CN109747655A - Steering instructions generation method and device for automatic driving vehicle - Google Patents
Steering instructions generation method and device for automatic driving vehicle Download PDFInfo
- Publication number
- CN109747655A CN109747655A CN201711082002.XA CN201711082002A CN109747655A CN 109747655 A CN109747655 A CN 109747655A CN 201711082002 A CN201711082002 A CN 201711082002A CN 109747655 A CN109747655 A CN 109747655A
- Authority
- CN
- China
- Prior art keywords
- steering instructions
- ambient image
- sample time
- current sample
- feature
- 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
Landscapes
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the present application discloses the steering instructions generation method and device for automatic driving vehicle.One specific embodiment of this method includes: to obtain the collected ambient image of current sample time；Extract the feature of the collected ambient image of current sample time；The feature of feature and history samples moment collected ambient image based on the collected ambient image of current sample time is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.The steering instructions that the embodiment realizes current time generation are homogeneously associated with current environment image and history environment image, and the accuracy rate of steering instructions decision can be improved.
Description
Technical field
The invention relates to field of computer technology, and in particular to field of artificial intelligence, more particularly, to
The steering instructions generation method and device of automatic driving vehicle.
Background technique
With the improvement of living standards, the quantity of motor vehicles is also more and more, so that the appearance of people is more and more simpler.
However due to the inappropriate driving of vehicle driving personnel, such as traffic problems caused by driving when intoxicated etc. are for example
Traffic accident is increasingly severe.In order to improve the traffic problems due to caused by the improper driving of driver, vehicle is driven automatically
Sailing technology becomes when next research hotspot.
Realizing that the steering instructions of Vehicular automatic driving are mainly to generate according to the running environment at current time at this stage
Steering instructions.
Summary of the invention
The purpose of the embodiment of the present application is to propose a kind of steering instructions generation method and dress for automatic driving vehicle
It sets.
In a first aspect, the embodiment of the present application provides a kind of steering instructions generation method for automatic driving vehicle, it should
Method includes: to obtain the collected ambient image of current sample time；Extract the collected ambient image of current sample time
Feature；Feature and history samples moment collected ambient image based on the collected ambient image of current sample time
Feature is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.
Second aspect, the embodiment of the present application provide a kind of steering instructions generating means for automatic driving vehicle, should
Device includes: acquiring unit, is configured to obtain the collected ambient image of current sample time；Extraction unit is configured to
Extract the feature of the collected ambient image of current sample time；Matching unit is configured to acquire based on current sample time
The feature of the feature of the ambient image arrived and history samples moment collected ambient image is from preset multiple steering instructions
Middle selection and the matched steering instructions of current sample time.
The third aspect, the embodiment of the present application provide a kind of driving control devices, the driving control devices include: one or
Multiple processors；Storage device, for storing one or more programs, when said one or multiple programs are by said one or more
When a processor executes, so that said one or multiple processors realize the method as described in first aspect.
Fourth aspect, the embodiment of the present application provide a kind of computer readable storage medium, are stored thereon with computer journey
Sequence, which is characterized in that the method as described in first aspect is realized when the computer program is executed by processor.
Steering instructions generation method and device provided by the embodiments of the present application for automatic driving vehicle passes through comprehensive point
Analyse the feature and history samples moment collected ambient image of the collected ambient image of current sample time extracted
Feature chooses from preset multiple steering instructions and the matched steering instructions of current sample time, so that current time is raw
At steering instructions be homogeneously associated with current environment image and history environment image, the accurate of steering instructions decision can be improved
Rate.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is that this application can be applied to exemplary system architecture figures therein；
Fig. 2 is the process according to one embodiment of the steering instructions generation method for automatic driving vehicle of the application
Figure；
Fig. 3 is illustrated according to a principle of the steering instructions generation method for automatic driving vehicle of the application
Figure；
Fig. 4 is the structure according to one embodiment of the steering instructions generating means for automatic driving vehicle of the application
Schematic diagram；
Fig. 5 is adapted for the structural schematic diagram for the computer system for realizing the driving control devices of the embodiment of the present application.
Specific embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, part relevant to related invention is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 is shown can be using the application for the steering instructions generation method of automatic driving vehicle or for automatic
Drive the exemplary system architecture 100 of the embodiment of the steering instructions generating means of vehicle.
As shown in Figure 1, system architecture 100 may include automatic driving vehicle (such as automatic driving vehicle) 101.
Driving control devices 1011, network 1012 and image capture device 1013 are installed in automatic driving vehicle 101.Net
Network 1012 between driving control devices 1011 and image capture device 1013 to provide the medium of communication link.Network 1012
It may include various connection types, such as wired, wireless communication link or fiber optic cables etc..
The intelligent control of the responsible automatic driving vehicle 101 of driving control devices (also known as vehicle-mounted brain) 1011.Drive control
Control equipment 1011 can be the controller being separately provided, such as programmable logic controller (PLC) (Programmable Logic
Controller, PLC), single-chip microcontroller, Industrial Control Computer etc.；It is also possible to by other with input/output end port, and there is fortune
The equipment for calculating the electronic device composition of control function；It can also be the computer equipment for being equipped with vehicle drive control class application.
The equipment such as can be video camera, camera of image capture device 1013.
Driving control devices 1011 receive the image that image capture device 1013 acquires by network 1012.Image is carried out
Processing obtains steering instructions to control the driving of automatic driving vehicle.
It should be noted that for knowing the steering instructions generation method of automatic driving vehicle provided by the embodiment of the present application
It is generally executed by driving control devices 1011, correspondingly, the steering instructions generating means for knowing automatic driving vehicle are generally set
It is placed in driving control devices 1011.
It should be understood that the number of driving control devices, network and image capture device in Fig. 1 is only schematical.
According to needs are realized, any number of driving control devices, network and image capture device can have.
It may include transverse driving control and longitudinal drive control in the Driving control of automatic driving vehicle.Laterally drive
Control i.e. course changing control is sailed, longitudinal drive control for example may include velocity of displacement control and Acceleration Control.Correspondingly, it controls
Instruction may include transverse driving instruction and longitudinal drive instruction.Longitudinal drive instruction for example may include specific displacement, speed
Degree, acceleration instruction value.Transverse driving instruction such as may include for example forward, to the left, to the right steering angle instruction control.
The scheme of the embodiment of the present application will schematically be described so that transverse driving instructs as an example below.But those skilled in the art should
Understand, the method in the embodiment of the present application is equally applicable to the generation of longitudinal drive instruction.
With continued reference to Fig. 2, it illustrates according to the steering instructions generation method for automatic driving vehicle of the application
The process 200 of one embodiment.This is used for the steering instructions generation method of automatic driving vehicle, comprising the following steps:
Step 201, the collected ambient image of current sample time is obtained.
In the present embodiment, the electronic equipment for the steering instructions generation method operation of automatic driving vehicle thereon
(such as driving control devices 1011 shown in FIG. 1) can be by wired connection mode or radio connection from Image Acquisition
Equipment (such as image capture device 1013 shown in FIG. 1) obtains the collected ambient image of current sample time.
Here environment can for example refer to environment locating in automatic driving vehicle driving process, for example, may include
But be not limited to driving lane shape, road sign, spacing, barrier etc..
In application scenes, above-mentioned image capture device can be spaced (such as 1/30 second) acquisition to schedule
Ambient image in automatic driving vehicle driving process.And above-mentioned electronic equipment is sent by collected ambient image.Every time
It can be considered as a sampling instant at the time of acquiring ambient image.
Step 202, the feature of the collected ambient image of current sample time is extracted.
In the present embodiment, based on the collected ambient image of current sample time obtained in step 201, above-mentioned electronics
Equipment (such as driving control devices shown in FIG. 1) can be used various image analysis processing methods and acquire to current sample time
The environment map arrived carries out various analysis processing, to extract the feature of the collected ambient image of current sample time.Above-mentioned image
Analysis and processing method for example can include but is not limited to fourier transform method, window Fourier transform method, Wavelet Transform, minimum
At least one of square law, edge histogram method, neural network method etc..Above-mentioned fourier transform method, window Fourier become
Changing method, Wavelet Transform, least square method, edge histogram method, neural network method is the public affairs studied and applied extensively at present
Know technology, details are not described herein again.
Step 203, feature and history samples moment based on the collected ambient image of current sample time collect
Ambient image feature chosen from preset multiple steering instructions with the matched steering instructions of current sample time.
It in the present embodiment, can for the electronic equipment of the steering instructions generation method operation of automatic driving vehicle thereon
To carry out feature extraction to the collected ambient image of each sampling instant obtained from image capture device.In addition, above-mentioned electricity
Sub- equipment can also store the feature of the ambient image at multiple history samples moment of electronic equipment acquisition.
In application scenes, when can also be specified one or more samplings before current sample time
It carves.
In application scenes, above-mentioned electronic equipment (each period for example can be 1 second) can obtain figure at times
As the acquisition collected ambient image of equipment.Each reception period may include multiple sampling instants.Completed a upper period
The collected ambient image of image capture device is obtained, and extract the collected ambient image of each sampling instant of the period
After feature, before obtaining first collected ambient image of sampling instant of subsequent period, above-mentioned electronic equipment can be deleted
Except the ambient image feature of extracted upper period each sampling instant.So, the above-mentioned historical juncture may include
Each sampling instant before the current sample time in the period is obtained where current sample time.
Above-mentioned electronic equipment can be collected according to the feature and historical juncture of current time collected ambient image
Ambient image feature selected from above-mentioned preset multiple steering instructions with the matched steering instructions of current sample time.Here
Steering instructions for example may include advancing, turning left, turn right etc. the specific crosswise joint instruction such as to turn to.It can be according to actual
It needs to refine above-mentioned crosswise joint instruction, for example, left/right deflects 15 ° etc..
Here, above-mentioned electronic equipment can be according to the feature of the ambient image locating for current time and historical juncture environment
The feature of image generates and the matched steering instructions of current sample time.
Such as there is obstacle in right ahead section in the ambient image of current sample time that gets of above-mentioned electronic equipment
Object and in the ambient image at the history samples moment that electronic equipment is got left side have moving traffic, then above-mentioned electronics is set
It is standby to make and drive according to the feature of the historical juncture ambient image of the feature and extraction of the current time ambient image of extraction
Sail the steering instructions that direction is deflected to right front.
The collected ring of current sample time that the method provided by the above embodiment of the application is extracted by comprehensive reference
The feature of border image and the feature of history samples moment collected ambient image are chosen from preset multiple steering instructions
With the matched steering instructions of current sample time so that the current time steering instructions and the current environment image that generate and going through
History ambient image is homogeneously associated with, and the accuracy rate of Driving Decision-making can be improved.
With continued reference to FIG. 3, it illustrates the steering instructions generation methods for automatic driving vehicle according to the application
Another embodiment flow chart.This is used for the steering instructions generation method of automatic driving vehicle, comprising the following steps:
Step 301, the collected ambient image of current sample time is obtained.
Step 302, the collected ambient image of current sample time is extracted using convolutional neural networks trained in advance
Feature.
In the present embodiment, the electronic equipment for the steering instructions generation method operation of automatic driving vehicle thereon
(such as driving control devices 1011 shown in FIG. 1) can be used the convolutional neural networks trained in advance being disposed therein and extract
The collected ambient image of current sample time.
Above-mentioned convolutional neural networks (Convolutional Neural Network, CNN) are depth artificial neural networks
One kind, usual convolutional neural networks may include multiple feature extraction layers (also known as convolutional layer) and multiple down-sampling layers (again
Title pond layer, Pooling layers).Wherein, feature extraction layer replaces connection with down-sampling layer.Each feature extraction layer can wrap
Include at least one convolution kernel.For a feature extraction layer, rolled up using a convolution kernel of this layer and the output of preceding layer
Product obtains a characteristic pattern.Down-sampling layer asks part flat for the convolution results of the output to feature extraction layer connected to it
Equal and dimension-reduction treatment.It wherein, include multiple weights in the convolution kernel in feature extraction layer.Weight in convolution kernel can be by multiple
Sample training obtains.It is shared using local weight when each convolution kernel of convolutional neural networks is to image zooming-out characteristic pattern,
The complexity of convolutional neural networks can be reduced.
So, the ambient image at collected current time can be input to convolutional neural networks, process is above-mentioned
The characteristic pattern of the processing of convolutional neural networks, output can be indicated with three-dimensional matrice.Such as convolutional neural networks output
Three-dimensional matrice characteristic pattern may include A B row C column matrix.It include B × C element in each B row C column matrix.Here, A >=1,
And A is positive integer.B >=1, and it is positive integer that B, which is positive integer, C >=1, and C,.
It is understood that in each sampling instant, convolutional neural networks can be used extracts the sampling instant and adopt
The feature of the ambient image collected.
The characteristic pattern that convolutional neural networks extract the collected ambient image of each sampling instant can be used.
In addition, the value of all the points in features described above figure can also be converted to one-dimensional characteristic vector by convolutional neural networks,
Using the one-dimensional characteristic vector as the output of convolutional neural networks.Such as the output of above-mentioned convolutional neural networks including A B row C
The characteristic pattern of the three-dimensional matrice of column element can be converted to including A × B × the one-dimensional characteristic vector of C component.
Step 303, the shot and long term trained by the feature of the collected ambient image of current sample time and in advance remembers net
The output feature vector of network last moment output inputs shot and long term memory network, obtains the output feature vector at current time.
Here, shot and long term memory network (Long Short-Term Memory, LSTM) is a kind of Recognition with Recurrent Neural Network, can
To include multiple hidden layers.It may include the state of historical juncture in hidden layer.LSTM can participate in the state of historical juncture hidden layer
Into the calculating process at current time, it is suitable for being spaced and postpone relatively long event in processing and predicted time sequence.
In the present embodiment, the input of LSTM can be the feature and upper one of the ambient image of current sample time acquisition
The output feature vector of moment LSTM output.The feature of the ambient image of current sample time acquisition can be by one-dimensional vector table
Show.Optionally, the output feature that the feature of current time collected ambient image and last moment LSTM can be exported
Vector is combined into input of the one-dimensional vector as current time LSTM.The output of LSTM can be one including multiple elements
Dimensional vector.
Step 304, the output feature vector at current time is handled using the full articulamentum of training in advance, according to place
Reason result is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.
Full articulamentum can play the role of classifier, and the distributed nature expression that can be inputted is mapped to sample
Label space.In actual use, full articulamentum can be realized by convolution.
In the present embodiment, above-mentioned sample labeling space can be for example driving order space.Drive order space for example
It may include multiple specific driving orders, such as drive disc spins and ordered to the driving of the crosswise joints automatic driving vehicle such as angle
It enables.
Optionally, above-mentioned steps 304 can be realized by following sub-step:
Sub-step 3041, for each of multiple steering instructions steering instructions, above-mentioned full articulamentum execution is driven as follows
It sails instruction value calculating operation: calculating each element (also referred to as component) numerical value and each element point in the output feature vector at current time
Not corresponding preset weights product cumulative and；The sum of default offset parameter cumulative and corresponding with the steering instructions is used as should
The corresponding steering instructions value of steering instructions.
Above-mentioned steps 3041 can be realized by following formula:
Wherein, a1, a2, a3 and a4 can be four driving such as move forward, turn left and turn right and turn round respectively
Instruct corresponding steering instructions value.X1, x2, x3 ..., xm can be shot and long term memory network output one-dimensional vector.b1,
B2, b3, b4, which can be, to be corresponded respectively to move forward, turn left and four steering instructions of turning right and turn round are corresponding
Default offset parameter.
For preset weight matrix.
Each preset weights W in above-mentioned weight matrix11、W12、…、W4mAnd each default offset parameter can be by more
A sample training obtains.Formula (1) is only illustrated by taking 4 steering instructions as an example, in actual use, can according to need
Transverse driving instruction is carried out refinement and obtains each transverse driving using the above method to instruct corresponding steering instructions value.
Sub-step 3042 is chosen the maximum steering instructions of steering instructions value and is referred to as the matched driving of current sample time
It enables.
Such as set and each steering instructions value is calculated according to formula (1) be respectively as follows: a1=90%, a2=50%, a3=30%,
A4=5%, then choose the corresponding steering instructions-of steering instructions value a1 move forward (namely steering direction not being deflected) work
For with current time matched steering instructions.
From figure 3, it can be seen that the driving for automatic driving vehicle compared with the example shown in Fig. 2, in the present embodiment
Instruction generation method highlights the feature that image is extracted using convolutional neural networks, and utilizes shot and long term memory network by upper a period of time
The ambient image feature extracted with current time of output vector for carving shot and long term memory network is driven collectively as current time is analyzed
The reference data of instruction is sailed, finally generates steering instructions using full articulamentum.It is suitable for currently driving so as to quickly generate
Sail the steering instructions at moment.
It is worth noting that stating convolutional neural networks, shot and long term memory network and full articulamentum in use to environment
It further include to above-mentioned convolutional neural networks, shot and long term memory network and Quan Lian before image carries out processing generation steering instructions
Connect the process that layer is trained.It can be by convolutional neural networks, shot and long term memory network and full articulamentum as a whole
Model, using multiple sample datas to the entirety formed by convolutional neural networks, shot and long term memory network and full articulamentum
Model be trained.
With further reference to Fig. 4, as the realization to method shown in above-mentioned each figure, this application provides one kind for driving automatically
One embodiment of the steering instructions generating means of vehicle is sailed, the Installation practice is corresponding with embodiment of the method shown in Fig. 2,
The device specifically can be applied in various electronic equipments.
As shown in figure 4, the steering instructions generating means 400 for automatic driving vehicle of the present embodiment include: to receive list
First 401, extraction unit 402, matching unit 403.Wherein, acquiring unit 401 are configured to acquisition current sample time and collect
Ambient image；Extraction unit 402 is configured to extract the feature of the collected ambient image of current sample time；Matching is single
Member 403, is configured to feature based on the collected ambient image of current sample time and the history samples moment is collected
The feature of ambient image is chosen and the matched steering instructions of current sample time from preset multiple steering instructions
In the present embodiment, for the receiving unit 401 of the steering instructions generating means 400 of automatic driving vehicle, extraction
The specific processing of unit 402 and matching unit 403 and its brought technical effect can refer to respectively to be walked in Fig. 2 corresponding embodiment
Rapid 201, the related description of step 202 and step 203, details are not described herein.
In some optional implementations of the present embodiment, extraction unit 402 is further configured to: using training in advance
Convolutional neural networks extract the collected ambient image of current sample time feature.
In some optional implementations of the present embodiment, extraction unit 402 is further configured to: extracting current time
The characteristic pattern of collected ambient image；Characteristic pattern is converted into one-dimensional characteristic vector.
In some optional implementations of the present embodiment, matching unit 403 is further configured to: when by present sample
Carve the feature of collected ambient image and the output feature vector of shot and long term memory network last moment output trained in advance
Shot and long term memory network is inputted, the output feature vector at current time is obtained；Using in advance training full articulamentum to it is current when
The output feature vector at quarter is handled, and is chosen from preset multiple steering instructions according to processing result and current sample time
Matched steering instructions.
In some optional implementations of the present embodiment, matching unit 403 is further configured to: for multiple driving
Each of instruction steering instructions, execute following steering instructions value calculating operation: calculating the output feature vector at current time
Middle each element numerical value preset weights product corresponding with each element cumulative and；It will be cumulative and corresponding with the steering instructions
The sum of default offset parameter is as the corresponding steering instructions value of the steering instructions；The maximum steering instructions of steering instructions value are chosen to make
For the matched steering instructions of current sample time.
Below with reference to Fig. 5, it illustrates the departments of computer science for the driving control devices for being suitable for being used to realize the embodiment of the present application
The structural schematic diagram of system 500.Control equipment shown in Fig. 5 is only an example, should not function to the embodiment of the present application and
Use scope brings any restrictions.
As shown in figure 5, computer system 500 includes central processing unit (CPU, Central Processing Unit)
501, it can be according to the program being stored in read-only memory (ROM, Read Only Memory) 502 or from storage section
506 programs being loaded into random access storage device (RAM, Random Access Memory) 503 and execute various appropriate
Movement and processing.In RAM 503, also it is stored with system 500 and operates required various programs and data.CPU 501,ROM
502 and RAM 503 is connected with each other by bus 504.Input/output (I/O, Input/Output) interface 505 is also connected to
Bus 504.
I/O interface 505 is connected to lower component: the storage section 506 including hard disk etc.；And including such as LAN (local
Net, Local Area Network) card, modem etc. network interface card communications portion 507.Communications portion 507 passes through
Communication process is executed by the network of such as internet.Driver 508 is also connected to I/O interface 505 as needed.Detachable media
509, such as disk, CD, magneto-optic disk, semiconductor memory etc., are mounted on as needed on driver 508, in order to from
The computer program read thereon is mounted into storage section 506 as needed.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed from network by communications portion 507, and/or from detachable media
509 are mounted.When the computer program is executed by central processing unit (CPU) 501, limited in execution the present processes
Above-mentioned function.It should be noted that computer-readable medium described herein can be computer-readable signal media or
Computer readable storage medium either the two any combination.Computer readable storage medium for example can be --- but
Be not limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.
The more specific example of computer readable storage medium can include but is not limited to: have one or more conducting wires electrical connection,
Portable computer diskette, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only deposit
Reservoir (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory
Part or above-mentioned any appropriate combination.In this application, computer readable storage medium, which can be, any include or stores
The tangible medium of program, the program can be commanded execution system, device or device use or in connection.And
In the application, computer-readable signal media may include in a base band or the data as the propagation of carrier wave a part are believed
Number, wherein carrying computer-readable program code.The data-signal of this propagation can take various forms, including but not
It is limited to electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be computer
Any computer-readable medium other than readable storage medium storing program for executing, the computer-readable medium can send, propagate or transmit use
In by the use of instruction execution system, device or device or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc., Huo Zheshang
Any appropriate combination stated.Computer readable storage medium for example may be-but not limited to-electricity, magnetic, optical, electromagnetic,
The system of infrared ray or semiconductor, device or device, or any above combination.Computer readable storage medium has more
The example of body can include but is not limited to: have the electrical connections of one or more conducting wires, portable computer diskette, hard disk, with
Machine accesses memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, just
Take formula compact disc read-only memory (CD-ROM), light storage device, magnetic memory device or above-mentioned any appropriate group
It closes.In this application, it includes or the tangible medium of storage program that the program can be with that computer readable storage medium, which can be any,
It is commanded execution system, device or device use or in connection.And in this application, computer-readable signal
Medium may include in a base band or as the data-signal that carrier wave a part is propagated, wherein carrying computer-readable journey
Sequence code.The data-signal of this propagation can take various forms, including but not limited to electromagnetic signal, optical signal or above-mentioned
Any appropriate combination.Computer-readable signal media can also be any computer other than computer readable storage medium
Readable medium, the computer-readable medium can be sent, propagated or transmitted for by instruction execution system, device or device
Using or program in connection.The program code for including on computer-readable medium can be with any suitable medium
Transmission, including but not limited to: wireless, electric wire, optical cable, RF etc. or above-mentioned any appropriate combination.
Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use
The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor packet
Include acquiring unit, extraction unit and matching unit.Wherein, the title of these units is not constituted under certain conditions to the unit
The restriction of itself, for example, acquiring unit is also described as " for obtaining the collected ambient image of current sample time
Unit ".
As on the other hand, present invention also provides a kind of computer-readable medium, which be can be
Included in device described in above-described embodiment；It is also possible to individualism, and without in the supplying device.Above-mentioned calculating
Machine readable medium carries one or more program, when said one or multiple programs are executed by the device, so that should
Device: the collected ambient image of current sample time is obtained；Extract the feature of the collected ambient image of current sample time；
The feature of feature and history samples moment collected ambient image based on the collected ambient image of current sample time
It is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (12)
1. a kind of steering instructions generation method for automatic driving vehicle, which comprises
Obtain the collected ambient image of current sample time；
Extract the feature of the collected ambient image of current sample time；
Feature and history samples moment collected ambient image based on the collected ambient image of current sample time
Feature is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.
2. the method according to claim 1, wherein the spy for extracting current time collected ambient image
Sign, comprising:
The feature of the collected ambient image of current sample time is extracted using convolutional neural networks trained in advance.
3. according to the method described in claim 2, it is characterized in that, the collected ambient image of the extraction current sample time
Feature, comprising:
Extract the characteristic pattern of the current time collected ambient image；
The characteristic pattern is converted into one-dimensional characteristic vector.
4. the method according to claim 1, wherein described be based on the collected ambient image of current sample time
Feature and history samples moment collected ambient image feature chosen from preset multiple steering instructions with it is current
The matched steering instructions of sampling instant, comprising:
Shot and long term memory network last moment trained by the feature of the collected ambient image of current sample time and in advance is defeated
Output feature vector out inputs the shot and long term memory network, obtains the output feature vector at current time；
Using in advance training full articulamentum the output feature vector at the current time is handled, according to processing result from
It is chosen and the matched steering instructions of current sample time in preset multiple steering instructions.
5. according to the method described in claim 4, it is characterized in that, the full articulamentum using training in advance is to described current
The output feature vector at moment is handled, when chosen from preset multiple steering instructions according to processing result with present sample
Carve matched steering instructions, comprising:
For each of multiple steering instructions steering instructions, executes following steering instructions value calculating operation: working as described in calculating
In the output feature vector at preceding moment each element numerical value preset weights product corresponding with each element cumulative and；It will be described
The sum of default offset parameter cumulative and corresponding with the steering instructions is as the corresponding steering instructions value of the steering instructions；
The maximum steering instructions of steering instructions value are chosen as the matched steering instructions of current sample time.
6. a kind of steering instructions generating means for automatic driving vehicle, described device include:
Acquiring unit is configured to obtain the collected ambient image of current sample time；
Extraction unit is configured to extract the feature of the collected ambient image of current sample time；
Matching unit is configured to feature and history samples moment based on the collected ambient image of current sample time and adopts
The feature of the ambient image collected is chosen and the matched steering instructions of current sample time from preset multiple steering instructions.
7. device according to claim 6, which is characterized in that the extraction unit is further configured to:
The feature of the collected ambient image of current sample time is extracted using convolutional neural networks trained in advance.
8. device according to claim 7, which is characterized in that the extraction unit is further configured to:
Extract the characteristic pattern of the current time collected ambient image；
The characteristic pattern is converted into one-dimensional characteristic vector.
9. device according to claim 6, which is characterized in that the matching unit is further configured to:
Shot and long term memory network last moment trained by the feature of the collected ambient image of current sample time and in advance is defeated
Output feature vector out inputs the shot and long term memory network, obtains the output feature vector at current time；
Using in advance training full articulamentum the output feature vector at the current time is handled, according to processing result from
It is chosen and the matched steering instructions of current sample time in preset multiple steering instructions.
10. device according to claim 9, which is characterized in that the matching unit is further configured to:
For each of multiple steering instructions steering instructions, executes following steering instructions value calculating operation: working as described in calculating
In the output feature vector at preceding moment each element numerical value preset weights product corresponding with each element cumulative and；It will be described
The sum of default offset parameter cumulative and corresponding with the steering instructions is as the corresponding steering instructions value of the steering instructions；
The maximum steering instructions of steering instructions value are chosen as the matched steering instructions of current sample time.
11. a kind of driving control devices, comprising:
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as method as claimed in any one of claims 1 to 5.
12. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
Such as method as claimed in any one of claims 1 to 5 is realized when execution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711082002.XA CN109747655B (en) | 2017-11-07 | 2017-11-07 | Driving instruction generation method and device for automatic driving vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711082002.XA CN109747655B (en) | 2017-11-07 | 2017-11-07 | Driving instruction generation method and device for automatic driving vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109747655A true CN109747655A (en) | 2019-05-14 |
| CN109747655B CN109747655B (en) | 2021-10-15 |
Family
ID=66399598
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711082002.XA Active CN109747655B (en) | 2017-11-07 | 2017-11-07 | Driving instruction generation method and device for automatic driving vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109747655B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188683A (en) * | 2019-05-30 | 2019-08-30 | 北京理工大学 | A kind of automatic Pilot control method based on CNN-LSTM |
| CN110221611A (en) * | 2019-06-11 | 2019-09-10 | 北京三快在线科技有限公司 | A kind of Trajectory Tracking Control method, apparatus and automatic driving vehicle |
| CN111414852A (en) * | 2020-03-19 | 2020-07-14 | 驭势科技(南京)有限公司 | Image prediction and vehicle behavior planning method, device and system and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20140078088A (en) * | 2012-12-17 | 2014-06-25 | (주) 청원통신 | A auto recognition system of vehicle number using CCTV camera |
| CN104484911A (en) * | 2014-10-29 | 2015-04-01 | 云南大学 | QoE-based customized automatic driving parameter optimal setting method |
| WO2016170764A1 (en) * | 2015-04-21 | 2016-10-27 | パナソニックＩｐマネジメント株式会社 | Driving assistance method and driving assistance device, driving control device, vehicle, and driving assistance program using such method |
| CN106844949A (en) * | 2017-01-18 | 2017-06-13 | 清华大学 | A kind of training method for realizing the controllable two-way LSTM models of locomotive section |
| CN107077792A (en) * | 2014-11-18 | 2017-08-18 | 日立汽车系统株式会社 | Drive-control system |
| CN107169567A (en) * | 2017-03-30 | 2017-09-15 | 深圳先进技术研究院 | The generation method and device of a kind of decision networks model for Vehicular automatic driving |
-
2017
- 2017-11-07 CN CN201711082002.XA patent/CN109747655B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20140078088A (en) * | 2012-12-17 | 2014-06-25 | (주) 청원통신 | A auto recognition system of vehicle number using CCTV camera |
| CN104484911A (en) * | 2014-10-29 | 2015-04-01 | 云南大学 | QoE-based customized automatic driving parameter optimal setting method |
| CN107077792A (en) * | 2014-11-18 | 2017-08-18 | 日立汽车系统株式会社 | Drive-control system |
| WO2016170764A1 (en) * | 2015-04-21 | 2016-10-27 | パナソニックＩｐマネジメント株式会社 | Driving assistance method and driving assistance device, driving control device, vehicle, and driving assistance program using such method |
| CN106844949A (en) * | 2017-01-18 | 2017-06-13 | 清华大学 | A kind of training method for realizing the controllable two-way LSTM models of locomotive section |
| CN107169567A (en) * | 2017-03-30 | 2017-09-15 | 深圳先进技术研究院 | The generation method and device of a kind of decision networks model for Vehicular automatic driving |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188683A (en) * | 2019-05-30 | 2019-08-30 | 北京理工大学 | A kind of automatic Pilot control method based on CNN-LSTM |
| CN110221611A (en) * | 2019-06-11 | 2019-09-10 | 北京三快在线科技有限公司 | A kind of Trajectory Tracking Control method, apparatus and automatic driving vehicle |
| CN110221611B (en) * | 2019-06-11 | 2020-09-04 | 北京三快在线科技有限公司 | Trajectory tracking control method and device and unmanned vehicle |
| CN111414852A (en) * | 2020-03-19 | 2020-07-14 | 驭势科技(南京)有限公司 | Image prediction and vehicle behavior planning method, device and system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109747655B (en) | 2021-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107571867B (en) | Method and apparatus for controlling automatic driving vehicle | |
| CN111860155B (en) | Lane line detection method and related equipment | |
| CN107697070B (en) | Driving behavior prediction method and device and unmanned vehicle | |
| CN107609502A (en) | Method and apparatus for controlling automatic driving vehicle | |
| CN106154834B (en) | Method and apparatus for controlling automatic driving vehicle | |
| US20190377965A1 (en) | Systems and Methods for Pipelined Processing of Sensor Data Using Hardware | |
| CN107571864B (en) | The collecting method and device of automatic driving vehicle | |
| CN109358614A (en) | Automatic Pilot method, system, device and readable storage medium storing program for executing | |
| CN107481292A (en) | The attitude error method of estimation and device of vehicle-mounted camera | |
| CN108216229A (en) | The vehicles, road detection and driving control method and device | |
| CN108537326A (en) | For the method for automatic driving vehicle, medium and system | |
| CN109508580A (en) | Traffic lights recognition methods and device | |
| CN107491072A (en) | Vehicle obstacle-avoidance method and apparatus | |
| US20210056388A1 (en) | Knowledge Transfer Between Different Deep Learning Architectures | |
| CN109747655A (en) | Steering instructions generation method and device for automatic driving vehicle | |
| CN111291697A (en) | Method and device for recognizing obstacle | |
| US11107228B1 (en) | Realistic image perspective transformation using neural networks | |
| CN109584299B (en) | Positioning method, positioning device, terminal and storage medium | |
| CN110077416B (en) | Decision tree-based driver intention analysis method and system | |
| CN110320883A (en) | A kind of Vehicular automatic driving control method and device based on nitrification enhancement | |
| CN101941438A (en) | Intelligent detection control device and method of safe interval | |
| CN110532883A (en) | On-line tracking is improved using off-line tracking algorithm | |
| CN111222522B (en) | Neural network training, road surface detection and intelligent driving control method and device | |
| CN110060230A (en) | Three-dimensional scenic analysis method, device, medium and equipment | |
| CN109376664A (en) | Machine learning training method, device, server and medium |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210305 Address after: Room a1905, 19 / F, building 2, No. 18, Kechuang 11th Street, Daxing District, Beijing, 100176 Applicant after: Beijing Jingdong Qianshi Technology Co.,Ltd. Address before: 101, 1st floor, building 2, yard 20, Suzhou street, Haidian District, Beijing 100080 Applicant before: Beijing Jingbangda Trading Co.,Ltd. Effective date of registration: 20210305 Address after: 101, 1st floor, building 2, yard 20, Suzhou street, Haidian District, Beijing 100080 Applicant after: Beijing Jingbangda Trading Co.,Ltd. Address before: 100080 Haidian District, Beijing, 65 Xing Shu Kou Road, 11C, west section of the western part of the building, 1-4 stories West 1-4 story. Applicant before: BEIJING JINGDONG SHANGKE INFORMATION TECHNOLOGY Co.,Ltd. Applicant before: BEIJING JINGDONG CENTURY TRADING Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |