CN108645375A - One kind being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method - Google Patents
One kind being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method Download PDFInfo
- Publication number
- CN108645375A CN108645375A CN201810571399.7A CN201810571399A CN108645375A CN 108645375 A CN108645375 A CN 108645375A CN 201810571399 A CN201810571399 A CN 201810571399A CN 108645375 A CN108645375 A CN 108645375A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- binocular
- target vehicle
- binocular camera
- visual angle
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C3/00—Measuring distances in line of sight; Optical rangefinders
Landscapes
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Measurement Of Optical Distance (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses the vehicle odometry methods based on vehicle-mounted binocular camera, which is characterized in that including：S1 carries out vehicle detection according to target vehicle texture information and priori road information in the visual angle of binocular camera side, and determines the position of vehicle and at a distance from this vehicle；S2, the similitude matching of target vehicle is carried out in the visual angle of the binocular camera other side, the higher matching position of similitude is chosen, determines the center offset of target vehicle in two-phase visual angle, and the physical location of target vehicle is determined according to center offset and at a distance from this vehicle；S3, the confidence interval that binocular camera is determined according to binocular camera parameter exports the physical location of the target vehicle determined according to center offset and at a distance from this vehicle in binocular confidence interval, outside confidence interval, then the reliability of vehicle detection is verified by the grader of off-line training.Using the present invention, the precision of vehicle odometry is improved.
Description
Technical field
The present invention relates to vehicle odometry methods, especially a kind of to be used for vehicle-mounted biocular systems rapid vehicle ranging optimization side
Method.
Background technology
Advanced DAS (Driver Assistant System) (ADAS) is one of the important technology of vehicle braking security developments, has gradually become each
State NCAP evaluates the important safety reference frame of vehicle safety grade.Preceding anti-collision early warning (FCW), lane departure warning (LDW) etc.
ADAS functions have become the standard configuration of existing middle and high end vehicle, for the ADAS systems such as FCW for front vehicles with
And automatic emergency brake system (AEB), not requiring nothing more than detect front vehicles, also requirement that sensory perceptual system can be promptly and accurately is
System can motion state of the Accurate Prediction front truck relative to this vehicle, to ensure the effect of early warning and DAS (Driver Assistant System).In addition, from
The realization of dynamic control loop requires vehicle to have the heightened perception ability to ambient enviroment, includes the phase to other traffic participants
Therefore to the judgement of movable information, sensory perceptual system is not only needed accurately to be detected based on single-frame images and calculates front vehicles
Relative position, it is also necessary to which motion state of the vehicle relative to vehicle is calculated based on sequential image.In addition, ADAS systems pair
It is higher in the accuracy and requirement of real-time of perception algorithm.
The existing distance measuring method based on binocular vision system is based primarily upon the disparity computation between two cameras to estimate in scene
The distance of barrier, such method are more demanding to processing unit computing capability, and require two cameras horizontal coaxial, optical axis
It is parallel, acquisition is synchronous and exposure parameter is consistent etc., practical application has larger limitation.
Invention content
It is an object of the invention to provide one kind being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method, improves apart from vehicle
Precision.
In order to solve the problems existing in the prior art, the present invention provides a kind of excellent for vehicle-mounted biocular systems rapid vehicle ranging
Change method, this method include：
S1 carries out vehicle inspection according to target vehicle texture information and priori road information in the visual angle of binocular camera side
Survey, and determine vehicle position and at a distance from this vehicle；
S2 carries out the similitude matching of target vehicle in the visual angle of the binocular camera other side, chooses higher of similitude
With position, the center offset of target vehicle in two-phase visual angle is determined, and the reality of target vehicle is determined according to center offset
Position and at a distance from this vehicle；
S3 determines the confidence interval of binocular camera according to binocular camera parameter, and in binocular confidence interval, output is in
Heart offset determine target vehicle physical location and at a distance from this vehicle, outside confidence interval, then pass through off-line training
Grader verifies the reliability of vehicle detection.
Preferably, step s1 is specifically included：
S11, obtains vehicle detecting information in the visual angle of binocular camera side, and according to plane longitudinal direction road model and
Binocular camera internal and external parameter determines vehicle detection ROI region；
S12 knows in the regions vehicle detection RIO according to the priori of road model and target vehicle width, height ratio
Know, ROI region sliding window search listing, and the cascade classifier obtained by off-line training is generated, in the image of sliding window list
Appearance is classified, and determines the respective image position of target vehicle in sliding window list；
S13, according to plane longitudinal direction road model and binocular camera internal and external parameter, by target vehicle candidate region bottom
Image coordinate inverse perspective mapping obtains the position of target vehicle to road surface coordinate system.
Preferably, step s1 further includes：According to edge filter, edge detection is carried out to vehicle detection ROI region, takes out side
The poor region of edge feature significance.
Preferably, the step s2 includes：
S21, determine the position of the target vehicle of binocular camera side visual angle detection binocular camera other side visual angle from
Related coefficient, used formula are：
R (x, y)=sum (IL(x ', y ')-IR(x+x ', y+y '))2)
Determining auto-correlation coefficient is compared by S22 with the lower threshold of preset auto-correlation coefficient, is chosen higher than pre-
If the vehicle location of the most auto-correlation coefficient of lower threshold, as the highest matching position of similarity；
S23 determines the center position of the highest matching position of similarity in two magazine parallax d of binocular, and will regard
Difference calculates the actual coordinate (X, Y, Z) being converted into camera coordinates system, and used algorithm is：
Wherein, stereofFor binocular camera focal length；stereobaselineFor binocular camera baseline length；(u0, v0) is double
Mesh picture centre point coordinates.
Preferably, the step s3 is specifically included：
S31 calculates binocular resolution of ranging according to binocular camera parameter, chooses binocular resolution of ranging and is more than 10% work
For confidence interval, resolution ratio define be apart from vehicle percentage error, calculation formula caused by unit parallax：
S32, using binocular distance measurement result as observation, updates target vehicle relative position in binocular confidence interval；
S33 is corresponded in the other side visual angle target vehicle of two camera of binocular in parallax region outside binocular confidence interval,
Made with monocular distance measurement result if consistent with testing result in two visual angle of binocular with sliding window search validation vehicle detection reliability
For observation, the relative position of target vehicle is updated；If testing result is inconsistent in two visual angle of binocular, predicted with state equation
As a result target vehicle relative position is updated, the state equation is：
xt+1=Axt+But+ωt
yt+1=Cxt+ρt
Wherein, x is system state amount, including the position of vehicle and change in location information, i.e. [X, Y, dX, dY] T；Y be
Overall view measures, and includes the location information of vehicle, i.e. [X, Y] T；ω, p are respectively process noise and measurement noise；Respectively by adjusting
Corresponding process noise matrix Q is completed with measurement noise matrix R1 and R2.
Improvement binocular obstacle distance measurement method proposed by the present invention, for vehicle mounted traffic scene vehicle detection and distance
The application of calculating has carried out corresponding optimization.Compared to traditional binocular algorithm, the invention avoids global or half global disparity meters
It calculates, efficiency of algorithm greatly promotes.First, using unity and coherence in writing information and priori road information to area-of-interest on the left of binocular camera
Vehicle detection is carried out in vision；For the alternative vehicle region detected, carry out giving vehicle mould in vision on the right side of binocular camera
The similitude matching of version；The highest matching position of similitude is chosen, calculates in two right side of camera visions and carries out the vehicle masterplate
Similitude matches；The highest matching position of similitude is chosen, the vehicle alternative area center offset in two camera perspectives is calculated,
To calculate the relative distance of the vehicle and this vehicle.The method arithmetic speed is fast, and is wanted to the calibration and installation of binocular camera
Ask relatively low, it is more preferable to the robustness of camera pitch angle variation relative to the monocular range estimation method based on road model, it improves
The precision of vehicle odometry.
Description of the drawings
Fig. 1 is a kind of a kind of flow of embodiment for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method of the invention
Schematic diagram；
Fig. 2 is a kind of a kind of flow of embodiment for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method of the invention
Schematic diagram.
Specific implementation mode
The present invention is described in detail below in conjunction with the accompanying drawings.
With reference to figure 1, which is a kind of a kind of embodiment for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method
Flow diagram, from the figure, it can be seen that this system can be divided into three layers, respectively input layer, image procossing at and output
Layer, wherein input layer are used to input the image of priori and forward sight system acquisition, priori be to road information it is assumed that
Here may include road model, track model and the high information of vehicle width, input using surface road model hypothesis
The hardware devices such as various sensors may be used in the acquisition of the priori of layer, and the corresponding hardware device of preceding viewing system can be
The hardware devices such as vehicle-mounted binocular camera.Image procossing layer for the ROI region generation of vehicle detection, online sliding window vehicle detection,
Monocular distance calculates, binocular parallax calculates and output result fusion.Therefore the corresponding hardware device of image procossing layer can be
Entire car controller is either individually used for the controller of vehicle odometry.Output layer is for exporting road front vehicles information, therefore
The corresponding hardware device of output layer can be image output device, audio output device etc..
Illustrate another aspect of the present invention below.
With reference to figure 2, which is a kind of a kind of embodiment for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method
Flow diagram, the flow include,
Step S1 carries out vehicle according to target vehicle texture information and priori road information in the visual angle of binocular camera side
Detection, and determine vehicle position and at a distance from this vehicle；
Step S2 carries out the similitude matching of target vehicle in the visual angle of the binocular camera other side, it is higher to choose similitude
Matching position, determine the center offset of target vehicle in two-phase visual angle, and target vehicle is determined according to center offset
Physical location and at a distance from this vehicle；
Step S3 determines the confidence interval of binocular camera according to binocular camera parameter, in binocular confidence interval, exports root
According to the physical location of the target vehicle of center offset determination and at a distance from this vehicle, outside confidence interval, then by instructing offline
Experienced grader verifies the reliability of vehicle detection.
When specific implementation, the flow of the embodiment of the present invention includes
Step S11 obtains the vehicle detecting information in the visual angle of binocular camera side, and according to plane longitudinal direction road mould
Type and binocular camera internal and external parameter, determine vehicle detection ROI region；When specific implementation, for example, vehicle front on real road
10*50 meters of rectangular area can be based on pinhole imaging system principle, and the ladder in corresponding image-region is calculated using above-mentioned parameter
Shape range.Wherein camera heights and pitch angle may be assumed that as constant, can also be integrated according to IMU sensors or image algorithm measured value
Real-time update.Camera internal and external parameter may include focal length, camera heights h, camera relative vehicle axis line deviation d and phase
Machine pitching angle theta.Wherein plane longitudinal direction road model is pre-set, just assumes that the road residing for vehicle is plane.
Step S12 is filtered ROI region significant characteristics：For example, using edge filter, to the areas vehicle detection ROI
Domain carries out edge detection, carries out binaryzation to edge detection results using threshold value, area-of-interest is generated using binaryzation result
ROI mask remove the poor region of edge feature conspicuousness from ROI region.Wherein binaryzation refers to result by some setting
It is 0 or 1 that threshold value, which divides output,.
Step S13, in the regions vehicle detection RIO, according to the priori of road model and target vehicle width, height ratio
Knowledge generates ROI region sliding window search listing, and list content includes sliding window lower-left angular coordinate, and sliding window width and sliding window are high
Wide ratio.Using the cascade classifier obtained by off-line training, classify to the picture material of sliding window list, determines in sliding window list
The respective image position of target vehicle；Vehicle candidate frame lower edge is optimized using vehicle bottom shadow information.It is specific real
Now, priori may include road model, vehicle width range, vehicle depth-width ratio range；Vehicle detection grader is offline
It is trained for：Using the adaboost graders of cascade haar features (or LBP features), training vehicle detecting algorithm is weak per level-one
The training process of grader is：Initialize the weights distribution (each sample assigns same weight coefficient) of training data, training
In the process, if the sample is accurately classified, the weight coefficient of the sample is reduced；Conversely, then improving respective weights coefficient.Instead
The multiple iteration above process, generates several Weak Classifier grades.It finally cascades each Weak Classifier and generates final strong classifier (increase error
The small Weak Classifier weight coefficient of rate reduces the big Weak Classifier weight coefficient of error rate).The vehicle acquired using test vehicle
Head positive sample image is no less than 12000, and negative sample image is no less than 20000, and it (refers to that will classify manually to be excavated using difficult example
The flase drop that device is difficult to is rejoined in training sample by screening and is trained) and Active Learning (refer to first with network
One cascade classifier of related training data pre-training of collection, the sample acquired using our camera apparatus of this grader pair
This progress detects roughly, according to testing result the positive negative sample of artificial screening, re -training grader) method training for promotion effect；
It is configurable parameter that grader, which cascades the number of plies,.
Step S14, in ROI region, according to plane longitudinal direction road model and binocular camera internal and external parameter, by target
Vehicle candidate region bottom image coordinate inverse perspective mapping obtains the position of target vehicle to road surface coordinate system.When specific implementation,
Camera internal and external parameter includes focal length, camera heights h, camera relative vehicle axis line deviation d and camera pitching angle theta；Road
The coordinate vertices of areal coordinate system can be defined on vehicle forefront midpoint, and longitudinal direction of car is the directions x, is laterally the directions y.
Step S15 determines the position of the target vehicle of binocular camera side visual angle detection at binocular camera other side visual angle
Auto-correlation coefficient.It, can be according to the target vehicle position detected in the visual angle candidate region of binocular camera side in practical application
It sets and the vehicle distances estimated value of monocular system output, calculates it and correspond to phase in disparity range in another camera estimated distance
Like property.The similitude can detect the auto-correlation coefficient of target vehicle template to judge by calculating, be used
R (x, y)=sum ((IL(x ', y ')-IR(x+x ', y+y ＇))2)
Formula be：
Determining auto-correlation coefficient is compared by step S16 with the lower threshold of preset auto-correlation coefficient, is chosen high
In the vehicle location of the most auto-correlation coefficient of predetermined lower threshold value, as the highest matching position of similarity；
Step S17 determines the center position of the highest matching position of similarity in two magazine parallax d of binocular, and
Disparity computation is converted into the actual coordinate (X, Y, Z) in camera coordinates system, used algorithm is：
Wherein, stereofFor binocular camera focal length；stereobaselineFor binocular camera baseline length；(u0, v0) is double
Mesh picture centre point coordinates.
Step S31 calculates binocular resolution of ranging according to binocular camera parameter, chooses binocular resolution of ranging and is more than
10% be used as confidence interval, resolution ratio define be apart from vehicle percentage error, calculation formula caused by unit parallax：
Step S18, using binocular distance measurement result as observation, updates target vehicle relative position in binocular confidence interval；
Step S19 corresponds to parallax region outside binocular confidence interval in the other side visual angle target vehicle of two camera of binocular
It is interior, with sliding window search validation vehicle detection reliability, if consistent with testing result in two visual angle of binocular, with monocular distance measurement result
As observation, the relative position of target vehicle is updated；If testing result is inconsistent in two visual angle of binocular, pre- with state equation
It surveys result and updates target vehicle relative position, the state equation is：
xt+1=Axt+But+ωt
yt+1=Cxt+ρt
Wherein, x is system state amount, including the position of vehicle and change in location information, i.e. [X, Y, dX, dY] T；Y be
Overall view measures, and includes the location information of vehicle, i.e. [X, Y] T；ω, ρ are respectively process noise and measurement noise；Respectively by adjusting
Corresponding process noise matrix Q is completed with measurement noise matrix R1 and R2.
Vehicle detection-determines image sense according to priori (road model, vehicle width range, vehicle depth-width ratio range)
Property interest region in sliding window search for size range, in the interesting image regions of setting, first with significant characteristics, primary dcreening operation is not
Containing notable unity and coherence in writing feature candidate region, the cascade classifier of off-line training is recycled, searches for the vehicle position within the scope of particular dimensions
It sets.According to fore-and-aft plane road model, the road target vehicle detected is calculated into row distance.It is surveyed compared to traditional binocular
Away from algorithm, the invention avoids global or half global disparities to calculate, and operation efficiency greatly promotes.In addition, the present invention is to binocular system
The installation of system and stated accuracy requirement are lower, and for the biocular systems of different focal length, the present invention still has the feasibility of application,
And traditional parallax calculation method is not suitable for such biocular systems.Compared to traditional monocular location algorithm, precision of the present invention
Higher, and it is more preferable for the robustness of vehicle pitch rate variation.
The present invention considers the respective quality of binocular range-measurement system, in the two range-measurement systems section high in respective confidence level
Interior optimum selecting distance measurement result improves the precision of ranging as system output.First, the present invention utilizes unity and coherence in writing information and priori
Road information carries out vehicle detection to area-of-interest in binocular camera left side perspective；For the alternative vehicle area detected
Domain carries out the similitude matching of the auto model in binocular camera right side perspective；Choose the highest matching position of similitude, meter
Vehicle alternative area center offset in calculation amount camera perspective, to calculate the relative distance of the vehicle and this vehicle.It improves
The precision of ranging.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (5)
1. one kind being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method, which is characterized in that including：
S1 carries out vehicle detection according to target vehicle texture information and priori road information in the visual angle of binocular camera side, and
Determine the position of vehicle and at a distance from this vehicle；
S2 carries out the similitude matching of target vehicle in the visual angle of the binocular camera other side, chooses the higher match bit of similitude
It sets, determines the center offset of target vehicle in two-phase visual angle, and determine the physical location of target vehicle according to center offset
With at a distance from this vehicle；
S3 determines the confidence interval of binocular camera according to binocular camera parameter, and in binocular confidence interval, output is inclined according to center
Shifting amount determine target vehicle physical location and at a distance from this vehicle, outside confidence interval, then pass through the classification of off-line training
Device verifies the reliability of vehicle detection.
2. according to claim 1 be used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method, which is characterized in that step
S1 is specifically included：
S11 obtains the vehicle detecting information in the visual angle of binocular camera side, and according to plane longitudinal direction road model and binocular
Camera internal and external parameter determines vehicle detection ROI region；
S12 is raw according to the priori of road model and target vehicle width, height ratio in the regions vehicle detection RIO
At ROI region sliding window search listing, and the cascade classifier obtained by off-line training, the picture material of sliding window list is carried out
Classification, determines the respective image position of target vehicle in sliding window list；
S13, according to plane longitudinal direction road model and binocular camera internal and external parameter, by target vehicle candidate region bottom image
Coordinate inverse perspective mapping obtains the position of target vehicle to road surface coordinate system.
3. according to claim 2 be used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method, which is characterized in that step
S1 further includes：According to edge filter, edge detection is carried out to vehicle detection ROI region, it is poor to take out edge feature conspicuousness
Region.
4. according to claim 1 be used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method, which is characterized in that described
Step s2 includes：
S21 determines auto-correlation of the position at binocular camera other side visual angle of the target vehicle of binocular camera side visual angle detection
Coefficient, used formula are：
R (x, y)=sum ((IL(x ', y ')-IR(x+x ＇, y+y ＇))2)
ILIndicate left view, IRIndicate that right view, x indicate sliding window in the search offset in the directions x, y expression sliding windows are in the directions y
Search for offset；X ', y ' pixel coordinate in left view is indicated respectively.
Determining auto-correlation coefficient is compared by S22 with the lower threshold of preset auto-correlation coefficient, is chosen to be higher than and be set in advance
The vehicle location for limiting the most auto-correlation coefficient of threshold value, as the highest matching position of similarity；
S23 determines the center position of the highest matching position of similarity in two magazine parallax d of binocular, and by parallaxometer
The actual coordinate (X, Y, Z) being converted into camera coordinates system is calculated, used algorithm is：
Wherein, stereoiFor binocular camera focal length；stereobaselineFor binocular camera baseline length；(u0, v0) is binocular figure
Inconocenter point coordinates.
5. being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method according to right 1, which is characterized in that the step
S3 is specifically included：
S31 calculates binocular resolution of ranging according to binocular camera parameter, and selection binocular resolution of ranging is used as more than 10% and sets
Believe section, resolution ratio define be apart from vehicle percentage error, calculation formula caused by unit parallax：
S32, using binocular distance measurement result as observation, updates target vehicle relative position in binocular confidence interval；
S33 is corresponded in the other side visual angle target vehicle of two camera of binocular in parallax region, outside binocular confidence interval with cunning
Window search validation vehicle detection reliability, if consistent with testing result in two visual angle of binocular, using monocular distance measurement result as sight
Measured value updates the relative position of target vehicle；If testing result is inconsistent in two visual angle of binocular, with state equation prediction result
Target vehicle relative position is updated, the state equation is：
xt+1=Axt+But+ωt
yt+1=Cxt+ρt
Wherein, x is system state amount, including the position of vehicle and change in location information, i.e. [X, Y, dX, dY] T；Y is systematic perspective
It measures, includes the location information of vehicle, i.e. [X, Y] T；ω, ρ are respectively process noise and measurement noise；It is corresponding by adjusting respectively
Process noise matrix Q completed with measurement noise matrix R1 and R2；A indicates state space matrices；B indicates input space square
Battle array；C indicates observation space matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810571399.7A CN108645375B (en) | 2018-06-05 | 2018-06-05 | Rapid vehicle distance measurement optimization method for vehicle-mounted binocular system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810571399.7A CN108645375B (en) | 2018-06-05 | 2018-06-05 | Rapid vehicle distance measurement optimization method for vehicle-mounted binocular system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108645375A true CN108645375A (en) | 2018-10-12 |
| CN108645375B CN108645375B (en) | 2020-11-17 |
Family
ID=63759496
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810571399.7A Active CN108645375B (en) | 2018-06-05 | 2018-06-05 | Rapid vehicle distance measurement optimization method for vehicle-mounted binocular system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108645375B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109711279A (en) * | 2018-12-08 | 2019-05-03 | 南京赫曼机器人自动化有限公司 | A kind of obstacle detection method for agricultural environment |
| CN110926408A (en) * | 2019-12-04 | 2020-03-27 | 北京中科慧眼科技有限公司 | Short-distance measuring method, device and system based on characteristic object and storage medium |
| CN110936937A (en) * | 2019-03-21 | 2020-03-31 | 湖南三一港口设备有限公司 | Automatic braking method and system |
| CN111447429A (en) * | 2020-04-02 | 2020-07-24 | 深圳普捷利科技有限公司 | Vehicle-mounted naked eye 3D display method and system based on binocular camera shooting |
| CN111982061A (en) * | 2020-07-15 | 2020-11-24 | 杭州晨安科技股份有限公司 | Distance measurement method based on different focal lengths of binocular fixed-focus cameras |
| CN112406901A (en) * | 2020-12-05 | 2021-02-26 | 深圳瑞为智能科技有限公司 | Binocular distance measuring method for vehicle blind area detection alarm device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101294801A (en) * | 2007-07-13 | 2008-10-29 | 东南大学 | Vehicle distance measuring method based on binocular vision |
| CN103245322A (en) * | 2013-04-10 | 2013-08-14 | 南京航空航天大学 | Distance measurement method and system based on binocular stereo vision |
| CN105718865A (en) * | 2016-01-15 | 2016-06-29 | 武汉光庭科技有限公司 | System and method for road safety detection based on binocular cameras for automatic driving |
| WO2018077218A1 (en) * | 2016-10-31 | 2018-05-03 | 杭州海康威视数字技术股份有限公司 | Depth measuring method and system |
| CN107990878A (en) * | 2017-11-08 | 2018-05-04 | 中国科学院西安光学精密机械研究所 | Distance measuring system and distance measuring method based on low-light-level binocular camera |
| CN108108680A (en) * | 2017-12-13 | 2018-06-01 | 长安大学 | A kind of front vehicle identification and distance measuring method based on binocular vision |
-
2018
- 2018-06-05 CN CN201810571399.7A patent/CN108645375B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101294801A (en) * | 2007-07-13 | 2008-10-29 | 东南大学 | Vehicle distance measuring method based on binocular vision |
| CN103245322A (en) * | 2013-04-10 | 2013-08-14 | 南京航空航天大学 | Distance measurement method and system based on binocular stereo vision |
| CN105718865A (en) * | 2016-01-15 | 2016-06-29 | 武汉光庭科技有限公司 | System and method for road safety detection based on binocular cameras for automatic driving |
| WO2018077218A1 (en) * | 2016-10-31 | 2018-05-03 | 杭州海康威视数字技术股份有限公司 | Depth measuring method and system |
| CN107990878A (en) * | 2017-11-08 | 2018-05-04 | 中国科学院西安光学精密机械研究所 | Distance measuring system and distance measuring method based on low-light-level binocular camera |
| CN108108680A (en) * | 2017-12-13 | 2018-06-01 | 长安大学 | A kind of front vehicle identification and distance measuring method based on binocular vision |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109711279A (en) * | 2018-12-08 | 2019-05-03 | 南京赫曼机器人自动化有限公司 | A kind of obstacle detection method for agricultural environment |
| CN109711279B (en) * | 2018-12-08 | 2023-06-20 | 南京赫曼机器人自动化有限公司 | Obstacle detection method for agricultural environment |
| CN110936937A (en) * | 2019-03-21 | 2020-03-31 | 湖南三一港口设备有限公司 | Automatic braking method and system |
| CN110936937B (en) * | 2019-03-21 | 2021-03-30 | 湖南三一港口设备有限公司 | Automatic braking method and system |
| CN110926408A (en) * | 2019-12-04 | 2020-03-27 | 北京中科慧眼科技有限公司 | Short-distance measuring method, device and system based on characteristic object and storage medium |
| CN111447429A (en) * | 2020-04-02 | 2020-07-24 | 深圳普捷利科技有限公司 | Vehicle-mounted naked eye 3D display method and system based on binocular camera shooting |
| CN111447429B (en) * | 2020-04-02 | 2021-03-19 | 深圳普捷利科技有限公司 | Vehicle-mounted naked eye 3D display method and system based on binocular camera shooting |
| CN111982061A (en) * | 2020-07-15 | 2020-11-24 | 杭州晨安科技股份有限公司 | Distance measurement method based on different focal lengths of binocular fixed-focus cameras |
| CN111982061B (en) * | 2020-07-15 | 2022-08-23 | 杭州晨安科技股份有限公司 | Distance measurement method based on different focal lengths of binocular fixed-focus cameras |
| CN112406901A (en) * | 2020-12-05 | 2021-02-26 | 深圳瑞为智能科技有限公司 | Binocular distance measuring method for vehicle blind area detection alarm device |
| CN112406901B (en) * | 2020-12-05 | 2022-03-29 | 深圳瑞为智能科技有限公司 | Binocular distance measuring method for vehicle blind area detection alarm device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108645375B (en) | 2020-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110322702B (en) | Intelligent vehicle speed measuring method based on binocular stereo vision system | |
| CN108645375A (en) | One kind being used for vehicle-mounted biocular systems rapid vehicle distance measurement optimization method | |
| CN110285793B (en) | Intelligent vehicle track measuring method based on binocular stereo vision system | |
| US11461912B2 (en) | Gaussian mixture models for temporal depth fusion | |
| CN104700414B (en) | A kind of road ahead pedestrian's fast ranging method based on vehicle-mounted binocular camera | |
| CN110569704B (en) | Multi-strategy self-adaptive lane line detection method based on stereoscopic vision | |
| Broggi et al. | The single frame stereo vision system for reliable obstacle detection used during the 2005 DARPA grand challenge on TerraMax | |
| Vaudrey et al. | Differences between stereo and motion behaviour on synthetic and real-world stereo sequences | |
| Bensrhair et al. | Stereo vision-based feature extraction for vehicle detection | |
| CN104318561B (en) | Method for detecting vehicle motion information based on integration of binocular stereoscopic vision and optical flow | |
| JP4919036B2 (en) | Moving object recognition device | |
| CN101408978B (en) | Method and apparatus for detecting barrier based on monocular vision | |
| JP6574611B2 (en) | Sensor system for obtaining distance information based on stereoscopic images | |
| CN107750364A (en) | Detected using the road vertically profiling of stable coordinate system | |
| JP6779365B2 (en) | Object detection device and vehicle | |
| CN109263637B (en) | Collision prediction method and device | |
| CN105335955A (en) | Object detection method and object detection apparatus | |
| JP2014138420A (en) | Depth sensing method and system for autonomous vehicle | |
| JP2004086779A (en) | Obstacle detection device and its method | |
| CN106370884A (en) | Vehicle speed measurement method based on binocular camera computer vision technology | |
| CN104574393A (en) | Three-dimensional pavement crack image generation system and method | |
| CN110220500B (en) | Binocular camera-based distance measurement method for unmanned driving | |
| CN109410264A (en) | A kind of front vehicles distance measurement method based on laser point cloud and image co-registration | |
| CN109791607A (en) | It is detected from a series of images of video camera by homography matrix and identifying object | |
| CN107492107A (en) | The object identification merged based on plane with spatial information and method for reconstructing |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 310051 1st and 6th floors, no.451 Internet of things street, Binjiang District, Hangzhou City, Zhejiang Province Patentee after: Zhejiang Zero run Technology Co.,Ltd. Address before: 310051 1st and 6th floors, no.451 Internet of things street, Binjiang District, Hangzhou City, Zhejiang Province Patentee before: ZHEJIANG LEAPMOTOR TECHNOLOGY Co.,Ltd. |