CN112606804B - Control method and control system for active braking of vehicle - Google Patents
Control method and control system for active braking of vehicle Download PDFInfo
- Publication number
- CN112606804B CN112606804B CN202011440480.5A CN202011440480A CN112606804B CN 112606804 B CN112606804 B CN 112606804B CN 202011440480 A CN202011440480 A CN 202011440480A CN 112606804 B CN112606804 B CN 112606804B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- biological
- obstacle
- biological obstacle
- information
- 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.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60T—VEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL; ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OF VEHICLES; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES
- B60T7/00—Brake-action initiating means
- B60T7/12—Brake-action initiating means for automatic initiation; for initiation not subject to will of driver or passenger
- B60T7/22—Brake-action initiating means for automatic initiation; for initiation not subject to will of driver or passenger initiated by contact of vehicle, e.g. bumper, with an external object, e.g. another vehicle, or by means of contactless obstacle detectors mounted on the vehicle
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60T—VEHICLE BRAKE CONTROL SYSTEMS OR PARTS THEREOF; BRAKE CONTROL SYSTEMS OR PARTS THEREOF, IN GENERAL; ARRANGEMENT OF BRAKING ELEMENTS ON VEHICLES IN GENERAL; PORTABLE DEVICES FOR PREVENTING UNWANTED MOVEMENT OF VEHICLES; VEHICLE MODIFICATIONS TO FACILITATE COOLING OF BRAKES
- B60T2201/00—Particular use of vehicle brake systems; Special systems using also the brakes; Special software modules within the brake system controller
- B60T2201/02—Active or adaptive cruise control system; Distance control
- B60T2201/022—Collision avoidance systems
Abstract
The invention discloses a control method and a control system for vehicle active braking, which comprises the following steps: acquiring near-infrared images in a set range around a vehicle, and extracting coordinate information of biological obstacles; acquiring point cloud data in a set range around a vehicle, determining a biological obstacle target, and acquiring the distance between the vehicle and the biological obstacle target; segmenting point cloud data by using coordinate information of the biological obstacle to obtain an interested area, and determining the motion trend of the biological obstacle by using the interested area; when the vehicle or the environment in the set range around the vehicle meets the preset braking condition, the distance between the vehicle and the biological obstacle target and the region of interest are used for judging the movement trend of the biological obstacle, and the vehicle is controlled to execute an active braking process. The invention can predict the movement trend of the biological obstacle and realize the active braking process of the biological obstacle which appears suddenly.
Description
Technical Field
The invention relates to the technical field of automobiles, in particular to a control method and a control system for active braking of a vehicle.
Background
When a vehicle runs at night, a radar arranged on the vehicle is generally adopted to detect a biological obstacle in front of the vehicle, whether a collision risk exists is judged according to the relative speed between the vehicle and the biological obstacle and the distance between the vehicle and the biological obstacle detected by the radar, and a driver adjusts the speed and the running route of the vehicle according to the collision risk to avoid the collision between the vehicle and the biological obstacle.
However, radar detection is not sensitive to biological obstacles, and when a plurality of obstacles are detected by radar, the type of the obstacle cannot be quickly distinguished, i.e., whether the obstacle is a biological obstacle such as a pedestrian, an animal, or the like, or another non-biological obstacle such as a vehicle, a bicycle, or an electric vehicle cannot be quickly judged. The reflected wave of the pedestrian is weak, and the resolution of the radar to the pedestrian is not high.
Moreover, for moving obstacles, the spatial position at each moment is different, and it cannot be determined whether the obstacles at different moments are the same obstacle or not, and the movement trend of the obstacles cannot be predicted; so as to achieve the function of early warning the pedestrian and the vehicle which appear suddenly.
Disclosure of Invention
The embodiment of the application provides a control method and a control system for vehicle active braking, which can predict the movement trend of a biological obstacle and realize an active braking process for the biological obstacle which appears suddenly.
The invention provides a control method for active braking of a vehicle, which comprises the following steps:
acquiring a near-infrared image in a set range around a vehicle, and extracting coordinate information of a biological obstacle in the near-infrared image;
acquiring point cloud data in a set range around the vehicle, determining a biological obstacle target according to the point cloud data, the biological obstacle coordinate information and preset biological obstacle geometric feature information, and acquiring the distance between the vehicle and the biological obstacle target;
segmenting the point cloud data by utilizing the coordinate information of the biological obstacle to obtain an interested area, and determining the motion trend of the biological obstacle by utilizing the interested area;
when the vehicle or the environment in the set range around the vehicle meets a preset braking condition, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to execute an active braking process.
Preferably, the acquiring a near-infrared image within a set range around a vehicle and extracting biological obstacle coordinate information in the near-infrared image includes:
acquiring a plurality of near-infrared images, and training the plurality of near-infrared images by using a convolutional neural network model to obtain an organism barrier recognition algorithm of the near-infrared images;
acquiring a near-infrared image to be recognized, judging whether an organism barrier exists in the near-infrared image to be recognized by utilizing the organism barrier recognition algorithm, and calculating the coordinate information of the organism barrier if the organism barrier exists.
Preferably, the biological obstacle recognition algorithm for training the plurality of near-infrared images by using the convolutional neural network model to obtain the near-infrared images includes:
selecting a plurality of near-infrared images containing biological body obstacles from the plurality of near-infrared images as a positive training sample;
selecting a plurality of near-infrared images without biological body obstacles from the plurality of near-infrared images as negative training samples;
selecting biological obstacle information from frames of the plurality of near-infrared images containing biological obstacles, and selecting non-biological obstacle information from frames of the plurality of near-infrared images not containing biological obstacles;
and training a target detection model by using the organism obstacle information and the non-organism obstacle information to obtain the organism obstacle recognition algorithm.
Preferably, the segmenting the point cloud data by using the coordinate information of the biological obstacle to obtain an area of interest, and determining the motion trend of the biological obstacle by using the area of interest includes:
organizing the coordinate information of the biological obstacle and the point cloud data at the same moment into a data set at a corresponding moment;
dividing the point cloud data by using the coordinate information of the biological obstacle in the same data set to obtain an interesting area corresponding to the data set;
combining the interested regions at adjacent time and adjacent space positions into a homogeneous data set;
and determining the movement trend of the biological obstacle by using the homogeneous data set, wherein the movement trend comprises the movement direction, the movement angle and the movement speed of the biological obstacle.
Preferably, when the vehicle or the environment within the set range around the vehicle meets a preset braking condition, the method for controlling the vehicle to perform an active braking process by determining the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest includes:
the method comprises the steps of obtaining illumination intensity information in a set range around a vehicle, judging whether the illumination intensity information is smaller than a preset illumination intensity value or not, or judging whether the vehicle speed is smaller than a preset vehicle speed value or not, and if the illumination intensity information is smaller than the preset illumination intensity value or the vehicle speed is smaller than the preset vehicle speed value, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to execute an active braking process.
Preferably, the determining a movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to perform an active braking process includes:
and judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target, the vehicle speed and the region of interest, calculating the collision time of the vehicle and the biological obstacle, judging whether the collision time is less than the preset time, and if so, controlling an automatic emergency braking system of the vehicle to execute an active braking process.
The present invention also provides a control system for active braking of a vehicle, comprising:
the near-infrared camera device is used for acquiring a near-infrared image in a set range around the vehicle;
the radar device is used for acquiring point cloud data in a set range around the vehicle;
the control device is respectively in communication connection with the near-infrared camera device and the distance measuring device and is used for extracting biological obstacle coordinate information in the near-infrared image, determining a biological obstacle target according to the point cloud data, the biological obstacle coordinate information and preset biological obstacle geometric characteristic information, and acquiring the distance between the vehicle and the biological obstacle target;
and the control device is also used for segmenting the point cloud data by utilizing the coordinate information of the biological obstacle to obtain an interested area, determining the motion trend of the biological obstacle by utilizing the interested area, and judging the motion trend of the biological obstacle by utilizing the distance between the vehicle and the target of the biological obstacle and the interested area when the vehicle or the environment in the set range around the vehicle meets the preset braking condition, so as to control the vehicle to execute the active braking process.
Preferably, the control device is configured to train a plurality of near-infrared images by using a convolutional neural network model to obtain a biological obstacle recognition algorithm for the near-infrared images, acquire a near-infrared image to be recognized, judge whether there is a biological obstacle in the near-infrared image to be recognized by using the biological obstacle recognition algorithm, and calculate coordinate information of the biological obstacle if there is a biological obstacle.
Preferably, the control device is configured to select a plurality of near-infrared images including a biological obstacle from the plurality of near-infrared images, select a plurality of near-infrared images not including a biological obstacle from the plurality of near-infrared images as a positive training sample, frame-select biological obstacle information from the plurality of near-infrared images including a biological obstacle as a negative training sample, frame-select non-biological obstacle information from the plurality of near-infrared images not including a biological obstacle, and train a target detection model using the biological obstacle information and the non-biological obstacle information to obtain the biological obstacle recognition algorithm.
Preferably, the control device is configured to obtain illumination intensity information within a set range around the vehicle, determine whether the illumination intensity information is smaller than a preset illumination intensity value, or determine whether the vehicle speed is smaller than a preset vehicle speed value, and if the illumination intensity information is smaller than the preset illumination intensity value, or the vehicle speed is smaller than the preset vehicle speed value, determine a movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and control the vehicle to execute an active braking process.
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
the method comprises the steps of acquiring a near-infrared image in a set range around a vehicle, extracting coordinate information of an organism obstacle in the near-infrared image, enabling the near-infrared image to be more ready to perceive the organism obstacle and the type of the organism obstacle relative to other signal data, determining an organism obstacle target according to point cloud data acquired by a radar device, the coordinate information of the organism obstacle and preset geometrical characteristic information of the organism obstacle, and acquiring the distance between the vehicle and the organism obstacle target; the point cloud data are segmented by utilizing the coordinate information of the biological obstacle to obtain an interested area, so that the subsequent data processing amount can be reduced, and the movement trend of the biological obstacle is determined by utilizing the interested area; when the vehicle or the environment in the set range around the vehicle meets the preset braking condition, the distance between the vehicle and the biological barrier target and the region of interest are used for judging the movement trend of the biological barrier, the vehicle is controlled to execute the active braking process, the collision time and the collision possibility can be predicted in advance according to the movement trend of the biological barrier, and the active braking process is realized on the biological barrier which suddenly appears, so that the active braking process can be better executed. The invention can sense and fuse the near infrared camera device and the radar device, so that the vehicle can sense the obstacles such as pedestrians, animals and the like more quickly and accurately when driving on roads in the night and the villages.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
FIG. 1 is a flow chart of a method for controlling active braking of a vehicle according to an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of a vehicle-mounted near-infrared security system according to another embodiment of the present invention;
FIG. 3 is a schematic view of the operation of an in-vehicle NIR security system according to another embodiment of the invention;
FIG. 4 is a schematic diagram of an obstacle sensing process according to another embodiment of the present invention;
FIG. 5 is a schematic diagram of data acquisition during obstacle sensing in another embodiment of the present invention;
FIG. 6 is a flow chart of pedestrian identification via near infrared imaging in another embodiment provided by the present invention;
FIG. 7 is a schematic diagram illustrating a pedestrian algorithm training process for near-infrared images according to another embodiment of the present invention;
FIG. 8 is a schematic diagram of a pedestrian detection process by lidar in another embodiment of the present invention;
FIG. 9 is a pedestrian point cloud aggregation-like process in another embodiment provided by the present invention;
FIG. 10 is a schematic diagram illustrating a process of fusing data acquired by a near infrared camera and data acquired by a lidar in another embodiment of the present invention;
FIG. 11 is a schematic diagram illustrating a detailed procedure of fusing a near-infrared image and point cloud data according to another embodiment of the present invention;
FIG. 12 is a schematic diagram of time synchronization of a near infrared image and point cloud data according to another embodiment of the present invention;
FIG. 13 is a schematic view illustrating a movement trend determination of a biological barrier target according to another embodiment of the present invention;
FIG. 14 is a schematic diagram of a collision warning process in another embodiment provided by the present invention;
FIG. 15 is a schematic diagram of a near infrared security system intervention AEB system in another embodiment provided by the present invention;
FIG. 16 is a schematic diagram of an early warning strategy in another embodiment provided by the present invention;
FIG. 17 is a schematic representation of the general steps of an active braking process in another embodiment provided by the present invention;
FIG. 18 is a schematic diagram illustrating the steps involved in an active braking process according to another embodiment of the present invention;
FIG. 19 is a functional block diagram of a control system for active braking of a vehicle according to an embodiment of the present invention.
Detailed Description
In order to make the present application more clearly understood by those skilled in the art to which the present application pertains, the following detailed description of the present application is made with reference to the accompanying drawings by way of specific embodiments.
A method for controlling active braking of a vehicle, as shown in fig. 1, the method comprising:
and S1, acquiring a near-infrared image in a set range around the vehicle, and extracting the coordinate information of the biological obstacle in the near-infrared image.
Step S1 specifically includes:
acquiring a plurality of near-infrared images, and training the plurality of near-infrared images by using a Convolutional Neural Network (CNN) model to obtain a biological obstacle recognition algorithm of the near-infrared images.
Acquiring a near-infrared image to be recognized, judging whether a biological obstacle exists in the near-infrared image to be recognized by utilizing a biological obstacle recognition algorithm, and calculating the coordinate information of the biological obstacle if the biological obstacle exists. Here, the biological obstacle may be a human or an animal.
The near-infrared image can be acquired by the control device through the near-infrared camera device, when the near-infrared image is acquired, the positioning signal of the vehicle positioning device can be acquired simultaneously, the current moment is determined by combining with the crystal oscillator error, and the near-infrared image with the time stamp is output. The biological obstacle coordinate information may be based on the biological obstacle coordinates (pitch, yaw, roll) in the near-infrared camera body coordinate system, where pitch is a rotation around the X-axis and is also called a pitch angle, yaw is a rotation around the Y-axis and is also called a yaw angle, and roll is a rotation around the Z-axis and is also called a roll angle.
The organism obstacle recognition algorithm for training a plurality of near-infrared images by using the convolutional neural network model to obtain the near-infrared images comprises the following steps:
selecting a plurality of near-infrared images containing biological obstacles from the plurality of near-infrared images as a positive training sample;
selecting a plurality of near-infrared images without biological body obstacles from the plurality of near-infrared images as negative training samples;
selecting biological obstacle information from a plurality of near-infrared images containing biological obstacles, and selecting non-biological obstacle information from a plurality of near-infrared images not containing biological obstacles;
training a target detection model by using the biological obstacle information and the non-biological obstacle information to obtain a biological obstacle recognition algorithm; wherein, the target detection model is a YoLO (You Only Look Once) model.
In one embodiment, 2n (n >1) near-infrared images are selected, n near-infrared images containing biological obstacles and n near-infrared images not containing biological obstacles are labeled, biological obstacles are selected as positive training samples from the n near-infrared images containing biological obstacles, non-biological obstacles are selected as negative training samples from the n near-infrared images not containing biological obstacles, and a target detection model is used to obtain a biological obstacle recognition algorithm.
And after the biological obstacle is identified in the near infrared image, the biological obstacle is selected by frame injection. And (3) calculating a camera coordinate system (x, y) of the near-infrared camera device corresponding to the center of the labeling frame to obtain a camera coordinate system of the biological obstacle, and converting the camera coordinate system of the biological obstacle into a corresponding body coordinate system (pitch, yaw, roll).
Here, fx and fy are values obtained by dividing the focal length f of the near-infrared imaging apparatus by dx and dy, that is, fx and fy are f/dx and f/dy, respectively.
u0、v0The number of horizontal and vertical pixels representing the difference between the coordinates of the central pixel of the near-infrared image and the coordinates of the pixels of the dots of the image is shown. The unit of p (x, y) is millimeters.
And S2, point cloud data in a set range around the vehicle are obtained, a biological obstacle target is determined according to the point cloud data, the biological obstacle coordinate information and preset biological obstacle geometric characteristic information, and the distance between the vehicle and the biological obstacle target is obtained.
Specifically, the point cloud data in the set area range in front of the vehicle can be obtained through a radar device (which can be a laser radar), the radar device can detect obstacles in the range of 0.1-150 m in front of the vehicle, and information such as the distance, the speed and the angle of the obstacles can be obtained through the radar device.
The point cloud data can be acquired by the control device through a radar device, and when the point cloud data is acquired, the positioning signals of the vehicle positioning device can be acquired at the same time, the current moment is determined by combining with the crystal oscillator error, and the point cloud data with the time stamp is output.
After point cloud data is acquired in real time, the point cloud data is cut by using a biological obstacle coordinate (pitch, yaw, roll) output by a near-infrared camera device, the point cloud data near the biological obstacle is reserved to reduce the data processing amount, and the point cloud data is clustered by combining physical characteristics (for example, the length, width and height range of the biological obstacle) of the biological obstacle, so that a single biological obstacle target is obtained.
The distance between the vehicle and the biological obstacle target can be obtained by the radar device through TOF (Time of flight), and the following formula is shown:
wherein R is a distance between the vehicle and the biological obstacle target, f is a counting pulse repetition frequency of the radar device, n is a number of counting pulses at which an echo pulse arrives at the time of measurement, and c is a velocity of propagation in the air, and is about 3.0x108m/s。
S3, segmenting the point cloud data by using the coordinate information of the biological obstacle to obtain a region of interest (ROI), and determining the movement trend of the biological obstacle by using the ROI.
Step S3 specifically includes:
organizing the coordinate information of the biological obstacle and the point cloud data at the same moment into a data set at a corresponding moment; that is, the coordinate information of the biological obstacle obtained from the near-infrared image is time-synchronized with the point cloud data obtained from the radar device. Point cloud data and object obstacle coordinate information at different moments are obtained, the point cloud data and the object obstacle coordinate information corresponding to the same timestamp are put into the same list and packaged into an Application Programming Interface (API) library.
And segmenting point cloud data by using coordinate information of the biological obstacle in the same data set to obtain an interested area corresponding to the data set, realizing target tracking of the biological obstacle and extracting target features.
Regions of interest at adjacent time instants and adjacent spatial positions are merged into a homogeneous dataset. Specifically, in adjacent time domains, the regions of interest in adjacent spatial positions at different times are linked and combined into a homogeneous data set. The time is the time when the near-infrared image is captured by the near-infrared imaging device, the spatial position is the coordinate position corresponding to the region of interest, and the region of interest is the corresponding region of the biological obstacle.
And determining the movement trend of the biological obstacle by using the homogeneous data set, wherein the movement trend comprises the movement direction, the movement angle and the movement speed of the biological obstacle.
According to the relative position of the biological obstacle at different times and the elapsed timeFor treating
The distance between the vehicle and the biological obstacle target is calculated and acquired, and the time required for collision between the vehicle and the biological obstacle can be calculated based on the distance and the relative speed between the vehicle and the biological obstacle.
By passing
Respectively calculating the moving speed of the organism barrier in the yaw angle direction, the moving speed of the organism barrier in the roll angle direction and the moving speed of the organism barrier in the pitch angle direction, wherein,
respectively the speed of yaw angle, roll angle, pitch angle, and y1And y2Respectively being biological obstacle at adjacent time t1And t2Corresponding coordinate in the Y-axis direction, r1And r2Respectively being biological obstacle at adjacent time t1And t2Corresponding Z-axis coordinate, p1And p2Respectively being biological obstacle at adjacent time t1And t2The coordinate Δ t in the X-axis direction is t2-t 1.
And S4, when the vehicle or the environment in the set range around the vehicle meets the preset braking condition, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to execute an active braking process.
Step S4 specifically includes:
the method comprises the steps of obtaining illumination intensity information in a set range around a vehicle, judging whether the illumination intensity information is smaller than a preset illumination intensity value or not, or judging whether the vehicle speed is smaller than a preset vehicle speed value or not, if the illumination intensity information is smaller than the preset illumination intensity value or the vehicle speed is smaller than the preset vehicle speed value, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and an interested area, and controlling the vehicle to execute an active braking process.
The method for judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest and controlling the vehicle to execute an active braking process comprises the following steps:
the method comprises the steps of judging the movement trend of the biological obstacle by using the distance between a vehicle and the biological obstacle target, the vehicle speed and the region of interest, calculating the collision time of the vehicle and the biological obstacle, judging whether the collision time is less than the preset time, and controlling an AEB (automatic emergency braking) active braking system of the vehicle to execute an active braking process if the collision time is less than the preset time.
In another embodiment provided by the present invention, the control method of vehicle active braking is applied to the vehicle-mounted near-infrared safety system shown in fig. 2, wherein the AEB active braking system is referred to as AEB system for short, and the AEB system includes ESP (i.e. vehicle body electronic stability system), ESC (i.e. vehicle body stability control system), EPS (i.e. electric power steering system), VCU (i.e. vehicle control unit), TCU (i.e. automatic transmission control unit), and other ECUs (i.e. electronic control unit).
The control method mainly comprises the steps of organism obstacle sensing, data fusion, collision early warning, active braking and the like shown in figure 3.
As shown in fig. 4, the obstacle sensing includes a near-infrared camera acquiring a near-infrared image, recognizing pedestrians and animals, and performing radar detection by a laser radar to obtain point cloud information.
As shown in fig. 5, when sensing an obstacle, the near-infrared camera acquires a real-time near-infrared image, the lidar senses and outputs real-time point cloud data, and the illumination sensor outputs illumination intensity; and transmitting the image data and the illumination intensity data to the intelligent terminal through an automobile bus.
Data time acquisition: the sensor receives a vehicle-mounted GPS (Global Positioning System) signal, determines the current moment by combining a crystal oscillator error, and outputs a near-infrared image with a timestamp and point cloud data.
As shown in fig. 6, by acquiring a dataset of near-infrared images: the infrared image of the pedestrian is in the image, and the pedestrian is not in the image. And training by using a CNN convolutional neural network model to obtain a pedestrian recognition algorithm of the near-infrared image. The near-infrared camera on the vehicle transmits a near-infrared image in real time, and whether a pedestrian exists in the image is judged through a pedestrian recognition algorithm. When a pedestrian is recognized, the position of the pedestrian in the image is framed, the position of the pedestrian is calculated, and the pedestrian coordinates (pitch, yaw, roll) based on the body coordinate system are output.
As shown in fig. 7, 2n near-infrared image data sets are selected, and n near-infrared images including pedestrians and n near-infrared images not including pedestrians are labeled. And selecting pedestrians in the n near-infrared images as positive training samples, selecting non-pedestrian targets in the n near-infrared images as negative training samples, and training by using a target detection model to obtain a near-infrared image pedestrian recognition algorithm.
As shown in fig. 8, the laser radar acquires point cloud data in real time, cuts the point cloud data using a pedestrian coordinate (pitch, yaw, roll) output by the near-infrared camera, and retains the point cloud data near the pedestrian coordinate. And (3) clustering the point cloud data clusters by combining physical geometric characteristics (such as length, width and height ranges) of the pedestrians to finally obtain a single pedestrian target.
As shown in fig. 9, the pedestrian point cloud is finally obtained through point cloud filtering, point cloud clustering and pedestrian physical geometric feature constraint, and a boundary box (bounding box) is established based on the pedestrian point cloud to finally obtain the pedestrian target.
As shown in fig. 10, after the near-infrared image is processed by using the convolutional neural network model, coordinate information of obstacles such as pedestrians and animals is obtained. The laser radar detects the obstacle and obtains the information of the distance, the speed, the angle and the like of the obstacle. And carrying out time synchronization and coordinate alignment on the obstacle information obtained by two different sensors. Clustering obstacles; obstacles with the same and similar coordinates are combined into the same data set.
As shown in fig. 11, the intelligent terminal obtains the infrared image, recognizes obstacles such as pedestrians and animals through the convolutional neural network model, and outputs coordinates (yaw, roll, pitch) of the obstacles. And (3) segmenting point cloud data by combining obstacle coordinates (yaw, roll and pitch) obtained by the near-infrared image, and determining the ROI by combining point cloud information. The point cloud information and the infrared image information of the ROI are processed and analyzed.
Time synchronization: the data (point cloud, infrared image) of the same time stamp is organized into a data set at this time.
As shown in fig. 12, point cloud and image data at different times are acquired; the point cloud and the near-infrared image data values corresponding to the same timestamp are put into the same List (List), and packed into an API library, and the near-infrared image shown in fig. 12 is subjected to signal transmission through a CAN (Controller Area Network) bus.
As shown in fig. 13, target tracking: in adjacent time domains, the ROIs adjacent to each other in the spatial position at different moments are linked, and are judged as the associated ROIs, namely, are combined into the homogeneous data set.
As shown in fig. 14, the intensity of illumination is monitored by an illumination sensor, and the near-infrared security system is turned on when the illumination is insufficient. And displaying the monitored obstacles such as pedestrians, animals and the like in real time through an HMI (Human Machine Interface) system.
If the vehicle speed exceeds 60kph, the driver is reminded in the modes of icon, character, sound and the like through the HMI, and the near-infrared safety system does not participate in the AEB active braking system.
As shown in fig. 15, the functional jump between the infrared system and the AEB active braking system is illustrated:
the near-infrared safety system intervenes in an AEB active braking system of the vehicle under the working condition that the illumination is insufficient (the illumination is less than 0.1LUX) or the vehicle speed is less than 60kph, and performs collision early warning and emergency braking functions in the AEB active braking system.
As shown in fig. 16, after the vehicle is powered on, the intelligent terminal reads the data of the illumination sensor, and if the illumination is less than 0.1LUX, the intelligent terminal sends a signal of 0x 1: the near-infrared safety system is inserted into an AEB active braking system; if the illumination intensity is greater than 0.1LUX, 0x0 is sent: infrared does not intervene in the AEB active braking system.
When a pedestrian or an animal is detected, the intelligent terminal sends out 0x1, the vehicle instrument displays infrared images of the pedestrian and the animal, and the buzzer sounds 3 sounds for reminding; otherwise 0x0, the vehicle meter is not displayed.
As shown in fig. 17, the near-infrared safety system determines a pedestrian or biological obstacle closest to the vehicle in conjunction with the obstacle information. The collision time is calculated by a formula (collision time is relative distance/relative speed), and AEB active braking is performed when the collision time is less than a set value of a user.
As shown in fig. 18, the illumination intensity in front of the vehicle is obtained, whether the illumination intensity is less than 0.1LUX (LUX) or not is determined, if the illumination intensity is less than 0.1LUX, it is determined that the illumination is insufficient, and if the illumination intensity is insufficient, the image analysis result obtained through the near-infrared image is inserted into the AEB active braking system of the vehicle, so as to perform the collision warning and emergency braking functions in the AEB active braking system.
And (3) collision early warning strategy:
the control device can be an intelligent terminal, after the vehicle is powered on, the intelligent terminal reads illumination intensity data collected by an illumination sensor, and if the illumination intensity is less than 0.1LUX, an instruction 0x1 is sent: the analysis and processing result of the near-infrared image intervenes in an AEB active braking system; if the illumination intensity is greater than 0.1LUX, then command 0x0 is issued: the analysis and processing result of the near infrared image does not intervene in an AEB active braking system.
When a pedestrian or an animal is detected, the intelligent terminal sends an instruction 0x1, the vehicle instrument displays near-infrared images of the pedestrian and the animal, and the buzzer sounds for 3-sound reminding; when no pedestrian or animal is detected, the command 0x0 is sent and the vehicle meter does not display the near infrared image.
An image analysis result obtained by the near-infrared image intervenes in an AEB active braking system strategy:
the method comprises the steps that an intelligent terminal obtains vehicle speed in real time, when the vehicle speed is greater than 60kph (kilometers per hour), an instruction 0x0 is sent to an AEB active braking system, and when the AEB active braking system carries out active braking early warning, the image analysis result of a near-infrared image is not considered; when the vehicle speed is less than 60kph, a command 0x1 is sent to the AEB active braking system, and when the AEB active braking system carries out active braking early warning, the image analysis result of the near-infrared image needs to be referred to, namely: and judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target, the vehicle speed and the region of interest, calculating the collision time of the vehicle and the biological obstacle, judging whether the collision time is less than the preset time, and if so, controlling an AEB active braking system of the vehicle to execute an active braking process.
The user may set a collision time threshold, for example, the collision time threshold may be set to 1 second, 2 seconds, 3 seconds, etc., and the collision time threshold parameter affects the sensitivity of the active braking system to alarm and brake. When the collision time threshold is set to be 1 second, the system is not easy to trigger early warning of collision, and when the collision time of the vehicle and the biological obstacle target reaches the set collision time threshold, the active braking process is executed.
When the collision time threshold is set to be 3 seconds, the active braking system is easy to trigger early warning of collision, and when the collision time of the vehicle and the biological obstacle target reaches the set collision time threshold, the active braking process is executed.
By the formula: the collision time is the relative distance between the vehicle and the obstacle/the relative speed between the vehicle and the obstacle, and the collision time between the vehicle and the obstacle is calculated. And when the calculated collision time is smaller than a collision time threshold set by a user, sending a command 0x1 to the AEB active braking system to execute active braking, otherwise, sending a command 0x0 to the AEB active braking system, and not executing the active braking by the AEB active braking system.
The active braking strategy of the AEB active braking system is: the method comprises the following modes of brake pre-filling, self-adaptive brake assistance, automatic alarm braking and full-force emergency braking. Since the AEB active braking strategy is mature and widely used, it will not be described further in this application.
The present invention also provides a control system for active braking of a vehicle, as shown in fig. 19, the control system including: the system comprises a near-infrared camera device 2, a radar device 3 and a control device 1, wherein the control device 1 can be the intelligent terminal, the near-infrared camera device 2 can be the near-infrared camera, and the radar device 3 can be the laser radar.
The near-infrared camera 2 is used for acquiring a near-infrared image within a set range around the vehicle.
The radar device 3 is used to acquire point cloud data in a set range around the vehicle.
The control device 1 is respectively in communication connection with the near-infrared camera device 2 and the distance measuring device, and is used for extracting the coordinate information of the biological obstacle in the near-infrared image, determining the biological obstacle target according to the point cloud data, the coordinate information of the biological obstacle and the preset geometric characteristic information of the biological obstacle, and acquiring the distance between the vehicle and the biological obstacle target.
And the control device 1 is further configured to segment the point cloud data by using the coordinate information of the biological obstacle to obtain an area of interest, determine a movement trend of the biological obstacle by using the area of interest, and determine the movement trend of the biological obstacle by using the distance between the vehicle and the target of the biological obstacle and the area of interest when the vehicle or an environment in a set range around the vehicle meets a preset braking condition, so as to control the vehicle to perform an active braking process.
The control device 1 is used for training a plurality of near-infrared images by using a convolutional neural network model to obtain a biological obstacle recognition algorithm of the near-infrared images, acquiring a near-infrared image to be recognized, judging whether a biological obstacle exists in the near-infrared image to be recognized or not by using the biological obstacle recognition algorithm, and calculating coordinate information of the biological obstacle if the biological obstacle exists in the near-infrared image to be recognized.
The control device 1 is configured to select a plurality of near-infrared images including a biological obstacle from a plurality of near-infrared images, select a plurality of near-infrared images not including the biological obstacle from the plurality of near-infrared images as a positive training sample, select biological obstacle information from the plurality of near-infrared images including the biological obstacle as a negative training sample, select non-biological obstacle information from the plurality of near-infrared images not including the biological obstacle, and train a target detection model using the biological obstacle information and the non-biological obstacle information to obtain a biological obstacle recognition algorithm.
The control device 1 is configured to organize biological obstacle coordinate information and point cloud data at the same time into data sets at corresponding times, divide the point cloud data by using the biological obstacle coordinate information in the same data set to obtain regions of interest corresponding to the data sets, merge the regions of interest at adjacent times and adjacent spatial positions into data sets of the same type, and determine a movement trend of the biological obstacle by using the data sets of the same type, where the movement trend includes a movement direction, a movement angle, and a movement speed of the biological obstacle.
The control device 1 is used for acquiring illumination intensity information in a set range around the vehicle, judging whether the illumination intensity information is smaller than a preset illumination intensity value or not, or judging whether the vehicle speed is smaller than a preset vehicle speed value or not, and if the illumination intensity information is smaller than the preset illumination intensity value or the vehicle speed is smaller than the preset vehicle speed value, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and an interested area, and controlling the vehicle to execute an active braking process.
The control device 1 is used for judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target, the vehicle speed and the region of interest, calculating the collision time of the vehicle and the biological obstacle, judging whether the collision time is less than the preset time, and if so, controlling the AEB of the vehicle to execute an active braking process.
In summary, the invention obtains the near-infrared image in the set range around the vehicle, and extracts the coordinate information of the biological obstacle in the near-infrared image, the near-infrared image is more ready to sense the biological obstacle and the type of the biological obstacle than other signal data, and then determines the biological obstacle target according to the point cloud data obtained by the radar device 3, the coordinate information of the biological obstacle and the preset geometric characteristic information of the biological obstacle, and obtains the distance between the vehicle and the biological obstacle target; the point cloud data are segmented by utilizing the coordinate information of the biological obstacle to obtain an interested area, so that the subsequent data processing amount can be reduced, and the movement trend of the biological obstacle is determined by utilizing the interested area; when the vehicle or the environment in the set range around the vehicle meets the preset braking condition, the distance between the vehicle and the biological obstacle target and the region of interest are used for judging the movement trend of the biological obstacle, the vehicle is controlled to execute the active braking process, and the collision time and the collision possibility can be predicted in advance according to the movement trend of the biological obstacle, so that the active braking process can be better executed. The invention can sense and fuse the near infrared camera device 2 and the radar device 3, so that the vehicle can sense the obstacles such as pedestrians, animals and the like more quickly and accurately when driving on roads in the night and the villages.
Furthermore, the invention also tracks biological obstacles: associating obstacles of adjacent time domains and space domains, and continuously tracking; the obstacles such as pedestrians, animals and the like can be pre-warned and pre-moved in real time by perceiving and fusing the near-infrared camera device 2 and the radar device 3.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (9)
1. A control method for active braking of a vehicle, comprising:
acquiring a near-infrared image in a set range around a vehicle, and extracting coordinate information of a biological obstacle in the near-infrared image;
acquiring point cloud data in a set range around the vehicle, determining a biological obstacle target according to the point cloud data, the biological obstacle coordinate information and preset biological obstacle geometric feature information, and acquiring the distance between the vehicle and the biological obstacle target;
segmenting the point cloud data by utilizing the coordinate information of the biological obstacle to obtain an interested area, and determining the motion trend of the biological obstacle by utilizing the interested area;
when the vehicle or the environment in the set range around the vehicle meets a preset braking condition, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to execute an active braking process; wherein:
the method for segmenting the point cloud data by using the coordinate information of the biological obstacle to obtain an area of interest and determining the motion trend of the biological obstacle by using the area of interest comprises the following steps:
organizing the coordinate information of the biological obstacle and the point cloud data at the same moment into a data set at a corresponding moment;
dividing the point cloud data by using the coordinate information of the biological obstacle in the same data set to obtain an interesting area corresponding to the data set;
combining the interested regions at adjacent time and adjacent space positions into a homogeneous data set;
and determining the movement trend of the biological obstacle by using the homogeneous data set, wherein the movement trend comprises the movement direction, the movement angle and the movement speed of the biological obstacle.
2. The method for controlling active braking of a vehicle according to claim 1, wherein the acquiring a near-infrared image within a set range around the vehicle and extracting biological obstacle coordinate information in the near-infrared image comprises:
acquiring a plurality of near-infrared images, and training the plurality of near-infrared images by using a convolutional neural network model to obtain an organism barrier recognition algorithm of the near-infrared images;
acquiring a near-infrared image to be recognized, judging whether an organism barrier exists in the near-infrared image to be recognized by utilizing the organism barrier recognition algorithm, and calculating the coordinate information of the organism barrier if the organism barrier exists.
3. The method for controlling active braking of a vehicle according to claim 2, wherein the biological obstacle recognition algorithm for training the plurality of near-infrared images by using the convolutional neural network model to obtain the near-infrared images comprises:
selecting a plurality of near-infrared images containing biological body obstacles from the plurality of near-infrared images as a positive training sample;
selecting a plurality of near-infrared images without biological body obstacles from the plurality of near-infrared images as negative training samples;
selecting biological obstacle information from frames of the plurality of near-infrared images containing biological obstacles, and selecting non-biological obstacle information from frames of the plurality of near-infrared images not containing biological obstacles;
and training a target detection model by using the organism obstacle information and the non-organism obstacle information to obtain the organism obstacle recognition algorithm.
4. The method for controlling active braking of a vehicle according to claim 1, wherein when the vehicle or an environment within a set range around the vehicle meets a preset braking condition, the vehicle is controlled to perform an active braking process by determining a movement trend of a biological obstacle by using a distance between the vehicle and the biological obstacle target and the region of interest, and the method comprises:
the method comprises the steps of obtaining illumination intensity information in a set range around a vehicle, judging whether the illumination intensity information is smaller than a preset illumination intensity value or not, or judging whether the vehicle speed is smaller than a preset vehicle speed value or not, and if the illumination intensity information is smaller than the preset illumination intensity value or the vehicle speed is smaller than the preset vehicle speed value, judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest, and controlling the vehicle to execute an active braking process.
5. The method for controlling the active braking of the vehicle according to claim 1, wherein the step of determining the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target and the region of interest and controlling the vehicle to perform the active braking process comprises the following steps:
and judging the movement trend of the biological obstacle by using the distance between the vehicle and the biological obstacle target, the vehicle speed and the region of interest, calculating the collision time of the vehicle and the biological obstacle, judging whether the collision time is less than the preset time, and if so, controlling an automatic emergency braking system of the vehicle to execute an active braking process.
6. A control system for active braking of a vehicle, comprising:
the near-infrared camera device is used for acquiring a near-infrared image in a set range around the vehicle;
the radar device is used for acquiring point cloud data in a set range around the vehicle;
the control device is respectively in communication connection with the near-infrared camera device and the distance measuring device and is used for extracting biological obstacle coordinate information in the near-infrared image, determining a biological obstacle target according to the point cloud data, the biological obstacle coordinate information and preset biological obstacle geometric characteristic information, and acquiring the distance between the vehicle and the biological obstacle target;
the control device is further configured to segment the point cloud data by using the coordinate information of the biological obstacle to obtain an area of interest, determine a movement trend of the biological obstacle by using the area of interest, and determine the movement trend of the biological obstacle by using the distance between the vehicle and the target of the biological obstacle and the area of interest when the vehicle or an environment in a set range around the vehicle meets a preset braking condition, so as to control the vehicle to perform an active braking process;
the method for segmenting the point cloud data by using the coordinate information of the biological obstacle to obtain an area of interest and determining the motion trend of the biological obstacle by using the area of interest comprises the following steps:
organizing the coordinate information of the biological obstacle and the point cloud data at the same moment into a data set at a corresponding moment;
dividing the point cloud data by using the coordinate information of the biological obstacle in the same data set to obtain an interesting area corresponding to the data set;
combining the interested regions at adjacent time and adjacent space positions into a homogeneous data set;
and determining the movement trend of the biological obstacle by using the homogeneous data set, wherein the movement trend comprises the movement direction, the movement angle and the movement speed of the biological obstacle.
7. The control system for active braking of a vehicle according to claim 6,
the control device is used for training a plurality of near-infrared images by using a convolutional neural network model to obtain an organism barrier recognition algorithm of the near-infrared images, acquiring a near-infrared image to be recognized, judging whether an organism barrier exists in the near-infrared image to be recognized or not by using the organism barrier recognition algorithm, and calculating the coordinate information of the organism barrier if the organism barrier exists.
8. The control system for active braking of a vehicle according to claim 7,
the control device is configured to select a plurality of near-infrared images including a biological obstacle from the plurality of near-infrared images, select a plurality of near-infrared images not including the biological obstacle from the plurality of near-infrared images as a positive training sample, frame-select biological obstacle information from the plurality of near-infrared images including the biological obstacle as a negative training sample, frame-select non-biological obstacle information from the plurality of near-infrared images not including the biological obstacle, and train a target detection model using the biological obstacle information and the non-biological obstacle information to obtain the biological obstacle recognition algorithm.
9. The control system for active braking of a vehicle according to claim 6,
the control device is used for acquiring illumination intensity information in a set range around the vehicle, judging whether the illumination intensity information is smaller than a preset illumination intensity value or not, or judging whether the vehicle speed is smaller than a preset vehicle speed value or not, if the illumination intensity information is smaller than the preset illumination intensity value or the vehicle speed is smaller than the preset vehicle speed value, utilizing the distance between the vehicle and the biological obstacle target and the movement trend of the biological obstacle judged by the region of interest, and controlling the vehicle to execute an active braking process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011440480.5A CN112606804B (en) | 2020-12-08 | 2020-12-08 | Control method and control system for active braking of vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011440480.5A CN112606804B (en) | 2020-12-08 | 2020-12-08 | Control method and control system for active braking of vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112606804A CN112606804A (en) | 2021-04-06 |
| CN112606804B true CN112606804B (en) | 2022-03-29 |
Family
ID=75232642
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011440480.5A Active CN112606804B (en) | 2020-12-08 | 2020-12-08 | Control method and control system for active braking of vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112606804B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113283311A (en) * | 2021-05-07 | 2021-08-20 | 恒大新能源汽车投资控股集团有限公司 | Animal driving-away method and device for vehicle |
| CN114446092B (en) * | 2022-01-19 | 2022-12-27 | 无锡学院 | S-shaped road simulated obstacle early warning method based on three-dimensional camera networking |
| CN115202350B (en) * | 2022-07-15 | 2023-06-09 | 盐城工学院 | Automatic conveying system of AGV dolly |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2077596A1 (en) * | 1992-09-04 | 1994-03-05 | Michel Nault | Vehicle safety device |
| CN102508246A (en) * | 2011-10-13 | 2012-06-20 | 吉林大学 | Method for detecting and tracking obstacles in front of vehicle |
| CN102930555A (en) * | 2011-08-11 | 2013-02-13 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for tracking interested areas in ultrasonic pictures |
| GB201321179D0 (en) * | 2013-12-02 | 2014-01-15 | Daimler Ag | Device and method to derive optimum speed of vehicle for comfortable ride |
| CN106599832A (en) * | 2016-12-09 | 2017-04-26 | 重庆邮电大学 | Method for detecting and recognizing various types of obstacles based on convolution neural network |
| CN110412986A (en) * | 2019-08-19 | 2019-11-05 | 中车株洲电力机车有限公司 | A kind of vehicle barrier detection method and system |
| CN110816527A (en) * | 2019-11-27 | 2020-02-21 | 奇瑞汽车股份有限公司 | Vehicle-mounted night vision safety method and system |
-
2020
- 2020-12-08 CN CN202011440480.5A patent/CN112606804B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2077596A1 (en) * | 1992-09-04 | 1994-03-05 | Michel Nault | Vehicle safety device |
| CN102930555A (en) * | 2011-08-11 | 2013-02-13 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for tracking interested areas in ultrasonic pictures |
| CN102508246A (en) * | 2011-10-13 | 2012-06-20 | 吉林大学 | Method for detecting and tracking obstacles in front of vehicle |
| GB201321179D0 (en) * | 2013-12-02 | 2014-01-15 | Daimler Ag | Device and method to derive optimum speed of vehicle for comfortable ride |
| CN106599832A (en) * | 2016-12-09 | 2017-04-26 | 重庆邮电大学 | Method for detecting and recognizing various types of obstacles based on convolution neural network |
| CN110412986A (en) * | 2019-08-19 | 2019-11-05 | 中车株洲电力机车有限公司 | A kind of vehicle barrier detection method and system |
| CN110816527A (en) * | 2019-11-27 | 2020-02-21 | 奇瑞汽车股份有限公司 | Vehicle-mounted night vision safety method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112606804A (en) | 2021-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112606804B (en) | Control method and control system for active braking of vehicle | |
| CN109927719B (en) | Auxiliary driving method and system based on obstacle trajectory prediction | |
| CN106485233B (en) | Method and device for detecting travelable area and electronic equipment | |
| EP2559016B1 (en) | Video based intelligent vehicle control system | |
| CN102765365B (en) | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision | |
| CN110356325B (en) | Urban traffic passenger vehicle blind area early warning system | |
| JP5410730B2 (en) | Automobile external recognition device | |
| CN108263279A (en) | The pedestrian detection and pedestrian impact avoiding device and method integrated based on sensor | |
| CN106503653A (en) | Area marking method, device and electronic equipment | |
| CN104260723B (en) | A kind of front vehicle motion state tracking prediction meanss and Forecasting Methodology | |
| CN109871787B (en) | Obstacle detection method and device | |
| CN108960035A (en) | The method and system of usage scenario detection pattern object of classification in perception scene figure | |
| EP3286056B1 (en) | System and method for a full lane change aid system with augmented reality technology | |
| US20230184560A1 (en) | Visual interface display method and apparatus, electronic device, and storage medium | |
| CN102792314A (en) | Cross traffic collision alert system | |
| CN108944920A (en) | It is generated in road vehicle application program and using the method and system of perception scene figure | |
| Ruder et al. | Highway lane change assistant | |
| EP2899669A1 (en) | Lane relative position estimation method and system for driver assistance systems | |
| CN110544368B (en) | Fatigue driving augmented reality early warning device and early warning method | |
| WO2021062596A1 (en) | Systems and methods for predicting a vehicle trajectory | |
| CN108944919A (en) | The method and system with the perception scene figure of focal zone is generated for motor vehicles | |
| CN108021899A (en) | Vehicle intelligent front truck anti-collision early warning method based on binocular camera | |
| CN110371018A (en) | Improve vehicle behavior using the information of other vehicle car lights | |
| CN113313182B (en) | Target identification method and terminal based on radar and video fusion | |
| CN111216718A (en) | Collision avoidance method, device and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |