CN109558768A - The detection method of automobile and road signs, device - Google Patents
The detection method of automobile and road signs, device Download PDFInfo
- Publication number
- CN109558768A CN109558768A CN201710876728.4A CN201710876728A CN109558768A CN 109558768 A CN109558768 A CN 109558768A CN 201710876728 A CN201710876728 A CN 201710876728A CN 109558768 A CN109558768 A CN 109558768A
- Authority
- CN
- China
- Prior art keywords
- road signs
- image
- interest
- detection
- result
- 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
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/582—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of traffic signs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
Abstract
The invention discloses the detection methods of a kind of automobile and road signs, device, wherein, detection method includes the following steps: obtaining road signs image by vehicle-mounted front camera, and pre-processes to road signs image, wherein, road signs image is gray level image;Area-of-interest is set in image after the pre-treatment, and edge detection is carried out to obtain the gradient information of edge detection point to area-of-interest;Loop truss is carried out to generate loop truss to area-of-interest according to the gradient information of edge detection point as a result, and judging in road signs image according to loop truss result with the presence or absence of circular traffic sign board.The detection method of the road signs, by the way that area-of-interest is arranged in the picture, the image detection time can be reduced, effective identification to circular traffic sign board can be realized by loop truss, and real-time is good, is influenced by the resolution ratio of image, illumination, weather etc. small.
Description
Technical field
The present invention relates to automotive field, in particular to a kind of detection method of road signs, a kind of road traffic mark
The detection device of will and a kind of automobile.
Background technique
Road traffic sign detection and identification belong to the important content that Vehicular intelligent drives field, and detection can be divided into detection-phase
And cognitive phase.In the related technology, as follows to the detection technique of traffic sign:
A kind of method for traffic sign detection based on edge color pair and Feature Selection device, specific steps include: to acquisition
To image pre-processed, enhance color contrast;Edge detection is carried out to pretreated image, extracts marginal information;
Edge color pair detection is carried out to edge point, removes the marginal point for not meeting traffic sign color;Remaining marginal point is carried out
Closing operation of mathematical morphology forms connected region, realizes the coarse positioning to traffic sign;Last design area feature and symmetrical feature
Secondary characteristics screening washer, is accurately positioned traffic sign.
A kind of method for traffic sign detection based on color segmentation and SVM classifier, specific steps include: to carry out to image
The color segmentation of rgb space obtains reddish yellow blue three-color segmented image;Morphological scale-space is carried out to the image after color segmentation
Obtain the profile of traffic sign;Extract profile sequence of points, carry out Fast Fourier Transform (FFT), obtain Fourier descriptor, by with
Nominal contour Fourier descriptor, which is compared, judges whether it is traffic sign profile;HOG feature is extracted to traffic sign and is adopted
Classified with the method based on SVM, and training classifier;Finally by the obtained SVM classifier of training to traffic sign into
Row classification and identification.
However, above two detection technique all relies on the color characteristic of traffic sign.In actual road conditions, due to day
Gas, illumination, brightness, the influence faded etc., will lead to the cross-color of traffic sign, and causing can not be accurately according to color characteristic point
Traffic sign profile is cut out, so as to cause the failure of detection.
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 that a kind of detection method of road signs, this method can carry out circular traffic sign board effective
Identification, and real-time is good, is influenced by the resolution ratio of image, illumination, weather etc. small.
Second object of the present invention is to propose a kind of non-transitorycomputer readable storage medium.
Third object of the present invention is to propose a kind of detection device of road signs.
Fourth 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 detection method of road signs,
The following steps are included: obtaining road signs image by vehicle-mounted front camera, and to the road signs image
It is pre-processed, wherein the road signs image is gray level image;Region of interest is set in image after the pre-treatment
Domain, and edge detection is carried out to obtain the gradient information of edge detection point to the area-of-interest;According to the edge detection
The gradient information of point carries out loop truss to the area-of-interest to generate loop truss as a result, and sentencing according to the loop truss result
Break in the road signs image with the presence or absence of circular traffic sign board.
The detection method of road signs according to an embodiment of the present invention can be obtained directly by vehicle-mounted front camera
The road signs image of gray processing needed for detection can reduce image detection by the way that area-of-interest is arranged in the picture
Time, and then effective identification to circular traffic sign board can be realized by loop truss, and real-time is good, by the resolution of image
The influence of rate, illumination, weather etc. is small.
Further, the invention proposes a kind of non-transitorycomputer readable storage medium, it is stored thereon with computer
Program, the program realize the detection method of above-mentioned road signs when being executed by processor.
The non-transitorycomputer readable storage medium of the embodiment of the present invention storing with above-mentioned road by executing thereon
The corresponding program of the detection method of traffic sign can effectively identify circular traffic sign board, and real-time is good, is schemed
The influence of resolution ratio, illumination, the weather of picture etc. is small.
In order to achieve the above objectives, third aspect present invention embodiment proposes a kind of detection device of road signs,
It include: acquisition module, for obtaining road signs image by vehicle-mounted front camera;Preprocessing module, for institute
It states road signs image to be pre-processed, wherein the road signs image is gray level image;Setup module is used
Area-of-interest is set in image after the pre-treatment;First detection module, for carrying out edge to the area-of-interest
Detection is to obtain the gradient information of edge detection point;Second detection module, for the gradient information according to the edge detection point
Loop truss is carried out to generate loop truss result to the area-of-interest;Judgment module, for being sentenced according to the loop truss result
Break in the road signs image with the presence or absence of circular traffic sign board.
The detection device of road signs according to an embodiment of the present invention can be obtained directly by vehicle-mounted front camera
The road signs image of gray processing needed for detection, is arranged area-of-interest by setup module in the picture, can reduce
The image detection time, and then effective identification to circular traffic sign board can be realized by the second detection module progress loop truss,
And real-time is good, is influenced by the resolution ratio of image, illumination, weather etc. small.
Further, the invention proposes a kind of automobiles comprising the detection device of above-mentioned road signs.
The automobile of the embodiment of the present invention can be to circular traffic sign using the detection device of above-mentioned road signs
Board is effectively identified, and real-time is good, is influenced by the resolution ratio of image, illumination, weather etc. small.
Detailed description of the invention
Fig. 1 is the flow chart according to the detection method of the road signs of the embodiment of the present invention;
Fig. 2 is the schematic diagram according to area-of-interest in the road signs image of an example of the present invention;
Fig. 3 is the schematic diagram according to subpoint in the loop truss of an example of the present invention;
Fig. 4 is the schematic diagram according to the road signs testing result of an example of the present invention;
Fig. 5 is the block diagram according to the detection device of the road signs of one embodiment of the invention;
Fig. 6 is the block diagram according to the detection device of the road signs of another embodiment of the present invention;
Fig. 7 is the block diagram according to the automobile of the 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.
The automobile of the embodiment of the present invention, the detection method and device of road signs described with reference to the accompanying drawing.
Fig. 1 is the flow chart according to the detection method of the road signs of the embodiment of the present invention.As shown in Figure 1, the road
The detection method of road traffic sign, comprising the following steps:
S101 obtains road signs image by vehicle-mounted front camera, and carries out to road signs image
Pretreatment.
Wherein, road signs image is gray level image.
Specifically, vehicle-mounted front camera can collect the video image frame of vehicle front, which is YUV
Format, and then Y channel data (i.e. gradation data) can be chosen, that is, road signs image is obtained, and retain original video figure
As frame.
For the detection convenient for road signs, road signs image can be pre-processed, such as by road traffic
Sign image zooms to default resolution ratio, to reduce substrate processing time.Wherein, default resolution ratio can be 640 × 360.
For example, for the road signs image of 1080P, image resolution ratio 1920x1080 zooms to 640
After × 360, area becomes the 1/9 of original image area;For 720P video, image resolution ratio is 1280 × 720, is zoomed to
After 640 × 360, area becomes the 1/4 of original image area, can effectively reduce the subsequent image processing time as a result,.
S102, is arranged area-of-interest in image after the pre-treatment, and to area-of-interest carry out edge detection with
To the gradient information of edge detection point.
Specifically, area-of-interest can be arranged according to the position feature of road and traffic sign plates.Due to road traffic mark
Will board is generally located on above both sides of the road or road, therefore, pretreated image can be divided into multiple images unit, and from
It is middle to select one or more as area-of-interest.Wherein, the size of elementary area can be set according to actual test situation.
For example, referring to Fig. 2, pretreated image is divided into nine elementary areas, i.e. each trisection of transverse and longitudinal coordinate, it can
Using the first row of dash area, the second row first row and the tertial elementary area of the second row as area-of-interest, i.e. in Fig. 2
Dash area.As a result, by setting area-of-interest, the area of detection of image can be reduced 4/9, convenient for improving detection effect
Rate.
In this embodiment, it is emerging to sense that intermediate value calculus of finite differences, Sobel operator, Prewitt operator, Canny operator etc. can be used
Interesting region carries out edge detection to obtain the gradient information of edge detection point.Wherein, the gradient information of edge detection point includes side
The horizontal direction gradient and vertical gradient of edge test point.
S103, according to the gradient information of edge detection point to area-of-interest carry out loop truss to generate loop truss as a result,
And judged in road signs image according to loop truss result with the presence or absence of circular traffic sign board.
Specifically, using the radial symmetry of round or ellipse, using rapidly radially symmetry algorithm to region of interest
Domain carries out loop truss to generate loop truss as a result, and judging the loop truss result.For example, it may be judged whether it is right to there is radial direction
The first vegetarian refreshments for claiming transformation results to be greater than preset value, if it is present judging there is round traffic mark in road signs image
Will board.Wherein, preset value can be demarcated according to actually detected situation.
The detection method of the road signs of the embodiment can be obtained directly needed for detection by vehicle-mounted front camera
The road signs image of gray processing can reduce the image detection time, in turn by the way that area-of-interest is arranged in the picture
Effective identification to circular traffic sign board can be realized by loop truss, and real-time is good, by the resolution ratio of image, illumination, day
The influence of gas etc. is small.
In one embodiment of the invention, using the ladder of edge detection point in median difference calculation of group dividing area-of-interest
Information g is spent, shown in specific formula for calculation such as following formula (1):
Wherein, gx=[P (i+1, j)-P (i-1, j)]/2, gy=[P (i, j+1)-P (i, j-1)]/2, (i, j) are to feel emerging
Edge detection point in interesting region.
Further, can be used rapidly radially symmetry algorithm according to the gradient information of edge detection point to area-of-interest into
Row loop truss is to generate loop truss result, wherein loop truss result includes first vegetarian refreshments and its corresponding radial symmetry transform result.
First vegetarian refreshments that radial symmetry transform result is greater than preset value is judged whether there is, is greater than if there is radial symmetry transform result pre-
If first vegetarian refreshments of value, then judge that there are circular traffic sign boards in road signs image.
Specifically, each edge detection point can be calculated by following formula (2) and (3) in gradient direction referring to Fig. 3
Upper two subpoint P corresponding with edge detection point+ve(P) and P-ve(P) position:
Wherein, g (p) indicates the horizontal gradient g (x) and vertical gradient g (y) of edge detection point P, | | g (p) | | it is edge
The L2 norm of test point gradient, | | g (p) | | the horizontal gradient of=sqrt (gx*gx+gy*gy), i.e. P point and vertical Royal Park square
With open radical sign;N is to carry out loop truss radius selected by symmetry transformation, and round function is to seek variable nearest integer.
And then the direction projection figure On and size perspective view of two following formula (4), (5) calculating subpoints can be passed through respectively
Mn:
Wherein, direction projection figure On and size perspective view Mn can regard two matrixes, the size of matrix and mapping to be checked as
Picture pixel size (i.e. the number of pixel) is identical, and if image to be detected is 360*240, then the size of On, Mn are also 360*
240, each of matrix point initial value is all 0.
Radial symmetry transform result SnIt can be calculated by following formula (6):
Sn=Fn*An (6)
Wherein,For radial symmetric intensity map, α is the stringent of radius
Parameter is spent, value can be 2, AnFor a dimensional Gaussian window.The radial symmetry transform result noise jamming being calculated as a result,
It is small.
It should be noted that | | On(p) | | indicate On(p) L1 norm or L2 norm, | | Mn(p) | | indicate Mn(p) L1
Norm or L2 norm.To simplify the calculation, | | On(p)||、||Mn(p) | | L1 norm can be used, that is, take absolute value, at this point, maxp
||On(p) | | after expression takes absolute value to each value in On matrix, then take the maximum value of element in matrix, maxp||Mn(p) | | table
Show after taking absolute value to value each in Mn matrix, then takes the maximum value of element in matrix.
In this embodiment, due to radial symmetry transform result SnFor a two-dimentional floating-point matrix, the element value in matrix is 0
Floating point values between~1, therefore the first vegetarian refreshments that radial symmetry transform result can be greater than 0.8 (i.e. preset value) is standby as the center of circle
Reconnaissance, and can judge there is round traffic in area-of-interest when there are first vegetarian refreshments that radial symmetry transform result is greater than 0.8
Sign board.Further, referring to Fig. 4, first vegetarian refreshments of the radial symmetry transform result greater than 0.8 is circular traffic sign board in quilt
Display position in detection image (Fig. 4 left-side images), i.e., the point institute that two colors are shallower in Fig. 4 image right dotted line frame are in place
It sets.
Further, in order to improve the accuracy of detection, the element of preset value can be greater than to radial symmetry transform result
Point is screened, in turn, the position of all elements point composition circular traffic sign board screened.
Specifically, the range formula that two centers of circle are alternatively put isFor
Calculation amount is reduced, simplified calculation can be used, i.e., if the corresponding center of circle (P is alternatively put in any two center of circle1(x),P1(y))、
(P2(x),P2(y)) and radius n meets following either condition, then retains the biggish first vegetarian refreshments of radial symmetry transform result:
(1)|P1(x)-P2(x)|+|P1(y)-P2(y) | < 0.6*n,
(2)|P1(x)-P2(x) | < 0.5*n,
(3)|P1(y)-P2(y)|<0.5*n。
It is vehicle-mounted preposition to take the photograph with driving direction, the variation of operating range it should be noted that while the car is driving
Collected road and traffic sign plates may deformation occurs for picture, and such as circular sign board becomes ellipse.At this point it is possible to
By changing loop truss radius n, with the detection of adaptive deformation sign board.
To sum up, the detection method of road signs according to an embodiment of the present invention, by the vehicle for directly acquiring yuv format
The Y channel data of front camera is carried, and zooms in and out pretreatment, and setting area-of-interest, reduces detection time, mentions
High operational efficiency;And loop truss is carried out using rapidly radially symmetry transformation, detection efficiency is on the one hand improved, place in real time is met
On the other hand reason demand can also have preferable recognition effect when detected image resolution ratio is lower, and be illuminated by the light and weather etc.
Influence it is smaller, incomplete to part or deformation circle row traffic sign also has preferable detection effect.
Further, the invention proposes a kind of non-transitorycomputer readable storage medium, it is stored thereon with computer
Program, the program realize the detection method of above-mentioned road signs when being executed by processor.
The non-transitorycomputer readable storage medium of the embodiment of the present invention storing with above-mentioned road by executing thereon
The corresponding program of the detection method of traffic sign can effectively identify circular traffic sign board, and real-time is good, is schemed
The influence of resolution ratio, illumination, the weather of picture etc. is small.
Fig. 5 is the block diagram according to the detection device of the road signs of one embodiment of the invention.As shown in figure 5,
The detection device 100 includes: to obtain module 10, preprocessing module 20, setup module 30, first detection module 40, second to detect
Module 50 and judgment module 60.
Wherein, module 10 is obtained to be used to obtain road signs image by vehicle-mounted front camera.Preprocessing module
20 for pre-processing road signs image, wherein road signs image is gray level image.Setup module 30
For area-of-interest to be arranged in image after the pre-treatment.First detection module 40 is used to carry out edge inspection to area-of-interest
It surveys to obtain the gradient information of edge detection point.Second detection module 50 is used for emerging to sense according to the gradient information of edge detection point
Interesting region carries out loop truss to generate loop truss result.Judgment module 60 is used to judge road signs according to loop truss result
It whether there is circular traffic sign board in image.
Specifically, vehicle-mounted front camera can collect the video image frame of vehicle front, which is YUV
Format, and then acquisition module 10 can choose Y channel data (i.e. gradation data), that is, obtain road signs image, and retain
Raw video image frame.
For the detection convenient for road signs, road signs image can be located in advance by preprocessing module 20
Reason, such as by road signs image scaling to default resolution ratio, to reduce substrate processing time.Wherein, default resolution ratio can
To be 640 × 360.
For example, for the road signs image of 1080P, image resolution ratio 1920x1080 zooms to 640
After × 360, area becomes the 1/9 of original image area;For 720P video, image resolution ratio is 1280 × 720, is zoomed to
After 640 × 360, area becomes the 1/4 of original image area, can effectively reduce the subsequent image processing time as a result,.
It is possible to further which region of interest is arranged according to the position feature of road and traffic sign plates by setup module 30
Domain.Since road and traffic sign plates are generally located on above both sides of the road or road, pretreated image can be divided into
Multiple images unit, and therefrom select one or more as area-of-interest.Wherein, the size of elementary area can be according to reality
Test case is set.
For example, referring to Fig. 2, pretreated image can be divided into nine image lists that three rows three arrange by setup module 30
Member, and set interested for the first row elementary area, the second row first row elementary area and the second row third column elementary area
Region, i.e. dash area in Fig. 2.As a result, by setting area-of-interest, the area of detection of image can be reduced 4/9, be convenient for
Improve detection efficiency.
In this embodiment, first detection module 40 can be used intermediate value calculus of finite differences, Sobel operator, Prewitt operator,
Canny operator etc. carries out edge detection to area-of-interest to obtain the gradient information of edge detection point.Wherein, edge detection point
Gradient information include edge detection point horizontal direction gradient and vertical gradient.
For example, first detection module 40 can be believed using the gradient of edge detection point in median difference calculation of group dividing area-of-interest
G is ceased, shown in specific formula for calculation such as following formula (1):
Wherein, gx=[P (i+1, j)-P (i-1, j)]/2, gy=[P (i, j+1)-P (i, j-1)]/2, (i, j) are to feel emerging
Edge detection point in interesting region.
Further, the second detection module 50 can utilize the radial symmetry of round or ellipse, using rapidly radially
Symmetry algorithm carries out loop truss to area-of-interest according to the gradient information of edge detection point to generate loop truss result, wherein
Loop truss result includes first vegetarian refreshments and its corresponding radial symmetry transform result.And then judgment module 60 can determine whether that there are diameters
It is greater than first vegetarian refreshments of preset value to symmetry transformation result, and in first vegetarian refreshments there are radial symmetry transform result greater than preset value
When, judge that there are circular traffic sign boards in road signs image.Wherein, preset value can be carried out according to actually detected situation
Calibration.
Specifically, each edge detection point can be calculated by following formula (2) and (3) in gradient direction referring to Fig. 3
Upper two subpoint P corresponding with edge detection point+ve(P) and P-ve(P) position:
Wherein, g (p) indicates the horizontal gradient g (x) and vertical gradient g (y) of edge detection point P, | | g (p) | | it is edge
The L2 norm of test point gradient, | | g (p) | | the horizontal gradient of=sqrt (gx*gx+gy*gy), i.e. P point and vertical Royal Park square
With open radical sign;N is to carry out loop truss radius selected by symmetry transformation, and round function is to seek variable nearest integer.
And then the direction projection figure On and size perspective view of two following formula (4), (5) calculating subpoints can be passed through respectively
Mn:
Wherein, direction projection figure On and size perspective view Mn can regard two matrixes, the size of matrix and mapping to be checked as
Picture pixel size (i.e. the number of pixel) is identical, and if image to be detected is 360*240, then the size of On, Mn are also 360*
240, each of matrix point initial value is all 0.
Radial symmetry transform result SnIt can be calculated by following formula (6):
Sn=Fn*An (6)
Wherein,For radial symmetric intensity map, α is the stringent of radius
Parameter is spent, value can be 2, AnFor a dimensional Gaussian window.The radial symmetry transform result noise jamming being calculated as a result,
It is small.
It should be noted that | | On(p) | | indicate On(p) L1 norm or L2 norm, | | Mn(p) | | indicate Mn(p) L1
Norm or L2 norm.To simplify the calculation, | | On(p)||、||Mn(p) | | L1 norm can be used, that is, take absolute value, at this point, maxp
||On(p) | | after expression takes absolute value to each value in On matrix, then take the maximum value of element in matrix, maxp||Mn(p) | | table
Show after taking absolute value to value each in Mn matrix, then takes the maximum value of element in matrix.
In this embodiment, due to radial symmetry transform result SnFor a two-dimentional floating-point matrix, the element value in matrix is 0
Floating point values between~1, therefore the first vegetarian refreshments that radial symmetry transform result can be greater than 0.8 (i.e. preset value) is standby as the center of circle
Reconnaissance, and can judge there is round traffic in area-of-interest when there are first vegetarian refreshments that radial symmetry transform result is greater than 0.8
Sign board.Further, referring to Fig. 4, first vegetarian refreshments of the radial symmetry transform result greater than 0.8 is circular traffic sign board in quilt
Display position in detection image (Fig. 4 left-side images), i.e., the point institute that two colors are shallower in Fig. 4 image right dotted line frame are in place
It sets.
In some embodiments of the invention, as shown in fig. 6, detection device 100 further includes screening module 70.Wherein, it sieves
Modeling block 70 is used to select the element that radial symmetry transform result is greater than preset value and be screened will screen obtained first vegetarian refreshments
The position for forming circular traffic sign board, can be improved the accuracy of detection as a result,.
Specifically, if the corresponding center of circle (P is alternatively put in any two center of circle1(x),P1(y))、(P2(x),P2(y)) and partly
Diameter n meets following either condition, and screening module 70 then retains the biggish first vegetarian refreshments of radial symmetry transform result:
(1)|P1(x)-P2(x)|+|P1(y)-P2(y) | < 0.6*n,
(2)|P1(x)-P2(x) | < 0.5*n,
(3)|P1(y)-P2(y)|<0.5*n。
It is vehicle-mounted preposition to take the photograph with driving direction, the variation of operating range it should be noted that while the car is driving
Collected road and traffic sign plates may deformation occurs for picture, and such as circular sign board becomes ellipse.At this point it is possible to
By changing loop truss radius n, with the detection of adaptive deformation sign board.
To sum up, the detection device of road signs according to an embodiment of the present invention directly acquires YUV by obtaining module
The Y channel data of the vehicle-mounted front camera of format, and pretreatment is zoomed in and out by preprocessing module, and pass through setting mould
Area-of-interest is arranged in block, reduces detection time, improves operational efficiency;And by the second detection module using rapidly radially
Symmetry transformation carries out loop truss, on the one hand improves detection efficiency, meets real-time processing requirement, on the other hand, in tested mapping
As can also there is preferable recognition effect when resolution ratio is lower, and it is illuminated by the light and the influence of weather etc. is smaller, it is incomplete to part or become
The circle row traffic sign of shape also has preferable detection effect.
Fig. 7 is the block diagram according to the automobile of the embodiment of the present invention.As shown in fig. 7, the automobile 1000 includes above-mentioned road
The detection device 100 of road traffic sign.
The automobile of the embodiment of the present invention can be to circular traffic sign using the detection device of above-mentioned road signs
Board is effectively identified, and real-time is good, is influenced by the resolution ratio of image, illumination, weather etc. small.
In addition, other of automobile according to an embodiment of the present invention constitute and its effect is to those skilled in the art
It is known, to reduce redundancy, it is not repeated herein.
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.
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.
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 custom logic 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, Lai Zhihang function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction
The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium ", which can be, any may include, stores, communicates, propagates or pass
Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of computer-readable medium include the following: there is the electricity of one or more wirings
It is interconnecting piece (electronic device), portable computer diskette box, random access memory (RAM), read-only memory (ROM), erasable
Editable read-only memory (EPROM or flash memory) and portable optic disk read-only storage (CDROM).In addition, calculating
Machine readable medium can even is that the paper that can print described program on it or other suitable media, because can for example pass through
Optical scanner is carried out to paper or other media, then edited, interpreted or handled with other suitable methods when necessary
Described program is electronically obtained, is then stored in computer storage.
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.Such as, if realized with hardware in another embodiment, following skill well known in the art can be used
Any one of art or their combination are realized: have for data-signal is realized the logic gates of logic function from
Logic circuit is dissipated, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries
It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium
In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.
It, can also be in addition, each functional unit in each embodiment of the present invention can integrate in a processing module
It is that each unit physically exists alone, can also be integrated in two or more units in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized and when sold or used as an independent product in the form of software function module, also can store in a computer
In read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above
The embodiment of the present invention is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as to limit of the invention
System, those skilled in the art can be changed above-described embodiment, modify, replace and become within the scope of the invention
Type.
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 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.It for the ordinary skill in the art, can be according to specific feelings
Condition understands the concrete meaning of above-mentioned term in the present invention.
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.
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 (16)
1. a kind of detection method of road signs, which comprises the following steps:
Road signs image is obtained by vehicle-mounted front camera, and the road signs image is located in advance
Reason, wherein the road signs image is gray level image;
Area-of-interest is set in image after the pre-treatment, and edge detection is carried out to obtain edge to the area-of-interest
The gradient information of test point;
Loop truss is carried out to generate loop truss as a result, simultaneously to the area-of-interest according to the gradient information of the edge detection point
Judged in the road signs image according to the loop truss result with the presence or absence of circular traffic sign board.
2. the detection method of road signs as described in claim 1, which is characterized in that described to the road traffic mark
Will image is pre-processed, comprising:
By the road signs image scaling to default resolution ratio.
3. the detection method of road signs as claimed in claim 1 or 2, which is characterized in that it is described after the pre-treatment
Area-of-interest is set in image, comprising:
Pretreated image is divided into multiple images unit, and will be schemed accordingly according to the position feature of circular traffic sign board
As unit is set as the area-of-interest.
4. the detection method of road signs as described in claim 1, which is characterized in that calculate edge according to the following formula
The gradient information g of test point:
Wherein, gx=[P (i+1, j)-P (i-1, j)]/2, gy=[P (i, j+1)-P (i, j-1)]/2, (i, j) are region of interest
Edge detection point in domain.
5. the detection method of road signs as described in claim 1, which is characterized in that described according to the edge detection
The gradient information of point carries out loop truss to the area-of-interest to generate loop truss as a result, and sentencing according to the loop truss result
Break in the road signs image with the presence or absence of circular traffic sign board, comprising:
Round inspection is carried out to the area-of-interest according to the gradient information of the edge detection point using rapidly radially symmetry algorithm
It surveys to generate loop truss result, wherein the loop truss result includes first vegetarian refreshments and its corresponding radial symmetry transform result;
Judge whether there is first vegetarian refreshments that radial symmetry transform result is greater than preset value;
It is greater than first vegetarian refreshments of the preset value if there is radial symmetry transform result, then judges the road signs image
In there are circular traffic sign boards.
6. the detection method of road signs as claimed in claim 5, which is characterized in that further include:
The element point for being greater than the preset value to radial symmetry transform result screens.
7. the detection method of road signs as claimed in claim 6, which is characterized in that radial symmetry transform result
When element point greater than the preset value is screened, if any two radial symmetry transform result is greater than the preset value
The corresponding center of circle (the P of first vegetarian refreshments1(x),P1(y))、(P2(x),P2(y)) and loop truss radius n meets following either condition, then protects
Stay the biggish first vegetarian refreshments of radial symmetry transform result:
(1)|P1(x)-P2(x)|+|P1(y)-P2(y) | < 0.6*n,
(2)|P1(x)-P2(x) | < 0.5*n,
(3)|P1(y)-P2(y)|<0.5*n。
8. a kind of non-transitorycomputer readable storage medium, is stored thereon with computer program, which is characterized in that the program quilt
The detection method such as road signs of any of claims 1-7 is realized when processor executes.
9. a kind of detection device of road signs characterized by comprising
Module is obtained, for obtaining road signs image by vehicle-mounted front camera;
Preprocessing module, for being pre-processed to the road signs image, wherein the road signs image
For gray level image;
Setup module, for area-of-interest to be arranged in image after the pre-treatment;
First detection module, for carrying out edge detection to the area-of-interest to obtain the gradient information of edge detection point;
Second detection module, for according to the gradient information of the edge detection point to the area-of-interest carry out loop truss with
Generate loop truss result;
Judgment module, for being judged in the road signs image according to the loop truss result with the presence or absence of round traffic
Sign board.
10. the detection device of road signs as claimed in claim 9, which is characterized in that the preprocessing module, specifically
For:
By the road signs image scaling to default resolution.
11. the detection device of the road signs as described in claim 9 or 10, which is characterized in that the setup module, tool
Body is used for:
Pretreated image is divided into multiple images unit, and will be schemed accordingly according to the position feature of circular traffic sign board
As unit is set as the area-of-interest.
12. the detection device of road signs as claimed in claim 9, which is characterized in that the first detection module root
The gradient information g of edge detection point is calculated according to following formula:
Wherein, gx=[P (i+1, j)-P (i-1, j)]/2, gy=[P (i, j+1)-P (i, j-1)]/2, (i, j) are region of interest
Edge detection point in domain.
13. the detection device of road signs as claimed in claim 9, which is characterized in that second detection module, tool
Body is used for:
Round inspection is carried out to the area-of-interest according to the gradient information of the edge detection point using rapidly radially symmetry algorithm
It surveys to generate loop truss result, wherein the loop truss result includes first vegetarian refreshments and its corresponding radial symmetry transform result;
Wherein, the judgment module judges whether there is first vegetarian refreshments that radial symmetry transform result is greater than preset value, and exists
When radial symmetry transform result is greater than first vegetarian refreshments of the preset value, judge there is round hand in the road signs image
Logical sign board.
14. the detection device of road signs as claimed in claim 13, which is characterized in that further include:
Screening module, the element point for being greater than preset value to radial symmetry transform result screen.
15. the detection device of road signs as claimed in claim 14, which is characterized in that the screening module is in diameter
When the element point for being greater than the preset value to symmetry transformation result is screened, if any two radial symmetry transform result is big
In the corresponding center of circle (P of first vegetarian refreshments of the preset value1(x),P1(y))、(P2(x),P2(y)) and loop truss radius n satisfaction is following
Either condition then retains the biggish first vegetarian refreshments of radial symmetry transform result:
(1)|P1(x)-P2(x)|+|P1(y)-P2(y) | < 0.6*n,
(2)|P1(x)-P2(x) | < 0.5*n,
(3)|P1(y)-P2(y)|<0.5*n。
16. a kind of automobile, which is characterized in that the detection including the road signs as described in any one of claim 9-15
Device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710876728.4A CN109558768A (en) | 2017-09-25 | 2017-09-25 | The detection method of automobile and road signs, device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710876728.4A CN109558768A (en) | 2017-09-25 | 2017-09-25 | The detection method of automobile and road signs, device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109558768A true CN109558768A (en) | 2019-04-02 |
Family
ID=65861951
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710876728.4A Pending CN109558768A (en) | 2017-09-25 | 2017-09-25 | The detection method of automobile and road signs, device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109558768A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112612861A (en) * | 2020-12-18 | 2021-04-06 | 广州方纬智慧大脑研究开发有限公司 | Method, device, equipment and medium for checking reasonability of road traffic sign line |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104766046A (en) * | 2015-02-06 | 2015-07-08 | 哈尔滨工业大学深圳研究生院 | Detection and recognition algorithm conducted by means of traffic sign color and shape features |
CN105469124A (en) * | 2015-11-20 | 2016-04-06 | 厦门雅迅网络股份有限公司 | Traffic sign classification method |
CN105740877A (en) * | 2014-12-09 | 2016-07-06 | 比亚迪股份有限公司 | Traffic sign recognition method and device, and vehicle |
US20160321511A1 (en) * | 2015-04-29 | 2016-11-03 | Mando Corporation | Method and apparatus for confirmation of relevant white inner circle in environment of circular traffic sign recognition |
-
2017
- 2017-09-25 CN CN201710876728.4A patent/CN109558768A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105740877A (en) * | 2014-12-09 | 2016-07-06 | 比亚迪股份有限公司 | Traffic sign recognition method and device, and vehicle |
CN104766046A (en) * | 2015-02-06 | 2015-07-08 | 哈尔滨工业大学深圳研究生院 | Detection and recognition algorithm conducted by means of traffic sign color and shape features |
US20160321511A1 (en) * | 2015-04-29 | 2016-11-03 | Mando Corporation | Method and apparatus for confirmation of relevant white inner circle in environment of circular traffic sign recognition |
CN105469124A (en) * | 2015-11-20 | 2016-04-06 | 厦门雅迅网络股份有限公司 | Traffic sign classification method |
Non-Patent Citations (4)
Title |
---|
GARETH LOY ET.AL: "A Fast Radial Symmetry Transform for Detecting Points of Interest", 《ECCV 2002》 * |
LEI ZHU ET.AL: "Detection and Recognition of Speed Limit Sign from Video", 《ACIIDS 2016》 * |
尹涛: "自然场景下交通标志牌的检测与识别", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
肖志涛 等: "基于径向对称变换的速度标识牌检测方法", 《计算机工程与设计》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112612861A (en) * | 2020-12-18 | 2021-04-06 | 广州方纬智慧大脑研究开发有限公司 | Method, device, equipment and medium for checking reasonability of road traffic sign line |
CN112612861B (en) * | 2020-12-18 | 2024-04-30 | 广州方纬智慧大脑研究开发有限公司 | Rationality checking method, device, equipment and medium for road traffic sign marking |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110246108B (en) | Image processing method, device and computer readable storage medium | |
CN104392205B (en) | A kind of recognition methods and system of improper license plate | |
CN110197589B (en) | Deep learning-based red light violation detection method | |
CN106056155B (en) | Superpixel segmentation method based on boundary information fusion | |
DE102011106050B4 (en) | Shadow removal in an image captured by a vehicle-based camera for detection of a clear path | |
CN106886987B (en) | A kind of train license plate binary image interfusion method | |
CN108921089A (en) | Method for detecting lane lines, device and system and storage medium | |
CN111898523A (en) | Remote sensing image special vehicle target detection method based on transfer learning | |
CN104809461B (en) | The licence plate recognition method and system of binding sequence image super-resolution rebuilding | |
EP2851841A2 (en) | System and method of alerting a driver that visual perception of pedestrian may be difficult | |
CN105469046B (en) | Based on the cascade vehicle model recognizing method of PCA and SURF features | |
CN109858310A (en) | Vehicles and Traffic Signs detection method | |
CN108090459B (en) | Traffic sign detection and identification method suitable for vehicle-mounted vision system | |
Shopa et al. | Traffic sign detection and recognition using OpenCV | |
CN103295021A (en) | Method and system for detecting and recognizing feature of vehicle in static image | |
CN101246593B (en) | Color image edge detection method and apparatus | |
CN110211518A (en) | Circuit device and electronic equipment | |
CA2815591C (en) | Method for detecting and recognising an object in an image, and an apparatus and a computer program therefor | |
CN112712057A (en) | Traffic signal identification method and device, electronic equipment and storage medium | |
CN108205649A (en) | Driver drives to take the state identification method and device of phone | |
CN107704833A (en) | A kind of front vehicles detection and tracking based on machine learning | |
CN111680580A (en) | Red light running identification method and device, electronic equipment and storage medium | |
DE102011005780B4 (en) | Method and device for determining a distance between a vehicle and an object | |
CN109558768A (en) | The detection method of automobile and road signs, device | |
JP5370375B2 (en) | Image recognition device |
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: 20190402 |