CN105718860B - Localization method and system based on driving safety map and binocular Traffic Sign Recognition - Google Patents
Localization method and system based on driving safety map and binocular Traffic Sign Recognition Download PDFInfo
- Publication number
- CN105718860B CN105718860B CN201610027611.4A CN201610027611A CN105718860B CN 105718860 B CN105718860 B CN 105718860B CN 201610027611 A CN201610027611 A CN 201610027611A CN 105718860 B CN105718860 B CN 105718860B
- Authority
- CN
- China
- Prior art keywords
- image
- vehicle
- red
- traffic sign
- red color
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
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 present invention provides a kind of localization method and system based on driving safety map and binocular Traffic Sign Recognition, by carrying out primary positioning to the vehicle in driving using positioning system in high-precision map；Vehicle forward image is acquired simultaneously, the traffic sign in image is detected and identified；And the coordinate of traffic sign is obtained in high-precision Map recognition, the spacing between vehicle and mark is measured, the coordinate for comparing traffic sign calculates the position of vehicle, realizes vehicle location.The present invention is on the basis of conventional navigation data, increase the acquisition of road signs, booster action is carried out using the positioning to vehicle of road sign, the marker coordinates and size identified by left and right lens camera, carry out the ranging between Vehicles and Traffic Signs, and the position of vehicle is calculated according to the spatial position coordinate for having had traffic sign in high-precision map, to provide the coordinate setting of sub-meter grade, foundation can be based on the topological network in lane.
Description
Technical field
The present invention relates to vehicle networking technical fields, and in particular to one kind is known based on driving safety map and binocular traffic sign
Other localization method and system.
Background technique
With the development of computer science and robot technology, intelligent driving vehicle is in military, civilian and scientific research etc.
All various aspects are widely used, it has concentrated structure, electronics, cybernetics and artificial intelligence etc. multi-disciplinary newest
Research achievement has broad application prospects.
For intelligent driving vehicle, vehicle location is its key components, and vehicle location is intelligent automobile decision
The basis of control.GPS (Global Positioning System, global positioning system are generally used at present to the positioning of vehicle
System) location technology realization.Vehicle determines the vehicle position according to GPS positioning technology, carries out trajectory planning and makes control
Decision.But since the position error that GPS positioning technology can reach is 10 meters or so, using GPS positioning technology to vehicle
It is very inaccurate for carrying out positioning；For vehicle driving in bridge, tunnel, or encounter bad weather such as heavy rain severe snow, GPS
It will appear no signal condition, vehicle can not thus be positioned.So if relying solely on GPS positioning mode to the vehicle
It is positioned, huge hidden danger will certainly be brought for the travel safety of intelligent vehicle, so how to be accurately positioned to vehicle
Make the technical problem that this field is anxious to be resolved.
Summary of the invention
In view of this, it is necessary to provide one kind target vehicle can be carried out it is pinpoint based on driving safety map and
The localization method and system of binocular Traffic Sign Recognition.
A kind of localization method based on driving safety map and binocular Traffic Sign Recognition, it is described to be based on driving safety map
And binocular Traffic Sign Recognition localization method the following steps are included:
S1, using positioning system the vehicle in driving is carried out in the high-precision map based on pavement of road it is primary fixed
Position；
S2, acquisition vehicle forward image, are detected and are identified to the traffic sign in image；
S3, the coordinate of traffic sign is obtained in high-precision Map recognition, measures the spacing between vehicle and mark, and compare
The coordinate of traffic sign calculates the position of vehicle, realizes vehicle location.
Preferably, the traffic sign includes speed(-)limit sign and emergency telephone.
Preferably, the speed(-)limit sign in image is detected and is identified, specifically include it is following step by step:
S21, acquisition vehicle forward image, are transformed into HSV color space from RGB color for the image of acquisition, and count
Calculate red color bitmaps and red color intensity；
S22, mean filter is carried out to red color bitmaps, obtains red mean value image；And mean filter is carried out to red color intensity,
Red color intensity mean value image is obtained, thus the red color bitmaps after being optimized；
S23, region growing is carried out to the red color bitmaps after optimization, be arranged minimum widith and height threshold to red area into
Row screening, the red color bitmaps after being screened；
S24, external convex polygon is asked to each red area, then subtracts red color bitmaps obtained in step S23, can obtain
To red inner zone bitmap；
S25, multilayer screening is carried out to red inner zone bitmap, obtains the speed(-)limit sign interior zone eventually detected；
S26, training classifier are inputted using histograms of oriented gradients as feature, linear SVM are generated, to speed limit
Mark is identified.
Preferably, the emergency telephone in image is detected and is identified, specifically include it is following step by step:
S21, acquisition vehicle forward image, are converted into gray level image by RGB image for the image of acquisition；
S22, using the straight line in LSD line detection algorithm detection image, to straight line angle, length, neighbouring relations, distance
Given threshold is filtered to remove noise straight line, obtains the straight line for meeting threshold condition；
S23, half rectangle of level and vertical half rectangle are found out, judges to be closed by half rectangle of level and vertical half rectangle
Rectangle, and judge whether there is peripheral rectangle；
S24, straight line, the rear judgement for carrying out connected domain are drawn to peripheral rectangle.Multiple screening conditions are set, multi-deck screen is carried out
Choosing, obtains the emergency call region eventually detected；
S25, training classifier are inputted using histograms of oriented gradients as feature, linear SVM are generated, to urgent
Phone logos are identified.
Preferably, classifier is trained using positive and negative sample image, the positive sample image is to remove red border
The interior zone of speed(-)limit sign afterwards, negative sample image are not include speed(-)limit sign but only by erroneous detection be speed limit by color detection
The region of mark.
A kind of positioning system based on driving safety map and binocular Traffic Sign Recognition, it is described to be based on driving safety map
And the positioning system of binocular Traffic Sign Recognition includes following functions module:
Primary locating module, for utilizing positioning system to the vehicle in driving in the high-precision map based on pavement of road
Carry out primary positioning；
Mark acquiring and identifying module is detected and is known to the traffic sign in image for acquiring vehicle forward image
Not；
Spacing is counter to push away locating module, for passing through the coordinate and size of high-precision Map recognition traffic sign, measures vehicle
Spacing between mark, and the coordinate for comparing traffic sign in high-precision map calculates the position of vehicle, realizes that vehicle is fixed
Position.
Preferably, the mark acquiring and identifying module includes speed(-)limit sign identification submodule and emergency telephone identification
Module.
Preferably, the speed(-)limit sign identification submodule includes following functions unit:
The image of acquisition is transformed by color space converting unit for acquiring vehicle forward image from RGB color
HSV color space, and calculate red color bitmaps and red color intensity；
Bitmap optimizes unit, for carrying out mean filter to red color bitmaps, obtains red mean value image；And to red color intensity
Mean filter is carried out, red color intensity mean value image is obtained, thus the red color bitmaps after being optimized；
Minimum widith and height threshold is arranged for carrying out region growing to the red color bitmaps after optimization in bitmap screening unit
Value screens red area, the red color bitmaps after being screened；
Bitmap deletes unit, for seeking external convex polygon to each red area, then subtracts in bitmap screening unit
Red inner zone bitmap can be obtained in the red color bitmaps arrived；
Interior zone acquiring unit is obtained and is eventually detected for carrying out multilayer screening to red inner zone bitmap
Speed(-)limit sign interior zone；
Speed(-)limit sign recognition unit generates linear for training classifier to input using histograms of oriented gradients as feature
Support vector machines identifies speed(-)limit sign.
Preferably, the emergency telephone identification submodule includes following functions unit:
The image of acquisition is converted into grayscale image by RGB image for acquiring vehicle forward image by image conversion unit
Picture；
Straight line filter element, for using the straight line in LSD line detection algorithm detection image, to straight line angle, length,
Neighbouring relations, apart from given threshold, be filtered to remove noise straight line, obtain the straight line for meeting threshold condition；
Rectangle searching unit passes through half rectangle of level and vertical half square for finding out half rectangle of level and vertical half rectangle
Shape judges to obtain enclosing square, and judges whether there is peripheral rectangle；
Mark region detection unit for drawing straight line, the rear judgement for carrying out connected domain to peripheral rectangle, and is arranged multiple
Screening conditions carry out multilayer screening, obtain the emergency telephone region eventually detected；
Emergency telephone recognition unit is generated for training classifier to input using histograms of oriented gradients as feature
Linear SVM identifies emergency telephone.
Preferably, vehicle forward image is acquired using binocular camera, the binocular camera carries out the image of acquisition
Matching and correction and the distance between vision measurement traffic sign and vehicle.
Localization method and system of the present invention based on driving safety map and binocular Traffic Sign Recognition, due to safety
The acquisition of driving map combined high precision navigation data and high-precision map location technology are prepared, traditional high-precision navigation
Data element is known as lane relevant information, the curvature and label of road and the guidance information in lane and constraint information etc., utilizes ground
Figure accounts for leading role to vehicle location.The present invention increases the acquisition of road signs on the basis of conventional navigation data,
Booster action is carried out to the positioning of vehicle using road sign, the marker coordinates that identify by left and right lens camera and big
It is small, the ranging between Vehicles and Traffic Signs is carried out, and according to the spatial position coordinate for having had traffic sign in high-precision map
The position of vehicle is calculated, to provide the coordinate setting of sub-meter grade, foundation can be provided and be based on based on the topological network in lane
The navigation application of lane grade, realizes vehicle location.
Compared with prior art, technical solution of the present invention has following obvious advantages:
1, positioning accuracy is high, and high-precision navigation data is utilized, and increases the acquisition for expanding element on road, positioning accurate
Degree is sub-meter grade；
2, security performance is strong, and map datum member is known as lane relevant information, and the curvature of road is with label and based on lane
Guidance information and constraint information, for vehicle driving trace planning and control decision safety guarantee is provided；
3, location efficiency is high, is 15 frames/second to speed(-)limit sign and emergency telephone identification frequency, binocular vision is utilized
Distance measurement function carries out real time correction to positioning result；
4, positioning is without careless mistake, and for the imperfect tunnel bridge opening of GPS signal and severe weather conditions, visual token positioning can
Effectively to make up.
Detailed description of the invention
Fig. 1 is the flow chart element of the localization method of the present invention based on driving safety map and binocular Traffic Sign Recognition
Figure；
Fig. 2 is the flow diagram that the speed(-)limit sign of the present invention in image is detected and identified；
Fig. 3 is the flow diagram that the emergency telephone of the present invention in image is detected and identified；
Fig. 4 is the module frame of the positioning system of the present invention based on driving safety map and binocular Traffic Sign Recognition
Figure；
Fig. 5 is the submodule block diagram for indicating acquiring and identifying module in Fig. 4；
Fig. 6 is the submodule block diagram of speed(-)limit sign identification submodule in Fig. 5；
Fig. 7 is the submodule block diagram of emergency telephone identification submodule in Fig. 5.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated, it should be understood that and the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
As shown in Figure 1, the embodiment of the present invention is provided and a kind of is determined based on driving safety map and binocular Traffic Sign Recognition
Position method, the localization method based on driving safety map and binocular Traffic Sign Recognition the following steps are included:
S1, using positioning system the vehicle in driving is carried out in the high-precision map based on pavement of road it is primary fixed
Position；
S2, acquisition vehicle forward image, are detected and are identified to the traffic sign in image；
S3, the coordinate of traffic sign is obtained in high-precision Map recognition, measures the spacing between vehicle and mark, and compare
The coordinate of traffic sign calculates the position of vehicle, realizes vehicle location.
Since the acquisition of safe driving map combined high precision navigation data and high-precision map location technology are prepared,
Traditional high-precision navigation data member is known as lane relevant information, the curvature and label of road and the guidance information and system in lane
About information etc. accounts for leading role to vehicle location using map.The present invention increases road on the basis of conventional navigation data
The acquisition of traffic sign carries out booster action using the positioning to vehicle of road sign, is identified by left and right lens camera
The marker coordinates and size come carry out the ranging between Vehicles and Traffic Signs, and according to there being traffic mark in high-precision map
The spatial position coordinate of will calculates the position of vehicle, to provide the coordinate setting of sub-meter grade, foundation can be based on lane
Topological network provides the navigation application based on lane grade, realizes vehicle location.
Specifically, the high-precision map is what the every terms of information based on pavement of road made, road can be accurately shown
The accurately shape of diagram data and road or fare, pavement marking etc., to be carried out on the high-precision map
Detailed path planning or vehicle control.
Target vehicle position detection that positioning system provides as a result, position of the instruction target vehicle on high-precision map,
Realize Primary Location.The positioning system mainly takes guideboard position control point, lane center, crossing central point to target carriage
It is positioned.
Vehicle forward image is acquired, the traffic sign in image is detected and identified；The traffic sign includes limit
Speed mark and emergency telephone.
Wherein, to the detection and identification of the speed(-)limit sign, mainly according to its red circular outline border, color and shape are utilized
Shape information is filtered judgement, detects to carry out identification classification with classifier behind speed(-)limit sign region.As shown in Fig. 2, specific packet
Include it is following step by step:
S21, acquisition vehicle forward image, are transformed into HSV color space from RGB color for the image of acquisition, and count
Calculate red color bitmaps and red color intensity；The function of the red color bitmaps R (x) calculates as follows:
The red color intensity IR(x) function calculates as follows:
S22, mean filter is carried out to red color bitmaps R (x), obtains red mean value imageTo red color intensity IR(x) into
Row mean filter obtains red color intensity mean value imageSpecific function calculates as follows:
Thus the red color bitmaps after being optimized；
S23, region growing is carried out to the red color bitmaps after optimization, be arranged minimum widith and height threshold to red area into
Row screening, the red color bitmaps after being screened；
S24, external convex polygon is asked to each red area, then subtracts red color bitmaps obtained in step S23, can obtain
To red inner zone bitmap；
S25, multilayer screening is carried out to red inner zone bitmap, obtains the speed(-)limit sign interior zone eventually detected；
Screening conditions are the area of interior zone, the width and height of interior zone, the aspect ratio of interior zone, the shape of interior zone
Deng.
S26, training classifier are inputted using histograms of oriented gradients as feature, linear SVM are generated, to speed limit
Mark is identified.Specifically classifier is trained using positive and negative sample image, the positive sample image is to remove red side
The interior zone of speed(-)limit sign behind boundary, negative sample image are not include speed(-)limit sign but be only limited by color detection by erroneous detection
The region of speed mark, the quantity of the positive and negative sample image each 1000.
The emergency telephone in image is detected and is identified, straight line is detected first, according to rectangular characteristic, mistake
Noise filtering straight line finally leaves the straight line of mark, after detecting mark region, then the identification indicated, it is identified with classifier
Phone logos out, as shown in figure 3, specifically include it is following step by step:
S21, acquisition vehicle forward image, are converted into gray level image by RGB image for the image of acquisition；
S22, using the straight line in LSD line detection algorithm detection image, to straight line angle, length, neighbouring relations, distance
Given threshold is filtered to remove noise straight line, obtains the straight line for meeting threshold condition；
Specifically, the threshold condition of the horizontal linear is -5 ° of 5 ° of < angle <；
The threshold condition of vertical straight line is 70 ° or < -70 ° of angle of angle >, and wherein sign indicates the side of straight line
To forming the straight line of the rectangle of emergency call substantially in level of approximation and vertical both direction；
The threshold condition of straight length is 10 < linelength < 130, and wherein the metering of length is as unit of pixel
The threshold condition of linear distance is linesdis < 8, and distance is also as unit of pixel；
The horizontal linear and vertical straight line that meet above-mentioned relation are respectively put into container a hor_arr and ver_arr.
S23, half rectangle of level and vertical half rectangle are found out, judges to be closed by half rectangle of level and vertical half rectangle
Rectangle, and judge whether there is peripheral rectangle；
Specifically, on the basis of horizontal linear hor_arr [i], finding out in the judgement of half rectangle of level and meeting distance relation
In the vertical straight line of linesdis < 8, and num1 is counted, as num1 >=2, then judges whether the vertical straight line that front is found out is full
The difference relationship of foot length degree is lengthdis < 8, meets distance relation distance > hor_arr [i] -5, that is, is judged as one
A half rectangle of level.And if two vertical edges of half rectangle of two of them level are all identical, the two half squares of level
Shape is an enclosing square, if half rectangle of level does not find half rectangle of level of pairing, one becomes one alone for it
Enclosing square.
The method for judging vertical half rectangle is consistent with the judgment method of half rectangle of level, but needs to add a judgement in front
Condition, if the vertical straight line occupied by half rectangle of level, the no longer reference line for vertical half rectangle.
If a rectangle is in another rectangle, and the horizontal sides of the rectangle of outside are greater than the horizontal sides of interior rectangle, outer square
The vertical edge of shape is greater than the vertical edge of interior rectangle, while the difference for meeting the center point coordinate of two rectangles meets: xdis < 10,
0 < ydis < 20, i.e. judgement have peripheral rectangle., whereas if to can not find a peripheral rectangle matching for a rectangle, then
It itself is peripheral rectangle.
S24, straight line is drawn to the peripheral rectangle of above-mentioned acquisition, carries out the judgement of connected domain, and multiple screening conditions are set,
Multilayer screening is carried out, the emergency call region eventually detected is obtained；The screening conditions are the area of connected domain, connected domain
Width and height, the aspect ratio of connected domain, the shape etc. of connected domain.
S25, training classifier are inputted using histograms of oriented gradients as feature, linear SVM are generated, to urgent
Phone logos are identified.Training method one in the training method of the classifier and the detection and identification process of speed(-)limit sign
It causes.
Further, the present invention is acquired vehicle front image using binocular camera, the binocular camera shooting function
It is enough the image of acquisition to be matched and corrected and the distance between vision measurement traffic sign and vehicle, to pass through traffic
The position of mark positioning target vehicle provides technical foundation.
According to the above-mentioned localization method based on driving safety map and binocular Traffic Sign Recognition, the present invention also provides one kind
Positioning system based on driving safety map and binocular Traffic Sign Recognition, as shown in figure 4, the positioning system includes following function
Energy module:
Primary locating module 10, for utilizing positioning system in driving in the high-precision map based on pavement of road
Vehicle carries out primary positioning；
Indicate acquiring and identifying module 20, for acquiring vehicle forward image, to the traffic sign in image carry out detection and
Identification；
Spacing is counter to push away locating module 30, for passing through the coordinate and size of high-precision Map recognition traffic sign, measures vehicle
Spacing between mark, and the coordinate for comparing traffic sign in high-precision map calculates the position of vehicle, realizes vehicle
Positioning.
Wherein, as shown in figure 5, the mark acquiring and identifying module includes speed(-)limit sign identification submodule 21 and emergency call
Landmark identification submodule 22.
As shown in fig. 6, the speed(-)limit sign identification submodule includes following functions unit:
Color space converting unit 211 turns the image of acquisition from RGB color for acquiring vehicle forward image
HSV color space is changed to, and calculates red color bitmaps and red color intensity；
Bitmap optimizes unit 212, for carrying out mean filter to red color bitmaps, obtains red mean value image；And to red
Intensity carries out mean filter, red color intensity mean value image is obtained, thus the red color bitmaps after being optimized；
Minimum widith and height is arranged for carrying out region growing to the red color bitmaps after optimization in bitmap screening unit 213
Threshold value screens red area, the red color bitmaps after being screened；
Bitmap deletes unit 214, for seeking external convex polygon to each red area, then subtracts in bitmap screening unit
Red inner zone bitmap can be obtained in obtained red color bitmaps；
Interior zone acquiring unit 215, for carrying out multilayer screening to red inner zone bitmap, acquisition is eventually detected
Speed(-)limit sign interior zone；
Speed(-)limit sign recognition unit 216 generates line for training classifier to input using histograms of oriented gradients as feature
Property support vector machines, identifies speed(-)limit sign.
As shown in fig. 7, the emergency telephone identification submodule includes following functions unit:
The image of acquisition is converted into gray scale by RGB image for acquiring vehicle forward image by image conversion unit 221
Image；
Straight line filter element 222, for using the straight line in LSD line detection algorithm detection image, to straight line angle, length
Degree, neighbouring relations, apart from given threshold, be filtered to remove noise straight line, obtain the straight line for meeting threshold condition；
Rectangle searching unit 223 passes through half rectangle of level and vertical half for finding out half rectangle of level and vertical half rectangle
Rectangle judges to obtain enclosing square, and judges whether there is peripheral rectangle；
Mark region detection unit 224 for drawing straight line, the rear judgement for carrying out connected domain to peripheral rectangle, and is arranged
Multiple screening conditions carry out multilayer screening, obtain the emergency telephone region eventually detected；
Emergency telephone recognition unit 225, it is raw for training classifier to input using histograms of oriented gradients as feature
Linear support vector machines, identifies emergency telephone.
Localization method and system of the present invention based on driving safety map and binocular Traffic Sign Recognition, due to safety
The acquisition of driving map combined high precision navigation data and high-precision map location technology are prepared, traditional high-precision navigation
Data element is known as lane relevant information, the curvature and label of road and the guidance information in lane and constraint information etc., utilizes ground
Figure accounts for leading role to vehicle location.The present invention increases the acquisition of road signs on the basis of conventional navigation data,
Booster action is carried out to the positioning of vehicle using road sign, the marker coordinates that identify by left and right lens camera and big
It is small, the ranging between Vehicles and Traffic Signs is carried out, and according to the spatial position coordinate for having had traffic sign in high-precision map
The position of vehicle is calculated, to provide the coordinate setting of sub-meter grade, foundation can be provided and be based on based on the topological network in lane
The navigation application of lane grade, realizes vehicle location.
Compared with prior art, technical solution of the present invention has following obvious advantages:
1, positioning accuracy is high, and high-precision navigation data is utilized, and increases the acquisition for expanding element on road, positioning accurate
Degree is sub-meter grade；
2, security performance is strong, and map datum member is known as lane relevant information, and the curvature of road is with label and based on lane
Guidance information and constraint information, for vehicle driving trace planning and control decision safety guarantee is provided；
3, location efficiency is high, is 15 frames/second to speed(-)limit sign and emergency telephone identification frequency, binocular vision is utilized
Distance measurement function carries out real time correction to positioning result；
4, positioning is without careless mistake, and for the imperfect tunnel bridge opening of GPS signal and severe weather conditions, visual token positioning can
Effectively to make up.
Apparatus above embodiment and embodiment of the method are one-to-one, the simple places of Installation practice, referring to method reality
Apply example.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with other
The difference of embodiment, the same or similar parts in each embodiment may refer to each other.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to functionality in the above description.This
A little functions are implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Specially
Industry technical staff can use different methods to achieve the described function each specific application, but this realization is not
It should be more than the scope of the present invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory, memory, read-only memory,
Electrically programmable ROM, electricity can sassafras except in programming ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical field institute it is public
In the storage medium for any other forms known.
The embodiment of the present invention is described with above attached drawing, but the invention is not limited to above-mentioned specific
Embodiment, the above mentioned embodiment is only schematical, rather than restrictive, those skilled in the art
Under the inspiration of the present invention, without breaking away from the scope protected by the purposes and claims of the present invention, it can also make very much
Form, all of these belong to the protection of the present invention.
Claims (4)
1. a kind of localization method based on driving safety map and binocular Traffic Sign Recognition, which is characterized in that described to be based on driving
Sail the localization method of safe map and binocular Traffic Sign Recognition the following steps are included:
S1, using positioning system primary positioning is carried out to the vehicle in driving in the high-precision map based on pavement of road；
S2, acquisition vehicle forward image, are detected and are identified to the traffic sign in image；
The traffic sign includes speed(-)limit sign and emergency telephone, wherein the speed(-)limit sign in image is examined
Survey and identification, specifically include it is following step by step:
A1, acquisition vehicle forward image, are transformed into HSV color space from RGB color for the image of acquisition, and calculate red
Bitmap and red color intensity；
A2, mean filter is carried out to red color bitmaps, obtains red mean value image；And mean filter is carried out to red color intensity, it obtains
Red color intensity mean value image, thus the red color bitmaps after being optimized；
A3, region growing is carried out to the red color bitmaps after optimization, minimum widith is set and height threshold sieves red area
Choosing, the red color bitmaps after being screened；
A4, external convex polygon is asked to each red area, then subtracts red color bitmaps obtained in step a3, red can be obtained
Interior zone bitmap；
A5, multilayer screening is carried out to red inner zone bitmap, obtains the speed(-)limit sign interior zone eventually detected；
A6, training speed(-)limit sign classifier are inputted using histograms of oriented gradients as feature, linear SVM are generated, to limit
Fast mark is identified；
Wherein, the detection and recognition methods of the emergency telephone, including it is following step by step:
B1, acquisition vehicle forward image, are converted into gray level image by RGB image for the image of acquisition；
B2, using the straight line in LSD line detection algorithm detection image, straight line angle, length, neighbouring relations, distance are set
Threshold value is filtered to remove noise straight line, obtains the straight line for meeting threshold condition；
B3, half rectangle of level and vertical half rectangle are found out, judge to obtain enclosing square by half rectangle of level and vertical half rectangle,
And judge whether there is peripheral rectangle；
B4, straight line is drawn to peripheral rectangle, the rear judgement for carrying out connected domain is arranged multiple screening conditions, carries out multilayer screening, obtain
Take the emergency call region eventually detected；
B5, training emergency telephone classifier are inputted using histograms of oriented gradients as feature, generate linear SVM,
Emergency telephone is identified；
S3, the coordinate of traffic sign is obtained in high-precision Map recognition, measures the spacing between vehicle and mark, and compare traffic
The coordinate of mark calculates the position of vehicle, realizes vehicle location.
2. a kind of localization method based on driving safety map and binocular Traffic Sign Recognition according to claim 1,
It is characterized in that, speed(-)limit sign classifier is trained using positive and negative sample image, the positive sample image is to remove red side
The interior zone of speed(-)limit sign behind boundary, negative sample image are not include speed(-)limit sign but be only limited by color detection by erroneous detection
The region of speed mark.
3. a kind of positioning system based on driving safety map and binocular Traffic Sign Recognition, which is characterized in that described to be based on driving
The positioning system for sailing safe map and binocular Traffic Sign Recognition includes following functions module:
Primary locating module, in the high-precision map based on pavement of road utilize positioning system to the vehicle in driving into
The primary positioning of row；
Mark acquiring and identifying module is detected and is identified to the traffic sign in image for acquiring vehicle forward image；
The mark acquiring and identifying module includes speed(-)limit sign identification submodule and emergency telephone identification submodule, wherein
The speed(-)limit sign identification submodule includes following functions unit:
The image of acquisition is transformed into HSV from RGB color for acquiring vehicle forward image by color space converting unit
Color space, and calculate red color bitmaps and red color intensity；
Bitmap optimizes unit, for carrying out mean filter to red color bitmaps, obtains red mean value image；And red color intensity is carried out
Mean filter obtains red color intensity mean value image, thus the red color bitmaps after being optimized；
Minimum widith and height threshold pair is arranged for carrying out region growing to the red color bitmaps after optimization in bitmap screening unit
Red area is screened, the red color bitmaps after being screened；
Bitmap deletes unit, for seeking external convex polygon to each red area, then subtracts obtained in bitmap screening unit
Red inner zone bitmap can be obtained in red color bitmaps；
Interior zone acquiring unit obtains the speed limit eventually detected for carrying out multilayer screening to red inner zone bitmap
Indicate interior zone；
Speed(-)limit sign recognition unit is generated for training speed(-)limit sign classifier to input using histograms of oriented gradients as feature
Linear SVM identifies speed(-)limit sign；
Wherein, the emergency telephone identification submodule includes following functions unit:
The image of acquisition is converted into gray level image by RGB image for acquiring vehicle forward image by image conversion unit；
Straight line filter element, for using the straight line in LSD line detection algorithm detection image, to straight line angle, length, adjacent
Relationship, apart from given threshold, be filtered to remove noise straight line, obtain the straight line for meeting threshold condition；
Rectangle searching unit is sentenced for finding out half rectangle of level and vertical half rectangle by half rectangle of level and vertical half rectangle
It is disconnected to obtain enclosing square, and judge whether there is peripheral rectangle；
For drawing straight line, the rear judgement for carrying out connected domain to peripheral rectangle, and multiple screenings are arranged in mark region detection unit
Condition carries out multilayer screening, obtains the emergency telephone region eventually detected；
Emergency telephone recognition unit, for training emergency telephone classifier defeated using histograms of oriented gradients as feature
Enter, generates linear SVM, emergency telephone is identified；
Spacing is counter to push away locating module, for passing through the coordinate and size of high-precision Map recognition traffic sign, measures vehicle and mark
Spacing between will, and the coordinate for comparing traffic sign in high-precision map calculates the position of vehicle, realizes vehicle location.
4. a kind of positioning system based on driving safety map and binocular Traffic Sign Recognition according to claim 3,
It is characterized in that, acquiring vehicle forward image using binocular camera, the binocular camera matches the image of acquisition
And correction and the distance between vision measurement traffic sign and vehicle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610027611.4A CN105718860B (en) | 2016-01-15 | 2016-01-15 | Localization method and system based on driving safety map and binocular Traffic Sign Recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610027611.4A CN105718860B (en) | 2016-01-15 | 2016-01-15 | Localization method and system based on driving safety map and binocular Traffic Sign Recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105718860A CN105718860A (en) | 2016-06-29 |
| CN105718860B true CN105718860B (en) | 2019-09-10 |
Family
ID=56147226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610027611.4A Active CN105718860B (en) | 2016-01-15 | 2016-01-15 | Localization method and system based on driving safety map and binocular Traffic Sign Recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105718860B (en) |
Families Citing this family (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106043170A (en) * | 2016-07-01 | 2016-10-26 | 蔡雄 | Police vehicle with autonomous navigation function |
| CN109791408B (en) * | 2016-09-27 | 2022-04-26 | 日产自动车株式会社 | Self-position estimation method and self-position estimation device |
| KR102395283B1 (en) * | 2016-12-14 | 2022-05-09 | 현대자동차주식회사 | Apparatus for controlling automatic driving, system having the same and method thereof |
| CN106530794B (en) * | 2016-12-28 | 2019-03-01 | 上海仪电数字技术股份有限公司 | The automatic identification and calibration method and system of carriage way |
| US10670416B2 (en) * | 2016-12-30 | 2020-06-02 | DeepMap Inc. | Traffic sign feature creation for high definition maps used for navigating autonomous vehicles |
| CN106651953B (en) * | 2016-12-30 | 2019-10-18 | 山东大学 | A kind of vehicle position and orientation estimation method based on traffic sign |
| CN108286977A (en) * | 2017-01-09 | 2018-07-17 | 北京四维图新科技股份有限公司 | A kind of offline update method of navigation data, device and hybrid navigation system |
| CN106980855B (en) * | 2017-04-01 | 2020-04-17 | 公安部交通管理科学研究所 | Traffic sign rapid identification and positioning system and method |
| CN107334469A (en) * | 2017-07-24 | 2017-11-10 | 北京理工大学 | Non-contact more people's method for measuring heart rate and device based on SVMs |
| CN109520495B (en) * | 2017-09-18 | 2022-05-13 | 财团法人工业技术研究院 | Navigation positioning device and navigation positioning method using same |
| CN109102711A (en) * | 2017-12-15 | 2018-12-28 | 蔚来汽车有限公司 | The method for determining road safety velocity information |
| CN108254776A (en) * | 2017-12-25 | 2018-07-06 | 东风汽车集团有限公司 | Tunnel placement system and method based on curb fluorescent reflection and binocular camera |
| CN109961643A (en) * | 2017-12-25 | 2019-07-02 | 博世汽车部件(苏州)有限公司 | Based reminding method, alarm set and mobile unit |
| CN110018503B (en) * | 2018-01-10 | 2023-04-14 | 上海汽车集团股份有限公司 | Vehicle positioning method and positioning system |
| CN110164166B (en) | 2018-02-11 | 2021-01-08 | 北京图森智途科技有限公司 | Vehicle positioning system, method and device |
| CN110162028A (en) | 2018-02-11 | 2019-08-23 | 北京图森未来科技有限公司 | A kind of vehicle location shutdown system, method and apparatus |
| CN108416045A (en) * | 2018-03-15 | 2018-08-17 | 斑马网络技术有限公司 | Location acquiring method, device, terminal device and the server of means of transportation |
| US20190316929A1 (en) * | 2018-04-17 | 2019-10-17 | Faraday&Future Inc. | System and method for vehicular localization relating to autonomous navigation |
| CN110388929B (en) * | 2018-04-20 | 2022-04-15 | 比亚迪股份有限公司 | Navigation map updating method, device and system |
| CN109029339B (en) * | 2018-05-09 | 2023-10-13 | 苏州天瞳威视电子科技有限公司 | Traffic sign distance measurement method and device based on vision |
| CN113170079A (en) * | 2018-07-30 | 2021-07-23 | 小马智行 | System and method for calibrating an in-vehicle camera |
| CN108871368B (en) * | 2018-08-06 | 2020-12-29 | 武汉中海庭数据技术有限公司 | Method and system for constructing transverse topological relation of high-precision map lane and memory |
| CN111046709B (en) * | 2018-10-15 | 2021-02-09 | 广州汽车集团股份有限公司 | Vehicle lane level positioning method and system, vehicle and storage medium |
| JP7155284B2 (en) * | 2018-10-24 | 2022-10-18 | パイオニア株式会社 | Measurement accuracy calculation device, self-position estimation device, control method, program and storage medium |
| CN109446973B (en) * | 2018-10-24 | 2021-01-22 | 中车株洲电力机车研究所有限公司 | Vehicle positioning method based on deep neural network image recognition |
| CN109635639B (en) * | 2018-10-31 | 2020-06-09 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and storage medium for detecting position of traffic sign |
| US20200149896A1 (en) * | 2018-11-09 | 2020-05-14 | GM Global Technology Operations LLC | System to derive an autonomous vehicle enabling drivable map |
| CN111380542B (en) * | 2018-12-28 | 2023-03-17 | 沈阳美行科技股份有限公司 | Vehicle positioning and navigation method and device and related system |
| CN109870689B (en) * | 2019-01-08 | 2021-06-04 | 武汉中海庭数据技术有限公司 | Lane-level positioning method and system based on matching of millimeter wave radar and high-precision vector map |
| CN109829401A (en) * | 2019-01-21 | 2019-05-31 | 深圳市能信安科技股份有限公司 | Traffic sign recognition method and device based on double capture apparatus |
| KR102604298B1 (en) * | 2019-01-28 | 2023-11-17 | 에스케이텔레콤 주식회사 | Apparatus and method for estimating location of landmark and computer recordable medium storing computer program thereof |
| CN110954112B (en) * | 2019-03-29 | 2021-09-21 | 北京初速度科技有限公司 | Method and device for updating matching relation between navigation map and perception image |
| CN111750882B (en) * | 2019-03-29 | 2022-05-27 | 北京魔门塔科技有限公司 | Method and device for correcting vehicle pose during initialization of navigation map |
| CN110334608B (en) * | 2019-06-12 | 2021-05-14 | 浙江亚太机电股份有限公司 | Real-time traffic speed limit sign identification method of vehicle-mounted image |
| CN110288614B (en) * | 2019-06-24 | 2022-01-11 | 睿魔智能科技(杭州)有限公司 | Image processing method, device, equipment and storage medium |
| CN110415299B (en) * | 2019-08-02 | 2023-02-24 | 山东大学 | Vehicle position estimation method based on set guideboard under motion constraint |
| CN110784680B (en) * | 2019-08-09 | 2021-02-02 | 中国第一汽车股份有限公司 | Vehicle positioning method and device, vehicle and storage medium |
| CN110501018B (en) * | 2019-08-13 | 2021-11-02 | 广东星舆科技有限公司 | Traffic sign information acquisition method for high-precision map production |
| CN113011212B (en) * | 2019-12-19 | 2024-04-05 | 北京四维图新科技股份有限公司 | Image recognition method and device and vehicle |
| CN113012422B (en) * | 2019-12-20 | 2023-03-21 | 百度在线网络技术（北京）有限公司 | Bus positioning method, device, equipment and storage medium |
| CN113124752B (en) * | 2021-03-01 | 2022-11-15 | 武汉理工大学 | System and method for positioning automobile based on roadside visual tag |
| CN113361348B (en) * | 2021-05-25 | 2022-04-08 | 郑州信大捷安信息技术股份有限公司 | Safe traffic sign error surveying method and system based on V2X |
| CN114485684B (en) * | 2021-12-31 | 2023-11-28 | 武汉中海庭数据技术有限公司 | Map road matching method and system based on ground element topological relation |
| CN115965926B (en) * | 2023-03-16 | 2023-06-02 | 四川京炜数字科技有限公司 | Vehicle-mounted road sign marking inspection system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101221049A (en) * | 2006-12-04 | 2008-07-16 | 韩国电子通信研究院 | Vehicle position estimation device and its method |
| CN102609702A (en) * | 2012-02-03 | 2012-07-25 | 中山大学 | Method and system for rapidly positioning road guide sign |
| CN103034836A (en) * | 2011-09-29 | 2013-04-10 | 株式会社理光 | Road sign detection method and device |
| CN104616021A (en) * | 2014-12-24 | 2015-05-13 | 深圳市腾讯计算机***有限公司 | Method and device for processing traffic sign symbols |
| CN105021201A (en) * | 2015-08-17 | 2015-11-04 | 武汉光庭信息技术有限公司 | System and method for reversely educing position of automobile by means of coordinates of traffic signboards |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7928905B2 (en) * | 2008-10-26 | 2011-04-19 | Mitac International Corp. | Method of using road signs to augment global positioning system (GPS) coordinate data for calculating a current position of a personal navigation device |
| JP2015114126A (en) * | 2013-12-09 | 2015-06-22 | 株式会社デンソー | Vehicle location detection device |
-
2016
- 2016-01-15 CN CN201610027611.4A patent/CN105718860B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101221049A (en) * | 2006-12-04 | 2008-07-16 | 韩国电子通信研究院 | Vehicle position estimation device and its method |
| CN103034836A (en) * | 2011-09-29 | 2013-04-10 | 株式会社理光 | Road sign detection method and device |
| CN102609702A (en) * | 2012-02-03 | 2012-07-25 | 中山大学 | Method and system for rapidly positioning road guide sign |
| CN104616021A (en) * | 2014-12-24 | 2015-05-13 | 深圳市腾讯计算机***有限公司 | Method and device for processing traffic sign symbols |
| CN105021201A (en) * | 2015-08-17 | 2015-11-04 | 武汉光庭信息技术有限公司 | System and method for reversely educing position of automobile by means of coordinates of traffic signboards |
Non-Patent Citations (1)
| Title |
|---|
| Light-weight Localization for Vehicles using Road Markings;Ananth Ranganathan 等;《2013 IEEE/RSJ International Conference on Intelligent Robots and Systems》;20131107;第921-927页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105718860A (en) | 2016-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105718860B (en) | Localization method and system based on driving safety map and binocular Traffic Sign Recognition | |
| CN105783936B (en) | For the road markings drawing and vehicle positioning method and system in automatic Pilot | |
| CN105718870B (en) | Based on the preceding roadmarking extracting method to camera in automatic Pilot | |
| CN102436644B (en) | Unstructured road detection method based on adaptive edge registration | |
| CN110178167B (en) | Intersection violation video identification method based on cooperative relay of cameras | |
| US9489586B2 (en) | Traffic sign recognizing apparatus and operating method thereof | |
| CN205992300U (en) | For providing the electronic installation of the guidance information using crossing recognition result | |
| Yoo et al. | Gradient-enhancing conversion for illumination-robust lane detection | |
| CN107463918A (en) | Lane line extracting method based on laser point cloud and image data fusion | |
| CN107643086A (en) | A kind of vehicle positioning method, apparatus and system | |
| CN105654073B (en) | A kind of speed automatic control method of view-based access control model detection | |
| CN109738910A (en) | A kind of curb detection method based on three-dimensional laser radar | |
| CN110057373A (en) | For generating the method, apparatus and computer storage medium of fine semanteme map | |
| RU2636120C2 (en) | Three-dimensional object detecting device | |
| CN102915433B (en) | Character combination-based license plate positioning and identifying method | |
| CN107463890A (en) | A kind of Foregut fermenters and tracking based on monocular forward sight camera | |
| Almagambetov et al. | Mobile standards-based traffic light detection in assistive devices for individuals with color-vision deficiency | |
| CN108364466A (en) | A kind of statistical method of traffic flow based on unmanned plane traffic video | |
| CN104616502A (en) | License plate identification and positioning system based on combined type vehicle-road video network | |
| CN102354457A (en) | General Hough transformation-based method for detecting position of traffic signal lamp | |
| CN109635737A (en) | Automobile navigation localization method is assisted based on pavement marker line visual identity | |
| KR102204037B1 (en) | Alignment error correction system for collected location information by triangle mesh composed rand-mark and specified object | |
| CN103366179A (en) | Top-down view classification in clear path detection | |
| CN109325390A (en) | A kind of localization method combined based on map with FUSION WITH MULTISENSOR DETECTION and system | |
| CN104613928A (en) | Automatic tracking and air measurement method for optical pilot balloon theodolite |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |