WO2019047644A1 - 用于控制无人驾驶车辆的方法和装置 - Google Patents

用于控制无人驾驶车辆的方法和装置 Download PDF

Info

Publication number
WO2019047644A1
WO2019047644A1 PCT/CN2018/098632 CN2018098632W WO2019047644A1 WO 2019047644 A1 WO2019047644 A1 WO 2019047644A1 CN 2018098632 W CN2018098632 W CN 2018098632W WO 2019047644 A1 WO2019047644 A1 WO 2019047644A1
Authority
WO
WIPO (PCT)
Prior art keywords
traffic sign
image
environment image
vehicle control
information
Prior art date
Application number
PCT/CN2018/098632
Other languages
English (en)
French (fr)
Inventor
唐坤
郁浩
闫泳杉
郑超
张云飞
姜雨
Original Assignee
百度在线网络技术（北京）有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 百度在线网络技术（北京）有限公司 filed Critical 百度在线网络技术（北京）有限公司
Publication of WO2019047644A1 publication Critical patent/WO2019047644A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology

Definitions

  • the present application relates to the field of computer technology, and in particular to the field of Internet technologies, and in particular, to a method and apparatus for controlling an unmanned vehicle.
  • Machine learning is a study of computers that acquire new knowledge and new skills and identify existing knowledge. By reorganizing existing knowledge structures, computers can continually improve their performance. With the development of machine learning, the use of machine learning to achieve the vehicle's unmanned functions is the main development direction of the vehicle driving field.
  • an embodiment of the present application provides a method for controlling an unmanned vehicle, the method comprising: acquiring an environment image of an unmanned vehicle; inputting an environment image into a pre-established traffic sign recognition model to obtain an environment The traffic sign category information of the traffic sign image in the image, wherein the traffic sign recognition model is used to represent the correspondence relationship between the environment image and the traffic sign category information; and based on the pre-established association relationship between the traffic sign category information and the vehicle control information, selecting and executing Vehicle control information to control unmanned vehicles.
  • an embodiment of the present application provides an apparatus for controlling an unmanned vehicle, the apparatus comprising: an acquiring unit, configured to acquire an environment image of the driverless vehicle; and a determining unit, configured to input the environment image to a traffic sign recognition model is established in advance, and the traffic sign category information of the traffic sign image in the environment image is obtained, wherein the traffic sign recognition model is used to represent the correspondence between the environment image and the traffic sign category information; and the execution unit is configured to be based on the pre-established The relationship between the traffic sign category information and the vehicle control information, and the vehicle control information is selected and executed to control the unmanned vehicle.
  • an embodiment of the present application provides an unmanned vehicle, including: one or more processors; a storage device for storing one or more programs; and when one or more programs are processed by one or more The apparatus is executed such that one or more processors implement the method as described in any one of the first aspects.
  • the embodiment of the present application provides a computer readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the method described in any one of the first aspects is implemented.
  • the method and device for controlling an unmanned vehicle obtaineds an environment image of an unmanned vehicle, and then inputs the environment image into a pre-established traffic sign recognition model to obtain a traffic sign image in the environment image. Traffic sign category information; finally, based on the association relationship between the pre-established traffic sign category information and the vehicle control information, the vehicle control information is selected and executed to control the unmanned vehicle, so that the vehicle can be controlled more accurately.
  • FIG. 1 is an exemplary system architecture diagram to which the present application can be applied;
  • FIG. 2 is a flow chart of one embodiment of a method for controlling an unmanned vehicle in accordance with the present application
  • FIG. 3 is a schematic diagram of an application scenario of a method for controlling an unmanned vehicle according to an embodiment of the present application
  • FIG. 4 is a flow chart of still another embodiment of a method for controlling an unmanned vehicle in accordance with the present application.
  • FIG. 5 is a schematic structural view of an embodiment of an apparatus for controlling an unmanned vehicle according to the present application.
  • FIG. 6 is a schematic structural diagram of a computer system suitable for implementing an electronic device of an embodiment of the present application.
  • FIG. 1 illustrates an exemplary system architecture 100 of an embodiment of a method for controlling an unmanned vehicle or a device for determining an unmanned vehicle to which the present application may be applied.
  • system architecture 100 can include driverless vehicle 101.
  • a driving control device 1011, a network 1012, and an image capturing device 1013 may be mounted on the driverless vehicle 101.
  • Network 1012 is used to provide a medium for the communication link between driving control device 1011 and image acquisition device 1013.
  • Network 1012 can include various types of connections, such as wired, wireless communication links, fiber optic cables, and the like.
  • the driving control device 1011 is responsible for the intelligent control of the unmanned vehicle.
  • the driving control device 1011 may be a separately set controller, such as a programmable logic controller (PLC), a single chip microcomputer, an industrial control machine, etc., or may have other input/output ports and have operation control functions.
  • PLC programmable logic controller
  • the method for controlling an unmanned vehicle provided by the embodiment of the present application is generally performed by the driving control device 1011. Accordingly, the device for controlling the unmanned vehicle is generally disposed in the driving control device 1011.
  • driving control devices 1011 and image acquisition devices 1013 in FIG. 1 is merely illustrative. Depending on the needs of the implementation, there may be any number of driving control devices 1011, image acquisition devices 1013. It should be noted that the image acquisition device may not be included in the system architecture.
  • the method for controlling an unmanned vehicle includes the following steps:
  • Step 201 Acquire an environment image of the driverless vehicle.
  • an image capturing device is installed in the unmanned vehicle, and the image capturing device can collect the left, right, front or rear environment images of the unmanned vehicle running in real time.
  • the image capture device can be a camera, a camera, or the like.
  • the environment image may be a picture or a video.
  • the environment image collected by the image capturing device may include many types of images, such as an image containing road conditions, and may include a traffic sign image.
  • an electronic device for example, the driving control device 1011 shown in FIG. 1 on which the method for controlling an unmanned vehicle is operated can be obtained from the image capturing device by means of a wired connection or a wireless connection.
  • the image to be identified may be an image of the environment surrounding the unmanned vehicle.
  • the image capture device may transmit the image to be recognized to the electronic device in the form of a single frame image.
  • Step 202 Input an environment image into a pre-established traffic sign recognition model to obtain traffic sign category information of the traffic sign image in the environment image.
  • an electronic device for example, the driving control device 1011 shown in FIG. 1 on which the method for controlling an unmanned vehicle runs may input the above-described environment image into a pre-established traffic sign recognition model to obtain Traffic sign category information of traffic sign images in environmental images.
  • the traffic sign category information may be information for characterizing the traffic sign category, such as a category of the traffic sign, a size of the traffic sign, a location of the traffic sign in the environment image, and the like.
  • the above traffic sign category information can be embodied in the form of a feature vector.
  • the traffic sign recognition model is used to represent the correspondence between the environment image and the traffic sign category information.
  • each environment image includes a traffic sign image
  • the traffic sign image may be, for example, a traffic light image, a forbidden U-turn image, a vehicle speed limit image, and the like.
  • one or more traffic signs having the same function may be used as a traffic sign category, such as a traffic sign type of a vehicle speed limit, a traffic sign type in which a vehicle is prohibited from turning around, and the like.
  • the traffic sign recognition model may be stored locally in the above electronic device. It should be noted that the above traffic sign recognition model can be established by other electronic devices.
  • the traffic sign recognition model of step 202 can be obtained by the following steps:
  • the historical environment image collection includes a plurality of historical environment images, which are usually acquired by an image acquisition device of an unmanned vehicle.
  • Step 203 Select and execute vehicle control information to control the unmanned vehicle based on the association relationship between the previously established traffic sign category information and the vehicle control information.
  • the electronic device for example, the driving control device shown in FIG. 1 on which the method for controlling the unmanned vehicle runs may be based on the association relationship between the pre-established traffic sign category information and the vehicle control information. Vehicle control information is selected and executed to control the above-mentioned unmanned vehicle.
  • the plurality of traffic sign category information and the plurality of vehicle control information may be pre-stored in the electronic device, and the relationship relationship information is stored, where the association relationship information is used to indicate the association between the traffic sign category information and the vehicle control information. relationship.
  • a traffic sign category information may be associated with a piece of vehicle control information, such as "speed limit 60 km/h" traffic sign category information, and the associated vehicle control information may be "control the speed of the unmanned vehicle at 60 km/h. Within.”
  • one type of traffic sign category information may be associated with a plurality of pieces of vehicle control information.
  • the vehicle control information may be selected based on the control strategy information corresponding to the traffic sign category information set in advance. For example, if the traffic sign category information of the traffic signal is encountered, the control strategy information may indicate “red light stop, green light line”, if the current encounter is a red light, then the control information indicating the brake is selected; if the current encounter is a green light , then select the control information indicating the driving.
  • FIG. 3 is a schematic diagram of an application scenario of a method for controlling an unmanned vehicle according to the present embodiment.
  • the driverless vehicle 301 is traveling on the road, and a traffic sign 302 is disposed in front of it, and the traffic sign 302 is a "speed limit 70 km/h" sign.
  • the camera provided in the driverless vehicle 301 can capture an environment image, and one of the environment images captured by the camera includes a traffic sign 302.
  • the camera can transmit the captured environmental image 303 to the driving control device 304 of the driverless vehicle.
  • the driving control device 304 can acquire an environmental image 303 of the driverless vehicle.
  • the driving control device 304 can input the above-described environment image 303 to the traffic sign recognition model to obtain the traffic sign category information 305 corresponding to the environment image 303.
  • the driving control device may select and execute the control information 306 according to the association relationship between the previously established traffic sign category information and the vehicle control information to control the driverless vehicle 301.
  • the speed of controlling an unmanned vehicle is limited to 70 km/h.
  • the method for controlling an unmanned vehicle obtaineds an environment image of an unmanned vehicle, and then inputs the environment image into a pre-established traffic sign recognition model to obtain a traffic sign of the traffic sign image in the environmental image.
  • the category information finally, based on the association relationship between the pre-established traffic sign category information and the vehicle control information, the vehicle control information is selected and executed to control the unmanned vehicle, so that the vehicle can be controlled more accurately.
  • the method flow 400 for controlling an unmanned vehicle includes the following steps:
  • Step 401 Acquire an environmental image of a human unmanned vehicle.
  • an electronic device e.g., the driving control device 1011 shown in Fig. 1 on which the method for controlling an unmanned vehicle operates can acquire an environmental image of the driverless vehicle.
  • Step 402 Input an environment image into a pre-established traffic sign recognition model to obtain traffic sign category information of the traffic sign image in the environment image.
  • an electronic device for example, the driving control device 1011 shown in FIG. 1 on which the method for controlling an unmanned vehicle runs may input the above-described environment image into a pre-established traffic sign recognition model to obtain Traffic sign category information of traffic sign images in environmental images.
  • the traffic sign category information of the traffic sign image in the environment image is a vector for indicating the type of the traffic sign.
  • the traffic sign category information may be a feature vector representing only a traffic sign type, the feature vector including a multi-dimensional component, wherein each dimension component may represent a traffic sign type.
  • the traffic sign image information may be a bitmap vector represented by "0" or "1”, and each traffic sign type may be one bit. The position corresponding to the traffic sign image type information may be "1", and the remaining positions are all set to "0".
  • the pre-established traffic sign recognition model may be a convolutional neural network.
  • the convolutional neural network is trained by a collection of historical environment images with traffic signs and traffic signs corresponding to each of the historical environment images in the collection of historical environment images.
  • the convolutional neural network includes a multi-layer pooling layer and a multi-layer link layer corresponding to the pooling layer. After inputting the environment image into the convolutional neural network, the convolutional neural network can evenly divide the above-mentioned environment image into a plurality of environment image sub-blocks, and average or pool the plurality of environment image sub-blocks in the pooling layer. After that, a vector characterizing the type of traffic sign is obtained.
  • Step 403 performing feature extraction on the environment image to obtain a feature vector of the environment image.
  • the electronic device may perform feature extraction on the environment image to obtain a feature vector of the environment image.
  • the features of the environmental image may include color features, texture features, pixel features, grayscale features, and the like.
  • the electronic device may extract a gray level co-occurrence matrix from the environment image by using a gray level co-occurrence matrix algorithm, and use the gray level co-occurrence matrix as an image feature.
  • the gray level co-occurrence matrix can be used to characterize the texture direction, adjacent interval, variation amplitude and other information in the image.
  • the electronic device may uniformly divide the environment image into a plurality of environment image sub-blocks, and then extract image features for each of the environment image sub-blocks, and then index the extracted image features to extract spatial relationship features of the environment images.
  • the foregoing electronic device may also be based on an arbitrary image feature extraction method such as a Hough transform, a random field structure model, a Fourier shape descriptor method, a structural image gray gradient direction matrix, or a plurality of image feature extraction methods. Any combination of the above) performs image feature extraction of the above environment image. Also, the manner of extracting image features is not limited to the above-mentioned manner.
  • feature extraction can be performed on the environment picture using a convolutional neural network.
  • the electronic device may generate an image matrix of the environment map image.
  • images can be analyzed and processed using matrix theory and matrix algorithms.
  • the row of the image matrix corresponds to the height of the image
  • the column of the image matrix corresponds to the width of the image
  • the elements of the image matrix correspond to the pixels of the image.
  • the elements of the image matrix may correspond to the grayscale values of the grayscale image
  • the elements of the image matrix correspond to the RGB of the color image (Red Green Blue, Red, green and blue) values.
  • the electronic device may input an image matrix of the environment image to the pre-trained convolutional neural network to obtain a feature vector of the environment image.
  • the feature vector of the environment image can be used to describe the features that the environment image has.
  • the convolutional neural network can be either AlexNet or GoogleNet. It should be noted that the convolutional neural network is a well-known prior art and will not be described here.
  • step 404 the feature vector and the traffic sign category information are input to the vehicle control model to obtain vehicle control information.
  • the electronic device may input the feature vector and the traffic sign category information into the vehicle control model to obtain the vehicle control information of the vehicle control model.
  • the vehicle control model is used to characterize the correspondence between the feature vector of the environment image, the traffic sign category information corresponding to the traffic sign image, and the vehicle control information.
  • the vehicle control model may be a convolutional neural network.
  • the feature vector and the traffic sign category information may be input to the convolutional neural network to obtain vehicle control information of the driverless vehicle.
  • an environment image is an environment image of a left turn, and the image also includes traffic sign category information of “speed limit 50 km/h”.
  • the electronic device may extract a feature vector of the environment image and a vector corresponding to the traffic sign category information indicating “speed limit 50 km/h”, and input the two vectors into the vehicle control model. In the middle, the vehicle control information is obtained as "left turn” and "the vehicle speed is kept within 50 km/h”.
  • the vehicle control model since the vehicle control model relates the input information to the output information, the vehicle control model can also be regarded as establishing the traffic sign category information, the feature vector of the environment image and the vehicle control information. connection relation.
  • Step 405 controlling the unmanned vehicle based on the vehicle control information.
  • the electronic device can control the unmanned vehicle.
  • the association relationship between the traffic sign category information, the environment image feature information, and the vehicle control information is established by using the vehicle control model.
  • the electronic device controlling the driverless vehicle can introduce more control information, thereby more accurately controlling the unmanned vehicle.
  • the vehicle control model may be stored locally on the electronic device. It should be noted that the above vehicle control model can be established by other electronic devices.
  • the vehicle control model of step 405 can be trained by the following steps:
  • each historical environment image in the historical environment image set respectively acquiring traffic sign category information of the historical environment image and vehicle control information corresponding to the historical environment image, and performing feature extraction on the historical environment image to obtain the The feature vector of the historical environment image.
  • the vehicle control model is trained.
  • the present application provides an embodiment of an apparatus for controlling an unmanned vehicle, the apparatus embodiment being in accordance with the method embodiment illustrated in FIG.
  • the device can be specifically applied to various electronic devices.
  • the apparatus 500 for controlling an unmanned vehicle described above in this embodiment includes an acquisition unit 501, a determination unit 502, and an execution unit 503.
  • the obtaining unit 501 is configured to acquire an environment image of the driverless vehicle
  • the determining unit 502 is configured to input the environment image into the pre-established traffic sign recognition model, and obtain traffic sign category information of the traffic sign image in the environment image, where
  • the traffic sign recognition model is used to represent the correspondence between the environment image and the traffic sign category information
  • the executing unit 503 is configured to select and execute the vehicle control information based on the association relationship between the pre-established traffic sign category information and the vehicle control information to control Driverless vehicles.
  • step 201 the specific processing of the taking unit 501, the determining unit 502, and the executing unit 503 and the technical effects thereof may be referred to the related descriptions of step 201, step 202, and step 203 in the corresponding embodiment of FIG. 2, respectively. This will not be repeated here.
  • the apparatus 500 for controlling an unmanned vehicle further includes a feature extraction unit (not shown) for performing feature extraction on the environment image to obtain an environment image.
  • Feature vector for performing feature extraction on the environment image to obtain an environment image.
  • the executing unit 503 is further configured to: input the feature vector and the traffic sign category information into the vehicle control model to obtain vehicle control information, where the vehicle control model is used to represent the feature vector, Correspondence between traffic sign category information and vehicle control information; control of unmanned vehicles based on vehicle control information.
  • the vehicle control model is obtained by: acquiring a historical environment image set with a traffic sign; and acquiring, for each historical environment image in the historical environment image set, respectively Traffic sign category information of the historical environment image and vehicle control information corresponding to the historical environment image, feature extraction of the historical environment image to obtain a feature vector of the historical environment image; based on each feature vector, corresponding traffic sign category information And the corresponding vehicle control information, training to get the vehicle control model.
  • FIG. 6 a block diagram of a computer system 600 suitable for use in implementing the electronic device of the embodiments of the present application is shown.
  • the electronic device shown in FIG. 6 is merely an example, and should not impose any limitation on the function and scope of use of the embodiments of the present application.
  • the computer system 600 includes a central processing unit (CPU) 601 that can be loaded into random access according to a program stored in a read only memory (ROM) 602 or from the storage portion 606.
  • CPU central processing unit
  • ROM read only memory
  • RAM Random Access Memory
  • various programs and data required for the operation of the system 600 are also stored.
  • the CPU 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604.
  • An input/output (I/O, Input/Output) interface 605 is also coupled to bus 604.
  • the following components are connected to the I/O interface 605: a storage portion 606 including a hard disk or the like; and a communication portion 607 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like.
  • the communication section 607 performs communication processing via a network such as the Internet.
  • Driver 608 is also coupled to I/O interface 605 as needed.
  • a removable medium 609 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory or the like, is mounted on the drive 608 as needed so that a computer program read therefrom is installed into the storage portion 606 as needed.
  • an embodiment of the present disclosure includes a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for executing the method illustrated in the flowchart.
  • the computer program can be downloaded and installed from the network via communication portion 607, and/or installed from removable media 609.
  • the central processing unit (CPU) 601 the above-described functions defined in the method of the present application are performed.
  • the computer readable medium described above may be a computer readable signal medium or a computer readable storage medium or any combination of the two.
  • the computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.
  • a computer readable storage medium may be any tangible medium that can contain or store a program, which can be used by or in connection with an instruction execution system, apparatus or device.
  • a computer readable signal medium may include a data signal that is propagated in the baseband or as part of a carrier, carrying computer readable program code. Such propagated data signals can take a variety of forms including, but not limited to, electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • the computer readable signal medium can also be any computer readable medium other than a computer readable storage medium, which can transmit, propagate, or transport a program for use by or in connection with the instruction execution system, apparatus, or device.
  • Program code embodied on a computer readable medium can be transmitted by any suitable medium, including but not limited to wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
  • each block of the flowchart or block diagram can represent a module, a program segment, or a portion of code that includes one or more of the logic functions for implementing the specified.
  • Executable instructions can also occur in a different order than that illustrated in the drawings. For example, two successively represented blocks may in fact be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented in a dedicated hardware-based system that performs the specified function or operation. Or it can be implemented by a combination of dedicated hardware and computer instructions.
  • the units involved in the embodiments of the present application may be implemented by software or by hardware.
  • the described unit may also be provided in the processor, for example, as a processor including an acquisition unit, a determination unit, and an execution unit.
  • the names of these units do not constitute a limitation on the unit itself under certain circumstances, for example, the acquisition unit may also be described as "a unit for acquiring an environmental image of an unmanned vehicle".
  • the present application also provides a computer readable medium, which may be included in the apparatus described in the above embodiments, or may be separately present and not incorporated into the apparatus.
  • the computer readable medium carries one or more programs that, when executed by the device, cause the device to: acquire an environmental image of the driverless vehicle; input the environmental image to a pre-established traffic sign recognition
  • the model obtains traffic sign category information of the traffic sign image in the environment image, wherein the traffic sign recognition model is used to represent the correspondence relationship between the environment image and the traffic sign category information; and the association relationship between the pre-established traffic sign category information and the vehicle control information , select and execute vehicle control information to control the unmanned vehicle.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Mathematical Physics (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Theoretical Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Human Computer Interaction (AREA)
  • Traffic Control Systems (AREA)

Abstract

本申请实施例公开了用于控制无人驾驶车辆的方法和装置。该方法的一具体实施方式包括：获取无人驾驶车辆的环境图像；将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息，其中，交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆。该实施方式可以准确的对无人驾驶车辆进行控制。

Description

用于控制无人驾驶车辆的方法和装置
本专利申请要求于2017年9月5日提交的、申请号为201710792748.3、申请人为百度在线网络技术(北京)有限公司、发明名称为“用于控制无人驾驶车辆的方法和装置”的中国专利申请的优先权，该申请的全文以引用的方式并入本申请中。
技术领域
本申请涉及计算机技术领域，具体涉及互联网技术领域，尤其涉及用于控制无人驾驶车辆的方法和装置。
背景技术
机器学习是一门研究计算机获取新知识和新技能，并识别现有知识的学问。通过重新组织已有的知识结构，计算机可以不断改善自身的性能。伴随着机器学习的发展，利用机器学习实现车辆的无人驾驶功能为车辆驾驶领域的主要发展方向。
然而，现有的应用于车辆无人驾驶领域的机器学习模型中，难以对图像中的细节进行训练，从而使得驾驶控制设备不能够精确的对车辆进行控制。
发明内容
本申请实施例的目的在于提出一种改进的用于控制无人驾驶车辆的方法和装置，来解决以上背景技术部分提到的技术问题。
第一方面，本申请实施例提供了一种用于控制无人驾驶车辆的方法，该方法包括：获取无人驾驶车辆的环境图像；将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息，其中，交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；基于预先建立的交通标志类别信息与车辆控 制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆。
第二方面，本申请实施例提供了一种用于控制无人驾驶车辆的装置，该装置包括：获取单元，用于获取无人驾驶车辆的环境图像；确定单元，用于将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息，其中，交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；执行单元，用于基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆。
第三方面，本申请实施例提供了一种无人驾驶车辆，包括：一个或多个处理器；存储装置，用于存储一个或多个程序；当一个或多个程序被一个或多个处理器执行，使得一个或多个处理器实现如第一方面中任一实现方式描述的方法。
第四方面，本申请实施例提供了一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现如第一方面中任一实现方式描述的方法。
本申请实施例提供的用于控制无人驾驶车辆的方法和装置，通过获取无人驾驶车辆的环境图像，然后将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息；最后基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆，从而可以更加精确的对车辆进行控制。
附图说明
通过阅读参照以下附图所作的对非限制性实施例所作的详细描述，本申请的其它特征、目的和优点将会变得更明显：
图1是本申请的可以应用于其中的示例性系统架构图；
图2是根据本申请的用于控制无人驾驶车辆的方法的一个实施例的流程图；
图3是根据本申请实施例的用于控制无人驾驶车辆的方法的一个应用场景的示意图；
图4是根据本申请的用于控制无人驾驶车辆的方法的又一个实施例的流程图；
图5是根据本申请的用于控制无人驾驶车辆的装置的一个实施例的结构示意图；
图6是适于用来实现本申请实施例的电子设备的计算机系统的结构示意图。
具体实施方式
下面结合附图和实施例对本申请作进一步的详细说明。可以理解的是，此处所描述的具体实施例仅仅用于解释相关发明，而非对该发明的限定。另外还需要说明的是，为了便于描述，附图中仅示出了与有关发明相关的部分。
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。
图1示出了可以应用本申请的用于控制无人驾驶车辆的方法或用于确定无人驾驶车辆的装置的实施例的示例性系统架构100。
如图1所示，系统架构100可以包括无人驾驶车辆101。无人驾驶车辆101上可以安装有驾驶控制设备1011、网络1012、图像采集装置1013。网络1012用以在驾驶控制设备1011和图像采集装置1013之间提供通信链路的介质。网络1012可以包括各种连接类型，例如有线、无线通信链路或者光纤电缆等等。
驾驶控制设备(又称为车载大脑)1011负责无人驾驶车辆的智能控制。驾驶控制设备1011可以是单独设置的控制器，例如可编程逻辑控制器(Programmable Logic Controller，PLC)、单片机、工业控制机等；也可以是由其他具有输入/输出端口，并具有运算控制功能的电子器件组成的设备；还可以是安装有车辆驾驶控制类应用的计算机设备。
需要说明的是，本申请实施例所提供的用于控制无人驾驶车辆的方法一般由驾驶控制设备1011执行，相应地，用于控制无人驾驶车辆的装置一般设置于驾驶控制设备1011中。
应该理解，图1中的驾驶控制设备1011和图像采集装置1013的数目仅仅是示意性的。根据实现需要，可以具有任意数目的驾驶控制设备1011、图像采集装置1013。需要说明的是，本系统架构中也可以不包括图像采集装置。
继续参考图2，其示出了根据本申请的用于控制无人驾驶车辆的方法的一个实施例的流程200。该用于控制无人驾驶车辆的方法，包括以下步骤：
步骤201，获取无人驾驶车辆的环境图像。
在本实施例中，无人驾驶车辆中安装有图像采集装置，该图像采集装置可以实时采集无人驾驶车辆运行的左方、右方、前方或后方的环境图像。该图像采集装置可以为摄像机、照相机等。其中，该环境图像可以为图片，也可以为视频。上述图像采集装置采集到的环境图像中可以包括很多类型的图像，例如可以是含有路况的图像、可以是含有交通标志图像。
在本实施例中，用于控制无人驾驶车辆的方法运行于其上的电子设备(例如图1所示的驾驶控制设备1011)可以通过有线连接或者无线连接的方式从上述图像采集装置中获取包含有交通标志图像的环境图像。待识别环境图像可以是无人驾驶车辆周围的环境的图像。图像采集装置可以以单帧图像的形式将待识别环境图像发送至上述电子设备。
步骤202，将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息。
在本实施例中，用于控制无人驾驶车辆的方法运行于其上的电子设备(例如图1所示的驾驶控制设备1011)可以将上述环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息。在这里，该交通标志类别信息可以为用于表征交通标志类别的信息，例如可以为交通标志的类别，交通标志的大小，交通标志在环境图像中的位置等。上述交通标志类别信息可以通过特征向量的形式体现。
在本实施例中，上述交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系。
在本实施例中，每一个环境图像中均包含有交通标志图像，该交通标志图像例如可以为红绿灯图像、禁止掉头图像、车辆时速限制图像等。在这里，一个或者多个具有相同功能的交通标志可以作为一种交通标志类别，例如车辆时速限制的交通标志类别、车辆禁止掉头的交通标志类别等。
在本实施例中，交通标志识别模型可以存储在上述电子设备本地。需要说明的是，可以由其它电子设备建立上述交通标志识别模型。
在本实施例的一些可选的实现方式中，步骤202的交通标志识别模型可以通过如下步骤训练得到：
首先，获取带有交通标志的历史环境图像集合。在这里，历史环境图像集合中包含有多个的历史环境图像，该历史环境图像通常是由无人驾驶车辆的图像采集装置采集的。
其次，对于历史环境图像集合中的每个历史环境图像，对该历史环境图像进行图像识别，确定历史交通标志信息，其中，历史交通标志信息包括交通标志的类别信息和交通标志在历史环境图像中的位置信息。最后，基于各个历史环境图像和相应的历史交通标志信息，训练得到交通标志识别模型。
步骤203，基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制所述无人驾驶车辆。
在本实施例中，用于控制无人驾驶车辆的方法运行于其上的电子设备(例如图1所示的驾驶控制设备)可以根据预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制上述无人驾驶车辆。
在本实施例中，上述电子设备中可以预先存储多种交通标志类别信息和多种车辆控制信息，并且存储关联关系信息，关联关系信息用于指示交通标志类别信息与车辆控制信息之间的关联关系。
作为示例，一个交通标志类别信息可以关联一条车辆控制信息，例如“限速60km/h”的交通标志类别信息，所关联的车辆控制信息可 以是“将无人驾驶车辆的车速控制在60km/h以内”。
作为示例，一种交通标志类别信息可以关联多条车辆控制信息。这种情况可以根据预先设置的与这种交通标志类别信息对应的控制策略信息，选取车辆控制信息。例如，遇到交通信号灯的交通标志类别信息，控制策略信息可以指示“红灯停，绿灯行”，如果当前遇到的是红灯，那么选取指示刹车的控制信息；如果当前遇到的是绿灯，那么选取指示行驶的控制信息。
继续参考图3，图3是根据本实施例的用于控制无人驾驶车辆的方法的应用场景的一个示意图。在图3的应用场景中，无人驾驶车辆301在路上行驶，其前方设置有一个交通标志302，该交通标志302为“限速70km/h”指示牌。无人驾驶车辆301中设置的摄像头可以采集环境图像，摄像机采集到的环境图像中的某一张包含有交通标志302。摄像机可以将采集的环境图像303传输至无人驾驶车辆的驾驶控制设备304。驾驶控制设备304可以获取无人驾驶车辆的环境图像303。驾驶控制设备304可以将上述环境图像303输入到交通标志识别模型，得到环境图像303对应的交通标志类别信息305。驾驶控制设备可以根据预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行控制信息306，以控制无人驾驶车辆301。例如，控制无人驾驶车辆的车速限制在70km/h以内。
本申请实施例提供的用于控制无人驾驶车辆的方法，通过获取无人驾驶车辆的环境图像，然后将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息；最后基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆，从而可以更加精确的对车辆进行控制。
进一步参考图4，其示出了用于控制无人驾驶车辆的方法的又一个实施例的流程400。该用于控制无人驾驶车辆的方法流程400，包括以下步骤：
步骤401，获取人无人驾驶车辆的环境图像。
在本实施例中，用于控制无人驾驶车辆的方法运行于其上的电子 设备(例如图1所示的驾驶控制设备1011)可以获取无人驾驶车辆的环境图像。
步骤402，将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息。
在本实施例中，用于控制无人驾驶车辆的方法运行于其上的电子设备(例如图1所示的驾驶控制设备1011)可以将上述环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息。
在本实施例中，环境图像中交通标志图像的交通标志类别信息为用于表示交通标志类型的向量。上述交通标志类别信息可以为仅表示交通标志类型的特征向量，该特征向量中包含有多维分量，其中每一维分量可以表征一个交通标志类型。作为示例，上述交通标志图像信息可以为一个利用“0”或“1”表示的位图向量，每一个交通标志类型可以为一个比特位。与该交通标志图像类型信息对应的位置可以为“1”，其余位置均设置为“0”。
在本实施例中，预先建立的交通标志识别模型可以为卷积神经网络。该卷积神经网络通过带有交通标志的历史环境图像集合以及与历史环境图像集合中每一个历史环境图像对应的交通标志训练而成。其中，卷积神经网络中包含有多层池化层以及与池化层对应的多层链路层。将环境图像输入到该卷积神经网络后，卷积神经网络可以将上述环境图像均匀地划分为若干环境图像子块，对该若干环境图像子块在池化层进行平均池化或最大池化后，得到表征交通标志类型的向量。
步骤403，对环境图像进行特征提取，得到环境图像的特征向量。
在本实施例中，基于步骤401中得到的环境图像，上述电子设备可以对该环境图像进行特征提取，得到环境图像的特征向量。在这里，环境图像的特征可以包括颜色特征、纹理特征、像素特征、灰度特征等。
在本实施例中，上述电子设备可以利用灰度共生矩阵算法，从上述环境图像中提取灰度共生矩阵，将上述灰度共生矩阵作为图像特征。实践中，灰度共生矩阵可以用于表征图像中的纹理方向、相邻间隔、变化幅度等信息。上述电子设备可以将上述环境图像均匀地划分为若干环境 图像子块，然后对每一个环境图像子块提取图像特征，之后为所提取的图像特征建立索引以提取上述环境图像的空间关系特征。
需要说明的是，上述电子设备还可以基于霍夫变换、随机场构造模型、傅里叶形状描述符法、构造图像灰度梯度方向矩阵等任意的图像特征提取方式(或者多种图像特征提取方式的任意结合)进行上述环境图像的图像特征的提取。并且，对图像特征的提取方式不限于以上提到的方式。
在本实施例中，可以利用卷积神经网络对环境图片进行特征提取。具体地，基于步骤中获取的环境图像，上述电子设备可以生成环境图图像的图像矩阵。实践中，图像可以采用矩阵理论和矩阵算法对图像进行分析和处理。其中，图像矩阵的行对应图像的高，图像矩阵的列对应图像的宽，图像矩阵的元素对应图像的像素。作为示例，在图像是灰度图像的情况下，图像矩阵的元素可以对应灰度图像的灰度值；在图像是彩色图像的情况下，图像矩阵的元素对应彩色图像的RGB(Red Green Blue，红绿蓝)值。接着，上述电子设备可以将环境图像的图像矩阵输入至预先训练的卷积神经网络，从而得到环境图像的特征向量。环境图像的特征向量可以用于描述环境图像所具有的特征。在这里，卷积神经网络可以是AlexNet，也可以是GoogleNet。值得注意的是，卷积神经网络为现有的公知技术，在此不再赘述。
步骤404，将特征向量与交通标志类别信息输入到车辆控制模型得到车辆控制信息。
在本实施例中，基于步骤202中得到的环境图像的特征向量，上述电子设备可以将特征向量与交通标志类别信息共同输入到车辆控制模型得到车辆控制模型的的车辆控制信息。
在本实施例中，车辆控制模型用于表征环境图像的特征向量、与上述交通标志图像对应的交通标志类别信息和车辆控制信息之间的对应关系。
在本实施例中，上述车辆控制模型可以为卷积神经网络。例如，可以将特征向量与交通标志类别信息输入到卷积神经网络得到无人驾驶车辆的车辆控制信息。
作为示例，某一环境图像为一左转弯口的环境图像，同时该图像中还包含有“限速50km/h”的交通标志类别信息。上述电子设备在获取到该环境图像后，可以提取该环境图像的特征向量以及用于表示“限速50km/h”的交通标志类别信息对应的向量，同时将该两个向量输入至车辆控制模型中，得到车辆控制信息为“左转弯”以及“车速保持50km/h以内”。
在这里值得注意的是，由于车辆控制模型将输入信息与输出信息之间相关联，因此，车辆控制模型也可以看作是建立交通标志类别信息、环境图像的特征向量与车辆控制信息之间的关联关系。
步骤405，基于车辆控制信息，对无人驾驶车辆进行控制。
本实施例中，基于步骤404得到的车辆控制信息，上述电子设备可以对无人驾驶车辆进行控制。
从图4中可以看出，与图2所示的实施例不同的是，本实施例中，通过利用车辆控制模型建立交通标志类别信息、环境图像特征信息与车辆控制信息之间的关联关系，从而可以使得控制无人驾驶车辆的电子设备可以引入更多的控制信息，从而更加精确的对无人驾驶车辆进行控制。
在本实施例中，车辆控制模型可以是存储在上述电子设备本地的。需要说明的是，可以由其它电子设备建立上述车辆控制模型。
在本实施例的一些可选的实现方式中，步骤405的车辆控制模型可以通过如下步骤训练得到：
首先，获取带有交通标志的历史环境图像集合。
其次，对于历史环境图像集合中的每张历史环境图像，分别获取该历史环境图像的交通标志类别信息以及与该历史环境图像相对应的车辆控制信息，对该历史环境图像进行特征提取，得到该历史环境图像的特征向量。
最后基于与各个历史环境图像对应的特征向量、交通标志类别信息和车辆控制信息，训练得到车辆控制模型。
进一步参考图5，作为对上述各图所示方法的实现，本申请提供 了一种用于控制无人驾驶车辆的装置的一个实施例，该装置实施例与图2所示的方法实施例相对应，该装置具体可以应用于各种电子设备中。
如图5所示，本实施例上述的用于控制无人驾驶车辆的装置500包括：获取单元501、确定单元502和执行单元503。其中，获取单元501，用于获取无人驾驶车辆的环境图像；确定单元502，用于将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息，其中，交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；执行单元503，用于基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆。
在本实施例中，取单元501、确定单元502和执行单元503的具体处理及其所带来的技术效果可分别参考图2对应实施例中步骤201、步骤202以及步骤203的相关说明，在此不再赘述。
在本实施例的一些可选的实现方式中，用于控制无人驾驶车辆的装置500还包括特征提取单元(未示出)，该特征提取单元用于对环境图像进行特征提取，得到环境图像的特征向量。
在本实施例的一些可选的实现方式中，执行单元503还用于：将特征向量与交通标志类别信息输入到车辆控制模型,得到车辆控制信息，其中，车辆控制模型用于表征特征向量、交通标志类别信息和车辆控制信息之间的对应关系；基于车辆控制信息，对无人驾驶车辆进行控制。
在本实施例的一些可选的实现方式中，上述车辆控制模型通过如下步骤训练得到：获取带有交通标志的历史环境图像集合；对于历史环境图像集合中的每张历史环境图像，分别获取该历史环境图像的交通标志类别信息以及与该历史环境图像相对应的车辆控制信息，对该历史环境图像进行特征提取，得到该历史环境图像的特征向量；基于各个特征向量、相应的交通标志类别信息和相应的车辆控制信息，训练得到车辆控制模型。
需要说明的是，本实施例提供的用于控制无人驾驶车辆的装置中 各单元的实现细节和技术效果可以参考本申请中其它实施例的说明，在此不再赘述。
下面参考图6，其示出了适于用来实现本申请实施例的电子设备的计算机系统600的结构示意图。图6示出的电子设备仅仅是一个示例，不应对本申请实施例的功能和使用范围带来任何限制。
如图6所示，计算机系统600包括中央处理单元(CPU，Central Processing Unit)601，其可以根据存储在只读存储器(ROM，Read Only Memory)602中的程序或者从存储部分606加载到随机访问存储器(RAM，Random Access Memory)603中的程序而执行各种适当的动作和处理。在RAM 603中，还存储有系统600操作所需的各种程序和数据。CPU 601、ROM 602以及RAM 603通过总线604彼此相连。输入/输出(I/O，Input/Output)接口605也连接至总线604。
以下部件连接至I/O接口605：包括硬盘等的存储部分606；以及包括诸如LAN(局域网，Local Area Network)卡、调制解调器等的网络接口卡的通信部分607。通信部分607经由诸如因特网的网络执行通信处理。驱动器608也根据需要连接至I/O接口605。可拆卸介质609，诸如磁盘、光盘、磁光盘、半导体存储器等等，根据需要安装在驱动器608上，以便于从其上读出的计算机程序根据需要被安装入存储部分606。
特别地，根据本公开的实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本公开的实施例包括一种计算机程序产品，其包括承载在计算机可读介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信部分607从网络上被下载和安装，和/或从可拆卸介质609被安装。在该计算机程序被中央处理单元(CPU)601执行时，执行本申请的方法中限定的上述功能。需要说明的是，本申请上述的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、 装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子可以包括但不限于：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本申请中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。而在本申请中，计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：无线、电线、光缆、RF等等，或者上述的任意合适的组合。
附图中的流程图和框图，图示了按照本申请各种实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，该模块、程序段、或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意，在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
描述于本申请实施例中所涉及到的单元可以通过软件的方式实现，也可以通过硬件的方式来实现。所描述的单元也可以设置在处理器中，例如，可以描述为：一种处理器包括获取单元、确定单元和执 行单元。其中，这些单元的名称在某种情况下并不构成对该单元本身的限定，例如，获取单元还可以被描述为“用于获取无人驾驶车辆的环境图像的单元”。
作为另一方面，本申请还提供了一种计算机可读介质，该计算机可读介质可以是上述实施例中描述的装置中所包含的；也可以是单独存在，而未装配入该装置中。上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被该装置执行时，使得该装置：获取无人驾驶车辆的环境图像；将环境图像输入到预先建立的交通标志识别模型，得到环境图像中交通标志图像的交通标志类别信息，其中，交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；基于预先建立的交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制无人驾驶车辆。
以上描述仅为本申请的较佳实施例以及对所运用技术原理的说明。本领域技术人员应当理解，本申请中所涉及的发明范围，并不限于上述技术特征的特定组合而成的技术方案，同时也应涵盖在不脱离上述发明构思的情况下，由上述技术特征或其等同特征进行任意组合而形成的其它技术方案。例如上述特征与本申请中公开的(但不限于)具有类似功能的技术特征进行互相替换而形成的技术方案。

Claims (10)

  1. 一种用于控制无人驾驶车辆的方法，其特征在于，所述方法包括：
    获取无人驾驶车辆的环境图像；
    将所述环境图像输入到预先建立的交通标志识别模型，得到所述环境图像中交通标志图像的交通标志类别信息，其中，所述交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；
    基于预先建立的所述交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制所述无人驾驶车辆。
  2. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    对所述环境图像进行特征提取，得到所述环境图像的特征向量。
  3. 根据权利要求2所述的方法，其特征在于，所述基于预先建立的所述交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制所述无人驾驶车辆，包括：
    将所述特征向量与所述交通标志类别信息输入到车辆控制模型,得到车辆控制信息，其中，所述车辆控制模型用于表征特征向量、交通标志类别信息和车辆控制信息之间的对应关系；
    基于所述车辆控制信息，对所述无人驾驶车辆进行控制。
  4. 根据权利要求1所述的方法，其特征在于，所述交通标志识别模型通过如下步骤训练得到：
    获取带有交通标志的历史环境图像集合；
    对于历史环境图像集合中的每个历史环境图像，对该历史环境图像进行图像识别，确定历史交通标志信息，其中，历史交通标志信息包括交通标志的类别信息和交通标志在历史环境图像中的位置信息；
    基于各个历史环境图像和相应的历史交通标志信息，训练得到交通标志识别模型。
  5. 根据权利要求3所述的方法，其特征在于，所述车辆控制模型通过如下步骤训练得到：
    获取带有交通标志的历史环境图像集合；
    对于历史环境图像集合中的每张历史环境图像，获取该历史环境图像的交通标志类别信息以及与该历史环境图像相对应的车辆控制信息，对该历史环境图像进行特征提取，得到该历史环境图像的特征向量；
    基于与各个历史环境图像对应的特征向量、交通标志类别信息和车辆控制信息，训练得到车辆控制模型。
  6. 一种用于控制无人驾驶车辆的装置，其特征在于，所述装置包括：
    获取单元，用于获取无人驾驶车辆的环境图像；
    确定单元，用于将所述环境图像输入到预先建立的交通标志识别模型，得到所述环境图像中交通标志图像的交通标志类别信息，其中，所述交通标志识别模型用于表征环境图像与交通标志类别信息的对应关系；
    执行单元，用于基于预先建立的所述交通标志类别信息与车辆控制信息的关联关系，选取并执行车辆控制信息，以控制所述无人驾驶车辆。
  7. 根据权利要求6所述的装置，其特征在于，所述装置还包括：
    特征提取单元，用于对所述环境图像进行特征提取，得到所述环境图像的特征向量。
  8. 根据权利要求7所述的装置，其特征在于，所述执行单元，还用于：
    将所述特征向量与所述交通标志类别信息输入到车辆控制模型,得到车辆控制信息，其中，所述车辆控制模型用于表征特征向量、交通标志类别信息和车辆控制信息之间的对应关系；
    基于所述车辆控制信息，对所述无人驾驶车辆进行控制。
  9. 一种无人驾驶车辆，其特征在于，所述设备包括：
    一个或多个处理器；
    存储装置，用于存储一个或多个程序；
    当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如权利要求1-5中任一所述的方法。
  10. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1-5中任一所述的方法。
PCT/CN2018/098632 2017-09-05 2018-08-03 用于控制无人驾驶车辆的方法和装置 WO2019047644A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710792748.3A CN107571867B (zh) 2017-09-05 2017-09-05 用于控制无人驾驶车辆的方法和装置
CN201710792748.3 2017-09-05

Publications (1)

Publication Number Publication Date
WO2019047644A1 true WO2019047644A1 (zh) 2019-03-14

Family

ID=61029828

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/098632 WO2019047644A1 (zh) 2017-09-05 2018-08-03 用于控制无人驾驶车辆的方法和装置

Country Status (2)

Country Link
CN (1) CN107571867B (zh)
WO (1) WO2019047644A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114971285A (zh) * 2022-05-25 2022-08-30 福州大学 一种驾驶环境信息负荷的评价方法及装置、设备、介质

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107571867B (zh) * 2017-09-05 2019-11-08 百度在线网络技术（北京）有限公司 用于控制无人驾驶车辆的方法和装置
US10699141B2 (en) * 2018-06-26 2020-06-30 Waymo Llc Phrase recognition model for autonomous vehicles
CN110969592B (zh) * 2018-09-29 2024-03-29 北京嘀嘀无限科技发展有限公司 图像融合方法、自动驾驶控制方法、装置和设备
CN109635639B (zh) * 2018-10-31 2020-06-09 百度在线网络技术（北京）有限公司 交通标识的位置检测方法、装置、设备和存储介质
CN109693672B (zh) * 2018-12-28 2020-11-06 百度在线网络技术（北京）有限公司 用于控制无人驾驶汽车的方法和装置
CN109886210B (zh) * 2019-02-25 2022-07-19 百度在线网络技术（北京）有限公司 一种交通图像识别方法、装置、计算机设备和介质
CN111753578A (zh) * 2019-03-27 2020-10-09 北京外号信息技术有限公司 光通信装置的识别方法和相应的电子设备
CN111775944B (zh) * 2019-04-04 2021-12-28 富泰华工业（深圳）有限公司 辅助驾驶装置、方法及计算机可读存储介质
CN110727269B (zh) * 2019-10-09 2023-06-23 陈浩能 车辆控制方法及相关产品
CN111223294A (zh) * 2019-11-12 2020-06-02 维特瑞交通科技有限公司 一种智能车辆引导控制方法及装置
CN112987707A (zh) * 2019-11-29 2021-06-18 北京京东乾石科技有限公司 一种车辆的自动驾驶控制方法及装置
CN111242046B (zh) * 2020-01-15 2023-08-29 江苏北斗星通汽车电子有限公司 一种基于图像检索的地面交通标志识别方法
CN113283269A (zh) * 2020-02-20 2021-08-20 上海博泰悦臻电子设备制造有限公司 用于标识地图的方法、电子设备和计算机存储介质
CN112699834B (zh) * 2021-01-12 2022-06-17 腾讯科技（深圳）有限公司 交通标识检测方法、装置、计算机设备和存储介质
CN113191255A (zh) * 2021-04-28 2021-07-30 浙江大学 一种基于移动机器人的交通标志识别方法
CN114244880B (zh) * 2021-12-16 2023-12-26 云控智行科技有限公司 智能网联驾驶云控功能的运行方法、装置、设备和介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1915725A (zh) * 2005-08-18 2007-02-21 本田汽车研究学院欧洲股份有限公司 驾驶员辅助系统
CN102745196A (zh) * 2012-07-18 2012-10-24 重庆邮电大学 基于粒计算的缩微智能车智能控制装置及方法
CN104850845A (zh) * 2015-05-30 2015-08-19 大连理工大学 一种基于非对称卷积神经网络的交通标志识别方法
CN106080590A (zh) * 2016-06-12 2016-11-09 百度在线网络技术（北京）有限公司 车辆控制方法和装置以及决策模型的获取方法和装置
WO2016178213A1 (en) * 2015-05-05 2016-11-10 B.G. Negev Technologies And Applications Ltd. Universal autonomous robotic driving system
US20170217435A1 (en) * 2016-01-30 2017-08-03 Bendix Commercial Vehicle Systems Llc System and Method for Providing a Speed Warning and Speed Control
CN107571867A (zh) * 2017-09-05 2018-01-12 百度在线网络技术（北京）有限公司 用于控制无人驾驶车辆的方法和装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102542260A (zh) * 2011-12-30 2012-07-04 中南大学 一种面向无人驾驶车的道路交通标志识别方法
CN103832434B (zh) * 2012-11-22 2016-06-29 中国移动通信集团公司 一种行车安全控制系统及方法
CN103971128B (zh) * 2014-05-23 2017-04-05 北京理工大学 一种面向无人驾驶车的交通标志识别方法
DE102015225900B3 (de) * 2015-12-18 2017-04-06 Continental Automotive Gmbh Verfahren und Vorrichtung zur kamerabasierten Verkehrszeichenerkennung in einem Kraftfahrzeug
CN105719499B (zh) * 2016-04-21 2018-06-01 百度在线网络技术（北京）有限公司 交通标志识别测试方法和装置
CN106218632B (zh) * 2016-07-19 2019-01-01 百度在线网络技术（北京）有限公司 用于控制无人驾驶车辆的方法和装置
CN107066965A (zh) * 2017-04-11 2017-08-18 北京汽车集团有限公司 检测交通标识的方法及装置

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1915725A (zh) * 2005-08-18 2007-02-21 本田汽车研究学院欧洲股份有限公司 驾驶员辅助系统
CN102745196A (zh) * 2012-07-18 2012-10-24 重庆邮电大学 基于粒计算的缩微智能车智能控制装置及方法
WO2016178213A1 (en) * 2015-05-05 2016-11-10 B.G. Negev Technologies And Applications Ltd. Universal autonomous robotic driving system
CN104850845A (zh) * 2015-05-30 2015-08-19 大连理工大学 一种基于非对称卷积神经网络的交通标志识别方法
US20170217435A1 (en) * 2016-01-30 2017-08-03 Bendix Commercial Vehicle Systems Llc System and Method for Providing a Speed Warning and Speed Control
CN106080590A (zh) * 2016-06-12 2016-11-09 百度在线网络技术（北京）有限公司 车辆控制方法和装置以及决策模型的获取方法和装置
CN107571867A (zh) * 2017-09-05 2018-01-12 百度在线网络技术（北京）有限公司 用于控制无人驾驶车辆的方法和装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114971285A (zh) * 2022-05-25 2022-08-30 福州大学 一种驾驶环境信息负荷的评价方法及装置、设备、介质

Also Published As

Publication number Publication date
CN107571867B (zh) 2019-11-08
CN107571867A (zh) 2018-01-12

Similar Documents

Publication Publication Date Title
WO2019047644A1 (zh) 用于控制无人驾驶车辆的方法和装置
WO2019047656A1 (zh) 用于控制无人驾驶车辆的方法和装置
CN108229489B (zh) 关键点预测、网络训练、图像处理方法、装置及电子设备
CN109492656B (zh) 用于输出信息的方法和装置
KR20200023708A (ko) 객체 검출 방법, 객체 검출을 위한 학습 방법 및 그 장치들
JP7284352B2 (ja) リアルタイムオブジェクト検出及び語意分割の同時行いシステム及び方法及び非一時的なコンピュータ可読媒体
JP2022521448A (ja) 交通画像認識方法、装置、コンピュータデバイスおよび媒体
CN113168510A (zh) 通过细化形状先验分割对象
CN112528878A (zh) 检测车道线的方法、装置、终端设备及可读存储介质
WO2020062433A1 (zh) 一种神经网络模型训练及通用接地线的检测方法
WO2015010451A1 (zh) 一种从单幅图像检测道路的方法
CN108765333B (zh) 一种基于深度卷积神经网络的深度图完善方法
CN112654998B (zh) 一种车道线检测方法和装置
CN110348463A (zh) 用于识别车辆的方法和装置
Jiang et al. Deep transfer learning enable end-to-end steering angles prediction for self-driving car
CN113205507A (zh) 一种视觉问答方法、系统及服务器
CN113326826A (zh) 网络模型的训练方法、装置、电子设备及存储介质
CN112009491B (zh) 一种基于交通元素视觉增强的深度学习的自动驾驶方法及系统
CN111382695A (zh) 用于检测目标的边界点的方法和装置
CN107527074A (zh) 用于车辆的图像处理方法和装置
CN109747655B (zh) 用于自动驾驶车辆的驾驶指令生成方法和装置
CN114596548A (zh) 目标检测方法、装置、计算机设备及计算机可读存储介质
CN111210411B (zh) 图像中灭点的检测方法、检测模型训练方法和电子设备
CN110633598B (zh) 用于确定环境图像中的行驶区域的方法和装置
CN114882372A (zh) 一种目标检测的方法及设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18853527

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 11/08/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18853527

Country of ref document: EP

Kind code of ref document: A1