to multiple ROI regions in the line image of every frame lane
Lane line carry out just localization process with obtain it is corresponding just localization region, and based on first localization region to the frame lane line image
In each ROI region in lane line scan for handling to obtain the lane line center of gravity in each ROI region, and then to more
A lane line center of gravity, which is fitted, to be handled to obtain the lane line of vehicle front in the line image of every frame lane, as a result, can not only be real
The identification of existing straight way, moreover it is possible to which the identification for realizing bend helps to improve the active safety of automobile.In addition, by obtaining
Lane line center of gravity is screened, and is verified to the recognition result of lane line, can be improved lane line identification accuracy,
Reliability and rapidity.
Fig. 4 is the block diagram of the identifying system of lane line according to an embodiment of the invention.As shown in figure 4, the lane
The identifying system 100 of line includes: solution frame module 10, segmentation module 20, first locating module 30, search module 40 and fitting module
50。
Wherein, solution frame module 10 is used to acquire the video information of vehicle front lane line, and carries out frame decoding to video information
Processing is to obtain M frame lane line image, wherein M is the integer greater than 1.Divide module 20 be used for every frame lane line image into
Row dividing processing is to obtain multiple ROI regions interested of the dynamic comprising lane line in the line image of every frame lane.First locating module
30 for carrying out first localization process to the lane line in the line image of every frame lane according to multiple ROI regions to obtain corresponding lane
The first localization region of line.Search module 40 be used for based in first localization region in each ROI region in the frame lane line image
Lane line scan for handling to obtain the lane line center of gravity in each ROI region.Fitting module 50 is used for multiple vehicles
Diatom center of gravity is fitted processing to obtain the lane line of vehicle front.
In another embodiment of the present invention, as shown in figure 5, the identifying system 100 of the lane line can also include sieve
Modeling block 60.Wherein, screening module 60 is used for before being fitted processing to multiple lane line centers of gravity, to multiple lane line weights
The heart carries out Screening Treatment.
In an embodiment of the present invention, fitting module is fitted multiple lane line centers of gravity by default lane line model
Processing, wherein default lane line model can be indicated by formula (4):
Y=0.5*kx2+mx+b (4)
Wherein, y is the lane line of vehicle front, and x is lane line center of gravity, and k, m, b are default lane line model parameter.
In an embodiment of the present invention, segmentation module 30 is specifically used for calculating each gray value in the line image of every frame lane
Maximum entropy, and obtain the maximum value in all maximum entropies；Using the corresponding gray value of the maximum value in all maximum entropies as the frame
The segmentation threshold of lane line image, and processing is split to the frame lane line image to obtain the frame lane according to segmentation threshold
Multiple ROI regions comprising lane line in line image.
Further, segmentation module 30 when multiple ROI regions comprising lane line, is gone back in obtaining every frame lane line image
For determining first in the line image of the every frame lane ROI region comprising left-lane line and comprising the ROI region of right-lane line
Separation, and the center of gravity of the center of gravity of above one ROI region comprising left-lane line and the ROI region comprising right-lane line
Separation of the midpoint as the next ROI region comprising left-lane line and the ROI region comprising right-lane line, dynamically to build
Found multiple ROI regions comprising left-lane line and the ROI region comprising right-lane line.
In an embodiment of the present invention, first locating module 30 is specifically used for calculating each area ROI in the line image of every frame lane
The lane line probability density in domain, and the maximum ROI region of lane line probability density is obtained, lane line probability density is maximum
ROI region is as localization region at the beginning of the lane line of the frame lane line image.
Further, as shown in fig. 6, the identifying system 100 of the lane line of the embodiment of the present invention can also include judging mould
Block 70, selecting module 80 and control module 90.
Wherein, judgment module 70 be used to judge present frame and previous frame lane line image recognition to lane line whether occur
Transition and can smooth curve be obtained according to changing value.Selecting module 80 is used for according to present frame and previous frame lane line image
The lane line recognized judges the vehicle of present frame lane line image there are transition and when cannot obtain smooth curve according to changing value
Diatom identification is invalid, abandons the Lane detection of present frame lane line image as a result, and carrying out the vehicle of next frame lane line image
Diatom identification.Control module 90 is used to that transition to be not present in the lane line of continuous m frame lane line image or can be obtained according to changing value
When obtaining smooth curve, control automobile enters lane line tracing mode, to improve the recognition speed of lane line.
It should be noted that the specific embodiment of the identifying system of the lane line of the embodiment of the present invention can be found in the present invention
The specific embodiment of the recognition methods of the lane line of above-described embodiment.
The identifying system of lane line according to an embodiment of the present invention divides every frame lane line image by dividing module
Processing is cut to obtain multiple ROI regions interested of the dynamic comprising lane line, and then by first locating module according to multiple areas ROI
Domain carries out first localization process to the lane line in the line image of every frame lane to obtain corresponding just localization region, and passes through search mould
Block scans for processing to the lane line in each ROI region in the frame lane line image based on first localization region to obtain often
Lane line center of gravity in a ROI region, and then processing is fitted to multiple lane line centers of gravity to obtain often by fitting module
The lane line of vehicle front, is not only able to achieve the identification of straight way as a result, in the line image of frame lane, moreover it is possible to realize the identification of bend,
Help to improve the active safety of automobile.In addition, being screened to obtained lane line center of gravity, and the identification to lane line
As a result it is verified, can be improved identification accuracy, reliability and the rapidity of lane line.
Fig. 7 is the block diagram of automobile according to an embodiment of the present invention.As shown in fig. 7, the automobile 1000 includes above-mentioned implementation
The identifying system 100 of the lane line of example.
The automobile of the embodiment of the present invention is not only able to achieve the identification of straight way using above-mentioned Lane detection system, moreover it is possible to
It enough realizes the identification of bend, and then is conducive to improve the active safety of vehicle.
It should be noted that in flow charts indicate or logic and/or step described otherwise above herein, for example,
It is considered the order list of the executable instruction for realizing logic function, may be embodied in any computer can
Read in medium, for instruction execution system, device or equipment (such as computer based system, including the system of processor or its
He can be from instruction execution system, device or equipment instruction fetch and the system executed instruction) it uses, or combine these instruction executions
System, device or equipment and use.For the purpose of this specification, " computer-readable medium " can be it is any may include, store,
Communicate, propagate, or transport program is for instruction execution system, device or equipment or combines these instruction execution systems, device or sets
The standby and device that uses.The more specific example (non-exhaustive list) of computer-readable medium include the following: have one or
The electrical connection section (electronic device) of multiple wirings, portable computer diskette box (magnetic device), random access memory (RAM), only
It reads memory (ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable light
Disk read-only memory (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or its
His suitable medium, because can then be edited for example by carrying out optical scanner to paper or other media, be interpreted or must
It is handled when wanting with other suitable methods electronically to obtain described program, is then stored in computer storage
In.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing specific logical function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, to execute function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
It should be appreciated that each section of the invention can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware
Any one of column technology or their combination are realized: having a logic gates for realizing logic function to data-signal
Discrete logic, with suitable combinational logic gate circuit specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
In the description of the present invention, it is to be understood that, term " center ", " longitudinal direction ", " transverse direction ", " length ", " width ",
" thickness ", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom" "inner", "outside", " up time
The orientation or positional relationship of the instructions such as needle ", " counterclockwise ", " axial direction ", " radial direction ", " circumferential direction " be orientation based on the figure or
Positional relationship is merely for convenience of description of the present invention and simplification of the description, rather than the device or element of indication or suggestion meaning must
There must be specific orientation, be constructed and operated in a specific orientation, therefore be not considered as limiting the invention.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or integral；It can be mechanical connect
It connects, is also possible to be electrically connected；It can be directly connected, can also can be in two elements indirectly connected through an intermediary
The interaction relationship of the connection in portion or two elements, unless otherwise restricted clearly.For those of ordinary skill in the art
For, the specific meanings of the above terms in the present invention can be understood according to specific conditions.
In the present invention unless specifically defined or limited otherwise, fisrt feature in the second feature " on " or " down " can be with
It is that the first and second features directly contact or the first and second features pass through intermediary mediate contact.Moreover, fisrt feature exists
Second feature " on ", " top " and " above " but fisrt feature be directly above or diagonally above the second feature, or be merely representative of
First feature horizontal height is higher than second feature.Fisrt feature can be under the second feature " below ", " below " and " below "
One feature is directly under or diagonally below the second feature, or is merely representative of first feature horizontal height less than second feature.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, modifies, replacement and variant.
Claims (15)
1. a kind of recognition methods of lane line, which comprises the following steps:
The video information of vehicle front lane line is acquired, and frame decoding processing is carried out to obtain M frame lane line to the video information
Image, wherein M is the integer greater than 1；
Processing is split to obtain multiple dynamic senses comprising lane line in the line image of every frame lane to every frame lane line image
Interest ROI region；
First localization process is carried out to the lane line in the line image of every frame lane according to multiple ROI regions to obtain corresponding just positioning
Region；
Processing is scanned for obtain to the lane line in each ROI region in the frame lane line image based on first localization region
Lane line center of gravity in each ROI region；
Processing is fitted to obtain the lane line of vehicle front described in the line image of every frame lane to multiple lane line centers of gravity.
2. the recognition methods of lane line as described in claim 1, which is characterized in that be fitted to multiple lane line centers of gravity
Before processing, further includes:
Screening Treatment is carried out to multiple lane line centers of gravity.
3. the recognition methods of lane line as claimed in claim 1 or 2, which is characterized in that by default lane line model to institute
It states multiple lane line centers of gravity and is fitted processing, wherein default lane line model are as follows:
Y=0.5*kx2+mx+b,
Wherein, y is the lane line of the vehicle front, and x is lane line center of gravity, and k, m, b are default lane line model parameter.
4. the recognition methods of lane line as described in claim 1, which is characterized in that described to divide every frame lane line image
Processing, which is cut, to obtain multiple ROI regions comprising lane line in the line image of every frame lane includes:
The maximum entropy of each gray value in the line image of every frame lane is calculated, and obtains the maximum value in all maximum entropies；
Using the corresponding gray value of maximum value in all maximum entropies as the segmentation threshold of the frame lane line image, and according to
It is multiple comprising lane line in the frame lane line image to obtain that the segmentation threshold is split processing to the frame lane line image
ROI region.
5. the recognition methods of lane line as claimed in claim 4, which is characterized in that multiple in obtaining every frame lane line image
When ROI region comprising lane line, further includes:
Determine first ROI region comprising left-lane line in the line image of every frame lane and the ROI region comprising right-lane line
Separation, and the center of gravity of the center of gravity of above one ROI region comprising left-lane line and the ROI region comprising right-lane line
Separation of the midpoint as the next ROI region comprising left-lane line and the ROI region comprising right-lane line, dynamically to build
Found multiple ROI regions comprising left-lane line and the ROI region comprising right-lane line.
6. the recognition methods of lane line as claimed in claim 4, which is characterized in that it is described according to multiple ROI regions to every frame
Lane line in the line image of lane carries out just localization process localization region at the beginning of to obtain corresponding lane line
The lane line probability density of each ROI region in the line image of every frame lane is calculated, and it is maximum to obtain lane line probability density
ROI region, using the maximum ROI region of lane line probability density as positioning area at the beginning of the lane line of the frame lane line image
Domain.
7. the recognition methods of lane line as described in claim 1, which is characterized in that further include:
Judge present frame and previous frame lane line image recognition to lane line transition whether occurs and can be obtained according to changing value
Obtain smooth curve；
If according to present frame and previous frame lane line image recognition to lane line cannot obtain there are transition and according to changing value
Smooth curve is obtained, then judges that the Lane detection of present frame lane line image is invalid, abandons the lane of present frame lane line image
Line recognition result, and carry out the Lane detection of next frame lane line image；
When transition is not present in the lane line of continuous m frame lane line image or can obtain smooth curve according to changing value, institute is controlled
It states automobile and enters lane line tracing mode, to improve the recognition speed of lane line.
8. a kind of identifying system of lane line characterized by comprising
Solve frame module, for acquiring the video information of vehicle front lane line, and to the video information carry out frame decoding processing with
Obtain M frame lane line image, wherein M is the integer greater than 1；
Divide module, is obtained for being split processing to every frame lane line image multiple comprising vehicle in the line image of every frame lane
The dynamic of diatom ROI region interested；
First locating module, for according to multiple ROI regions in the line image of every frame lane lane line carry out just localization process with
Obtain corresponding just localization region；
Search module, for being carried out based on first localization region to the lane line in each ROI region in the frame lane line image
Search process is to obtain the lane line center of gravity in each ROI region；
Fitting module, before being fitted processing to multiple lane line centers of gravity to obtain automobile described in the line image of every frame lane
The lane line of side.
9. the identifying system of lane line as claimed in claim 8, which is characterized in that further include:
Screening module, for being screened to multiple lane line centers of gravity before being fitted processing to multiple lane line centers of gravity
Processing.
10. the identifying system of lane line as claimed in claim 8 or 9, which is characterized in that the fitting module is by presetting vehicle
Road line model is fitted processing to the multiple lane line center of gravity, wherein default lane line model are as follows:
Y=0.5*kx2+mx+b,
Wherein, y is the lane line of the vehicle front, and x is lane line center of gravity, and k, m, b are default lane line model parameter.
11. the identifying system of lane line as claimed in claim 8, which is characterized in that the segmentation module is specifically used for:
The maximum entropy of each gray value in the line image of every frame lane is calculated, and obtains the maximum value in all maximum entropies；
Using the corresponding gray value of maximum value in all maximum entropies as the segmentation threshold of the frame lane line image, and according to
It is multiple comprising lane line in the frame lane line image to obtain that the segmentation threshold is split processing to the frame lane line image
ROI region.
12. the identifying system of lane line as claimed in claim 11, which is characterized in that the segmentation module is obtaining every frame vehicle
In road line image when multiple ROI regions comprising lane line, it is also used to:
Determine first ROI region comprising left-lane line in the line image of every frame lane and the ROI region comprising right-lane line
Separation, and the center of gravity of the center of gravity of above one ROI region comprising left-lane line and the ROI region comprising right-lane line
Separation of the midpoint as the next ROI region comprising left-lane line and the ROI region comprising right-lane line, dynamically to build
Found multiple ROI regions comprising left-lane line and the ROI region comprising right-lane line.
13. the identifying system of lane line as claimed in claim 11, which is characterized in that the just locating module is specifically used for:
The lane line probability density of each ROI region in the line image of every frame lane is calculated, and it is maximum to obtain lane line probability density
ROI region, using the maximum ROI region of lane line probability density as positioning area at the beginning of the lane line of the frame lane line image
Domain.
14. the identifying system of lane line as claimed in claim 8, which is characterized in that further include:
Judgment module, for judge present frame and previous frame lane line image recognition to lane line whether occur transition and according to
Can changing value obtain smooth curve；
Selecting module, for according to present frame and previous frame lane line image recognition to lane line there are transition and according to change
When change value cannot obtain smooth curve, judges that the Lane detection of present frame lane line image is invalid, abandon present frame lane line
The Lane detection of image is as a result, and carry out the Lane detection of next frame lane line image；
Control module is not present transition for the lane line in continuous m frame lane line image or can be obtained according to changing value smooth
It when curve, controls the automobile and enters lane line tracing mode, to improve the recognition speed of lane line.
15. a kind of automobile, which is characterized in that the automobile includes the knowledge of the lane line as described in any one of claim 8-14
Other system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710642211.9A CN109325388A (en) | 2017-07-31 | 2017-07-31 | Recognition methods, system and the automobile of lane line |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710642211.9A CN109325388A (en) | 2017-07-31 | 2017-07-31 | Recognition methods, system and the automobile of lane line |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109325388A true CN109325388A (en) | 2019-02-12 |
Family
ID=65245159
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710642211.9A Pending CN109325388A (en) | 2017-07-31 | 2017-07-31 | Recognition methods, system and the automobile of lane line |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325388A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110006440A (en) * | 2019-04-12 | 2019-07-12 | 北京百度网讯科技有限公司 | A kind of expression, device, electronic equipment and the storage medium of map relationship |
| CN111444778A (en) * | 2020-03-04 | 2020-07-24 | 武汉理工大学 | Lane line detection method |
| CN111797658A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Lane line recognition method and device, storage medium and electronic device |
| CN113869293A (en) * | 2021-12-03 | 2021-12-31 | 禾多科技（北京）有限公司 | Lane line recognition method and device, electronic equipment and computer readable medium |
| CN114782916A (en) * | 2022-04-11 | 2022-07-22 | 广州优创电子有限公司 | ADAS rear vehicle identification system carried by rearview mirror and based on multi-sensor fusion |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295420A (en) * | 2013-01-30 | 2013-09-11 | 吉林大学 | Method for recognizing lane line |
-
2017
- 2017-07-31 CN CN201710642211.9A patent/CN109325388A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295420A (en) * | 2013-01-30 | 2013-09-11 | 吉林大学 | Method for recognizing lane line |
Non-Patent Citations (1)
| Title |
|---|
| 王珂娜 等: "一种基于Meanshift和RANSAC的车道识别方法", 《计算机测量与控制》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111797658A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Lane line recognition method and device, storage medium and electronic device |
| CN110006440A (en) * | 2019-04-12 | 2019-07-12 | 北京百度网讯科技有限公司 | A kind of expression, device, electronic equipment and the storage medium of map relationship |
| CN110006440B (en) * | 2019-04-12 | 2021-02-05 | 北京百度网讯科技有限公司 | Map relation expression method and device, electronic equipment and storage medium |
| CN111444778A (en) * | 2020-03-04 | 2020-07-24 | 武汉理工大学 | Lane line detection method |
| CN111444778B (en) * | 2020-03-04 | 2023-10-17 | 武汉理工大学 | Lane line detection method |
| CN113869293A (en) * | 2021-12-03 | 2021-12-31 | 禾多科技（北京）有限公司 | Lane line recognition method and device, electronic equipment and computer readable medium |
| CN113869293B (en) * | 2021-12-03 | 2022-03-11 | 禾多科技（北京）有限公司 | Lane line recognition method and device, electronic equipment and computer readable medium |
| CN114782916A (en) * | 2022-04-11 | 2022-07-22 | 广州优创电子有限公司 | ADAS rear vehicle identification system carried by rearview mirror and based on multi-sensor fusion |
| CN114782916B (en) * | 2022-04-11 | 2024-03-29 | 广州优创电子有限公司 | ADAS rear-car recognition system based on multi-sensor fusion and carried on rearview mirror |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325388A (en) | Recognition methods, system and the automobile of lane line | |
| US8655081B2 (en) | Lane recognition system, lane recognition method, and lane recognition program | |
| JP4934701B2 (en) | Stereo image processing apparatus and stereo image processing method | |
| JP6184877B2 (en) | Vehicle external recognition device | |
| EP3007099A1 (en) | Image recognition system for a vehicle and corresponding method | |
| JP4930046B2 (en) | Road surface discrimination method and road surface discrimination device | |
| US8098933B2 (en) | Method and apparatus for partitioning an object from an image | |
| DE102009048699A1 (en) | Travel's clear path detection method for motor vehicle i.e. car, involves monitoring images, each comprising set of pixels, utilizing texture-less processing scheme to analyze images, and determining clear path based on clear surface | |
| DE102009050492A1 (en) | Travel's clear path detection method for motor vehicle i.e. car, involves monitoring images, each comprising set of pixels, utilizing texture-less processing scheme to analyze images, and determining clear path based on clear surface | |
| CN108052904B (en) | Method and device for acquiring lane line | |
| EP2580740A2 (en) | An illumination invariant and robust apparatus and method for detecting and recognizing various traffic signs | |
| DE102009048892A1 (en) | Clear traveling path detecting method for vehicle e.g. car, involves generating three-dimensional map of features in view based upon preferential set of matched pairs, and determining clear traveling path based upon features | |
| DE112012002885T5 (en) | Method and device for image-based vehicle detection and distance measurement | |
| DE102009050505A1 (en) | Clear path detecting method for vehicle i.e. motor vehicle such as car, involves modifying clear path based upon analysis of road geometry data, and utilizing clear path in navigation of vehicle | |
| KR101968349B1 (en) | Method for detecting lane boundary by visual information | |
| WO2019009110A1 (en) | Exterior recognition device for vehicles | |
| WO2014002692A1 (en) | Stereo camera | |
| WO2013116598A1 (en) | Low-cost lane marker detection | |
| CN110770741B (en) | Lane line identification method and device and vehicle | |
| CN109558765B (en) | Automobile and lane line detection method and device | |
| KR101236223B1 (en) | Method for detecting traffic lane | |
| JP5073700B2 (en) | Object detection device | |
| CN114693716A (en) | Driving environment comprehensive identification information extraction method oriented to complex traffic conditions | |
| WO2019085929A1 (en) | Image processing method, device for same, and method for safe driving | |
| CN110537206B (en) | Railway track recognition device, program, and railway track recognition method |
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: 20190212 |