CN107316486A - Pilotless automobile visual identifying system based on dual camera - Google Patents
Pilotless automobile visual identifying system based on dual camera Download PDFInfo
- Publication number
- CN107316486A CN107316486A CN201710559187.2A CN201710559187A CN107316486A CN 107316486 A CN107316486 A CN 107316486A CN 201710559187 A CN201710559187 A CN 201710559187A CN 107316486 A CN107316486 A CN 107316486A
- Authority
- CN
- China
- Prior art keywords
- region
- vehicle
- line
- module
- lane line
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
- G08G1/09623—Systems involving the acquisition of information from passive traffic signs by means mounted on the vehicle
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
- G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
- G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
- G08G1/096708—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control
- G08G1/096725—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control where the received information generates an automatic action on the vehicle control
Abstract
The present invention provides a kind of pilotless automobile visual identifying system based on dual camera.The pilotless automobile visual identifying system based on dual camera includes the first image capture module, the second image capture module, Lane detection module, vehicle identification module, Traffic Sign Recognition module, traffic lights detection and identification module.It is big that the pilotless automobile visual identifying system based on dual camera that the present invention is provided solves single camera processing data amount in the vehicle vision processing system of prior art, influences the technical problem of real-time.
Description
Technical field
The present invention relates to field of image recognition, and in particular to a kind of pilotless automobile visual identity based on dual camera
System.
Background technology
Existing intelligent vehicular visual processing system typically using single camera gathered data come to lane line, traffic sign,
The road informations such as traffic lights are identified.But in use, Lane detection wishes the most near region that camera is photographed
Domain is more near better from car, and the identification of traffic sign and traffic lights wishes that the region that camera is shot is more wide more remote better,
In order to be handled in real time.Need to shoot very big scope using single camera simultaneously, so need the shooting of higher pixel
Head, deals with data volume greatly, influences real-time.
The content of the invention
Single camera processing data amount is big in vehicle vision processing system to solve prior art, influences the skill of real-time
Art problem, the present invention provides a kind of pilotless automobile visual identifying system based on dual camera solved the above problems.
A kind of pilotless automobile visual identifying system based on dual camera, including the first image capture module, second
Image capture module, Lane detection module, vehicle identification module, Traffic Sign Recognition module, traffic lights detection are with knowing
Other module；
Described first image acquisition module, second image capture module are gathered and output image data, and the is designated as respectively
One view data, the second view data；
Described first image data input to the Lane detection module, the Lane detection module identifies described first
Lane line in view data, and fit track center line；
Described first image data are further input to the vehicle identification module, and the vehicle identification module is known according to the lane line
The scope for the lane line that other module is identified, identifies the vehicle in lane line and its residing region；
Second view data is inputted to the Traffic Sign Recognition module, and the Traffic Sign Recognition module identifies described
Traffic Sign Images in second view data, and it is fast and accurately classified；
Second view data is further input to traffic lights detection and identification module, the traffic lights detection with
Identification module is by identifying the color region of target area, so as to identify traffic lights and its dispaly state.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, the recognition methods of the Lane detection module specifically includes following steps：
Step 1-1：It is interested to intercept the possible residing trapezoid area of scope, i.e., one of lane line in described first image data
Region, is designated as lane line region；
Step 1-2：Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the lane line region successively, obtains and only has car
The image of diatom feature, is designated as lane line provincial characteristics；
Step 1-3：Canny rim detections are carried out to the lane line provincial characteristics, the profile of each object in image is obtained, is designated as
Lane line region contour；
Step 1-4：Hough line conversion is carried out to the lane line region contour, the line feature of each object, note in region is extracted to obtain
For region line feature；
Step 1-5：Interfering line is rejected according to the coordinate slope of the region line feature, line feature is obtained；
Step 1-6：The line feature is divided into by two lane lines in left and right by slope characteristics analysis, and it is merged,
Obtain and merge line；
Step 1-7：Centerline fit is carried out to the merging line, lane line center line is obtained.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, the recognition methods of the vehicle identification module specifically includes following steps：
Step 2-1：Intercept the possible residing scope of vehicle in described first image data, note consistent with the lane line region
For vehicle region；
Step 2-2：Carry out medium filtering, binary conversion treatment, contours extract successively to the vehicle region, obtain the vehicle area
All profiles in domain, are designated as vehicle region profile；
Step 2-3：Size to each profile in the vehicle region profile is screened, and rejects unrelated profile, obtains vehicle wheel
It is wide；
Step 2-4：The vehicle's contour is tracked by Kalman filtering, and passes through the classification based on probabilistic neural network
Device is classified to the vehicle's contour, so as to identify vehicle and its residing region.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, the recognition methods of the Traffic Sign Recognition module specifically includes following steps：
Step 3-1：Traffic sign in second view data possible residing scope, i.e. image border region are intercepted, is designated as
Mark region；
Step 3-2：Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the mark region successively, obtains and only has traffic
The image of flag sign, is designated as mark region feature；
Step 3-3：Gabor, HOG, ORB feature in the mark region feature are extracted, and it is merged, is marked
Will feature；
Step 3-4：Classification and Identification is carried out to the flag sign using SVMs.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, the traffic lights detection and the recognition methods of identification module specifically include following steps：
Step 4-1：Intercept traffic lights in second view data possible residing scope, i.e. image top region, note
For signal lamp region；
Step 4-2：The color space in the signal lamp region is converted into HSV from RGB；
Step 4-3：Carry out preliminary filtering process to the signal lamp region, and the training set of red, green, yellow three kinds of colors be set,
So as to carry out Gauss curve fitting to the distribution of color in the signal lamp region, color region is designated as；
Step 4-4：The color region is split using priori color threshold, signal lamp region is obtained；
Step 4-5：To the signal lamp region carry out connected domain detection, perceive each of which region bounding box；
Step 4-6：Each region in the signal lamp region is screened using certain condition, determines which part is
Candidate region；
Step 4-7：Repeating said steps 4-6, is extended using the same terms to the candidate region, obtains extension candidate regions
Domain；
Step 4-8：Classification and Identification is carried out to the extension candidate region using SVMs, traffic lights and its display shape is confirmed
State.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, in the step 4-6, each region in the signal lamp region is screened using following three necessary condition：
Condition 1：Colored pixels must account for more than the 70% of bounding box inner area；
Condition 2：Determination relation between the area of bounding box and position；
Condition 3：Meet the basic template of traffic lights.
A kind of preferred embodiment of the pilotless automobile visual identifying system based on dual camera provided in the present invention
In, in the step 4-8, the SVMs is trained using a large amount of traffic lights samples pictures, training is used
The extension candidate region is identified the good SVMs with HOG features.
The pilotless automobile visual identifying system based on dual camera provided compared to prior art, the present invention
Have the advantages that：
First, the pilotless automobile visual identifying system based on dual camera uses two cameras respectively to car nearby
Diatom and vehicle and the traffic mark board and signal lamp of distant place have carried out video acquisition, handle respectively.Compared to single camera, sheet
The pixel for inventing two cameras used is relatively low, and the data volume of view data is small, has effectively speeded arithmetic speed soon, can
Meet the requirement of real-time under really environment.
2nd, the pilotless automobile visual identifying system based on dual camera is entered by two passes to view data
The processing of row difference, identification target different during advancing sends the track that corresponding data signal realizes automatic driving vehicle
Recognition and tracking, for the active collision avoidance of surrounding vehicles, carries out corresponding wagon control according to traffic sign, passes through traffic signals
Lamp realizes the functions such as quick start and stop, and driving safety is ensure that by the high degree of accuracy of identification process.
3rd, vehicle identification module described in the pilotless automobile visual identifying system based on dual camera is using card
Kalman Filtering technology, is tracked to vehicle Probability Area, it is to avoid missing inspection caused by disturbing factor in identification process, greatly improves
Algorithm robustness.
Brief description of the drawings
Fig. 1 is the structure and identification stream for the pilotless automobile visual identifying system based on dual camera that the present invention is provided
Journey schematic diagram.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.
Referring to Fig. 1, being the structure for the pilotless automobile visual identifying system based on dual camera that the present invention is provided
And identification process schematic diagram.
The pilotless automobile visual identifying system 1 based on dual camera includes Lane detection module 11, vehicle
Identification module 12, Traffic Sign Recognition module 13, traffic lights detection with identification module 14, the first image capture module 15,
Second image capture module 16.
Described first image acquisition module 15 is gathered and output image located at the CMOS industrial cameras of 360,000 pixels
Data, are designated as the first view data.
Second image capture module 16 gathers located at the CMOS industrial cameras of 1,200,000 pixels and exports figure
As data, the second view data is designated as.
Described first image data input is to the Lane detection module 11, and 11 pairs of the Lane detection module is wherein
The center line of lane line is identified, and specific recognition methods comprises the following steps：
S11：Lane line region is intercepted：
It is area-of-interest, note to intercept the possible residing trapezoid area of scope, i.e., one of lane line in described first image data
For lane line region；
S12：Lane line is pre-processed：
Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the lane line region successively, with exclusive PCR, calculating is reduced
Amount, obtains the image for only having lane line feature, is designated as lane line provincial characteristics；
S13：Rim detection
Canny rim detections are carried out to the lane line provincial characteristics, the profile of each object in region is obtained, is designated as lane line area
Domain profile；
S14：Line feature extraction：
Hough line conversion is carried out to the lane line region contour, the line feature of each object in region is extracted to obtain, is designated as region line
Feature；
S15：Interfering line is rejected：
Interfering line is rejected according to the coordinate slope of the region line feature, line feature is obtained；
S16：Merge line：
The line feature is divided into by two lane lines in left and right by slope characteristics analysis, and it is merged, is merged
Line；
S17：Centerline fit：
Centerline fit is carried out to the merging line, lane line center line is obtained.
Obtain after the lane line center line, by comparing the center line of described first image data compared in the lane line
The deviation of line, you can obtain deviation of the current driving direction relative to lane line.So as to control Vehicular turn, this is reduced
Deviation, it is ensured that vehicle is travelled along lane line.
Described first image data are further input to the vehicle identification module 12, and the vehicle identification module 12 is to therein
Vehicle and its residing region are identified, and specific recognition methods comprises the following steps：
S21：Vehicle region is intercepted：
The possible residing scope of vehicle in described first image data is intercepted, it is consistent with the lane line region, it is designated as vehicle area
Domain；
S22：Vehicle is pre-processed：
Medium filtering, binary conversion treatment, contours extract are carried out successively to the vehicle region, with exclusive PCR, amount of calculation is reduced,
Profiles all in the vehicle region are obtained, vehicle region profile is designated as；
S23：Unrelated profile is rejected：
Size to each profile in the vehicle region profile is screened, and rejects unrelated profile, obtains vehicle's contour, but now
It only can determine that the position of objects in front, the as possible residing region of vehicle；
S24：Tracking and classification：
The vehicle's contour is tracked by Kalman filtering, and by the grader based on probabilistic neural network to described
Vehicle's contour is classified, so as to identify specific vehicle and its residing region.
When front there is obstacle vehicle or occur spacing it is excessively near when, you can identification sends information, and control vehicle enters
Row braking or avoidance.
Second view data is inputted to the Traffic Sign Recognition module 13,13 pairs of the Traffic Sign Recognition module
Traffic sign therein is identified, and specific recognition methods comprises the following steps：
S31：Mark region is intercepted：
Traffic sign in second view data possible residing scope, i.e. image border region are intercepted, mark region is designated as；
S32：Mark pretreatment：
Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the mark region successively, with exclusive PCR, calculating is reduced
Amount, obtains the image for only having traffic sign feature, is designated as mark region feature；
S33：Feature extraction：
Gabor, HOG, ORB feature in the mark region feature are extracted, and it is merged, flag sign is obtained；
S34：Classification：
Classification and Identification is carried out to the flag sign using SVMs.
After Classification and Identification is carried out to the flag sign, you can know the implication wherein respectively indicated, and control control according to this
Vehicle processed carries out the actions such as deceleration accordingly, u-turn.
Second view data is further input to the traffic lights detection and identification module 14, the traffic lights
Traffic lights therein and its dispaly state are identified with identification module 14 for detection, and specific recognition methods includes following
Step：
S41：Signal lamp region is intercepted：
Traffic lights in second view data possible residing scope, i.e. image top region are intercepted, signal lamp is designated as
Region；
S42：Form is changed：
The color space in the signal lamp region is converted into HSV from RGB；
S43：Color region is fitted：
Preliminary filtering process is carried out to the signal lamp region, and the training set of red, green, yellow three kinds of colors is set, so as to institute
The distribution of color for stating signal lamp region carries out Gauss curve fitting, is designated as color region；
S44：Color region is split：
Using red -30<=h<30th, yellow 30<=h<90th, green 90<=h<150 as hsv color spatial color differentiation threshold
Value, splits to the color region, obtains cut zone；
S45：Border is perceived：
To the cut zone carry out connected domain detection, perceive each of which region bounding box；
S46：Screening：
Each region in the signal lamp region is screened using using following three necessary condition：
Condition 1：Colored pixels must account for more than the 70% of bounding box inner area；
Condition 2：Within the specific limits, between bounding box there is the relative position relation determined in the area of bounding box；
Condition 3：Meet the basic template of traffic lights；
For example, a red area R has a width l and height h bounding box B, B is moved down into h and 2h creates new side
Boundary frame B1 and B2.If R is the candidate region of traffic lights, B1 and B2 mainly should be made up of dark pixels, and otherwise the region will
It is not qualified candidate region.
S47：Extension screening：
The S46 is repeated, the candidate region is extended using the same terms, it is to avoid miss effective information, extended
Candidate region；
S48：Classification and Identification：
Classification and Identification is carried out to the extension candidate region using the SVMs with HOG features trained, so that really
Recognize the current state of traffic lights and its display.
In traffic intersection, vehicle is then controlled to carry out corresponding row according to prompting when recognizing that traffic lights are green light
Sail, when being identified as red light with amber light, then send brake signal, control vehicle braking immediately.
The pilotless automobile visual identifying system based on dual camera provided compared to prior art, the present invention
1 has the advantages that：
First, the pilotless automobile visual identifying system 1 based on dual camera using two cameras respectively to nearby
Lane line and vehicle and the traffic mark board and signal lamp of distant place have carried out video acquisition, handle respectively.Compared to single camera,
The pixel for two cameras that the present invention is used is relatively low, and the data volume of view data is small, has effectively speeded arithmetic speed soon, energy
Enough requirements for meeting real-time under really environment.
2nd, the pilotless automobile visual identifying system 1 based on dual camera by two passes to view data
The processing distinguished, identification target different during advancing sends the car that corresponding data signal realizes automatic driving vehicle
Road recognition and tracking, for the active collision avoidance of surrounding vehicles, carries out corresponding wagon control according to traffic sign, is believed by traffic
Signal lamp realizes the functions such as quick start and stop.Driving safety is ensured by the high degree of accuracy of identification process.
3rd, vehicle identification module 12 is adopted described in the pilotless automobile visual identifying system 1 based on dual camera
With Kalman Filter Technology, vehicle Probability Area is tracked, it is to avoid missing inspection caused by disturbing factor in identification process, significantly
Improve algorithm robustness.
Embodiments of the invention are the foregoing is only, are not intended to limit the scope of the invention, it is every to utilize this hair
Equivalent structure or equivalent flow conversion that bright description is made, or directly or indirectly it is used in other related technology necks
Within domain, the scope of patent protection for being similarly included in the present invention.
Claims (7)
1. a kind of pilotless automobile visual identifying system based on dual camera, it is characterised in that：Including the first IMAQ
Module, the second image capture module, Lane detection module, vehicle identification module, Traffic Sign Recognition module, traffic lights
Detection and identification module；
Described first image acquisition module, second image capture module are gathered and output image data, and the is designated as respectively
One view data, the second view data；
Described first image data input to the Lane detection module, the Lane detection module identifies described first
Lane line in view data, and fit track center line；
Described first image data are further input to the vehicle identification module, and the vehicle identification module is known according to the lane line
The scope for the lane line that other module is identified, identifies the vehicle in lane line and its residing region；
Second view data is inputted to the Traffic Sign Recognition module, and the Traffic Sign Recognition module identifies described
Traffic Sign Images in second view data, and it is fast and accurately classified；
Second view data is further input to traffic lights detection and identification module, the traffic lights detection with
Identification module is by identifying the color region of target area, so as to identify traffic lights and its dispaly state.
2. the pilotless automobile visual identifying system according to claim 1 based on dual camera, it is characterised in that：Institute
The recognition methods for stating Lane detection module specifically includes following steps：
Step 1-1：It is interested to intercept the possible residing trapezoid area of scope, i.e., one of lane line in described first image data
Region, is designated as lane line region；
Step 1-2：Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the lane line region successively, obtains and only has car
The image of diatom feature, is designated as lane line provincial characteristics；
Step 1-3：Canny rim detections are carried out to the lane line provincial characteristics, the profile of each object in image is obtained, is designated as
Lane line region contour；
Step 1-4：Hough line conversion is carried out to the lane line region contour, the line feature of each object, note in region is extracted to obtain
For region line feature；
Step 1-5：Interfering line is rejected according to the coordinate slope of the region line feature, line feature is obtained；
Step 1-6：The line feature is divided into by two lane lines in left and right by slope characteristics analysis, and it is merged,
Obtain and merge line；
Step 1-7：Centerline fit is carried out to the merging line, lane line center line is obtained.
3. the pilotless automobile visual identifying system according to claim 2 based on dual camera, it is characterised in that：Institute
The recognition methods for stating vehicle identification module specifically includes following steps：
Step 2-1：Intercept the possible residing scope of vehicle in described first image data, note consistent with the lane line region
For vehicle region；
Step 2-2：Carry out medium filtering, binary conversion treatment, contours extract successively to the vehicle region, obtain the vehicle area
All profiles in domain, are designated as vehicle region profile；
Step 2-3：Size to each profile in the vehicle region profile is screened, and rejects unrelated profile, obtains vehicle wheel
It is wide；
Step 2-4：The vehicle's contour is tracked by Kalman filtering, and passes through the classification based on probabilistic neural network
Device is classified to the vehicle's contour, so as to identify vehicle and its residing region.
4. the pilotless automobile visual identifying system according to claim 1 based on dual camera, it is characterised in that：Institute
The recognition methods for stating Traffic Sign Recognition module specifically includes following steps：
Step 3-1：Traffic sign in second view data possible residing scope, i.e. image border region are intercepted, is designated as
Mark region；
Step 3-2：Gray processing, binaryzation, filtering, Morphological scale-space are carried out to the mark region successively, obtains and only has traffic
The image of flag sign, is designated as mark region feature；
Step 3-3：Gabor, HOG, ORB feature in the mark region feature are extracted, and it is merged, is marked
Will feature；
Step 3-4：Classification and Identification is carried out to the flag sign using SVMs.
5. the pilotless automobile visual identifying system according to claim 1 based on dual camera, it is characterised in that：Institute
State traffic lights detection and the recognition methods of identification module specifically includes following steps：
Step 4-1：Intercept traffic lights in second view data possible residing scope, i.e. image top region, note
For signal lamp region；
Step 4-2：The color space in the signal lamp region is converted into HSV from RGB；
Step 4-3：Carry out preliminary filtering process to the signal lamp region, and the training set of red, green, yellow three kinds of colors be set,
So as to carry out Gauss curve fitting to the distribution of color in the signal lamp region, color region is designated as；
Step 4-4：The color region is split using priori color threshold, signal lamp region is obtained；
Step 4-5：To the signal lamp region carry out connected domain detection, perceive each of which region bounding box；
Step 4-6：Each region in the signal lamp region is screened using certain condition, determines which part is
Candidate region；
Step 4-7：Repeating said steps 4-6, is extended using the same terms to the candidate region, obtains extension candidate regions
Domain；
Step 4-8：Classification and Identification is carried out to the extension candidate region using SVMs, traffic lights and its display shape is confirmed
State.
6. the pilotless automobile visual identifying system according to claim 5 based on dual camera, it is characterised in that：Institute
State in step 4-6, each region in the signal lamp region is screened using following three necessary condition：
Condition 1：Colored pixels must account for more than the 70% of bounding box inner area；
Condition 2：Determination relation between the area of bounding box and position；
Condition 3：Meet the basic template of traffic lights.
7. the pilotless automobile visual identifying system according to claim 5 based on dual camera, it is characterised in that：Institute
State in step 4-8, the SVMs is trained using a large amount of traffic lights samples pictures, the band trained is used
The extension candidate region is identified the SVMs for having HOG features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710559187.2A CN107316486A (en) | 2017-07-11 | 2017-07-11 | Pilotless automobile visual identifying system based on dual camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710559187.2A CN107316486A (en) | 2017-07-11 | 2017-07-11 | Pilotless automobile visual identifying system based on dual camera |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107316486A true CN107316486A (en) | 2017-11-03 |
Family
ID=60177596
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710559187.2A Pending CN107316486A (en) | 2017-07-11 | 2017-07-11 | Pilotless automobile visual identifying system based on dual camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107316486A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107978165A (en) * | 2017-12-12 | 2018-05-01 | 南京理工大学 | Intersection identifier marking and signal lamp Intellisense method based on computer vision |
| CN108172003A (en) * | 2017-11-30 | 2018-06-15 | 广州华夏职业学院 | Traffic lights car system for prompting |
| CN108875589A (en) * | 2018-05-25 | 2018-11-23 | 武汉东智科技股份有限公司 | A kind of video detecting method of road area |
| CN109116846A (en) * | 2018-08-29 | 2019-01-01 | 五邑大学 | A kind of automatic Pilot method, apparatus, computer equipment and storage medium |
| CN109389046A (en) * | 2018-09-11 | 2019-02-26 | 昆山星际舟智能科技有限公司 | Round-the-clock object identification and method for detecting lane lines for automatic Pilot |
| CN109781739A (en) * | 2019-03-04 | 2019-05-21 | 杭州晶耐科光电技术有限公司 | Automobile finish surface appearance defects automatic detection system and method |
| CN109883432A (en) * | 2019-02-21 | 2019-06-14 | 百度在线网络技术（北京）有限公司 | Location determining method, device, equipment and computer readable storage medium |
| CN110182206A (en) * | 2019-05-06 | 2019-08-30 | 深圳市华芯技研科技有限公司 | A kind of pilotless automobile visual identifying system based on dual camera |
| CN110619782A (en) * | 2019-10-10 | 2019-12-27 | 武汉蛋玩科技有限公司 | Vision-based traffic knowledge cognitive dynamic sand table system for education |
| CN110647863A (en) * | 2019-09-30 | 2020-01-03 | 重庆元韩汽车技术设计研究院有限公司 | Visual signal acquisition and analysis system for intelligent driving |
| WO2020024844A1 (en) * | 2018-07-28 | 2020-02-06 | 上海商汤智能科技有限公司 | Smart driving control method and apparatus, vehicle, electronic device, medium, and product |
| CN111251994A (en) * | 2018-11-30 | 2020-06-09 | 华创车电技术中心股份有限公司 | Method and system for detecting objects around vehicle |
| CN111783604A (en) * | 2020-06-24 | 2020-10-16 | 中国第一汽车股份有限公司 | Vehicle control method, device and equipment based on target identification and vehicle |
| CN112818956A (en) * | 2021-03-19 | 2021-05-18 | 张秀霞 | Intelligent traffic data acquisition system and acquisition method thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102745196A (en) * | 2012-07-18 | 2012-10-24 | 重庆邮电大学 | Intelligent control device and method for granular computing-based micro intelligent vehicle |
| US20140139670A1 (en) * | 2012-11-16 | 2014-05-22 | Vijay Sarathi Kesavan | Augmenting adas features of a vehicle with image processing support in on-board vehicle platform |
| US20150360692A1 (en) * | 2012-03-26 | 2015-12-17 | Google Inc. | Robust Method for Detecting Traffic Signals and their Associated States |
| EP3007099A1 (en) * | 2014-10-10 | 2016-04-13 | Application Solutions (Electronics and Vision) Ltd. | Image recognition system for a vehicle and corresponding method |
| WO2016130719A2 (en) * | 2015-02-10 | 2016-08-18 | Amnon Shashua | Sparse map for autonomous vehicle navigation |
| CN107200018A (en) * | 2016-03-16 | 2017-09-26 | Lg电子株式会社 | Drive assistance device and its control method |
-
2017
- 2017-07-11 CN CN201710559187.2A patent/CN107316486A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150360692A1 (en) * | 2012-03-26 | 2015-12-17 | Google Inc. | Robust Method for Detecting Traffic Signals and their Associated States |
| CN102745196A (en) * | 2012-07-18 | 2012-10-24 | 重庆邮电大学 | Intelligent control device and method for granular computing-based micro intelligent vehicle |
| US20140139670A1 (en) * | 2012-11-16 | 2014-05-22 | Vijay Sarathi Kesavan | Augmenting adas features of a vehicle with image processing support in on-board vehicle platform |
| EP3007099A1 (en) * | 2014-10-10 | 2016-04-13 | Application Solutions (Electronics and Vision) Ltd. | Image recognition system for a vehicle and corresponding method |
| WO2016130719A2 (en) * | 2015-02-10 | 2016-08-18 | Amnon Shashua | Sparse map for autonomous vehicle navigation |
| CN107200018A (en) * | 2016-03-16 | 2017-09-26 | Lg电子株式会社 | Drive assistance device and its control method |
Non-Patent Citations (2)
| Title |
|---|
| 朱永珍等: "基于HSV色彩空间与形状特征的交通灯自动识别", 《电视技术》 * |
| 王海涛: "车前障碍物检测及车道线识别研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108172003B (en) * | 2017-11-30 | 2020-11-27 | 广州华夏职业学院 | Traffic signal lamp in-vehicle reminding system |
| CN108172003A (en) * | 2017-11-30 | 2018-06-15 | 广州华夏职业学院 | Traffic lights car system for prompting |
| CN107978165A (en) * | 2017-12-12 | 2018-05-01 | 南京理工大学 | Intersection identifier marking and signal lamp Intellisense method based on computer vision |
| CN108875589A (en) * | 2018-05-25 | 2018-11-23 | 武汉东智科技股份有限公司 | A kind of video detecting method of road area |
| WO2020024844A1 (en) * | 2018-07-28 | 2020-02-06 | 上海商汤智能科技有限公司 | Smart driving control method and apparatus, vehicle, electronic device, medium, and product |
| CN109116846A (en) * | 2018-08-29 | 2019-01-01 | 五邑大学 | A kind of automatic Pilot method, apparatus, computer equipment and storage medium |
| CN109116846B (en) * | 2018-08-29 | 2022-04-05 | 五邑大学 | Automatic driving method, device, computer equipment and storage medium |
| CN109389046A (en) * | 2018-09-11 | 2019-02-26 | 昆山星际舟智能科技有限公司 | Round-the-clock object identification and method for detecting lane lines for automatic Pilot |
| CN109389046B (en) * | 2018-09-11 | 2022-03-29 | 昆山星际舟智能科技有限公司 | All-weather object identification and lane line detection method for automatic driving |
| CN111251994B (en) * | 2018-11-30 | 2021-08-24 | 华创车电技术中心股份有限公司 | Method and system for detecting objects around vehicle |
| CN111251994A (en) * | 2018-11-30 | 2020-06-09 | 华创车电技术中心股份有限公司 | Method and system for detecting objects around vehicle |
| CN109883432A (en) * | 2019-02-21 | 2019-06-14 | 百度在线网络技术（北京）有限公司 | Location determining method, device, equipment and computer readable storage medium |
| CN109883432B (en) * | 2019-02-21 | 2020-12-29 | 百度在线网络技术（北京）有限公司 | Position determination method, device, equipment and computer readable storage medium |
| CN109781739A (en) * | 2019-03-04 | 2019-05-21 | 杭州晶耐科光电技术有限公司 | Automobile finish surface appearance defects automatic detection system and method |
| CN110182206A (en) * | 2019-05-06 | 2019-08-30 | 深圳市华芯技研科技有限公司 | A kind of pilotless automobile visual identifying system based on dual camera |
| CN110647863A (en) * | 2019-09-30 | 2020-01-03 | 重庆元韩汽车技术设计研究院有限公司 | Visual signal acquisition and analysis system for intelligent driving |
| CN110647863B (en) * | 2019-09-30 | 2024-02-02 | 重庆元韩汽车技术设计研究院有限公司 | Visual signal acquisition and analysis system for intelligent driving |
| CN110619782A (en) * | 2019-10-10 | 2019-12-27 | 武汉蛋玩科技有限公司 | Vision-based traffic knowledge cognitive dynamic sand table system for education |
| CN111783604A (en) * | 2020-06-24 | 2020-10-16 | 中国第一汽车股份有限公司 | Vehicle control method, device and equipment based on target identification and vehicle |
| CN112818956A (en) * | 2021-03-19 | 2021-05-18 | 张秀霞 | Intelligent traffic data acquisition system and acquisition method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107316486A (en) | Pilotless automobile visual identifying system based on dual camera | |
| CN107766821B (en) | Method and system for detecting and tracking full-time vehicle in video based on Kalman filtering and deep learning | |
| CN104361350B (en) | A kind of traffic mark identifying system | |
| US8750567B2 (en) | Road structure detection and tracking | |
| CN105981042B (en) | Vehicle detection system and method | |
| CN104021378B (en) | Traffic lights real-time identification method based on space time correlation Yu priori | |
| CN103984950B (en) | A kind of moving vehicle brake light status recognition methods for adapting to detection on daytime | |
| CN109190523B (en) | Vehicle detection tracking early warning method based on vision | |
| CN101950350B (en) | Clear path detection using a hierachical approach | |
| CN107609472A (en) | A kind of pilotless automobile NI Vision Builder for Automated Inspection based on vehicle-mounted dual camera | |
| CN101929867B (en) | Clear path detection using road model | |
| Chen et al. | Nighttime brake-light detection by Nakagami imaging | |
| CN110160536A (en) | Device and method is provided using the guidance information of crossing recognition result | |
| EP3537348A1 (en) | Heterogeneous convolutional neural network for multi-problem solving | |
| CN103077384A (en) | Method and system for positioning and recognizing vehicle logo | |
| CN105654073A (en) | Automatic speed control method based on visual detection | |
| CN110688907A (en) | Method and device for identifying object based on road light source at night | |
| JP2020064583A (en) | Vehicle detection method, nighttime vehicle detection method based on dynamic light intensity, and system for the same | |
| CN106985819A (en) | The determination methods and device of vehicle driving intention | |
| CN103927548A (en) | Novel vehicle collision avoiding brake behavior detection method | |
| Lin et al. | Adaptive IPM-based lane filtering for night forward vehicle detection | |
| Biswas et al. | Detection and classification of speed limit traffic signs | |
| Sushmitha et al. | Multiple car detection, recognition and tracking in traffic | |
| CN104268859A (en) | Image preprocessing method for night lane line detection | |
| Moizumi et al. | Traffic light detection considering color saturation using in-vehicle stereo camera |
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: 20171103 |