WO2019095735A1 - 用于车辆的信息输出方法和装置 - Google Patents

用于车辆的信息输出方法和装置 Download PDF

Info

Publication number
WO2019095735A1
WO2019095735A1 PCT/CN2018/099164 CN2018099164W WO2019095735A1 WO 2019095735 A1 WO2019095735 A1 WO 2019095735A1 CN 2018099164 W CN2018099164 W CN 2018099164W WO 2019095735 A1 WO2019095735 A1 WO 2019095735A1
Authority
WO
WIPO (PCT)
Prior art keywords
historical
image information
lane line
dimensional
dimensional image
Prior art date
Application number
PCT/CN2018/099164
Other languages
English (en)
French (fr)
Inventor
饶先拓
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 WO2019095735A1 publication Critical patent/WO2019095735A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/34—Route searching; Route guidance
    • G01C21/36—Input/output arrangements for on-board computers
    • G01C21/3626—Details of the output of route guidance instructions
    • G01C21/3635—Guidance using 3D or perspective road maps
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/34—Route searching; Route guidance
    • G01C21/36—Input/output arrangements for on-board computers
    • G01C21/3626—Details of the output of route guidance instructions
    • G01C21/3658—Lane guidance

Definitions

  • the embodiments of the present invention relate to the field of computer technologies, and in particular, to the field of vehicle navigation technologies, and in particular, to an information output method and apparatus for a vehicle.
  • Augmented Reality is a technique for calculating the position and angle of camera images in real time and adding corresponding images, videos, and 3D models.
  • the embodiment of the present application proposes an information output method and apparatus for a vehicle.
  • an embodiment of the present application provides an information output method for a vehicle, where the vehicle includes a camera, the method includes: acquiring, by a camera, a current environment image including a current lane line; determining to identify a current from the current environment image. Whether the credibility of the lane line is greater than or equal to a preset credibility threshold, wherein the credibility is determined based on at least one of the following: the degree of obstruction of the obstacle to the lane line and the camera parameter of the camera; in response to the credibility
  • the two-dimensional image information of the current lane line in the current environment image is input into the pre-established space model to obtain three-dimensional image information of the current lane line, and the obtained three-dimensional image information is output.
  • the vehicle further includes a display screen, the method further comprising: inputting the two-dimensional image information of the current navigation guide line displayed on the display screen into the spatial model in response to the reliability being less than the confidence threshold The three-dimensional image information of the current navigation guide line; determining the three-dimensional image information of the current lane line based on the three-dimensional image information of the lane line in the first preset historical time period and the three-dimensional image information of the current navigation guide line, and outputting the determined three-dimensional image information .
  • the spatial model is obtained by acquiring two-dimensional image information of a historical navigation guide line displayed on the display screen during the second preset historical time period, and two-dimensional image information of the historical navigation guide line. Entering three-dimensional image information of the historical navigation guide line in the initial space model; in response to determining that the reliability of the lane line is greater than or equal to the credibility threshold from the historical environment image in the second preset historical time period, the history is The two-dimensional image information of the historical lane line in the environment image is input into the initial space model to obtain three-dimensional image information of the historical lane line; the two-dimensional image information based on the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image of the historical navigation guide line Information, determine the loss function of the initial space model; adjust the parameters of the initial space model based on the loss function, and use the parameter-adjusted initial space model as the space model.
  • determining a loss function of the initial spatial model based on the two-dimensional image information of the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image information of the historical navigation guide line includes: determining a two-dimensional history lane line Whether the two-dimensional historical lane line indicated by the image information is a straight line and whether the three-dimensional historical lane line indicated by the three-dimensional image information of the historical lane line is a straight line; in response to determining that the two-dimensional historical lane line is a straight line, and the three-dimensional historical lane line is not For a straight line, determine the first loss function of the initial space model.
  • determining a loss function of the initial spatial model based on the two-dimensional image information of the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image information of the historical navigation guide line includes: responsive to determining the three-dimensional historical lane The line is a straight line and the three-dimensional historical lane lines are at least two, further determining whether at least two three-dimensional historical lane lines are parallel; and determining a second loss function of the initial space model in response to determining that at least two three-dimensional historical lane lines are not parallel.
  • determining a loss function of the initial spatial model based on the two-dimensional image information of the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image information of the historical navigation guide line includes: responsive to determining the three-dimensional historical lane The line is a straight line, further determining whether the direction of travel of the vehicle guided by the three-dimensional historical lane line is consistent with the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line indicated by the three-dimensional image information of the historical navigation guide line; in response to determining the three-dimensional historical lane line The guided vehicle traveling direction is inconsistent with the vehicle traveling direction guided by the three-dimensional historical navigation guide line, and the third loss function of the initial space model is determined.
  • an embodiment of the present application provides an information output device for a vehicle, where the vehicle includes a camera, the device includes: an acquiring unit configured to acquire, by using a camera, a current environment image including a current lane line; and a determining unit, The configuration is configured to determine whether the credibility of the current lane line is recognized from the current environment image is greater than or equal to a preset credibility threshold, wherein the credibility is determined based on at least one of the following: an obstacle obstructing the lane line a degree and a camera parameter of the camera; the first output unit is configured to input the two-dimensional image information of the current lane line in the current environment image into the pre-established space model to obtain the current lane in response to the credibility being greater than or equal to the credibility threshold The three-dimensional image information of the line outputs the obtained three-dimensional image information.
  • the vehicle further includes a display screen
  • the apparatus further comprising: an input unit configured to respond to the confidence level to be less than a confidence threshold, to display a two-dimensional image of the current navigation guide line displayed on the display screen Obtaining three-dimensional image information of the current navigation guide line in the information input space model; the second output unit is configured to determine the current based on the three-dimensional image information of the lane line in the first preset historical time period and the three-dimensional image information of the current navigation guide line The three-dimensional image information of the lane line outputs the determined three-dimensional image information.
  • the apparatus further includes a spatial model establishing unit, the spatial model establishing unit includes: a first input module configured to acquire two historical navigation guide lines displayed on the display screen during the second preset historical time period Dimension image information, the two-dimensional image information of the historical navigation guide line is input into the initial space model to obtain three-dimensional image information of the historical navigation guide line; and the second input module is configured to be responsive to determining that the second preset historical time period is The historical environment image identifies that the credibility of the lane line is greater than or equal to the credibility threshold, and inputs the two-dimensional image information of the historical lane line in the historical environment image into the initial space model to obtain the three-dimensional image information of the historical lane line; the determining module, Configuring a two-dimensional image information based on a historical lane line, three-dimensional image information of a historical lane line, and three-dimensional image information of a historical navigation guide line to determine a loss function of the initial space model; an adjustment module configured to adjust an initial based on the loss function The parameters of the
  • the determining module includes: a first determining submodule configured to determine whether the two-dimensional historical lane line indicated by the two-dimensional image information of the historical lane line is a straight line and the three-dimensional image information of the historical lane line is indicated Whether the three-dimensional historical lane line is a straight line; the second determining sub-module is configured to determine a first loss function of the initial space model in response to determining that the two-dimensional historical lane line is a straight line and the three-dimensional historical lane line is not a straight line.
  • the determining module includes: a third determining submodule configured to further determine at least two three-dimensional historical lane lines in response to determining that the three-dimensional historical lane line is a straight line and the three-dimensional historical lane line is at least two Whether parallel; a fourth determining sub-module configured to determine a second loss function of the initial spatial model in response to determining that at least two three-dimensional historical lane lines are not parallel.
  • the determining module includes: a fifth determining submodule configured to further determine a three-dimensional historical lane line guided by the three-dimensional historical lane line and a three-dimensional history guide line in response to determining that the three-dimensional historical lane line is a straight line Whether the traveling direction of the vehicle guided by the three-dimensional historical navigation guide line indicated by the image information is consistent; the sixth determining sub-module configured to be guided in response to determining the traveling direction of the vehicle guided by the three-dimensional historical lane line and the three-dimensional historical navigation guide line The vehicle travel direction is inconsistent, and the third loss function of the initial space model is determined.
  • the information output method and apparatus for a vehicle provided by the present application, by using a camera of a vehicle to acquire a current environment image including a current lane line, determining whether the credibility of the current lane line is greater than or equal to a preset from the current environment image.
  • the credibility threshold if the credibility is greater than or equal to the credibility threshold, input the two-dimensional image information of the current lane line into a pre-established space model to obtain three-dimensional image information of the current lane line, and output the obtained
  • the three-dimensional image information effectively utilizes the current environment image including the current lane line, so that the accuracy of the three-dimensional image information of the output lane line is higher, thereby further improving the safety of driving the vehicle.
  • 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 an information output method for a vehicle according to the present application.
  • FIG. 3 is a schematic diagram of an application scenario of an information output method for a vehicle according to the present application.
  • FIG. 4 is a flow chart of still another embodiment of an information output method for a vehicle according to the present application.
  • FIG. 5 is a schematic structural view of an embodiment of an information output device for a vehicle according to the present application.
  • FIG. 6 is a schematic structural diagram of a computer system suitable for implementing the terminal device of the embodiment of the present application.
  • FIG. 1 illustrates an exemplary system architecture 100 of an embodiment of an information output method for a vehicle or an information output device for a vehicle to which the present application may be applied.
  • the system architecture 100 can include terminal devices 101, 102, 103, a network 104, a server 105, and an image information collection device 106.
  • the network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105.
  • Network 104 may include various types of connections, such as wired, wireless communication links, fiber optic cables, and the like.
  • the terminal devices 101, 102, 103 interact with the server 105 via the network 104 to receive or transmit messages and the like.
  • Various communication client applications such as navigation applications, map applications, music video applications, and the like, may be installed on the terminal devices 101, 102, and 103.
  • the terminal device 101, 102, 103 can acquire the current environment image including the current lane line by the image information collecting device 106; after that, it can be determined whether the credibility of the current lane line is greater than or equal to the preset value from the current environment image.
  • a reliability threshold in response to the credibility being greater than or equal to the credibility threshold, inputting two-dimensional image information of the current lane line in the current environment image into a pre-established space model to obtain three-dimensional image information of the current lane line, and outputting
  • the obtained three-dimensional image information may be presented on the display of the terminal devices 101, 102, 103 on the three-dimensional image determined by the obtained three-dimensional image information, or the obtained three-dimensional image information may be transmitted to the server 105.
  • the terminal devices 101, 102, 103 may be various electronic devices having a display screen and a camera and supporting information interaction, including but not limited to in-vehicle terminals, smart phones, tablets, e-book readers, laptop portable computers, and desktop computers. and many more.
  • Server 105 may be a server that provides various services, such as a backend server that provides support for spatial models in terminal devices 101, 102, 103.
  • the background server can also receive the three-dimensional image information output by the terminal devices 101, 102, and 103.
  • the information output method for the vehicle provided by the embodiment of the present application may be performed by the terminal devices 101, 102, and 103. Accordingly, the information output device for the vehicle may be disposed at the terminal device 101, 102, and 103. in.
  • FIG. 1 the number of terminal devices, networks, servers, and image information collection devices in FIG. 1 is merely illustrative. Depending on the implementation needs, there can be any number of terminal devices, networks, servers, and image information collection devices.
  • the information output method for a vehicle includes the following steps:
  • step 201 the current environment image including the current lane line is acquired by the camera.
  • an electronic device on which the information output method for the vehicle runs can acquire a current environment image including the current lane line through the camera mounted on the vehicle.
  • the lane line may also be referred to as a road lane line, and the lane line may include a guiding lane line and a variable guiding lane line, wherein the guiding lane line is a lane marking line guiding direction, which is used to indicate that the vehicle entering the section at the intersection should be according to the indicated Driving in the direction, there is a guide lane line at the traffic intersection with large traffic flow, which is used to clarify the direction of traffic and slow down the traffic pressure.
  • the intersection allows for right and straight (ie, right and straight merge into one lane), and some intersections allow U-turn and left turn (ie, U-turn and left-turn merge into one lane).
  • the current environment image may be an image of the road ahead environment of the vehicle at the current time acquired by the camera during the running of the vehicle.
  • Step 202 Determine whether the credibility of the current lane line is greater than or equal to a preset credibility threshold from the current environment image.
  • the electronic device may first determine the credibility of the current lane line from the current environment image.
  • the above electronic device can determine the above-mentioned reliability based on the degree of occlusion of the obstacle line on the lane line. Specifically, the highest value (for example, 0.9) and the lowest value (for example, 0.1) of the reliability may be first set.
  • the credibility may be the set minimum value; if the obstacle If half of the lane line is blocked, the confidence level may be the intermediate value between the set highest value and the lowest value (for example, the intermediate value between the highest value of 0.9 and the lowest value of 0.1 is 0.5); if the lane line is not obstructed If the object is occluded, the credibility can be the highest value set.
  • the electronic device may determine the credibility based on the camera parameters of the camera, and the camera parameter may include at least one of the following: resolution, texture reduction, dynamic speed, and auto focus speed.
  • the resolution generally refers to the ability of the camera to analyze the image, that is, the number of pixels of the image sensor of the camera;
  • the texture reduction degree refers to the degree to which the camera can truly restore the texture of the small object;
  • the dynamic speed can refer to the reaction speed of the camera, that is, the display Whether the image displayed inside is synchronized with the image captured by the camera without a certain lag time difference;
  • autofocus is a method of determining the focus adjustment according to the image sharpness evaluation algorithm (the focal length becomes longer or shorter), thereby making the image sharpest Excellent, and the auto focus speed refers to the degree to which the image sharpness is optimized by adjusting the focal length.
  • the electronic device may first search for a first score corresponding to the resolution of the camera in a correspondence table between the preset resolution and the first score, and may correspond to the preset texture reduction degree and the second score. Finding a second score corresponding to the texture reduction degree of the camera in the relationship table, and searching for a third score corresponding to the dynamic speed of the camera in the correspondence table between the preset dynamic speed and the third score, which may be preset Find a fourth score corresponding to the autofocus speed of the camera in the correspondence table between the autofocus speed and the fourth score; afterwards, the weights corresponding to the resolution, the texture reduction degree, the dynamic speed, and the autofocus speed may be separately obtained; Finally, the first product is obtained by multiplying the first score by the weight corresponding to the resolution, and the second product is obtained by multiplying the second score by the weight corresponding to the texture reduction degree, and the third score is multiplied by the weight corresponding to the dynamic speed. The third product is obtained by multiplying the fourth score by the weight corresponding to the autofocus speed to obtain the fourth
  • the electronic device may further determine the credibility by integrating the degree of obstruction of the obstacle to the lane line and the camera parameters of the camera. Specifically, the electronic device may determine the first credibility by using the foregoing method for determining the degree of occlusion based on the obstacle to the lane line, and determine the second method by using the camera-based camera parameter to determine the credibility. Afterwards, the weights corresponding to the first credibility and the second credibility are respectively obtained; and finally, the first credibility is multiplied by the weight corresponding to the first credibility, and the second The credibility is multiplied by the weight corresponding to the second credibility, and the weighted sum obtained by adding the products is determined as the credibility of identifying the current lane line from the current environment image.
  • the electronic device may determine whether the credibility is greater than or equal to a preset credibility threshold (for example, 0.7). If the credibility is greater than or equal to the credibility threshold, step 203 may be performed.
  • a preset credibility threshold for example, 0.7
  • Step 203 Input two-dimensional image information of the current lane line in the current environment image into the pre-established space model to obtain three-dimensional image information of the current lane line, and output the obtained three-dimensional image information.
  • the two-dimensional image of the current lane line in the current environment image may be The information is input into a pre-established spatial model to obtain three-dimensional image information of the current lane line, and the obtained three-dimensional image information is output.
  • a two-dimensional image generally refers to a planar image that includes height information and width information without including depth information, and therefore, the two-dimensional image information may include height information and width information, or two-dimensional point coordinates included in the image coordinate system.
  • the digital image captured by the camera can be stored as an array in the computer. The value of each element (pixel) in the array is the brightness (grayscale) of the image point.
  • a Cartesian coordinate system u-v is defined on the image, and the coordinates (u, v) of each pixel are the number of columns and the number of rows of the pixel in the array, respectively.
  • a three-dimensional image generally refers to a stereoscopic image including height information, width information, and depth information, and thus, the three-dimensional image information may include height information, width information, and depth information, or include three-dimensional point coordinates in a world coordinate system.
  • the spatial model may be a model that converts two-dimensional image information of an object into three-dimensional image information by using a three-dimensional reconstruction technique, and may also be used to convert two-dimensional point coordinates of the object in the image coordinate system to a world coordinate system.
  • the spatial model may be a conversion matrix of the two-dimensional image information determined by the technician based on the camera parameters of the camera that collects the image into three-dimensional image information; or the technician may be based on the shooting of objects at various angles and distances.
  • the three-dimensional reconstruction technique refers to acquiring a two-dimensional image of a scene object through a camera, analyzing and processing the information of the two-dimensional image, and combining computer vision knowledge to derive three-dimensional image information of the object in the real environment.
  • the electronic device may generate a three-dimensional image of the current lane line by using the obtained three-dimensional image information, and superimpose the generated three-dimensional image into the navigation map of the electronic device.
  • the vehicle on which the electronic device operates may further include a sensor.
  • the above spatial model can be obtained by the following steps:
  • the electronic device may acquire two-dimensional image information of a historical navigation guide line displayed on the display screen in a second preset historical time period (for example, in the past one day), and then the second navigation guide line may be
  • the dimensional image information is input into the initial spatial model to obtain three-dimensional image information of the historical navigation guide line
  • the initial spatial model may be a model that converts the two-dimensional image information of the object into three-dimensional image information by using a three-dimensional reconstruction technique, and may also be used for
  • the two-dimensional point coordinates of the object in the image coordinate system are converted to the three-dimensional point coordinates in the world coordinate system.
  • the electronic device may initialize parameters of the initial space model based on camera parameters of the camera.
  • a transformation matrix for converting a three-dimensional image into a two-dimensional image may be determined in the heart, the optical axis, etc., and the initial spatial model may be used to convert the two-dimensional image into a three-dimensional image, so that the initial spatial model may be determined based on the transformation matrix parameter.
  • the electronic device may determine that the credibility of the lane line is recognized from the historical environment image in the second preset historical time period, and the method for determining the credibility is identified from the current environment image as described above
  • the method for determining the credibility of the current lane line is basically the same, and will not be described again here.
  • the electronic device may input the two-dimensional image information of the historical lane line in the historical environment image into the initial space model in response to determining that the reliability of the lane line is determined from the historical environment image is greater than or equal to the reliability threshold. Three-dimensional image information of the historical lane line.
  • the electronic device may determine a loss function of the initial space model based on the two-dimensional image information of the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image information of the historical navigation guide line.
  • the function can be used to estimate the degree of inconsistency between the predicted value f(x) of the model and the true value Y. It is a non-negative real-valued function, usually expressed by L(Y, f(x)).
  • the loss function may include a logarithmic loss function (logical regression), a square loss function (Ordinary Least Squares), an exponential loss function (Adaboost), and the like. How to use the logarithmic loss function, the squared loss function and the exponential loss function to solve the function optimal solution is a common knowledge of the current research and application, and will not be repeated here.
  • the electronic device may first determine whether the two-dimensional historical lane line indicated by the two-dimensional image information of the historical lane line is a straight line, and may determine a three-dimensional image of the historical lane line. Whether the three-dimensional historical lane line indicated by the information is a straight line. In response to determining that the two-dimensional historical lane line is a straight line and the three-dimensional historical lane line is not a straight line, the electronic device may determine a first loss function of the initial spatial model.
  • the two-dimensional historical lane line is a straight line, it can be predicted that the three-dimensional historical lane line should also be a straight line, and the true three-dimensional historical lane line is not a straight line, then the degree of inconsistency between the true three-dimensional historical lane line and the straight line can be determined, thereby The first loss function of the initial space model described above is determined.
  • the electronic device in response to determining that the three-dimensional historical lane line is a straight line and the three-dimensional historical lane line is at least two, the electronic device may further determine the at least two three-dimensional historical lane lines. Are they parallel to each other? Specifically, the electronic device may determine whether the slopes of the at least two three-dimensional historical lane lines are the same. If the same, the at least two three-dimensional historical lane lines are parallel to each other. If not, the at least two three-dimensional historical lane lines are not parallel. In response to determining that the at least two three-dimensional historical lane lines are not parallel, the electronic device can determine a second loss function of the initial spatial model.
  • At least two three-dimensional historical lane lines it can be predicted that the at least two three-dimensional historical lane lines should be parallel to each other, and the true at least two three-dimensional historical lane lines are not parallel, so that at least two three-dimensional historical lanes can be determined.
  • the degree of non-parallel between the lines (the difference between the slopes of the historical lane lines), thereby determining the second loss function of the initial space model described above.
  • the electronic device in response to determining that the three-dimensional historical lane line is a straight line, may further determine a three-dimensional historical lane line and a three-dimensional direction of the historical navigation guide line. Whether the traveling direction of the vehicle guided by the three-dimensional historical navigation guide line indicated by the image information is consistent.
  • the electronic device may determine a third loss function of the initial space model in response to determining that the direction of travel of the vehicle guided by the three-dimensional historical lane line does not coincide with the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line.
  • the three-dimensional historical lane line is a straight line, it can be predicted that the direction of travel of the vehicle guided by the three-dimensional historical lane line should be consistent with the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, and the direction of travel of the vehicle guided by the real three-dimensional historical lane line
  • the inconsistency of the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line may determine the degree of inconsistency between the direction of travel of the vehicle guided by the three-dimensional historical lane line and the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, thereby determining the initial space.
  • the third loss function of the model may determine the degree of inconsistency between the direction of travel of the vehicle guided by the three-dimensional historical lane line and the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, thereby determining the initial space.
  • the electronic device may also combine the actual driving parameters (traveling distance, steering angle, etc.) of the vehicle in the second preset historical time period to determine the actual driving direction of the vehicle and the driving direction of the vehicle guided by the three-dimensional historical navigation guide line. Inconsistent extent.
  • the electronic device may adjust the parameters of the initial spatial model based on the loss function of the initial spatial model, and may use the parameter-adjusted initial spatial model as the spatial model.
  • the electronic device may adjust the parameters of the initial space model based on the degree of inconsistency between the three-dimensional historical lane line and the straight line to minimize the degree of inconsistency between the three-dimensional historical lane line and the straight line; the electronic device may also be based on The degree of non-parallel between the at least two three-dimensional historical lane lines is adjusted to adjust the parameters of the initial spatial model such that the degree of non-parallelism between the at least two three-dimensional historical lane lines is minimized; the electronic device may also be based on the three-dimensional history described above.
  • the degree of inconsistency between the direction of travel of the vehicle guided by the lane line and the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, and the parameters of the initial space model are adjusted such that the direction of travel of the vehicle guided by the three-dimensional historical lane line and the three-dimensional historical navigation guide line The degree of inconsistency in the direction of travel of the guided vehicle is minimized.
  • the electronic device may use the parameter-adjusted initial space model as the space model.
  • FIG. 3 is a schematic diagram of an application scenario of an information output method for a vehicle according to the present embodiment.
  • the in-vehicle terminal 301 first acquires the current environment image 303 including the current lane line through the camera 302; thereafter, since the lane line in front of the vehicle is not obstructed by the obstacle, the in-vehicle terminal 301 determines from the current environment image 303 The reliability of the current lane line is determined to be 0.9, and the reliability 0.9 is determined to be greater than the preset credibility threshold 0.7; then, the in-vehicle terminal 301 acquires the two-dimensional image information 304 of the current lane line to represent the lane line.
  • the coordinate values of the respective two-dimensional coordinate points, and the two-dimensional image information 304 is input into the pre-established spatial model 305 to obtain the three-dimensional image information 306 of the current lane line as the coordinate values of the three-dimensional coordinate points of the lane line;
  • the in-vehicle terminal 301 can output three-dimensional image information 306 of the current lane line.
  • the above-mentioned embodiment of the present application provides a method for determining three-dimensional image information of a current lane line by using a current environment image including a current lane line, so that three-dimensional image information of a lane line with higher accuracy is output, thereby further improving safety of driving of the vehicle. .
  • the flow 400 of the information output method for a vehicle includes the following steps:
  • step 401 the current environment image including the current lane line is acquired by the camera.
  • Step 402 Determine whether the credibility of the current lane line is greater than or equal to a preset credibility threshold from the current environment image.
  • Step 403 Input two-dimensional image information of the current lane line in the current environment image into the pre-established space model to obtain three-dimensional image information of the current lane line, and output the obtained three-dimensional image information.
  • steps 401-403 are substantially the same as the operations of steps 201-203, and details are not described herein again.
  • Step 404 Input two-dimensional image information of the current navigation guide line displayed on the display screen into the space model to obtain three-dimensional image information of the current navigation guide line.
  • the vehicle on which the electronic device runs may further include a display screen. If the electronic device determines that the credibility of the current lane line is less than the credibility threshold from the current environment image, the two-dimensional image information of the current navigation guide line displayed on the display screen may be input to the foregoing The three-dimensional image information of the current navigation guide line is obtained in the space model.
  • the navigation guide line may be a route guide indicating a destination direction, or may be a to-be-traveled trajectory of the vehicle within a preset time period (for example, 10 seconds).
  • Step 405 Determine three-dimensional image information of the current lane line based on the three-dimensional image information of the lane line in the first preset historical time period and the three-dimensional image information of the current navigation guide line, and output the determined three-dimensional image information.
  • the electronic device may be based on the three-dimensional image information of the lane line within the first preset historical time period (eg, 10 seconds before the current time to the current time) and obtained in step 404.
  • the three-dimensional image information of the current navigation guide line determines the three-dimensional image information of the current lane line, and can output the determined three-dimensional image information.
  • the electronic device may determine location information of the historical lane line in the three-dimensional image information of the lane line in the first preset historical time period, and if there are at least two lane lines, determine the distance between the historical lane lines.
  • the length of the current navigation guide line and the turning angle can be obtained from the three-dimensional image information of the current navigation guide line; finally, the electronic device can determine the position indicated by the position information of the historical lane line as the start of the current lane line.
  • the position, the length of the current navigation guide line and the turning angle can be determined as the length of the current lane line and the turning angle, respectively, and the distance between the historical lane lines can be determined as the distance between the current lane lines.
  • the electronic device may generate a three-dimensional image of the current lane line by using the determined three-dimensional image information, and superimpose the generated three-dimensional image into the navigation map of the electronic device.
  • the flow 400 of the information output method for a vehicle in the present embodiment highlights the credibility of identifying the current lane line from the current environment image as compared with the embodiment corresponding to FIG. 2.
  • the step of determining the three-dimensional image information of the current lane line when less than the confidence threshold.
  • the solution described in this embodiment can determine the three-dimensional image information of the current lane line based on the three-dimensional image information of the historical lane line and the three-dimensional image information of the current navigation guide line, thereby identifying the current lane line from the current environment image.
  • the reliability is low, the three-dimensional image information of the more accurate lane line is output.
  • the present application provides an embodiment of an information output device for a vehicle, the device embodiment corresponding to the method embodiment shown in FIG.
  • the device can be specifically applied to various electronic devices.
  • the information output apparatus 500 for a vehicle of the present embodiment includes an acquisition unit 501, a determination unit 502, and a first output unit 503.
  • the obtaining unit 501 is configured to acquire a current environment image including a current lane line by using a camera; and the determining unit 502 is configured to determine whether the credibility of the current lane line is greater than or equal to a preset credibility from the current environment image.
  • the reliability is determined based on at least one of: an occlusion degree of the obstacle to the lane line and a camera parameter of the camera; the first output unit 503 is configured to respond to the reliability being greater than or equal to the confidence threshold,
  • the two-dimensional image information of the current lane line in the current environment image is input into a pre-established space model to obtain three-dimensional image information of the current lane line, and the obtained three-dimensional image information is output.
  • the specific processing of the obtaining unit 501, the determining unit 502, and the first output unit 503 of the information output apparatus 500 for the vehicle may refer to step 201, step 202, and step 203 in the corresponding embodiment of FIG.
  • the vehicle may further include a display screen.
  • the above information output apparatus 500 for a vehicle may further include an input unit (not shown) and a second output unit (not shown). If the determining unit 502 determines that the credibility of the current lane line from the current environment image is less than the credibility threshold, the input unit may display a two-dimensional image of the current navigation guide line displayed on the display screen. The information is input into the above spatial model to obtain three-dimensional image information of the current navigation guide line.
  • the navigation guide line may be a route guide indicating a destination direction, or may refer to a to-be-traveled trajectory of the vehicle within a preset time period.
  • the second output unit may determine the three-dimensional image information of the current lane line based on the three-dimensional image information of the lane line in the first preset historical time period and the three-dimensional image information of the current navigation guide line obtained in step 404, and may output the determined 3D image information. Specifically, the second output unit may determine location information of the historical lane line in the three-dimensional image information of the lane line in the first preset historical time period, and if there are at least two lane lines, determine the history lane line The distance of the current navigation guide line and the turning angle can be obtained from the three-dimensional image information of the current navigation guide line; finally, the second output unit can determine the position indicated by the position information of the historical lane line as the current lane.
  • the starting position of the line can determine the length of the current navigation guide line and the turning angle as the length of the current lane line and the turning angle, respectively, and can determine the distance between the historical lane lines as the distance between the current lane lines.
  • the second output unit may generate a three-dimensional image of the current lane line using the determined three-dimensional image information, and superimpose the generated three-dimensional image into the navigation map.
  • the information output apparatus 500 for a vehicle described above may further include a space model establishing unit (not shown).
  • the space model establishing unit may include a first input module (not shown), a second input module (not shown), a determining module (not shown), and an adjusting module (not shown) .
  • the above spatial model building unit may be to establish a spatial model by the following steps:
  • the first input module may acquire two-dimensional image information of the historical navigation guide line displayed on the display screen in the second preset historical time period, and then input the two-dimensional image information of the historical navigation guide line into the initial
  • the three-dimensional image information of the historical navigation guide line is obtained in the spatial model, wherein the initial spatial model may be a model for converting two-dimensional image information of the object into three-dimensional image information by using a three-dimensional reconstruction technique, or may be used for the object in the image coordinate system.
  • the next two-dimensional point coordinates are converted to three-dimensional point coordinates in the world coordinate system.
  • the first input module may initialize parameters of the initial space model based on camera parameters of the camera.
  • the three-dimensional point coordinates of the object in the world coordinate system are converted into two-dimensional point coordinates in the image coordinate system, and it can be determined from the above camera parameters.
  • the three-dimensional image is converted into a transformation matrix of the two-dimensional image, and the initial spatial model described above can be used to convert the two-dimensional image into a three-dimensional image, so that the parameters of the initial spatial model can be determined based on the transformation matrix described above.
  • the second input module may determine that the credibility of the lane line is identified from the historical environment image in the second preset historical time period, and the method for determining the credibility is the same as the current environment image as explained above
  • the method for determining the credibility of the current lane line is basically the same, and will not be described herein.
  • the second input module may input the two-dimensional image information of the historical lane line in the historical environment image into the initial space model in response to determining that the reliability of the lane line is determined from the historical environment image is greater than or equal to the credibility threshold.
  • the three-dimensional image information of the historical lane line is obtained.
  • the determining module may determine the loss function of the initial space model based on the two-dimensional image information of the historical lane line, the three-dimensional image information of the historical lane line, and the three-dimensional image information of the historical navigation guide line, and the loss function may be used.
  • the loss function can include a log loss function, a square loss function, an exponential loss function, and the like. How to use the logarithmic loss function, the squared loss function and the exponential loss function to solve the function optimal solution is a common knowledge of the current research and application, and will not be repeated here.
  • the adjustment module may adjust the parameters of the initial spatial model based on the loss function of the initial spatial model, and may use the parameter-adjusted initial spatial model as the spatial model.
  • the adjusting module may adjust the parameters of the initial space model based on the degree of inconsistency between the three-dimensional historical lane line and the straight line to minimize the degree of inconsistency between the three-dimensional historical lane line and the straight line; the adjusting module may also be based on The degree of non-parallel between the at least two three-dimensional historical lane lines is adjusted to adjust the parameters of the initial spatial model to minimize the degree of non-parallel between the at least two three-dimensional historical lane lines; the adjustment module may also be based on the three-dimensional history described above.
  • the degree of inconsistency between the direction of travel of the vehicle guided by the lane line and the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, and the parameters of the initial space model are adjusted such that the direction of travel of the vehicle guided by the three-dimensional historical lane line and the three-dimensional historical navigation guide line The degree of inconsistency in the direction of travel of the guided vehicle is minimized.
  • the above adjustment module may use the parameter-adjusted initial space model as the space model.
  • the determining module may include a first determining submodule (not shown) and a second determining submodule (not shown).
  • the first determining sub-module may first determine whether the two-dimensional historical lane line indicated by the two-dimensional image information of the historical lane line is a straight line, and may determine whether the three-dimensional historical lane line indicated by the three-dimensional image information of the historical lane line is straight line.
  • the second determining sub-module may determine a first loss function of the initial spatial model.
  • the two-dimensional historical lane line is a straight line, it can be predicted that the three-dimensional historical lane line should also be a straight line, and the true three-dimensional historical lane line is not a straight line, then the degree of inconsistency between the true three-dimensional historical lane line and the straight line can be determined, thereby The first loss function of the initial space model described above is determined.
  • the determining module may include a third determining submodule (not shown) and a fourth determining submodule (not shown).
  • the third determining sub-module may further determine whether the at least two three-dimensional historical lane lines are parallel to each other. Specifically, the third determining sub-module may determine whether the slopes of the at least two three-dimensional historical lane lines are the same. If they are the same, the at least two three-dimensional historical lane lines are parallel to each other. If not, the at least two three-dimensional history are The lane lines are not parallel.
  • the fourth determining sub-module may determine a second loss function of the initial spatial model. If there are at least two three-dimensional historical lane lines, it can be predicted that the at least two three-dimensional historical lane lines should be parallel to each other, and the true at least two three-dimensional historical lane lines are not parallel, so that at least two three-dimensional historical lanes can be determined. The extent to which the lines are not parallel, thereby determining the second loss function of the initial space model described above.
  • the determining module may include a fifth determining submodule (not shown) and a sixth determining submodule (not shown).
  • the fifth determining sub-module may further determine a three-dimensional historical navigation guide line indicated by the three-dimensional historical lane line and the three-dimensional image information indicated by the three-dimensional image information of the historical navigation guide line. Whether the direction of travel of the guided vehicles is the same.
  • the sixth determining sub-module may determine a third loss function of the initial spatial model in response to determining that the direction of travel of the vehicle guided by the three-dimensional historical lane line does not coincide with the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line.
  • the three-dimensional historical lane line is a straight line, it can be predicted that the direction of travel of the vehicle guided by the three-dimensional historical lane line should be consistent with the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, and the direction of travel of the vehicle guided by the real three-dimensional historical lane line
  • the inconsistency of the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line may determine the degree of inconsistency between the direction of travel of the vehicle guided by the three-dimensional historical lane line and the direction of travel of the vehicle guided by the three-dimensional historical navigation guide line, thereby determining the initial space.
  • the sixth determining sub-module may also determine the degree of inconsistency between the actual driving direction of the vehicle and the traveling direction of the vehicle guided by the three-dimensional historical navigation guide line in combination with the actual driving parameter of the vehicle in the second preset historical time period.
  • FIG. 6 a block diagram of a computer system 600 suitable for use in implementing a terminal device in accordance with an embodiment of the present invention 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.
  • computer system 600 includes a central processing unit (CPU) 601 that can be loaded into a program in random access memory (RAM) 603 according to a program stored in read only memory (ROM) 602 or from storage portion 608. And perform various appropriate actions and processes.
  • RAM random access memory
  • 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) interface 605 is also coupled to bus 604.
  • the following components are connected to the I/O interface 605: an input portion 606 including a keyboard, a mouse, etc.; an output portion 607 including, for example, a liquid crystal display (LCD) and a speaker; a storage portion 608 including a hard disk or the like; and including, for example, a LAN card, a modem
  • the communication portion 609 of the network interface card performs communication processing via a network such as the Internet.
  • Driver 610 is also coupled to I/O interface 605 as needed.
  • a removable medium 611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory or the like, is mounted on the drive 610 as needed so that a computer program read therefrom is installed into the storage portion 608 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 609, and/or installed from removable media 611.
  • 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 invention 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 a first output unit.
  • the names of these units do not in any way constitute a limitation on the unit itself.
  • the acquisition unit may also be described as "a unit that acquires a current environmental image of the current lane line through the camera.”
  • 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, when the one or more programs are executed by the device, causing the device to: acquire a current environment image including a current lane line through a camera; and determine to identify from the current environment image Whether the credibility of the current lane line is greater than or equal to a preset credibility threshold, wherein the credibility is determined based on at least one of the following: the degree of obstruction of the obstacle to the lane line and the camera parameter of the camera; The degree is greater than or equal to the credibility threshold, and the two-dimensional image information of the current lane line in the current environment image is input into the pre-established space model to obtain three-dimensional image information of the current lane line, and the obtained three-dimensional image information is output.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Traffic Control Systems (AREA)
  • Navigation (AREA)

Abstract

本申请实施例公开了用于车辆的信息输出方法和装置。该车辆包括摄像头，该方法的一具体实施方式包括：通过摄像头获取包含当前车道线的当前环境图像；确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和摄像头的摄像头参数；响应于可信度大于等于可信度阈值，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的三维图像信息。该实施方式可以输出准确度更高的车道线的三维图像信息，进一步提高了车辆驾驶的安全性。

Description

用于车辆的信息输出方法和装置
相关申请的交叉引用
本专利申请要求于2017年11月16日提交的、申请号为201711139956.X、申请人为百度在线网络技术(北京)有限公司、发明名称为“用于车辆的信息输出方法和装置”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
本申请实施例涉及计算机技术领域，具体涉及车辆导航技术领域，尤其涉及用于车辆的信息输出方法和装置。
背景技术
目前，汽车已成为人们生活中越来越重要的交通工具，为了更好提高汽车行驶过程中的安全性和驾驶员的舒适性，汽车智能化已经成为目前汽车发展的主流。例如，可以将增强现实(Augmented Reality，AR)技术应用于车辆导航邻域，增强现实技术是一种实时地计算摄影机影像的位置及角度并加上相应图像、视频、3D模型的技术。
发明内容
本申请实施例提出了用于车辆的信息输出方法和装置。
第一方面，本申请实施例提供了一种用于车辆的信息输出方法，该车辆包括摄像头，该方法包括：通过摄像头获取包含当前车道线的当前环境图像；确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和摄像头的摄像头参数；响应于可信度大于等于可信度阈值，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输 出所得到的三维图像信息。
在一些实施例中，该车辆还包括显示屏，该方法还包括：响应于可信度小于可信度阈值，将显示屏上所显示的当前导航引导线的二维图像信息输入空间模型中得到当前导航引导线的三维图像信息；基于第一预设历史时间段内车道线的三维图像信息和当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，输出所确定的三维图像信息。
在一些实施例中，空间模型是通过如下步骤得到的：获取第二预设历史时间段内在显示屏上显示过的历史导航引导线的二维图像信息，将历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图像信息；响应于确定出从第二预设历史时间段内的历史环境图像中识别出车道线的可信度大于等于可信度阈值，将历史环境图像中历史车道线的二维图像信息输入初始空间模型中得到历史车道线的三维图像信息；基于历史车道线的二维图像信息、历史车道线的三维图像信息和历史导航引导线的三维图像信息，确定初始空间模型的损失函数；基于损失函数，调整初始空间模型的参数，并将参数调整后的初始空间模型作为空间模型。
在一些实施例中，基于历史车道线的二维图像信息、历史车道线的三维图像信息和历史导航引导线的三维图像信息，确定初始空间模型的损失函数，包括：确定历史车道线的二维图像信息所指示的二维历史车道线是否为直线以及历史车道线的三维图像信息所指示的三维历史车道线是否为直线；响应于确定出二维历史车道线为直线、而三维历史车道线不为直线，确定初始空间模型的第一损失函数。
在一些实施例中，基于历史车道线的二维图像信息、历史车道线的三维图像信息和历史导航引导线的三维图像信息，确定初始空间模型的损失函数，包括：响应于确定出三维历史车道线为直线、且三维历史车道线为至少两条，进一步确定至少两条三维历史车道线是否平行；响应于确定出至少两条三维历史车道线不平行，确定初始空间模型的第二损失函数。
在一些实施例中，基于历史车道线的二维图像信息、历史车道线 的三维图像信息和历史导航引导线的三维图像信息，确定初始空间模型的损失函数，包括：响应于确定出三维历史车道线为直线，进一步确定三维历史车道线所引导的车辆行驶方向与历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导的车辆行驶方向是否一致；响应于确定出三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向不一致，确定初始空间模型的第三损失函数。
第二方面，本申请实施例提供了一种用于车辆的信息输出装置，该车辆包括摄像头，该装置包括：获取单元，配置用于通过摄像头获取包含当前车道线的当前环境图像；确定单元，配置用于确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和摄像头的摄像头参数；第一输出单元，配置用于响应于可信度大于等于可信度阈值，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的三维图像信息。
在一些实施例中，该车辆还包括显示屏，该装置还包括：输入单元，配置用于响应于可信度小于可信度阈值，将显示屏上所显示的当前导航引导线的二维图像信息输入空间模型中得到当前导航引导线的三维图像信息；第二输出单元，配置用于基于第一预设历史时间段内车道线的三维图像信息和当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，输出所确定的三维图像信息。
在一些实施例中，该装置还包括空间模型建立单元，空间模型建立单元包括：第一输入模块，配置用于获取第二预设历史时间段内在显示屏上显示过的历史导航引导线的二维图像信息，将历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图像信息；第二输入模块，配置用于响应于确定出从第二预设历史时间段内的历史环境图像中识别出车道线的可信度大于等于可信度阈值，将历史环境图像中历史车道线的二维图像信息输入初始空间模型中得到历史车道线的三维图像信息；确定模块，配置用于基于历史车道线 的二维图像信息、历史车道线的三维图像信息和历史导航引导线的三维图像信息，确定初始空间模型的损失函数；调整模块，配置用于基于损失函数，调整初始空间模型的参数，并将参数调整后的初始空间模型作为空间模型。
在一些实施例中，确定模块，包括：第一确定子模块，配置用于确定历史车道线的二维图像信息所指示的二维历史车道线是否为直线以及历史车道线的三维图像信息所指示的三维历史车道线是否为直线；第二确定子模块，配置用于响应于确定出二维历史车道线为直线、而三维历史车道线不为直线，确定初始空间模型的第一损失函数。
在一些实施例中，确定模块，包括：第三确定子模块，配置用于响应于确定出三维历史车道线为直线、且三维历史车道线为至少两条，进一步确定至少两条三维历史车道线是否平行；第四确定子模块，配置用于响应于确定出至少两条三维历史车道线不平行，确定初始空间模型的第二损失函数。
在一些实施例中，确定模块，包括：第五确定子模块，配置用于响应于确定出三维历史车道线为直线，进一步确定三维历史车道线所引导的车辆行驶方向与历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导的车辆行驶方向是否一致；第六确定子模块，配置用于响应于确定出三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向不一致，确定初始空间模型的第三损失函数。
本申请提供的用于车辆的信息输出方法和装置，通过利用车辆的摄像头获取包含当前车道线的当前环境图像，确定从上述当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，若上述可信度大于等于上述可信度阈值，将上述当前车道线的二维图像信息输入预先建立的空间模型中得到上述当前车道线的三维图像信息，并输出所得到的三维图像信息，从而有效利用了包含当前车道线的当前环境图像，使得输出的车道线的三维图像信息准确度更高，进一步提高了车辆驾驶的安全性。
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车辆的信息输出方法的一个实施例的流程图；
图3是根据本申请的用于车辆的信息输出方法的一个应用场景的示意图；
图4是根据本申请的用于车辆的信息输出方法的又一个实施例的流程图；
图5是根据本申请的用于车辆的信息输出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终端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车辆的信息输出方法或用于车辆的信息输出方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可以包括终端设备101、102、103，网络104、服务器105和图像信息采集装置106。网络104用以在终端设备101、102、103和服务器105之间提供通信链路的介质。网络104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。
终端设备101、102、103通过网络104与服务器105交互，以接 收或发送消息等。终端设备101、102、103上可以安装有各种通讯客户端应用，例如导航类应用、地图类应用、音乐视频类应用等。终端设备101、102、103可以通过图像信息采集装置106获取包含当前车道线的当前环境图像；之后，可以确定从上述当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值；响应于上述可信度大于等于上述可信度阈值，将上述当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到上述当前车道线的三维图像信息，输出所得到的三维图像信息，并可以在终端设备101、102、103的显示器上呈现由所得到的三维图像信息所确定的三维图像，也可以向服务器105发送所得到的三维图像信息。
终端设备101、102、103可以是具有显示屏与摄像头并且支持信息交互的各种电子设备，包括但不限于车载终端、智能手机、平板电脑、电子书阅读器、膝上型便携计算机和台式计算机等等。
服务器105可以是提供各种服务的服务器，例如对终端设备101、102、103中的空间模型提供支持的后台服务器。后台服务器也可以接收终端设备101、102、103输出的三维图像信息。
需要说明的是，本申请实施例所提供的用于车辆的信息输出方法可以由终端设备101、102、103执行，相应地，用于车辆的信息输出装置可以设置于终端设备101、102、103中。
应该理解，图1中的终端设备、网络、服务器和图像信息采集装置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终端设备、网络、服务器和图像信息采集装置。
继续参考图2，示出了根据本申请的用于车辆的信息输出方法的一个实施例的流程200。该用于车辆的信息输出方法，包括以下步骤：
步骤201，通过摄像头获取包含当前车道线的当前环境图像。
在本实施例中，用于车辆的信息输出方法运行于其上的电子设备(例如图1所示的车载终端)可以通过上述车辆上所安装的摄像头获取包含当前车道线的当前环境图像。车道线也可以称为道路车道线，车道线可以包括导向车道线和可变导向车道线，其中，导向车道线是引导方向的车道标线，用来指示车辆在路口驶入段应按所指方向行驶， 在车流量大的交通路口一般画有导向车道线，用于明确行车方向，减缓交通压力；若车辆进入画有可变导向车道线的车道后可以有不止一个车道走向，例如，有的路口允许右转和直行(即右转和直行合并为一个车道)，有的路口允许掉头和左转(即掉头和左转合并为一个车道)。上述当前环境图像可以是车辆在行驶过程中通过上述摄像头所获取的当前时刻的车辆前方道路环境图像。
步骤202，确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值。
在本实施例中，上述电子设备可以首先确定从上述当前环境图像中识别出当前车道线的可信度。上述电子设备可以基于障碍物对车道线的遮挡程度确定上述可信度。具体地，可以首先设置可信度的最高值(例如，0.9)与最低值(例如，0.1)，若障碍物将车道线完全遮挡，则可信度可以为所设置的最低值；若障碍物遮挡住车道线的一半，则可信度可以为所设置的最高值与最低值之间的中间值(例如，最高值0.9与最低值0.1之间的中间值0.5)；若车道线未被障碍物所遮挡，则可信度可以为所设置的最高值。
在本实施例中，上述电子设备也可以基于上述摄像头的摄像头参数确定上述可信度，上述摄像头参数可以包括以下至少一项：分辨率、纹理还原度、动态速度、自动对焦速度。分辨率一般是指摄像头解析图像的能力，也即摄像头的影像传感器的像素数；纹理还原度是指摄像头对细小物体的纹理能够真实还原的程度；动态速度可以是指摄像头的反应速度，即显示器内显示的图像是否与摄像头拍摄的图像同步，而不至于存在一定滞后时间差；自动对焦是根据图像清晰度评价算法来确定焦距调节的方式(焦距变长还是变短)，从而使得图像清晰度最优，而自动对焦速度是指通过调节焦距使得图像清晰度达到最优的快慢程度。具体地，上述电子设备可以首先在预设的分辨率与第一分数的对应关系表中查找上述摄像头的分辨率所对应的第一分数，可以在预设的纹理还原度与第二分数的对应关系表中查找上述摄像头的纹理还原度所对应的第二分数，可以在预设的动态速度与第三分数的对应关系表中查找上述摄像头的动态速度所对应的第三分数，可以在预设 的自动对焦速度与第四分数的对应关系表中查找上述摄像头的自动对焦速度所对应的第四分数；之后，可以分别获取分辨率、纹理还原度、动态速度和自动对焦速度所对应的权重；最后，利用上述第一分数乘以分辨率对应的权重得到第一乘积，利用上述第二分数乘以纹理还原度对应的权重得到第二乘积，利用上述第三分数乘以动态速度对应的权重得到第三乘积，利用上述第四分数乘以自动对焦速度对应的权重得到第四乘积，并将上述第一乘积、上述第二乘积、上述第三乘积和上述第四乘积之和确定为上述可信度。
在本实施例中，上述电子设备还可以综合障碍物对车道线的遮挡程度以及上述摄像头的摄像头参数确定上述可信度。具体地，上述电子设备可以利用上述基于障碍物对车道线的遮挡程度确定可信度方法确定出第一可信度，并可以利用上述基于摄像头的摄像头参数确定可信度的方法确定出第二可信度；之后，分别获取上述第一可信度和上述第二可信度所对应的权重；最后，利用上述第一可信度乘以第一可信度对应的权重，利用上述第二可信度乘以第二可信度对应的权重，并将乘积相加所得的加权和确定为从上述当前环境图像中识别出当前车道线的可信度。
之后，上述电子设备可以确定上述可信度是否大于等于预设的可信度阈值(例如，0.7)，若上述可信度大于等于上述可信度阈值，则可以执行步骤203。
步骤203，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的三维图像信息。
在本实施例中，若上述电子设备确定出从上述当前环境图像中识别出当前车道线的可信度大于等于上述可信度阈值，则可以将上述当前环境图像中当前车道线的二维图像信息输入到预先建立的空间模型中得到当前车道线的三维图像信息，并输出所得到的三维图像信息。二维图像通常是指包含高度信息和宽度信息而不包含深度信息的平面图像，因此，二维图像信息可以包括高度信息和宽度信息，或者包括在图像坐标系下的二维点坐标。摄像机采集的数字图像在计算机内可 以存储为数组，数组中的每一个元素(像素)的值即是图像点的亮度(灰度)。在图像上定义直角坐标系u-v，每一像素的坐标(u，v)分别是该像素在数组中的列数和行数。三维图像通常是指包含高度信息、宽度信息和深度信息的立体图像，因此，三维图像信息可以包括高度信息、宽度信息和深度信息，或者包括在世界坐标系下的三维点坐标。
在本实施例中，空间模型可以是利用三维重建技术将物体的二维图像信息转换成三维图像信息的模型，也可以用于将物体在图像坐标系下的二维点坐标转换到世界坐标系下的三维点坐标。作为示例，空间模型可以是技术人员基于采集图像的摄像头的摄像头参数而确定出的二维图像信息转换成三维图像信息的转换矩阵；也可以是技术人员基于对各个角度、各个距离的物体的拍摄结果而预先制定的、存储有物体的二维图像信息与三维图像信息的对应关系表。三维重建技术指的是通过摄像机获取场景物体的二维图像，并对此二维图像的信息进行分析处理，再结合计算机视觉知识推导出现实环境中物体的三维图像信息。
在本实施例中，上述电子设备可以利用所得到的三维图像信息生成当前车道线的三维图像，并将生成的三维图像叠加到上述电子设备的导航地图中。
在本实施例的一些可选的实现方式中，上述电子设备运行于其上的车辆还可以包括传感器。上述空间模型可以是通过如下步骤得到的：
首先，上述电子设备可以获取第二预设历史时间段(例如，过去1天)内在上述显示屏上显示过的历史导航引导线的二维图像信息，之后，可以将上述历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图像信息，其中，上述初始空间模型可以是利用三维重建技术将物体的二维图像信息转换成三维图像信息的模型，也可以用于将物体在图像坐标系下的二维点坐标转换到世界坐标系下的三维点坐标。上述电子设备可以基于上述摄像头的摄像头参数，对初始空间模型的参数进行初始化。由于摄像机在对三维物体进行拍摄时得到的是该物体的二维图像，即将物体在世界坐标系下的三维点坐标转换成图像坐标系下的二维点坐标，从上述摄像头参数(焦 距、光心、光轴等)中可以确定将三维图像转换成二维图像的转换矩阵，而上述初始空间模型可以用于将二维图像转换成三维图像，从而可以基于上述转换矩阵确定上述初始空间模型的参数。
之后，上述电子设备可以确定从上述第二预设历史时间段内的历史环境图像中识别出车道线的可信度，上述可信度的确定方法与上文所阐述的从当前环境图像中识别当前车道线的可信度的确定方法基本相同，在此不再赘述。响应于确定出从历史环境图像中识别出车道线的可信度大于等于上述可信度阈值，上述电子设备可以将上述历史环境图像中历史车道线的二维图像信息输入上述初始空间模型中得到历史车道线的三维图像信息。
然后，上述电子设备可以基于上述历史车道线的二维图像信息、上述历史车道线的三维图像信息和上述历史导航引导线的三维图像信息，确定上述初始空间模型的损失函数(loss function)，损失函数可以是用来估量模型的预测值f(x)与真实值Y的不一致程度，它是一个非负实值函数，通常使用L(Y，f(x))表示。损失函数可以包括对数损失函数(逻辑回归)、平方损失函数(最小二乘法，Ordinary Least Squares)、指数损失函数(Adaboost)等等。如何利用对数损失函数、平方损失函数和指数损失函数求解函数最优解是目前广泛研究和应用的公知常识，在此不再赘述。
在本实施例的一些可选的实现方式中，上述电子设备可以首先确定上述历史车道线的二维图像信息所指示的二维历史车道线是否为直线，并可以确定上述历史车道线的三维图像信息所指示的三维历史车道线是否为直线。响应于确定出上述二维历史车道线为直线、而上述三维历史车道线不为直线，上述电子设备可以确定上述初始空间模型的第一损失函数。若二维历史车道线为直线，则可以预测三维历史车道线也应为直线，而真实的三维历史车道线不为直线，则可以确定真实的三维历史车道线与直线之间的不一致程度，从而确定上述初始空间模型的第一损失函数。
在本实施例的一些可选的实现方式中，响应于确定出上述三维历史车道线为直线、且上述三维历史车道线为至少两条，上述电子设备 可以进一步确定上述至少两条三维历史车道线是否相互平行。具体地，上述电子设备可以确定上述至少两条三维历史车道线的斜率是否相同，若相同，则上述至少两条三维历史车道线相互平行，若不相同，则上述至少两条三维历史车道线不平行。响应于确定出上述至少两条三维历史车道线不平行，上述电子设备可以确定上述初始空间模型的第二损失函数。若存在至少两条三维历史车道线，则可以预测上述至少两条三维历史车道线应为互相平行，而真实的至少两条三维历史车道线不平行，则可以确定真实的至少两条三维历史车道线之间不平行的程度(历史车道线的斜率之间的差值)，从而确定上述初始空间模型的第二损失函数。
在本实施例的一些可选的实现方式中，响应于确定出上述三维历史车道线为直线，上述电子设备可以进一步确定上述三维历史车道线所引导的车辆行驶方向与上述历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导的车辆行驶方向是否一致。响应于确定出上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向不一致，上述电子设备可以确定上述初始空间模型的第三损失函数。若上述三维历史车道线为直线，则可以预测三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向应该一致，而真实的三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向不一致，则可以确定上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度，从而确定上述初始空间模型的第三损失函数。上述电子设备也可以结合在上述第二预设历史时间段内车辆的实际行驶参数(行驶距离、转向角度等)，确定车辆的实际行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度。
最后，上述电子设备可以基于上述初始空间模型的损失函数，调整上述初始空间模型的参数，并可以将参数调整后的初始空间模型作为上述空间模型。具体地，上述电子设备可以基于上述三维历史车道线与直线之间的不一致程度，调整上述初始空间模型的参数以使得三 维历史车道线与直线之间的不一致程度达到最小；上述电子设备也可以基于上述至少两条三维历史车道线之间不平行的程度，调整上述初始空间模型的参数以使得上述至少两条三维历史车道线之间不平行的程度达到最小；上述电子设备还可以基于上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度，调整上述初始空间模型的参数以使得三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向的不一致程度达到最小。上述电子设备可以将参数调整后的初始空间模型作为上述空间模型。
继续参见图3，图3是根据本实施例的用于车辆的信息输出方法的应用场景的一个示意图。在图3的应用场景中，车载终端301首先通过摄像头302获取包含当前车道线的当前环境图像303；之后，由于车辆前方的车道线未被障碍物所遮挡，车载终端301确定从当前环境图像303中识别出当前车道线的可信度为0.9，并确定出可信度0.9大于预设的可信度阈值0.7；而后，车载终端301获取到当前车道线的二维图像信息304为表征车道线的各个二维坐标点的坐标值，并将二维图像信息304输入到预先建立的空间模型305中得到当前车道线的三维图像信息306为表征车道线的各个三维坐标点的坐标值；最后，车载终端301可以输出当前车道线的三维图像信息306。
本申请的上述实施例提供的方法通过利用包含当前车道线的当前环境图像确定当前车道线的三维图像信息，使得输出准确度更高的车道线的三维图像信息，进一步提高了车辆驾驶的安全性。
进一步参考图4，其示出了用于车辆的信息输出方法的又一个实施例的流程400。该用于车辆的信息输出方法的流程400，包括以下步骤：
步骤401，通过摄像头获取包含当前车道线的当前环境图像。
步骤402，确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值。
步骤403，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的 三维图像信息。
在本实施例中，步骤401-403的操作与步骤201-203的操作基本相同，在此不再赘述。
步骤404，将显示屏上所显示的当前导航引导线的二维图像信息输入空间模型中得到当前导航引导线的三维图像信息。
在本实施例中，上述电子设备运行于其上的车辆还可以包括显示屏。若上述电子设备确定出从上述当前环境图像中识别出当前车道线的可信度小于上述可信度阈值，则可以将上述显示屏上所显示的当前导航引导线的二维图像信息输入到上述空间模型中得到当前导航引导线的三维图像信息。导航引导线可以是标示出目的地方向的路线引导，也可以是指车辆在预设时间段(例如，10秒)内的待行驶轨迹。
步骤405，基于第一预设历史时间段内车道线的三维图像信息和当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，输出所确定的三维图像信息。
在本实施例中，上述电子设备可以基于第一预设历史时间段(例如，当前时刻之前的10秒起到当前时刻所形成的时间段)内车道线的三维图像信息和在步骤404中得到的当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，并可以输出所确定的三维图像信息。具体地，上述电子设备可以在上述第一预设历史时间段内车道线的三维图像信息中确定历史车道线的位置信息，若存在至少两条车道线，则可以确定历史车道线之间的距离；之后，可以从当前导航引导线的三维图像信息中获取当前导航引导线的长度和转弯角度；最后，上述电子设备可以将历史车道线的位置信息所指示的位置确定为当前车道线的起始位置，可以将当前导航引导线的长度和转弯角度分别确定为当前车道线的长度和转弯角度，并可以将历史车道线之间的距离确定为当前车道线之间的距离。
在本实施例中，上述电子设备可以利用所确定出的三维图像信息生成当前车道线的三维图像，并将生成的三维图像叠加到上述电子设备的导航地图中。
从图4中可以看出，与图2对应的实施例相比，本实施例中的用 于车辆的信息输出方法的流程400突出了若从当前环境图像中识别出当前车道线的可信度小于可信度阈值时确定当前车道线的三维图像信息的步骤。由此，本实施例描述的方案可以基于历史车道线的三维图像信息和当前导航引导线的三维图像信息确定当前车道线的三维图像信息，从而在从当前环境图像中识别出当前车道线的可信度较低时，输出较为准确的车道线的三维图像信息。
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车辆的信息输出装置的一个实施例，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，本实施例的用于车辆的信息输出装置500包括：获取单元501、确定单元502和第一输出单元503。其中，获取单元501配置用于通过摄像头获取包含当前车道线的当前环境图像；确定单元502配置用于确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和摄像头的摄像头参数；第一输出单元503配置用于响应于可信度大于等于可信度阈值，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的三维图像信息。
在本实施例中，用于车辆的信息输出装置500的获取单元501、确定单元502和第一输出单元503的具体处理可以参考图2对应实施例中的步骤201、步骤202和步骤203。
在本实施例的一些可选的实现方式中，上述车辆还可以包括显示屏。上述用于车辆的信息输出装置500还可以包括输入单元(图中未示出)和第二输出单元(图中未示出)。若上述确定单元502确定出从上述当前环境图像中识别出当前车道线的可信度小于上述可信度阈值，则上述输入单元可以将上述显示屏上所显示的当前导航引导线的二维图像信息输入到上述空间模型中得到当前导航引导线的三维图像信息。导航引导线可以是标示出目的地方向的路线引导，也可以是指车辆在预设时间段内的待行驶轨迹。上述第二输出单元可以基于第一预设历史时间段内车道线的三维图像信息和在步骤404中得到当前导航引导 线的三维图像信息，确定当前车道线的三维图像信息，并可以输出所确定的三维图像信息。具体地，上述第二输出单元可以在上述第一预设历史时间段内车道线的三维图像信息中确定历史车道线的位置信息，若存在至少两条车道线，则可以确定历史车道线之间的距离；之后，可以从当前导航引导线的三维图像信息中获取当前导航引导线的长度和转弯角度；最后，上述第二输出单元可以将历史车道线的位置信息所指示的位置确定为当前车道线的起始位置，可以将当前导航引导线的长度和转弯角度分别确定为当前车道线的长度和转弯角度，并可以将历史车道线之间的距离确定为当前车道线之间的距离。上述第二输出单元可以利用所确定出的三维图像信息生成当前车道线的三维图像，并将生成的三维图像叠加到导航地图中。
在本实施例的一些可选的实现方式中，上述用于车辆的信息输出装置500还可以包括空间模型建立单元(图中未示出)。上述空间模型建立单元可以包括第一输入模块(图中未示出)、第二输入模块(图中未示出)、确定模块(图中未示出)和调整模块(图中未示出)。
上述空间模型建立单元可以是通过如下步骤建立空间模型：
首先，上述第一输入模块可以获取第二预设历史时间段内在上述显示屏上显示过的历史导航引导线的二维图像信息，之后，可以将上述历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图像信息，其中，上述初始空间模型可以是利用三维重建技术将物体的二维图像信息转换成三维图像信息的模型，也可以用于将物体在图像坐标系下的二维点坐标转换到世界坐标系下的三维点坐标。上述第一输入模块可以基于上述摄像头的摄像头参数，对初始空间模型的参数进行初始化。由于摄像机在对三维物体进行拍摄时得到的是该物体的二维图像，即将物体在世界坐标系下的三维点坐标转换成图像坐标系下的二维点坐标，从上述摄像头参数中可以确定将三维图像转换成二维图像的转换矩阵，而上述初始空间模型可以用于将二维图像转换成三维图像，从而可以基于上述转换矩阵确定上述初始空间模型的参数。
之后，上述第二输入模块可以确定从上述第二预设历史时间段内 的历史环境图像中识别出车道线的可信度，上述可信度的确定方法与上文所阐述的从当前环境图像中识别当前车道线的可信度的确定方法基本相同，在此不再赘述。响应于确定出从历史环境图像中识别出车道线的可信度大于等于上述可信度阈值，上述第二输入模块可以将上述历史环境图像中历史车道线的二维图像信息输入上述初始空间模型中得到历史车道线的三维图像信息。
然后，上述确定模块可以基于上述历史车道线的二维图像信息、上述历史车道线的三维图像信息和上述历史导航引导线的三维图像信息，确定上述初始空间模型的损失函数，损失函数可以是用来估量模型的预测值f(x)与真实值Y的不一致程度，它是一个非负实值函数，通常使用L(Y，f(x))表示。损失函数可以包括对数损失函数、平方损失函数、指数损失函数等等。如何利用对数损失函数、平方损失函数和指数损失函数求解函数最优解是目前广泛研究和应用的公知常识，在此不再赘述。
最后，上述调整模块可以基于上述初始空间模型的损失函数，调整上述初始空间模型的参数，并可以将参数调整后的初始空间模型作为上述空间模型。具体地，上述调整模块可以基于上述三维历史车道线与直线之间的不一致程度，调整上述初始空间模型的参数以使得三维历史车道线与直线之间的不一致程度达到最小；上述调整模块也可以基于上述至少两条三维历史车道线之间不平行的程度，调整上述初始空间模型的参数以使得上述至少两条三维历史车道线之间不平行的程度达到最小；上述调整模块还可以基于上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度，调整上述初始空间模型的参数以使得三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向的不一致程度达到最小。上述调整模块可以将参数调整后的初始空间模型作为上述空间模型。
在本实施例的一些可选的实现方式中，上述确定模块可以包括第一确定子模块(图中未示出)和第二确定子模块(图中未示出)。上述第一确定子模块可以首先确定上述历史车道线的二维图像信息所指示 的二维历史车道线是否为直线，并可以确定上述历史车道线的三维图像信息所指示的三维历史车道线是否为直线。响应于确定出上述二维历史车道线为直线、而上述三维历史车道线不为直线，上述第二确定子模块可以确定上述初始空间模型的第一损失函数。若二维历史车道线为直线，则可以预测三维历史车道线也应为直线，而真实的三维历史车道线不为直线，则可以确定真实的三维历史车道线与直线之间的不一致程度，从而确定上述初始空间模型的第一损失函数。
在本实施例的一些可选的实现方式中，上述确定模块可以包括第三确定子模块(图中未示出)和第四确定子模块(图中未示出)。响应于确定出上述三维历史车道线为直线、且上述三维历史车道线为至少两条，上述第三确定子模块可以进一步确定上述至少两条三维历史车道线是否相互平行。具体地，上述第三确定子模块可以确定上述至少两条三维历史车道线的斜率是否相同，若相同，则上述至少两条三维历史车道线相互平行，若不相同，则上述至少两条三维历史车道线不平行。响应于确定出上述至少两条三维历史车道线不平行，上述第四确定子模块可以确定上述初始空间模型的第二损失函数。若存在至少两条三维历史车道线，则可以预测上述至少两条三维历史车道线应为互相平行，而真实的至少两条三维历史车道线不平行，则可以确定真实的至少两条三维历史车道线之间不平行的程度，从而确定上述初始空间模型的第二损失函数。
在本实施例的一些可选的实现方式中，上述确定模块可以包括第五确定子模块(图中未示出)和第六确定子模块(图中未示出)。响应于确定出上述三维历史车道线为直线，上述第五确定子模块可以进一步确定上述三维历史车道线所引导的车辆行驶方向与上述历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导的车辆行驶方向是否一致。响应于确定出上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向不一致，上述第六确定子模块可以确定上述初始空间模型的第三损失函数。若上述三维历史车道线为直线，则可以预测三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向应该一致，而真 实的三维历史车道线所引导的车辆行驶方向与三维历史导航引导线所引导的车辆行驶方向不一致，则可以确定上述三维历史车道线所引导的车辆行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度，从而确定上述初始空间模型的第三损失函数。上述第六确定子模块也可以结合在上述第二预设历史时间段内车辆的实际行驶参数，确定车辆的实际行驶方向与上述三维历史导航引导线所引导的车辆行驶方向的不一致程度。
下面参考图6，其示出了适于用来实现本发明实施例的终端设备的计算机系统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)601，其可以根据存储在只读存储器(ROM)602中的程序或者从存储部分608加载到随机访问存储器(RAM)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)接口605也连接至总线604。
以下部件连接至I/O接口605：包括键盘、鼠标等的输入部分606；包括诸如液晶显示器(LCD)以及扬声器等的输出部分607；包括硬盘等的存储部分608；以及包括诸如LAN卡、调制解调器等的网络接口卡的通信部分609。通信部分609经由诸如因特网的网络执行通信处理。驱动器610也根据需要连接至I/O接口605。可拆卸介质611，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610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608。
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609从网络上被下载和安装，和/或从可拆卸介质611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，图示了按照本发明各种实施例的系统、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。
描述于本发明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、确定单元和第一输出单元。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“通过摄像头获取包含当前车道线的当前环境图像的单元”。
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：通过摄像头获取包含当前车道线的当前环境图像；确定从当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和摄像头的摄像头参数；响应于可信度大于等于可信度阈值，将当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到当前车道线的三维图像信息，输出所得到的三维图像信息。
以上描述仅为本发明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，本发明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。例如上述特征与本发明中公开的(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 (14)

  1. 一种用于车辆的信息输出方法，其中，所述车辆包括摄像头，所述方法包括：
    通过所述摄像头获取包含当前车道线的当前环境图像；
    确定从所述当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，所述可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和所述摄像头的摄像头参数；
    响应于所述可信度大于等于所述可信度阈值，将所述当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到所述当前车道线的三维图像信息，输出所得到的三维图像信息。
  2. 根据权利要求1所述的方法，其中，所述车辆还包括显示屏，所述方法还包括：
    响应于所述可信度小于所述可信度阈值，将所述显示屏上所显示的当前导航引导线的二维图像信息输入所述空间模型中得到所述当前导航引导线的三维图像信息；
    基于第一预设历史时间段内车道线的三维图像信息和所述当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，输出所确定的三维图像信息。
  3. 根据权利要求1或2所述的方法，其中，所述空间模型是通过如下步骤得到的：
    获取第二预设历史时间段内在所述显示屏上显示过的历史导航引导线的二维图像信息，将所述历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图像信息；
    响应于确定出从所述第二预设历史时间段内的历史环境图像中识别出车道线的可信度大于等于所述可信度阈值，将所述历史环境图像中历史车道线的二维图像信息输入所述初始空间模型中得到历史车道线的三维图像信息；
    基于所述历史车道线的二维图像信息、所述历史车道线的三维图像信息和所述历史导航引导线的三维图像信息，确定所述初始空间模型的损失函数；
    基于所述损失函数，调整所述初始空间模型的参数，并将参数调整后的初始空间模型作为空间模型。
  4. 根据权利要求3所述的方法，其中，所述基于所述历史车道线的二维图像信息、所述历史车道线的三维图像信息和所述历史导航引导线的三维图像信息，确定所述初始空间模型的损失函数，包括：
    确定所述历史车道线的二维图像信息所指示的二维历史车道线是否为直线以及所述历史车道线的三维图像信息所指示的三维历史车道线是否为直线；
    响应于确定出所述二维历史车道线为直线、而所述三维历史车道线不为直线，确定所述初始空间模型的第一损失函数。
  5. 根据权利要求3所述的方法，其中，所述基于所述历史车道线的二维图像信息、所述历史车道线的三维图像信息和所述历史导航引导线的三维图像信息，确定所述初始空间模型的损失函数，包括：
    响应于确定出所述三维历史车道线为直线、且所述三维历史车道线为至少两条，进一步确定所述至少两条三维历史车道线是否平行；
    响应于确定出所述至少两条三维历史车道线不平行，确定所述初始空间模型的第二损失函数。
  6. 根据权利要求3所述的方法，其中，所述基于所述历史车道线的二维图像信息、所述历史车道线的三维图像信息和所述历史导航引导线的三维图像信息，确定所述初始空间模型的损失函数，包括：
    响应于确定出所述三维历史车道线为直线，进一步确定所述三维历史车道线所引导的车辆行驶方向与所述历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导的车辆行驶方向是否一致；
    响应于确定出所述三维历史车道线所引导的车辆行驶方向与所述 三维历史导航引导线所引导的车辆行驶方向不一致，确定所述初始空间模型的第三损失函数。
  7. 一种用于车辆的信息输出装置，其中，所述车辆包括摄像头，所述装置包括：
    获取单元，配置用于通过所述摄像头获取包含当前车道线的当前环境图像；
    确定单元，配置用于确定从所述当前环境图像中识别出当前车道线的可信度是否大于等于预设的可信度阈值，其中，所述可信度是基于以下至少一项确定的：障碍物对车道线的遮挡程度和所述摄像头的摄像头参数；
    第一输出单元，配置用于响应于所述可信度大于等于所述可信度阈值，将所述当前环境图像中当前车道线的二维图像信息输入预先建立的空间模型中得到所述当前车道线的三维图像信息，输出所得到的三维图像信息。
  8. 根据权利要求7所述的装置，其中，所述车辆还包括显示屏，所述装置还包括：
    输入单元，配置用于响应于所述可信度小于所述可信度阈值，将所述显示屏上所显示的当前导航引导线的二维图像信息输入所述空间模型中得到所述当前导航引导线的三维图像信息；
    第二输出单元，配置用于基于第一预设历史时间段内车道线的三维图像信息和所述当前导航引导线的三维图像信息，确定当前车道线的三维图像信息，输出所确定的三维图像信息。
  9. 根据权利要求7或8所述的装置，其中，所述装置还包括空间模型建立单元，所述空间模型建立单元包括：
    第一输入模块，配置用于获取第二预设历史时间段内在所述显示屏上显示过的历史导航引导线的二维图像信息，将所述历史导航引导线的二维图像信息输入初始空间模型中得到历史导航引导线的三维图 像信息；
    第二输入模块，配置用于响应于确定出从所述第二预设历史时间段内的历史环境图像中识别出车道线的可信度大于等于所述可信度阈值，将所述历史环境图像中历史车道线的二维图像信息输入所述初始空间模型中得到历史车道线的三维图像信息；
    确定模块，配置用于基于所述历史车道线的二维图像信息、所述历史车道线的三维图像信息和所述历史导航引导线的三维图像信息，确定所述初始空间模型的损失函数；
    调整模块，配置用于基于所述损失函数，调整所述初始空间模型的参数，并将参数调整后的初始空间模型作为空间模型。
  10. 根据权利要求9所述的装置，其中，所述确定模块，包括：
    第一确定子模块，配置用于确定所述历史车道线的二维图像信息所指示的二维历史车道线是否为直线以及所述历史车道线的三维图像信息所指示的三维历史车道线是否为直线；
    第二确定子模块，配置用于响应于确定出所述二维历史车道线为直线、而所述三维历史车道线不为直线，确定所述初始空间模型的第一损失函数。
  11. 根据权利要求9所述的装置，其中，所述确定模块，包括：
    第三确定子模块，配置用于响应于确定出所述三维历史车道线为直线、且所述三维历史车道线为至少两条，进一步确定所述至少两条三维历史车道线是否平行；
    第四确定子模块，配置用于响应于确定出所述至少两条三维历史车道线不平行，确定所述初始空间模型的第二损失函数。
  12. 根据权利要求9所述的装置，其中，所述确定模块，包括：
    第五确定子模块，配置用于响应于确定出所述三维历史车道线为直线，进一步确定所述三维历史车道线所引导的车辆行驶方向与所述历史导航引导线的三维图像信息所指示的三维历史导航引导线所引导 的车辆行驶方向是否一致；
    第六确定子模块，配置用于响应于确定出所述三维历史车道线所引导的车辆行驶方向与所述三维历史导航引导线所引导的车辆行驶方向不一致，确定所述初始空间模型的第三损失函数。
  13. 一种电子设备，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1-6中任一所述的方法。
  14. 一种计算机可读存储介质，其上存储有计算机程序，其中，该程序被处理器执行时实现如权利要求1-6中任一所述的方法。
PCT/CN2018/099164 2017-11-16 2018-08-07 用于车辆的信息输出方法和装置 WO2019095735A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711139956.XA CN107830869B (zh) 2017-11-16 2017-11-16 用于车辆的信息输出方法和装置
CN201711139956.X 2017-11-16

Publications (1)

Publication Number Publication Date
WO2019095735A1 true WO2019095735A1 (zh) 2019-05-23

Family

ID=61651848

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/099164 WO2019095735A1 (zh) 2017-11-16 2018-08-07 用于车辆的信息输出方法和装置

Country Status (2)

Country Link
CN (1) CN107830869B (zh)
WO (1) WO2019095735A1 (zh)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107830869B (zh) * 2017-11-16 2020-12-11 百度在线网络技术（北京）有限公司 用于车辆的信息输出方法和装置
CN109460739A (zh) * 2018-11-13 2019-03-12 广州小鹏汽车科技有限公司 车道线检测方法及装置
CN109684944B (zh) * 2018-12-10 2021-09-03 百度在线网络技术（北京）有限公司 障碍物检测方法、装置、计算机设备和存储介质
CN109703569B (zh) * 2019-02-21 2021-07-27 百度在线网络技术（北京）有限公司 一种信息处理方法、装置及存储介质
CN109765902B (zh) * 2019-02-22 2022-10-11 阿波罗智能技术(北京)有限公司 无人车驾驶参考线处理方法、装置及车辆
CN112183415A (zh) * 2020-09-30 2021-01-05 上汽通用五菱汽车股份有限公司 车道线处理方法、车辆和可读存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102806913A (zh) * 2011-05-31 2012-12-05 德尔福电子(苏州)有限公司 一种新型的车道线偏离检测方法及装置
CN105260699A (zh) * 2015-09-10 2016-01-20 百度在线网络技术（北京）有限公司 一种车道线数据的处理方法及装置
CN107025432A (zh) * 2017-02-28 2017-08-08 合肥工业大学 一种高效的车道线检测跟踪方法及系统
CN107830869A (zh) * 2017-11-16 2018-03-23 百度在线网络技术（北京）有限公司 用于车辆的信息输出方法和装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10282997B2 (en) * 2015-07-20 2019-05-07 Dura Operating, Llc System and method for generating and communicating lane information from a host vehicle to a vehicle-to-vehicle network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102806913A (zh) * 2011-05-31 2012-12-05 德尔福电子(苏州)有限公司 一种新型的车道线偏离检测方法及装置
CN105260699A (zh) * 2015-09-10 2016-01-20 百度在线网络技术（北京）有限公司 一种车道线数据的处理方法及装置
CN107025432A (zh) * 2017-02-28 2017-08-08 合肥工业大学 一种高效的车道线检测跟踪方法及系统
CN107830869A (zh) * 2017-11-16 2018-03-23 百度在线网络技术（北京）有限公司 用于车辆的信息输出方法和装置

Also Published As

Publication number Publication date
CN107830869A (zh) 2018-03-23
CN107830869B (zh) 2020-12-11

Similar Documents

Publication Publication Date Title
WO2019095735A1 (zh) 用于车辆的信息输出方法和装置
US11042762B2 (en) Sensor calibration method and device, computer device, medium, and vehicle
CN111626208B (zh) 用于检测小目标的方法和装置
US11632536B2 (en) Method and apparatus for generating three-dimensional (3D) road model
US9378424B2 (en) Method and device for detecting road region as well as method and device for detecting road line
US10891795B2 (en) Localization method and apparatus based on 3D color map
JP7245275B2 (ja) 軌道計画モデルの訓練方法と装置、電子機器、コンピュータ記憶媒体及びコンピュータプログラム
CN107941226B (zh) 用于生成车辆的方向引导线的方法和装置
CN109725634A (zh) 用于自动驾驶车辆的使用分色镜的3d lidar系统
JP2018530825A (ja) 非障害物エリア検出のためのシステムおよび方法
CN113377888B (zh) 训练目标检测模型和检测目标的方法
CN109974733A (zh) 用于ar导航的poi显示方法、装置、终端和介质
CN113483774B (zh) 导航方法、装置、电子设备及可读存储介质
CN114170826B (zh) 自动驾驶控制方法和装置、电子设备和存储介质
WO2022206414A1 (zh) 三维目标检测方法及装置
CN114919584A (zh) 机动车定点目标测距方法、装置及计算机可读存储介质
CN113126120B (zh) 数据标注方法、装置、设备、存储介质以及计算机程序产品
CN115578432B (zh) 图像处理方法、装置、电子设备及存储介质
US11741671B2 (en) Three-dimensional scene recreation using depth fusion
US11619498B2 (en) Verification method and device for modeling route, unmanned vehicle, and storage medium
CN115375739A (zh) 车道线生成方法、设备和介质
CN114386481A (zh) 一种车辆感知信息融合方法、装置、设备及存储介质
JP7324792B2 (ja) 位置情報を生成するための方法及び装置
CN111383337A (zh) 用于识别对象的方法和装置
US11620831B2 (en) Register sets of low-level features without data association

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: 18878637

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18878637

Country of ref document: EP

Kind code of ref document: A1