CN116001799A - Vehicle line pressing driving early warning method and device, electronic equipment and storage medium - Google Patents

Vehicle line pressing driving early warning method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN116001799A
CN116001799A CN202211551915.2A CN202211551915A CN116001799A CN 116001799 A CN116001799 A CN 116001799A CN 202211551915 A CN202211551915 A CN 202211551915A CN 116001799 A CN116001799 A CN 116001799A
Authority
CN
China
Prior art keywords
vehicle
early warning
line
tire
distance
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211551915.2A
Other languages
Chinese (zh)
Inventor
陆强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Network Technology Shanghai Co Ltd
Original Assignee
International Network Technology Shanghai Co Ltd
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 International Network Technology Shanghai Co Ltd filed Critical International Network Technology Shanghai Co Ltd
Priority to CN202211551915.2A priority Critical patent/CN116001799A/en
Publication of CN116001799A publication Critical patent/CN116001799A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Traffic Control Systems (AREA)

Abstract

The invention provides a vehicle line pressing driving early warning method, a device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground; inputting a vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line; based on the first distance and the tire deflection angle, whether the vehicle runs along the line within the preset early warning advance time or not is judged by combining the current motion state of the vehicle, and early warning is carried out in advance. Compared with the prior art, the vehicle warning device can realize early judgment and early warning only when the vehicle really presses the line. In addition, the line pressing driving prediction model does not need to separately predict the lane lines, so that the calculation amount of the model is saved, and the prediction speed of the model is increased.

Description

Vehicle line pressing driving early warning method and device, electronic equipment and storage medium
Technical Field
The invention relates to the technical field of automatic driving of vehicles, in particular to a vehicle line pressing driving early warning method, a device, electronic equipment and a storage medium.
Background
The auxiliary driving refers to providing auxiliary support for a driver in the process of driving the vehicle, so that the driver can drive the vehicle to run on a road more easily and safely, and the lane keeping auxiliary system, the automatic parking auxiliary system, the lane line pressing early warning auxiliary system and the like are all vehicle auxiliary driving. The lane line pressing early warning auxiliary system can remind a driver or intervene in the steering system to adjust the driving route in time when the system is arranged on the line pressing of the vehicle by identifying the mark line of the driving road of the vehicle through the camera.
In the prior art, an image is generally input into a target detection model capable of predicting the positions of a lane line and a tire, and whether to press the tire is judged according to the position distance between the lane line and the tire. In the scheme, on one hand, the target detection model needs to predict the lane line, but the calculation amount of the predicted lane line is usually higher, so that the calculation amount of the model is increased; on the other hand, whether the line is pressed or not at the driving moment corresponding to the input image is judged, and the warning is only carried out until the vehicle really presses the line, so that the early judgment and early warning cannot be realized. Compared with the prior art, the method only judges whether the line is pressed at the driving moment corresponding to the input image, and the warning is not carried out until the vehicle really presses the line, so that the early judgment and early warning can not be realized.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a vehicle line pressing driving early warning method, a device, electronic equipment and a storage medium, and aims to solve the problem that a model in the prior art cannot realize early judgment.
The invention provides a vehicle line pressing driving early warning method, which comprises the following steps:
acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
inputting the vehicle driving image into a line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
and judging whether the vehicle runs on a line in a preset early warning advance time and early warning in advance according to the first distance and the tire deflection angle and the current motion state of the vehicle.
According to the vehicle line pressing driving early warning method provided by the invention, based on the first distance and the tire deflection angle, the current motion state of the vehicle is combined to judge whether the vehicle runs in line pressing and early warns in advance within the preset early warning advance time, and the vehicle line pressing driving early warning method comprises the following steps:
acquiring the current running speed of the vehicle through a speed sensor of the vehicle;
the cosine value of the tire deflection angle is obtained through cosine function calculation on the tire deflection angle;
multiplying the cosine value by the running speed and the early warning advance time to obtain a second distance;
and judging the second distance based on the first distance, if the second distance is larger than or equal to the first distance, indicating that the vehicle is driven by the internal pressure line in the early warning time, sending early warning information, and if not, not generating line pressing driving.
According to the vehicle line pressing driving early warning method provided by the invention, the line pressing early warning decision data further comprises: a category of positional relationship of a tire ground contact point with an adjacent lane line, the category of positional relationship comprising: a first positional relationship category, a second positional relationship category, and a third positional relationship category, wherein the first positional relationship category represents that the adjacent lane line is to the left of the tire ground point, the second positional relationship category represents that the adjacent lane line is to the right of the tire ground point, and the third positional relationship category represents that the tire ground point is on the adjacent lane line;
correspondingly, when the vehicle runs in a line pressing mode within the early warning advance time, the vehicle further comprises: and judging whether the vehicle is left wheel wire pressing early warning or right wheel wire pressing early warning or wire pressing warning based on the position relation type.
According to the vehicle line-pressing driving early warning method provided by the invention, before the vehicle driving image is input into the line-pressing driving prediction model, the method further comprises the following steps: training the line pressing driving prediction model specifically comprises the following steps:
acquiring a vehicle running sample image and a corresponding target true value thereof;
and training the line pressing driving prediction model according to a loss value corresponding to the target true value, so as to obtain the line pressing driving prediction model for generating line pressing early warning decision data.
According to the vehicle line-pressing driving early warning method provided by the invention, the target true value corresponding to the vehicle driving sample image is used as a label, and the line-pressing driving prediction model to be trained is trained, and the method comprises the following steps:
marking the center point of each target frame in the vehicle driving sample image to generate a position label;
generating a corresponding center probability map based on the position label, wherein the center probability map comprises probabilities that each pixel point in the vehicle driving sample image belongs to a center point of a target frame;
training the line pressing driving prediction model to be trained based on the center probability map to obtain the center point position of each target frame, wherein the center point position of each target frame comprises a tire center point position and a tire grounding point position.
According to the vehicle line pressing driving early warning method provided by the invention, when the line pressing driving prediction model is trained, the method further comprises the following steps:
acquiring a tire center point position and a tire ground point position based on the vehicle running sample image;
and optimizing the internal parameters of the line pressing driving prediction model based on the tire center point position and the tire grounding point position.
According to the vehicle line pressing driving early warning method provided by the invention, the line pressing driving prediction model comprises the following steps: a two-dimensional feature extraction layer, a first sub-model, and a second sub-model, wherein an output of the first sub-model serves as an input of the second sub-model;
the two-dimensional feature extraction layer is used for extracting two-dimensional features of the vehicle running image;
the first sub-model is used for outputting a tire center point position and a tire grounding point position according to the two-dimensional characteristics;
the second sub-model is used for outputting a first distance, a tire deflection angle and a position relation category according to the two-dimensional characteristics.
The invention also provides a vehicle line pressing driving early warning device, which comprises:
the image acquisition module is used for acquiring a vehicle running image which is shot by the vehicle body looking around camera and contains vehicle tires and the ground;
the early warning data acquisition module is used for inputting the vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, and the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
and the line pressing early warning judging module is used for judging whether the vehicle runs in line pressing and early warning in advance in a preset early warning advance time by combining the current motion state of the vehicle based on the first distance and the tire deflection angle.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the vehicle line pressing driving early warning method according to any one of the above when executing the program.
The invention also provides a non-transitory computer readable storage medium, on which is stored a computer program which, when executed by a processor, implements a vehicle line-pressing driving pre-warning method as described in any one of the above.
The invention provides a vehicle line pressing driving early warning method, a device, electronic equipment and a storage medium, wherein a vehicle running image which is shot by a vehicle body looking-around camera and contains a vehicle tire and the ground is acquired; inputting a vehicle driving image into a line pressing driving prediction model to obtain line pressing early warning decision data comprising a first distance and a tire deflection angle, wherein the first distance is the distance between a tire grounding point and an adjacent lane line; based on the first distance and the tire deflection angle, whether the vehicle runs along the line or not can be judged by combining the current motion state of the vehicle within the preset early warning advance time. Compared with the prior art, the vehicle warning device can realize early judgment and early warning only when the vehicle really presses the line. In addition, the line pressing driving prediction model can directly predict the distance between the grounding point of the tire and the adjacent lane line according to the vehicle driving image. Compared with the existing model, the method has the advantages that the positions of the lane lines and the tires can be further predicted only by predicting the lane lines independently, so that the calculated amount of the model is saved, and the prediction speed of the model is increased.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the invention, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a vehicle line-pressing driving early warning method provided by the invention;
FIG. 2 is a second flow chart of a method for warning the line-pressing driving of a vehicle according to the present invention;
fig. 3 is a schematic structural diagram of a line pressing driving prediction model in the vehicle line pressing driving early warning method provided by the invention;
FIG. 4 is a third flow chart of a method for warning the line-pressing driving of a vehicle according to the present invention;
fig. 5 is a schematic structural diagram of a vehicle line-pressing driving early warning device provided by the invention;
fig. 6 is a schematic structural diagram of an electronic device provided by the present invention.
Reference numerals:
21: an image acquisition module; 22: the early warning data acquisition module; 23: and the line pressing early warning judging module.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
A neural network may be composed of neural units, and is understood to mean, in particular, a neural network having an input layer, an hidden layer, and an output layer, where in general, the first layer is the input layer, the last layer is the output layer, and the middle layers are all hidden layers. Among them, the neural network with many hidden layers is called deep neural network (deep neural network, DNN).
In an intelligent driving scene, a car body looking-around camera is generally used for near-field sensing, and a fish-eye camera is generally adopted and is arranged around the car body. A typical look-around system consists of four fisheye cameras, each with a perceived field angle exceeding 180 degrees; a 360 degree perceived field of view is formed by stitching the four fields of view. Four fisheye cameras are sufficient to cover a 360 degree near field region, with major application scenarios including auto park, traffic congestion mitigation assistance, driving assistance, and the like.
Example 1
Referring to fig. 1, the method for early warning of vehicle line ball driving provided in this embodiment includes:
step S1: acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
in this step, the camera for looking around the vehicle body includes four fisheye cameras which can be installed at the front left side, front right side, rear left side and rear right side of the vehicle, respectively, so as to collect running images including tires and ground in the movement of the vehicle around the vehicle by 360 degrees.
Step S2: inputting a vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
in the step, the line pressing driving prediction model adopts a deep learning model, and line pressing driving is performed in advanceThe geodetic model further generates a top view or a bird's eye view comprising only the tire and the ground by cutting or cropping each constant area image output from the pan-around camera device. Determining a plurality of target frame areas from the aerial view, including a tire target frame area, a tire ground contact point and a target frame area adjacent to the lane line, and obtaining a distance between the tire ground contact point and the adjacent lane line, namely a first distance d 1 And acquiring the tire deflection angle α from the tire target frame area.
Step S3: judging whether the vehicle runs on a line in a preset early warning advance time or not and early warning in advance according to the first distance and the tire deflection angle and the current motion state of the vehicle;
specifically, according to the vehicle line pressing driving early warning method provided by the invention, a vehicle driving image which is shot by a vehicle body looking around camera and contains a vehicle tire and the ground is acquired; inputting a vehicle driving image into a line pressing driving prediction model (depth neural network) after training, and predicting to obtain line pressing early warning decision data comprising a first distance and a tire deflection angle, wherein the first distance is the distance between a tire grounding point and an adjacent lane line; based on the first distance and the tire deflection angle, whether the vehicle runs along the line or not can be judged by combining the current motion state of the vehicle within the preset early warning advance time. Compared with the prior art, the invention can realize early judgment and early warning only when the vehicle really presses the line. In addition, the prediction model can directly predict the distance between the grounding point of the tire and the adjacent lane line according to the running image of the vehicle. Compared with the existing model, the method has the advantages that the positions of the lane lines and the tires can be further predicted only by predicting the lane lines independently, so that the calculated amount of the model is saved, and the prediction speed of the model is increased.
Preferably, referring to fig. 2, based on the first distance and the tire deflection angle, determining whether the vehicle runs on a line and early warns in advance in a preset early warning advance time according to the current motion state of the vehicle includes:
step S31: acquiring the current running speed V of the vehicle through a speed sensor of the vehicle;
step S32: the cosine value cos alpha of the tire deflection angle is obtained through cosine function calculation on the tire deflection angle alpha;
step S33: multiplying the cosine value cos alpha by the running speed V and the early warning advance time t to obtain a second distance d 2 ；
Step S34: based on the first distance d 1 For a second distance d 2 Judging if the second distance d 2 Greater than or equal to the first distance d 1 The vehicle is driven by a pressure line within the early warning advance time t, early warning information is sent out, and if not, the vehicle is not driven by the pressure line;
specifically, based on the tire deflection angle α, the current running speed V of the vehicle, and the early warning advance time t, the tire deflection angle α is calculated by the formula d 2 Calculating the driving distance of the vehicle at the early warning advance time t as the second distance d 2 Based on the first distance d 1 For a second distance d 2 Judging if the second distance d 2 Greater than or equal to the first distance d 1 The vehicle is driven by the internal pressure line in the early warning advance time t, the warning is triggered, and the vehicle-mounted equipment (such as V2X equipment) is used for warning the driver that the line pressing driving is about to happen on devices such as an instrument panel and a vehicle screen through voice and characters, so that the driver is reminded to correct the driving track in time to avoid dangerous actions such as line pressing driving. The warning device can also be used for reminding in a mode of simulating high-frequency flicker light effect of the warning lamp and the like when the driver drives the line or the line immediately by displaying green on the part of the front windshield and the rearview mirror to indicate that the line is not pressed, displaying blue to indicate that the line is about to be pressed and displaying red to indicate that the line is pressed. By the method, the line pressing driving behavior of the vehicle at a certain time in the future can be prejudged in advance, the driver is reminded to take the operation to avoid line pressing, the probability of violation and accident occurrence is reduced, and the safety performance of vehicle driving is improved.
Preferably, the wire pressing early warning decision data further includes: the tire grounding point and the adjacent lane line are in a position relation category comprising: a first positional relationship category, a second positional relationship category, and a third positional relationship category, wherein the first positional relationship category represents that the adjacent lane line is to the left of the tire ground point, the second positional relationship category represents that the adjacent lane line is to the right of the tire ground point, and the third positional relationship category represents that the tire ground point is on the adjacent lane line;
correspondingly, when the vehicle runs in a line pressing mode within the early warning advance time, the vehicle further comprises: and judging whether the vehicle is left wheel wire pressing early warning or right wheel wire pressing early warning or wire pressing warning based on the position relation type.
Specifically, three positional relationship types of the ground point and the adjacent lane line can be obtained by further judging the target frame areas of the ground point and the adjacent lane line of the tire at different positions photographed by the fisheye cameras at different positions. When the line pressing driving behavior of the pre-judging vehicle occurs at a certain time in the future, if the first position relation type is acquired, the vehicle is judged to be a left wheel line pressing early warning, the information reminds the left wheel of the vehicle of being about to line pressing, if the second position relation type is acquired, the vehicle is judged to be a left wheel line pressing early warning, the information reminds the right wheel of the vehicle of being about to line pressing, and if the third position relation type is acquired, the vehicle is judged to be a line pressing warning, and the information reminds the vehicle to be in line pressing running. Through judging the position relation between the grounding point of the tire and the adjacent lane lines, the content of the early warning information is enriched, the driver is assisted to further make corresponding measures, and the reliability and the professionality of the early warning information are improved.
In this embodiment, before inputting the vehicle running image into the line ball driving prediction model, the method further includes: training a line ball driving prediction model, specifically includes:
acquiring a vehicle running sample image and a corresponding target true value thereof;
and training the line pressing driving prediction model to be trained according to a loss value of the target true value, so as to obtain the line pressing driving prediction model for generating line pressing early warning decision data.
Preferably, taking a target true value corresponding to a vehicle driving sample image as a label, training a line pressing driving prediction model to be trained, including:
marking the center point of each target frame in the vehicle running sample image to generate a position label;
generating a corresponding center probability map based on the position label, wherein the center probability map comprises probabilities that all pixel points in a vehicle running sample image belong to a target frame center point;
training the line pressing driving prediction model to be trained based on the central probability map to obtain the central point position of each target frame, wherein the central point position of each target frame comprises the tire central point position and the tire grounding point position.
Preferably, when training the line pressing driving prediction model, the method further comprises:
acquiring a tire center point position and a tire ground point position based on a vehicle running sample image;
and optimizing the internal parameters of the line pressing driving prediction model based on the tire center point position and the tire grounding point position.
Specifically, a vehicle running sample image and a corresponding target true value thereof are obtained, a line pressing driving prediction model is trained through a probability icon to obtain the center point position of a tire target frame, the size of the probability map is 1/4 of the size of an input image through downsampling of a top view or a bird's eye view only comprising the tire and the ground, the probability values of all pixels are initialized to be 0, and then the probability map is manufactured around the center point of each target frame. Let the center point coordinate of each target frame (here, car) in the training label be O, the width and height be h, w respectively, then regard O as the center, radius be (h+w)/4 as the circle, the distance from the pixel point p in the circle to the point O is recorded as L, then the probability at this point is 1-L/r. And training a line pressing driving prediction model by using a vehicle driving sample image with a probability map label of a target true value, acquiring a tire target frame area, training a tire center point to obtain a loss function, and calculating by using focal loss. The tire grounding point position is also predicted by a probability map mode, the manufacturing method is the same as the above, the center point is the tire grounding point, the width w1 and the height h1 of a tire target frame are used, the radius is (h1+w1)/4 to be a circle, a line pressing driving prediction model is trained by using a vehicle running sample image with a probability map label with a target true value, a tire grounding point target frame area is obtained, a loss function for tire grounding point training is obtained, and focus loss focal calculation is adopted. The first distance and the tire deflection angle are predicted in a probability map mode, the manufacturing method is the same as above, the first distance is at a pixel position between a tire grounding point and an adjacent lane line, a training label value is set to be a true value of the distance between the tire grounding point and the adjacent lane line, and other pixel positions are set to be 0; the tire deflection angle is set to the true value of the deflection angle of the tire at the pixel position corresponding to the tire grounding point, the other pixel positions are set to 0, and the loss function for training is calculated by adopting the smooth L1 loss. The loss function for the position relation type prediction training adopts cross entropy loss calculation.
Further, the tire center point position and the tire grounding point position are predicted based on the vehicle running sample image and are trained based on supervised learning, so that the robustness of the model to feature extraction is improved, and the training effect is improved. According to regression prediction of the tire center point position and the tire grounding point position, parameters of the line pressing driving prediction model are optimized according to the predicted value, and the accuracy of the model prediction is improved.
Referring to fig. 3, in this embodiment, the line ball driving prediction model includes: the system comprises a two-dimensional feature extraction layer, a first sub-model and a second sub-model, wherein the output of the first sub-model is used as the input of the second sub-model;
the two-dimensional feature extraction layer is used for extracting two-dimensional features of the vehicle driving image;
the first sub-model is used for outputting a tire center point position and a tire grounding point position according to the two-dimensional characteristics;
the second sub-model is used for outputting a first distance, a tire deflection angle and a position relation category according to the two-dimensional characteristics.
Based on the monocular 3D detection frame prediction model in fig. 3, as shown in fig. 4, step S2 includes:
step S21: inputting a vehicle running image into a two-dimensional feature extraction layer to obtain two-dimensional features;
step S22: inputting the two-dimensional characteristics into a first sub-model to obtain a tire center point position and a tire grounding point position;
step S23: and inputting the two-dimensional characteristics into a second sub-model to obtain a first distance, a tire deflection angle and a position relation type.
Example two
Referring to fig. 5, this embodiment provides a vehicle line ball driving early warning device, including:
an image acquisition module 21 for acquiring a vehicle running image including a vehicle tire and a ground taken by a vehicle body looking around camera;
the early warning data acquisition module 22 is configured to input a vehicle driving image to the line pressing driving prediction model, and obtain line pressing early warning decision data output by the line pressing driving prediction model, where the line pressing early warning decision data includes: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
the line pressing early warning judging module 23 is configured to judge whether the vehicle is in line pressing driving and early warning in advance within a preset early warning advance time according to the first distance and the tire deflection angle and in combination with the current motion state of the vehicle.
Optionally, the vehicle line pressing driving early warning device provided in this embodiment further includes: the model training module is specifically used for:
inputting a vehicle running image into a two-dimensional feature extraction layer to obtain two-dimensional features;
inputting the two-dimensional characteristics into a first sub-model, introducing the tire center point position output by the first sub-model and the corresponding training label into a preset first loss function, introducing the output tire grounding point position and the corresponding training label into a preset second loss function, and completing the training of the first sub-model when the first loss function and the second loss function reach convergence;
inputting the two-dimensional characteristics into a second sub-model, bringing the first distance output by the second sub-model together with the corresponding training label into a preset third loss function, bringing the output tire deflection angle together with the corresponding training label into a preset fourth loss function, and bringing the output position relation category together with the corresponding training label into a preset fifth loss function, wherein the second sub-model training is completed when the third loss function, the fourth loss function and the fifth loss function reach convergence.
Preferably, the tire center point position and the tire ground point position output by the first sub-model are input into the second sub-model, and the internal parameters of the second sub-model are optimized.
Optionally, the early warning data acquisition module 22 is based on a deep learning neural network, including: the two-dimensional feature extraction layer, the first sub-model and the second sub-model;
the two-dimensional feature extraction layer is used for extracting two-dimensional features of the vehicle driving image;
the first sub-model is used for outputting a tire center point position and a tire grounding point position according to the two-dimensional characteristics;
the second sub-model is used for outputting a first distance, a tire deflection angle and a position relation category according to the two-dimensional characteristics.
Optionally, the early warning data acquisition module 22 is specifically configured to:
inputting the two-dimensional image into a two-dimensional feature extraction layer to obtain two-dimensional features;
inputting the two-dimensional characteristics into a first sub-model to obtain a tire center point position and a tire grounding point position;
and inputting the two-dimensional characteristics into a second sub-model to obtain a first distance, a tire deflection angle and a position relation type.
Optionally, the wire pressing early warning judging module 23 is a logic judging module, and the main body for executing the logic judging function may be a vehicle-mounted ECU controller, and the relevant instructions, such as a control instruction, an early warning instruction, and the like, are generated by judging according to the input data, and specifically include: a first acquisition unit for acquiring a current running speed of the vehicle through a speed sensor of the vehicle; the first calculation unit is used for calculating the tire deflection angle through a cosine function to obtain a cosine value of the tire deflection angle; the second calculation unit is used for multiplying the cosine value with the running speed and the early warning advance time to obtain a second distance; the judging unit is used for judging the second distance based on the first distance, if the second distance is larger than or equal to the first distance, the vehicle is driven in a pressure line in early warning time and an early warning instruction is generated, early warning information is sent out, and if not, line pressing driving does not occur; and the early warning unit is used for reminding a driver according to the early warning information generated by the early warning instruction.
The implementation process of the functions and actions of each module in the above device is specifically detailed in the implementation process of the corresponding steps in the above method, so relevant parts only need to be referred to in the description of the method embodiments, and are not repeated here. The system embodiments described above are merely illustrative, and some or all of the modules may be selected according to actual needs to achieve the objectives of the inventive arrangements.
Example III
As shown in fig. 6, the present embodiment provides an electronic apparatus including: processor 310, communication interface (Communications Interface) 320, memory 330 and communication bus 340, wherein processor 310, communication interface 320, memory 330 accomplish communication with each other through communication bus 340. The processor 310 may invoke logic instructions in the memory 330, the processor 310 executing a vehicle line ball drive warning method comprising:
acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
inputting a vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
based on the first distance and the tire deflection angle, whether the vehicle runs along the line within the preset early warning advance time or not is judged by combining the current motion state of the vehicle, and early warning is carried out in advance.
Further, the logic instructions in the memory 330 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions to cause a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being storable on a non-transitory computer readable storage medium, the computer program, when executed by a processor, being capable of executing the methods described above to provide a vehicle line ball driving warning method, the method comprising:
acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
inputting a vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
based on the first distance and the tire deflection angle, whether the vehicle runs along the line within the preset early warning advance time or not is judged by combining the current motion state of the vehicle, and early warning is carried out in advance.
Example IV
The present embodiment provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a vehicle line ball driving early warning method as described above, the method comprising:
acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
inputting a vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
based on the first distance and the tire deflection angle, whether the vehicle runs along the line within the preset early warning advance time or not is judged by combining the current motion state of the vehicle, and early warning is carried out in advance.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solutions may be embodied essentially or in part in the form of a software product, which may be stored in a computer-readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the method described in the various embodiments or portions of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. The vehicle line pressing driving early warning method is characterized by comprising the following steps of:
acquiring a vehicle running image which is shot by a vehicle body looking-around camera and contains vehicle tires and ground;
inputting the vehicle driving image into a line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, wherein the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
and judging whether the vehicle runs on a line in a preset early warning advance time and early warning in advance according to the first distance and the tire deflection angle and the current motion state of the vehicle.
2. The vehicle line ball driving early warning method according to claim 1, wherein the determining whether the vehicle is in line ball driving and early warning in advance in a preset early warning advance time based on the first distance and the tire deflection angle in combination with the current motion state of the vehicle includes:
acquiring the current running speed of the vehicle through a speed sensor of the vehicle;
the cosine value of the tire deflection angle is obtained through cosine function calculation on the tire deflection angle;
multiplying the cosine value by the running speed and the early warning advance time to obtain a second distance;
and judging the second distance based on the first distance, if the second distance is larger than or equal to the first distance, indicating that the vehicle is driven by the internal pressure line in the early warning time, sending early warning information, and if not, not generating line pressing driving.
3. The vehicle line ball driving pre-warning method of claim 1, wherein the line ball pre-warning decision data further comprises: a category of positional relationship of a tire ground contact point with an adjacent lane line, the category of positional relationship comprising: a first positional relationship category, a second positional relationship category, and a third positional relationship category, wherein the first positional relationship category represents that the adjacent lane line is to the left of the tire ground point, the second positional relationship category represents that the adjacent lane line is to the right of the tire ground point, and the third positional relationship category represents that the tire ground point is on the adjacent lane line;
correspondingly, when the vehicle runs in a line pressing mode within the early warning advance time, the vehicle further comprises: and judging whether the vehicle is left wheel wire pressing early warning or right wheel wire pressing early warning or wire pressing warning based on the position relation type.
4. The vehicle tandem drive early warning method according to claim 1, characterized by further comprising, before inputting the vehicle running image to the tandem drive prediction model: training the line pressing driving prediction model specifically comprises the following steps:
acquiring a vehicle running sample image and a corresponding target true value thereof;
and training the line pressing driving prediction model according to a loss value corresponding to the target true value, so as to obtain the line pressing driving prediction model for generating line pressing early warning decision data.
5. The method for warning the line ball driving of the vehicle according to claim 4, wherein training the line ball driving prediction model to be trained by using the target true value corresponding to the vehicle driving sample image as a tag comprises:
marking the center point of each target frame in the vehicle driving sample image to generate a position label;
generating a corresponding center probability map based on the position label, wherein the center probability map comprises probabilities that each pixel point in the vehicle driving sample image belongs to a center point of a target frame;
training the line pressing driving prediction model to be trained based on the center probability map to obtain the center point position of each target frame, wherein the center point position of each target frame comprises a tire center point position and a tire grounding point position.
6. The vehicle tandem driving early warning method according to claim 5, wherein when training the tandem driving prediction model, further comprising:
acquiring a tire center point position and a tire ground point position based on the vehicle running sample image;
and optimizing the internal parameters of the line pressing driving prediction model based on the tire center point position and the tire grounding point position.
7. The vehicle tandem driving early warning method according to claim 1, wherein the tandem driving prediction model includes: a two-dimensional feature extraction layer, a first sub-model, and a second sub-model, wherein an output of the first sub-model serves as an input of the second sub-model;
the two-dimensional feature extraction layer is used for extracting two-dimensional features of the vehicle running image;
the first sub-model is used for outputting a tire center point position and a tire grounding point position according to the two-dimensional characteristics;
the second sub-model is used for outputting a first distance, a tire deflection angle and a position relation category according to the two-dimensional characteristics.
8. The utility model provides a vehicle line ball driving early warning device which characterized in that includes:
the image acquisition module is used for acquiring a vehicle running image which is shot by the vehicle body looking around camera and contains vehicle tires and the ground;
the early warning data acquisition module is used for inputting the vehicle driving image into the line pressing driving prediction model to obtain line pressing early warning decision data output by the line pressing driving prediction model, and the line pressing early warning decision data comprises: a first distance and a tire deflection angle, the first distance being a distance between a tire ground contact point and an adjacent lane line;
and the line pressing early warning judging module is used for judging whether the vehicle runs in line pressing and early warning in advance in a preset early warning advance time by combining the current motion state of the vehicle based on the first distance and the tire deflection angle.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the vehicle line ball drive warning method of any one of claims 1-7 when the program is executed.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the vehicle line ball drive warning method according to any one of claims 1 to 7.
CN202211551915.2A 2022-12-05 2022-12-05 Vehicle line pressing driving early warning method and device, electronic equipment and storage medium Pending CN116001799A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211551915.2A CN116001799A (en) 2022-12-05 2022-12-05 Vehicle line pressing driving early warning method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211551915.2A CN116001799A (en) 2022-12-05 2022-12-05 Vehicle line pressing driving early warning method and device, electronic equipment and storage medium

Publications (1)

Publication Number Publication Date
CN116001799A true CN116001799A (en) 2023-04-25

Family

ID=86018517

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211551915.2A Pending CN116001799A (en) 2022-12-05 2022-12-05 Vehicle line pressing driving early warning method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN116001799A (en)

Similar Documents

Publication Publication Date Title
CN112965504B (en) Remote confirmation method, device and equipment based on automatic driving and storage medium
JP7332726B2 (en) Detecting Driver Attention Using Heatmaps
AU2018397461B2 (en) Multiple operating modes to expand dynamic range
US9352689B2 (en) Driver assistance apparatus capable of diagnosing vehicle parts and vehicle including the same
JP5316713B2 (en) Lane departure prevention support apparatus, lane departure prevention method, and storage medium
CN102785661B (en) Lane departure control system and lane departure control method
CN112349144B (en) Monocular vision-based vehicle collision early warning method and system
CN107203738A (en) Vehicle lane boundary alignment
US10005473B2 (en) Stereo camera, vehicle driving auxiliary device having same, and vehicle
CN104875681A (en) Dynamic vehicle-mounted camera control method based on application scenarios
CN109002797B (en) Vehicle lane change detection method, device, storage medium and computer equipment
KR20200043391A (en) Image processing, image processing method and program for image blur correction
CN109389060B (en) Vision-based vehicle surrounding collision early warning method
CN112109633A (en) Lane departure early warning method, device, equipment, vehicle and storage medium
CN114418895A (en) Driving assistance method and device, vehicle-mounted device and storage medium
CN112084232A (en) Vehicle driving risk assessment method and device based on visual field information of other target vehicles
CN112937445B (en) 360-degree vehicle safety auxiliary method and vehicle-mounted system
CN111351474B (en) Vehicle moving target detection method, device and system
CN112654998B (en) Lane line detection method and device
CN114290990A (en) Obstacle early warning system and method for vehicle A-column blind area and signal processing device
US10977503B2 (en) Fault isolation for perception systems in autonomous/active safety vehicles
CN110727269A (en) Vehicle control method and related product
CN116486351A (en) Driving early warning method, device, equipment and storage medium
CN116001799A (en) Vehicle line pressing driving early warning method and device, electronic equipment and storage medium
CN115082894A (en) Distance detection method, vehicle high beam control method, device, medium and vehicle

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination