CN108008727A - A kind of pilotless automobile that can be run at high speed - Google Patents
A kind of pilotless automobile that can be run at high speed Download PDFInfo
- Publication number
- CN108008727A CN108008727A CN201711306123.8A CN201711306123A CN108008727A CN 108008727 A CN108008727 A CN 108008727A CN 201711306123 A CN201711306123 A CN 201711306123A CN 108008727 A CN108008727 A CN 108008727A
- Authority
- CN
- China
- Prior art keywords
- pixel
- classification
- submodule
- movement
- detection
- 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.)
- Withdrawn
Links
- 238000001514 detection method Methods 0.000 claims description 46
- 238000011156 evaluation Methods 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 7
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000465 moulding Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0246—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means
Abstract
The present invention provides a kind of pilotless automobile that can be run at high speed, including EMS, global position system and central control system, the EMS is used to be detected vehicle periphery moving target during galloping, and testing result is sent to central control system, the global position system is used to determine the real-time road condition information on optimal path and the path, and optimal path and real-time road are sent to central control system, the central control system is used to be controlled vehicle according to the information that EMS and global position system are sent.Beneficial effects of the present invention are:Realize running at high speed for pilotless automobile.
Description
Technical field
The present invention relates to unmanned vehicle technology field, and in particular to a kind of pilotless automobile that can be run at high speed.
Background technology
Current automobile be essentially all it is manned, it is manned to bring the problem of very much, such as the wave of human resources
Expense, traffic accident incidence remain high, and pilotless automobile comes into being, and still, existing pilotless automobile can not be right
Moving target is effectively detected, and causes travel speed slow.
The content of the invention
A kind of in view of the above-mentioned problems, the present invention is intended to provide pilotless automobile that can be run at high speed.
The purpose of the present invention is realized using following technical scheme:
Provide a kind of pilotless automobile that can be run at high speed, including EMS, global position system and
Central control system, the EMS are used to be detected vehicle periphery moving target during galloping, and
Testing result is sent to central control system, the global position system is used to determine the reality on optimal path and the path
When traffic information, and optimal path and real-time road are sent to central control system, the central control system is used for basis
The information that EMS and global position system are sent is controlled vehicle.
Beneficial effects of the present invention are:Realize running at high speed for pilotless automobile.
Brief description of the drawings
Using attached drawing, the invention will be further described, but the embodiment in attached drawing does not form any limit to the present invention
System, for those of ordinary skill in the art, without creative efforts, can also obtain according to the following drawings
Other attached drawings.
Fig. 1 is the structure diagram of the present invention;
Reference numeral:
EMS 1, global position system 2, central control system 3.
Embodiment
The invention will be further described with the following Examples.
Referring to Fig. 1, a kind of pilotless automobile that can be run at high speed of the present embodiment, including EMS 1, defend
Star alignment system 2 and central control system 3, the EMS 1 are used to move vehicle periphery during galloping
Target is detected, and testing result is sent to central control system 3, and the global position system 2 is used to determine optimal road
Real-time road condition information on footpath and the path, and optimal path and real-time road are sent to central control system 3, in described
Centre control system 3 is used to be controlled vehicle according to the information that EMS 1 and global position system 2 are sent.
The present embodiment realizes running at high speed for pilotless automobile.
Preferably, the EMS 1 includes the first segmentation module, the second sort module, the 3rd modeling module, the
Four detection modules and the 5th evaluation module, the first segmentation module are used to being partitioned into movement pixel in scene, and described second
Sort module is used to movement pixel being divided into two class of dynamic background pixel and moving target pixel, and the 3rd modeling module is used
In building gauss hybrid models to dynamic background pixel and moving target pixel, the 4th detection module is used to be mixed according to Gauss
Molding type is detected moving target, and the 5th evaluation module is used to assess detection result.
This preferred embodiment realizes the detection of moving target and the assessment to detection result when running at high speed.
Preferably, the first segmentation module is used for the movement pixel being partitioned into scene, is specially:If EUi(x, y) and
QK (x, y) represents the i-th two field picture of video sequence and current background image respectively, then the movement pixel of the i-th two field picture is:
In above-mentioned formula, EHi(x, y) represents the movement pixel of the i-th two field picture, FNiRepresent given threshold.
This preferred embodiment obtains the movement pixel of image by the first segmentation module, can quickly and accurately obtain field
Movement pixel in scape.
Preferably, second sort module includes the first classification submodule, the second classification submodule and compressive classification
Module, the first classification submodule are used for a classification results for obtaining movement pixel, and the second classification submodule is used for
The secondary classification of movement pixel is obtained as a result, the compressive classification submodule is obtained according to a classification results and secondary classification result
Take the final classification result of movement pixel.
The first classification submodule is used for a classification results for obtaining movement pixel, is specially:In video image sequence
In row, two regions are selected:There is the region of dynamic background and have the region of moving target, in the two selected regions, adopt
The motion vector of each movement pixel is calculated with Block- matching mode, is painted according to the range weight r of motion vector and angle component θ
Motion vector histogram processed;Dynamic background picture is divided into by pixel is moved according to the relation of motion vector histogram and movement pixel
Element and two class of moving target pixel, as a classification results;
The second classification submodule is used to obtain the secondary classification of movement pixel as a result, being specially:If move in pixel
Dynamic background pixel is HX1, moving target pixel be HX2, make AYl,i=(rl,i,θl,i) represent the in the i-th two field picture movement pixel
The motion vector of l pixel polar form, and AYl,iBelong to moving target pixel HX2Conditional probability pHX2|AYl,i) be:
Then movement pixel is that the condition of moving target pixel is:pi(AYl,i|HX2)pi(HX2) > pi(AYl,i|HX1)pi
(HX1);
In above-mentioned formula, CAi-1Represent image EUi-1Quilt
It is detected as the number of pixels of moving target, Ai-1Represent EUi-1It is detected as the total number of movement pixel;
Using ineligible pixel as dynamic background pixel, secondary classification result is obtained;
The compressive classification submodule obtains final point of movement pixel according to a classification results and secondary classification result
Class is as a result, be specially:Move pixel in the same manner for secondary classification result and a classification results, using a classification results as
Final classification as a result, for the secondary classification result movement pixel different from a classification results, using secondary classification result as
Final classification result.
With continuous lifting of the people to requirement of real-time, for fast-changing dynamic scene, traditional video detection
Method usually seems unable to do what one wishes.In the faster scene of target speed, non-targeted in the scene and motion mode of target
It is absolutely different.This preferred embodiment proposes the moving object detection strategy based on movable information, specifically, due to fortune
Dynamic information can effectively be stated by motion vector, according to the difference pair between moving target and the motion vector of dynamic background
Movement pixel is classified, and helps to realize the accurate detection of moving target.
Preferably, the 5th evaluation module includes one-time detection assessment submodule, secondary detection assesses submodule and comprehensive
Check and evaluation submodule is closed, the one-time detection assessment submodule is used for the first assessed value for obtaining detection result, described secondary
Check and evaluation submodule is used for the second assessed value for obtaining detection result, and the comprehensive detection assessment submodule is used for according to first
Assessed value and the second assessed value carry out comprehensive assessment to detection result;
The one-time detection assessment submodule is used for the first assessed value for obtaining detection result:
In above-mentioned formula, YW represents the number of moving target, YW1Represent the moving target number detected, EM1Represent the
One assessed value;
The secondary detection assessment submodule is used for the second assessed value for obtaining detection result:
In above-mentioned formula, YW2Represent the number that the moving target detected is overlapped with real moving target, EM2Represent the
Two assessed values;
The comprehensive detection assessment submodule is used to carry out detection result according to the first assessed value and the second assessed value comprehensive
Close assessment:Calculate comprehensive assessment value:
In above-mentioned formula, EM represents comprehensive assessment value;Comprehensive assessment value is bigger, represents that detection result is better.
This preferred embodiment realizes the assessment of moving object detection effect by the 5th evaluation module, ensure that detection is accurate
True property, specifically, the first assessed value considers the comprehensive of moving object detection, the second assessed value considers moving object detection
Accuracy, comprehensive assessment value then assesses detection result according to the first assessed value and the second assessed value, obtained assessment
As a result it is more accurate, so as to ensure that pilotless automobile is run at high speed.
The pilotless automobile trip that the present invention can run at high speed is selected, selectes departure place, 5 destinations is chosen and carries out
Experiment, is respectively destination 1, destination 2, destination 3, destination 4, destination 5, unites to driving time and driving cost
Meter, is compared, generation is had the beneficial effect that shown in table compared with pilotless automobile:
Driving time shortens | Drive cost reduction | |
Destination 1 | 29% | 27% |
Destination 2 | 27% | 26% |
Destination 3 | 26% | 26% |
Destination 4 | 25% | 24% |
Destination 5 | 24% | 22% |
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected
The limitation of scope is protected, although being explained with reference to preferred embodiment to the present invention, those of ordinary skill in the art should
Work as understanding, can be to technical scheme technical scheme is modified or replaced equivalently, without departing from the reality of technical solution of the present invention
Matter and scope.
Claims (7)
1. a kind of pilotless automobile that can be run at high speed, it is characterised in that including EMS, global position system
And central control system, the EMS are used to be detected vehicle periphery moving target during galloping,
And send testing result to central control system, the global position system is used to determine on optimal path and the path
Real-time road condition information, and optimal path and real-time road are sent to central control system, the central control system is used for root
The information sent according to EMS and global position system is controlled vehicle.
2. the pilotless automobile according to claim 1 that can be run at high speed, it is characterised in that the environment measuring system
System includes the first segmentation module, the second sort module, the 3rd modeling module, the 4th detection module and the 5th evaluation module, described
First segmentation module is used for the movement pixel being partitioned into scene, and second sort module is used to movement pixel being divided into dynamic
Two class of state background pixel and moving target pixel, the 3rd modeling module are used for dynamic background pixel and moving target pixel
Gauss hybrid models are built, and the 4th detection module is used to moving target is detected according to gauss hybrid models, described
5th evaluation module is used to assess detection result.
3. the pilotless automobile according to claim 2 that can be run at high speed, it is characterised in that the first segmentation mould
Block is used for the movement pixel being partitioned into scene, is specially:If EUi(x, y) and QK (x, y) represent the i-th of video sequence respectively
Two field picture and current background image, then the movement pixel of the i-th two field picture be:
In above-mentioned formula, EHi(x, y) represents the movement pixel of the i-th two field picture, FNiRepresent given threshold.
4. the pilotless automobile according to claim 3 that can be run at high speed, it is characterised in that the second classification mould
Block includes the first classification submodule, the second classification submodule and compressive classification submodule, and the first classification submodule is used to obtain
A classification results of movement pixel are taken, the secondary classification that the second classification submodule is used to obtain movement pixel is as a result, institute
State the final classification result that compressive classification submodule obtains movement pixel according to a classification results and secondary classification result;
The first classification submodule is used for a classification results for obtaining movement pixel, is specially:In sequence of video images,
Select two regions:There is the region of dynamic background and have the region of moving target, in the two selected regions, using block
The motion vector of each movement pixel is calculated with mode, movement is drawn according to the range weight r of motion vector and angle component θ
Histogram vector;Dynamic background pixel and fortune are divided into by pixel is moved according to the relation of motion vector histogram and movement pixel
Two class of moving-target pixel, as a classification results.
5. the pilotless automobile according to claim 4 that can be run at high speed, it is characterised in that the second classification
Module is used to obtain the secondary classification of movement pixel as a result, being specially:If it is HX to move dynamic background pixel in pixel1, movement
Object pixel is HX2, make AYl,i=(rl,i,θl,i) represent that the i-th two field picture moves the fortune of l-th of pixel polar form in pixel
Moving vector, and AYl,iBelong to moving target pixel HX2Conditional probability p (HX2|AYl,i) be:
Then movement pixel is that the condition of moving target pixel is:pi(AYl,i|HX2)pi(HX2) > pi(AYl,i|HX1)pi(HX1);
In above-mentioned formula,CAi-1Represent image EUi-1It is tested
Survey the number of pixels for moving target, Ai-1Represent EUi-1It is detected as the total number of movement pixel;
Using ineligible pixel as dynamic background pixel, secondary classification result is obtained;
The compressive classification submodule obtains the final classification knot of movement pixel according to a classification results and secondary classification result
Fruit, is specially:Pixel is moved in the same manner for secondary classification result and a classification results, using a classification results as final
Classification results, for the secondary classification result movement pixel different from a classification results, using secondary classification result as final
Classification results.
6. the pilotless automobile according to claim 5 that can be run at high speed, it is characterised in that the 5th assessment mould
Block includes one-time detection assessment submodule, secondary detection assessment submodule and comprehensive detection assessment submodule, the one-time detection
Assessment submodule is used for the first assessed value for obtaining detection result, and the secondary detection assessment submodule is used to obtain detection result
The second assessed value, comprehensive detection assessment submodule be used for according to the first assessed value and the second assessed value to detection result into
Row comprehensive assessment;
The one-time detection assessment submodule is used for the first assessed value for obtaining detection result:
In above-mentioned formula, YW represents the number of moving target, YW1Represent the moving target number detected, EM1Represent that first comments
Valuation;
The secondary detection assessment submodule is used for the second assessed value for obtaining detection result:
In above-mentioned formula, YW2Represent the number that the moving target detected is overlapped with real moving target, EM2Represent that second comments
Valuation.
7. the pilotless automobile according to claim 6 that can be run at high speed, it is characterised in that the comprehensive detection is commented
Estimate submodule to be used to carry out comprehensive assessment to detection result according to the first assessed value and the second assessed value:Calculate comprehensive assessment value:
In above-mentioned formula, EM represents comprehensive assessment value;Comprehensive assessment value is bigger, represents that detection result is better.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711306123.8A CN108008727A (en) | 2017-12-11 | 2017-12-11 | A kind of pilotless automobile that can be run at high speed |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711306123.8A CN108008727A (en) | 2017-12-11 | 2017-12-11 | A kind of pilotless automobile that can be run at high speed |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108008727A true CN108008727A (en) | 2018-05-08 |
Family
ID=62057745
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711306123.8A Withdrawn CN108008727A (en) | 2017-12-11 | 2017-12-11 | A kind of pilotless automobile that can be run at high speed |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108008727A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108650736A (en) * | 2018-05-29 | 2018-10-12 | 深圳信息职业技术学院 | A kind of environment adjustable type intelligent lighting lamp |
CN114715197A (en) * | 2022-06-10 | 2022-07-08 | 深圳市爱云信息科技有限公司 | Automatic driving safety method and system based on digital twin DaaS platform |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080136934A1 (en) * | 2006-12-12 | 2008-06-12 | Industrial Technology Research Institute | Flame Detecting Method And Device |
CN104943684A (en) * | 2014-03-31 | 2015-09-30 | 比亚迪股份有限公司 | Pilotless automobile control system and automobile with same |
CN206031278U (en) * | 2016-09-07 | 2017-03-22 | 西安科技大学 | Self -driving car control system |
CN106846356A (en) * | 2017-01-13 | 2017-06-13 | 广东万安科技股份有限公司 | A kind of moving target foreground detection method of Bayes's full probability Combined estimator model |
CN107139917A (en) * | 2017-04-27 | 2017-09-08 | 江苏大学 | It is a kind of based on mix theory pilotless automobile crosswise joint system and method |
CN107161141A (en) * | 2017-03-08 | 2017-09-15 | 深圳市速腾聚创科技有限公司 | Pilotless automobile system and automobile |
CN107272687A (en) * | 2017-06-29 | 2017-10-20 | 深圳市海梁科技有限公司 | A kind of driving behavior decision system of automatic Pilot public transit vehicle |
CN107330922A (en) * | 2017-07-04 | 2017-11-07 | 西北工业大学 | Video moving object detection method of taking photo by plane based on movable information and provincial characteristics |
-
2017
- 2017-12-11 CN CN201711306123.8A patent/CN108008727A/en not_active Withdrawn
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080136934A1 (en) * | 2006-12-12 | 2008-06-12 | Industrial Technology Research Institute | Flame Detecting Method And Device |
CN104943684A (en) * | 2014-03-31 | 2015-09-30 | 比亚迪股份有限公司 | Pilotless automobile control system and automobile with same |
CN206031278U (en) * | 2016-09-07 | 2017-03-22 | 西安科技大学 | Self -driving car control system |
CN106846356A (en) * | 2017-01-13 | 2017-06-13 | 广东万安科技股份有限公司 | A kind of moving target foreground detection method of Bayes's full probability Combined estimator model |
CN107161141A (en) * | 2017-03-08 | 2017-09-15 | 深圳市速腾聚创科技有限公司 | Pilotless automobile system and automobile |
CN107139917A (en) * | 2017-04-27 | 2017-09-08 | 江苏大学 | It is a kind of based on mix theory pilotless automobile crosswise joint system and method |
CN107272687A (en) * | 2017-06-29 | 2017-10-20 | 深圳市海梁科技有限公司 | A kind of driving behavior decision system of automatic Pilot public transit vehicle |
CN107330922A (en) * | 2017-07-04 | 2017-11-07 | 西北工业大学 | Video moving object detection method of taking photo by plane based on movable information and provincial characteristics |
Non-Patent Citations (1)
Title |
---|
ZHANG WEI等: "Moving vehicles segmentation based on Bayesian framework for Gaussian motion model", 《PATTERN RECOGNITION LETTERS》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108650736A (en) * | 2018-05-29 | 2018-10-12 | 深圳信息职业技术学院 | A kind of environment adjustable type intelligent lighting lamp |
CN114715197A (en) * | 2022-06-10 | 2022-07-08 | 深圳市爱云信息科技有限公司 | Automatic driving safety method and system based on digital twin DaaS platform |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107576960B (en) | Target detection method and system for visual radar space-time information fusion | |
Ali et al. | Real-time multiple damage mapping using autonomous UAV and deep faster region-based neural networks for GPS-denied structures | |
WO2021135554A1 (en) | Method and device for planning global path of unmanned vehicle | |
WO2020151109A1 (en) | Three-dimensional target detection method and system based on point cloud weighted channel feature | |
WO2022206942A1 (en) | Laser radar point cloud dynamic segmentation and fusion method based on driving safety risk field | |
CN103150559B (en) | Head recognition and tracking method based on Kinect three-dimensional depth image | |
CN110008843A (en) | Combine cognitive approach and system based on the vehicle target of cloud and image data | |
TW202020811A (en) | Systems and methods for correcting a high -definition map based on detection of obstructing objects | |
CN105809687A (en) | Monocular vision ranging method based on edge point information in image | |
CN104320617B (en) | A kind of round-the-clock video frequency monitoring method based on deep learning | |
CN110490114A (en) | Target detection barrier-avoiding method in a kind of unmanned plane real-time empty based on depth random forest and laser radar | |
He et al. | Integrated moment-based LGMD and deep reinforcement learning for UAV obstacle avoidance | |
CN108008727A (en) | A kind of pilotless automobile that can be run at high speed | |
CN103942786B (en) | The self adaptation block objects detection method of unmanned plane visible ray and infrared image | |
Nose et al. | A study on a lane keeping system using CNN for online learning of steering control from real time images | |
US11798289B2 (en) | Streaming object detection and segmentation with polar pillars | |
Lee et al. | Object detection using vision and LiDAR sensor fusion for multi-channel V2X system | |
Moon et al. | Obstacle detecting system for unmanned ground vehicle using laser scanner and vision | |
Zhou et al. | Fast road detection and tracking in aerial videos | |
CN102123234A (en) | Unmanned aerial vehicle reconnaissance video grading motion compensation method | |
Chou et al. | A parallel convolutional neural network architecture for stereo vision estimation | |
CN108053647A (en) | A kind of intelligent traffic monitoring system | |
Sevi et al. | Rail Tracking and Detection with Drone in Gazebo Environment | |
Manikandan et al. | Ad hoc-obstacle avoidance-based navigation system using Deep Reinforcement Learning for self-driving vehicles | |
Shrivastava et al. | CubifAE-3D: Monocular camera space cubification for auto-encoder based 3D object detection |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180508 |