CN111081061B - Collision early warning method and device - Google Patents
Collision early warning method and device Download PDFInfo
- Publication number
- CN111081061B CN111081061B CN201811231736.4A CN201811231736A CN111081061B CN 111081061 B CN111081061 B CN 111081061B CN 201811231736 A CN201811231736 A CN 201811231736A CN 111081061 B CN111081061 B CN 111081061B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- alarm
- target
- time
- obtaining
- 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
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
Abstract
The invention discloses a collision early warning method and device, and belongs to the field of intelligent control. The method comprises the following steps: acquiring state information, inputting the state information into a collision early warning model, and outputting action information corresponding to the state information; when the action information indicates alarming, alarming is carried out; when a target instruction is received in a target time period after alarming, acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle; and updating the collision early warning model according to the time difference and the speed information of the vehicle. The invention considers the driving habit of the driver and updates the model according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm opportunity, is more in line with the expectation of the driver, and the accuracy of collision early warning is improved.
Description
Technical Field
The invention relates to the field of intelligent control, in particular to a collision early warning method and device.
Background
Vehicle collision is a common traffic accident, and on a congested urban road, if the distance between a vehicle and a front target object is too close, a collision event is easy to occur, so that a collision early warning method is urgently needed to warn a driver when a potential collision danger exists.
At present, a collision early warning method generally measures a distance between a vehicle and a front target object and a vehicle speed of the vehicle, and calculates collision time according to the measured distance and the vehicle speed, wherein the collision time is time required for the vehicle to collide with the front target object; and alarming the driver according to the comparison result of the distance and the safety vehicle distance threshold value and the comparison result of the collision time and the collision time threshold value.
According to the technology, the collision early warning is realized by setting the alarm thresholds such as the safe vehicle distance threshold, the collision time threshold and the like, the alarm time for all drivers is the same, but the recognition degree of the alarm time is different for different drivers due to the difference of driving habits, the setting of the alarm threshold is difficult to meet the driving habits of different drivers, and the accuracy of the collision early warning is low.
Disclosure of Invention
The embodiment of the invention provides a collision early warning method and device, which can solve the problem of low accuracy of collision early warning in the related art. The technical scheme is as follows:
in a first aspect, a collision warning method is provided, where the method includes:
acquiring state information, inputting the state information into a collision early warning model, and outputting action information corresponding to the state information;
when the action information indicates alarming, alarming;
when a target instruction is received in a target time period after alarming, acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle, wherein the target instruction is triggered by the response operation of a driver to the alarming;
and updating the collision early warning model according to the time difference and the speed information of the vehicle.
In one possible implementation manner, the obtaining the status information includes:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, the target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and obtaining the time required by the collision between the vehicle and the front target object according to the relative distance and the relative speed.
In one possible implementation manner, the obtaining the relative distance according to internal reference and external reference of an imaging device mounted on the vehicle and coordinate information of a target point imaged in an image captured by the imaging device includes:
applying the following formula to obtain the relative distance:
wherein Δ s is the relative distance, HCIs the mounting height of the image pickup apparatus, y is the ordinate of the target point imaged in the image taken by the image pickup apparatus, v0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRIs the tilt angle of the camera device.
In one possible implementation, the obtaining the relative speed of the vehicle and the front target according to the focal length, the relative distance, the actual dimension of the front target, a target time interval, and a dimension variation of the front target imaged in the image in the target time interval includes:
applying the following formula to obtain the relative velocity:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, wgΔ t is the target time interval, which is the actual dimension of the vehicle.
In one possible implementation, the action information includes a weight of alarm and a weight of no alarm;
correspondingly, when the action information indicates an alarm, the alarming comprises:
and when the weight value of the alarm in the action information is greater than the weight value of the non-alarm, alarming.
In one possible implementation manner, the updating the collision warning model according to the time difference and the speed information of the vehicle includes:
obtaining a return value of the alarm according to the time difference and the speed information of the vehicle, wherein the return value is used for evaluating the recognition degree of the driver to the alarm;
and updating the collision early warning model based on the return value.
In one possible implementation, the speed information of the vehicle is an acceleration of the vehicle,
correspondingly, the obtaining of the return value of the alarm according to the time difference and the speed information of the vehicle includes:
and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
In one possible implementation, the speed information of the vehicle includes a first speed of the vehicle at the warning time point, and a second speed of the vehicle at the target instruction reception time point;
correspondingly, the obtaining of the return value of the alarm according to the time difference and the speed information of the vehicle includes:
obtaining the acceleration of the vehicle according to the first speed, the second speed and the time difference;
and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
In one possible implementation manner, the obtaining the reward value of the present alarm according to the time difference and the acceleration of the vehicle includes:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the reward function is:
wherein R istIs the reported value, tbIs the receiving time point, t, of the target instructionwIs the alarm time point, t0The reaction time of the driver to the warning is denoted as a, and the acceleration of the vehicle is denoted as a.
In one possible implementation manner, after the alarming when the action information indicates the alarming, the method further includes:
and when the target instruction is not received in the target time period after alarming, assigning the time difference as the time difference corresponding to the target time period, and executing the steps of acquiring speed information and updating a collision early warning model.
In one possible implementation manner, after the outputting the action information corresponding to the state information, the method further includes:
and when the action information indicates that no alarm is given and the target instruction is received in a target time period, assigning the time difference as a target numerical value, and executing the steps of acquiring speed information and updating a collision early warning model.
In a second aspect, there is provided a collision warning apparatus, the apparatus comprising:
the acquisition module is used for acquiring state information, inputting the state information into the collision early warning model and outputting action information corresponding to the state information;
the alarm module is used for alarming when the action information indicates alarming;
the acquisition module is further used for acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle when the target instruction is received in the target time period after alarming, wherein the target instruction is triggered by the response operation of a driver to the alarming;
and the updating module is used for updating the collision early warning model according to the time difference and the speed information of the vehicle.
In one possible implementation, the obtaining module is configured to:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, the target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and obtaining the time required by the collision between the vehicle and the front target object according to the relative distance and the relative speed.
In one possible implementation, the obtaining module is configured to:
applying the following formula to obtain the relative distance:
wherein Δ s is the relative distance, HCIs the mounting height of the image pickup apparatus, y is the ordinate of the target point imaged in the image taken by the image pickup apparatus, v0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRIs the tilt angle of the camera device.
In one possible implementation, the obtaining module is configured to:
applying the following formula to obtain the relative velocity:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, wgΔ t is the target time interval, which is the actual dimension of the vehicle.
In one possible implementation, the action information includes a weight of alarm and a weight of no alarm;
correspondingly, the alarm module is used for alarming when the alarm weight in the action information is greater than the non-alarm weight.
In a possible implementation manner, the updating module is configured to obtain a return value of the current alarm according to the time difference and the speed information of the vehicle, where the return value is used to evaluate an approval degree of the driver for the current alarm; and updating the collision early warning model based on the return value.
In a possible implementation manner, the speed information of the vehicle is an acceleration of the vehicle, and accordingly, the updating module is configured to obtain a return value of the present alarm according to the time difference and the acceleration of the vehicle.
In one possible implementation, the speed information of the vehicle includes a first speed of the vehicle at the warning time point, and a second speed of the vehicle at the target instruction reception time point;
correspondingly, the updating module is used for obtaining the acceleration of the vehicle according to the first speed, the second speed and the time difference; and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
In one possible implementation, the update module is configured to:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the reward function is:
wherein R istIs the reported value, tbIs the receiving time point, t, of the target instructionwIs the alarm time point, t0The reaction time of the driver to the warning is denoted as a, and the acceleration of the vehicle is denoted as a.
In a possible implementation manner, the obtaining module is further configured to assign the time difference to a time difference corresponding to the target time period and execute a step of obtaining speed information when the target instruction is not received within the target time period after the alarm;
the updating module is further configured to perform the step of updating the collision warning model.
In a possible implementation manner, the obtaining module is further configured to assign the time difference as a target numerical value and execute the step of obtaining the speed information when the action information indicates that no alarm is given and the target instruction is received within a target time period;
the updating module is further configured to perform the step of updating the collision warning model.
In a third aspect, a computer device is provided, comprising a processor and a memory; the memory is used for storing at least one instruction; the processor is configured to execute at least one instruction stored in the memory to implement the method steps of any one of the implementation manners of the first aspect.
In a fourth aspect, a computer-readable storage medium is provided, in which at least one instruction is stored, and the at least one instruction, when executed by a processor, implements the method steps of any one of the implementations of the first aspect.
The technical scheme provided by the embodiment of the invention has the beneficial effects that at least:
the method comprises the steps of obtaining real-time state information of a vehicle, obtaining corresponding action information by using a collision early warning model, and updating the collision early warning model according to the time difference between the warning time point and the time point of response operation of a driver and the speed information of the vehicle after warning after each warning. The scheme considers the driving habit of the driver and updates the model according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm opportunity, is more in line with the expectation of the driver, and the accuracy of collision early warning is improved.
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 will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart of a collision warning method according to an embodiment of the present invention;
fig. 2 is a flowchart of a collision warning method according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a monocular distance measuring method according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a reinforcement learning process according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a collision warning apparatus according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a computer device 600 according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
Fig. 1 is a flowchart of a collision warning method according to an embodiment of the present invention. Referring to fig. 1, the method includes:
101. the method comprises the steps of obtaining state information, inputting the state information into a collision early warning model, and outputting action information corresponding to the state information, wherein the state information is used for describing at least one of the relative motion condition of a vehicle and a front target object, the running state of the vehicle and the current road surface condition, the collision early warning model is used for outputting the action information according to the input state information, and the action information is used for indicating whether to give an alarm or not.
102. And when the action information indicates an alarm, alarming.
103. And when a target instruction is received in a target time period after alarming, acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle, wherein the target instruction is triggered by the response operation of the driver to the alarming.
104. And updating the collision early warning model according to the time difference and the speed information of the vehicle.
According to the method provided by the embodiment of the invention, the real-time state information of the vehicle is obtained, the corresponding action information is obtained by utilizing the collision early warning model, and after each alarm, the collision early warning model is updated according to the time difference between the alarm time point and the time point of response operation of the driver and the speed information of the vehicle after the alarm. The scheme considers the driving habit of the driver and updates the model according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm opportunity, is more in line with the expectation of the driver, and the accuracy of collision early warning is improved.
In one possible implementation, the obtaining the status information includes:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, a target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and obtaining the time required by the collision of the vehicle and the front target object according to the relative distance and the relative speed.
In one possible implementation, the obtaining a relative distance according to internal reference and external reference of the camera device mounted on the vehicle and coordinate information of a target point imaged in an image captured by the camera device includes:
the relative distance is obtained using the following equation:
where Δ s is the relative distance, HCY is the ordinate of the target point imaged in the image captured by the image capture apparatus, v is the mounting height of the image capture apparatus0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRFor tilting of the cameraAnd (4) an angle.
In one possible implementation, the obtaining the relative speed of the vehicle and the front target according to the focal length, the relative distance, the actual dimension of the front target, a target time interval, and a dimension variation of the front target imaged in the image in the target time interval includes:
the relative velocity is obtained using the following equation:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, and w is the focal lengthgΔ t is the target time interval, which is the actual scale of the vehicle.
In one possible implementation, the action information includes a weight of the alarm and a weight of the non-alarm;
correspondingly, when the action information indicates an alarm, the alarm is given, and the method comprises the following steps:
and when the weight value of the alarm in the action information is greater than the weight value of the non-alarm, alarming.
In one possible implementation, the updating the collision warning model according to the time difference and the speed information of the vehicle includes:
obtaining a return value of the alarm according to the time difference and the speed information of the vehicle, wherein the return value is used for evaluating the recognition degree of the driver to the alarm;
and updating the collision early warning model based on the return value.
In one possible implementation, the speed information of the vehicle is an acceleration of the vehicle,
correspondingly, obtaining the return value of the alarm according to the time difference and the speed information of the vehicle includes:
and obtaining the return value of the alarm at this time according to the time difference and the acceleration of the vehicle.
In one possible implementation, the speed information of the vehicle includes a first speed of the vehicle at the warning time point, and a second speed of the vehicle at the target command reception time point;
correspondingly, obtaining the return value of the alarm according to the time difference and the speed information of the vehicle includes:
obtaining the acceleration of the vehicle according to the first speed, the second speed and the time difference;
and obtaining the return value of the alarm at this time according to the time difference and the acceleration of the vehicle.
In one possible implementation manner, the obtaining the reward value of the alarm according to the time difference and the acceleration of the vehicle includes:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the return function is:
wherein R istIs the reported value, tbIs the reception time point, t, of the target instructionwFor the alarm time point, t0The reaction time of the driver to the warning is denoted as a as the acceleration of the vehicle.
In one possible implementation, when the action information indicates an alarm, after the alarm is performed, the method further includes:
and when the target instruction is not received in the target time period after alarming, assigning the time difference as the time difference corresponding to the target time period, and executing the steps of acquiring the speed information and updating the collision early warning model.
In a possible implementation manner, after the outputting the action information corresponding to the state information, the method further includes:
and when the action information indicates that no alarm is given and the target instruction is received in the target time period, assigning the time difference as a target numerical value, and executing the steps of acquiring the speed information and updating the collision early warning model.
All the above-mentioned optional technical solutions can be combined arbitrarily to form the optional embodiments of the present invention, and are not described herein again.
Fig. 2 is a flowchart of a collision warning method according to an embodiment of the present invention. Referring to fig. 2, the method includes:
201. state information describing at least one of a relative movement situation of the vehicle and a preceding object, a driving state of the vehicle, and a current road surface condition is acquired.
In the embodiment of the invention, the state information can be acquired in real time during the running process of the vehicle. In one possible implementation, the state information includes a time required for the vehicle to collide with the preceding object, a relative distance of the vehicle to the preceding object, speed information of the vehicle, and a road surface adhesion coefficient. The time required by the collision of the vehicle and the front target object and the relative distance between the vehicle and the front target object are used for describing the relative motion condition of the vehicle and the front target object, the speed information of the vehicle is used for describing the running state of the vehicle, and the road surface adhesion coefficient is used for describing the current road surface condition.
In one possible implementation manner, for the speed information of the vehicle, the speed information of the vehicle may include a speed of the vehicle and an acceleration of the vehicle, and the vehicle may measure the speed of the vehicle through a vehicle speed sensor, obtain the acceleration of the vehicle according to a speed change condition of the vehicle in a period of time, or directly measure the acceleration of the vehicle through an acceleration sensor. Aiming at the road adhesion coefficient, the vehicle can respectively acquire the real-time rainfall and the real-time temperature of the road through the rainfall sensor and the temperature sensor, and then the road adhesion coefficient corresponding to the real-time rainfall and the real-time temperature is obtained according to the real-time rainfall and the real-time temperature and the corresponding relation between the pre-calibrated rainfall and temperature and the road adhesion coefficient.
For the relative distance between the vehicle and the front target object, the vehicle can obtain the relative distance according to internal parameters of the camera device installed on the vehicle, the external parameters and coordinate information of a target point imaged in an image shot by the camera device, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitch angle of the camera device.
Wherein the image pickup apparatus may be a monocular camera. Referring to fig. 3, fig. 3 is a schematic diagram of a monocular distance measuring method according to an embodiment of the present invention, as shown in fig. 3, the camera is taken as the origin of the coordinate system, and the internal reference of the camera is known (including the focal length f and the image center point (u)0,v0) Lens distortion coefficient), assuming that the camera is mounted in the horizontal direction and has no deflection, the external parameters to be calibrated are as follows: mounting height H of cameraCAngle of pitch alphaR)。
The target point needing distance measurement is P, the imaging coordinate of P in the image is y, and the included angle between the connecting line from P to the central point of the camera and the road surface is alphaSThen, there are:
αS＝αC+αR (1)
the perspective imaging relationship is as follows:
and the ranging result (relative distance) is:
Δs＝HC×tan(αS) (3)
combining equations (1) to (3), the vehicle may apply equation (4) below to obtain the relative distance:
where Δ s is the relative distance, HCY is the ordinate of the target point imaged in the image captured by the image capture apparatus, v is the mounting height of the image capture apparatus0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRIs the tilt angle of the image pickup apparatus.
For the time required by the collision between the vehicle and the front target object, the vehicle can obtain the relative speed between the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, the target time interval and the scale variation of the front target object in the image in the target time interval; and obtaining the time required by the collision of the vehicle and the front target object according to the relative distance and the relative speed. Wherein, the actual dimension of the front target object may refer to the actual width of the front target object.
The relative speed is calculated according to the relative distance between the vehicle and the front target object and the scale variation of the front target object in the image, and the actual scale of the front target object is assumed to be wgThe scale of imaging in the image is w, human eyes observe an object, and the size of the object is always large and small, namely w is inversely proportional to deltas; the images of objects with large physical dimensions, i.e. w and w, are also largegIs in direct proportion.
Assuming that the actual dimensions of the front target are known (w)g2 m), the interval time of the scale change calculation is Δ t, the scale change amount within the time Δ t is Δ w, and the vehicle can apply the following formula (5) according to the imaging scale of the front target object to obtain the relative speed:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, and w is the focal lengthgΔ t is the target time interval, which is the actual scale of the vehicle.
Further, the vehicle may calculate the time to collision TTC based on the relative distance and the relative speed, applying equation (6) below:
TTC＝Δs/Δv (6)
in the related art, the alarm is determined to be carried out by comparing the magnitude relation between the real-time measurement value and the alarm threshold, for example, comparing the Time To Collision (TTC) and the time to collision threshold, so that the alarm threshold needs to be set, for example, the time to collision threshold is generally set between 1.5 seconds and 2.7 seconds, specifically, the alarm threshold is adjusted according to the different safe vehicle distance between the vehicle and the front target object, and the alarm is more sensitive when the safe vehicle distance is smaller and the alarm threshold is larger. It is difficult to set a reasonable alarm threshold value for different road conditions, and further considering different driving habits of different drivers, the driving experience is high, and the setting of the alarm threshold value is more difficult. The efficiency and accuracy of collision early warning by setting an alarm threshold are low. In addition, rule-based alarm policies, when the scene is very variable, the created rules cannot guarantee sufficient coverage, and when more new rules are added, the old rules must be cancelled or overwritten, which makes the alarm system vulnerable.
According to the technical scheme provided by the embodiment of the invention, the alarm threshold value is not required to be set, and a reinforcement learning mode is used for analyzing the influence brought by the operation of the driver after the alarm, so that the alarm strategy is optimized. The reinforcement learning means learning an optimal strategy through exploring the environment, so that the ontology can make a decision or an action according to the current state in a specific environment, and the maximum return is obtained.
Referring to fig. 4, fig. 4 is a schematic diagram of a reinforcement learning process provided by an embodiment of the present invention, and as shown in fig. 4, the reinforcement learning includes an environment part and a collision warning model, which may be a multi-layer neural network model, wherein the environment part includes front target perception, vehicle control and driver operation. The specific names involved in reinforcement learning are defined as follows:
policy (mapping of environmental state to action): the multilayer neural network model outputs corresponding action information according to the state information;
a state set S: { TTC, Δ s, v, μ, α }, wherein TTC is alarm time, Δ s is relative distance, v is vehicle speed, μ is road adhesion coefficient, and α is vehicle acceleration;
action set A: { alarm (weight), no alarm (weight) }, for example, the alarm weight may be 1, the no alarm weight is 0, of course, 1 and 0 are only an example of the weight, and in fact, the weight may have a larger numerical range;
the optimal strategy is as follows: the return obtained under the strategy is maximum, and the braking opportunity is most correct;
probability of state transition: if the brake is not reminded in time, the TTC with a certain probability is increased or decreased, and the state transition probability can be applied to the reinforcement learning process.
Specifically, the specific process of performing collision warning by using the collision warning model refers to step 202 and the subsequent steps.
202. And inputting the state information into a collision early warning model, and outputting action information corresponding to the state information, wherein the collision early warning model is used for outputting the action information according to the input state information, and the action information is used for indicating whether to alarm or not.
In one possible implementation, the obtaining process of the collision warning model includes: for example, the vehicle may maintain a database in which historical state information of the vehicle and corresponding actual action information are recorded, that is, state information of the vehicle at each historical time and actual actions performed by the vehicle under each state information, including alarming and non-alarming. A plurality of historical state information is extracted from the database to serve as sample state information, actual action information corresponding to the plurality of historical state information is obtained, and then a collision early warning model is obtained through training based on the plurality of historical state information and the corresponding actual action information, and the collision early warning model obtained through training has the capacity of outputting the action information according to the input state information.
In the embodiment of the present invention, in terms of the reinforcement learning process shown in fig. 4, after the vehicle acquires the state information, the acquired state information may be input into the collision early warning model in real time, and the output result of the collision early warning model is used as the action information corresponding to the state information.
203. And when the action information indicates an alarm, alarming.
In one possible implementation, the action information may include both the alarm and non-alarm actions, and the weight of each action, i.e., the weight of the alarm and the weight of the non-alarm. The vehicle can determine whether to alarm according to the weight values of the two actions. For example, when the weight value of the alarm in the action information is greater than the weight value of the non-alarm, the alarm is performed. The alarm mode can be the alarm sound, or the control of the alarm lamp to flash, or the seat to shake for alarm.
204. And when a target instruction is received in a target time period after alarming, acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle, wherein the target instruction is triggered by the response operation of the driver to the alarming.
In the embodiment of the invention, if the vehicle gives an alarm according to the action information, the driver can judge whether the collision danger exists at present according to the actual situation, and if the driver considers that the collision danger exists, the driver can respond to the alarm according to the alarm and trigger the vehicle to receive the target instruction. For example, the response operation may be a braking operation, and the target command may be a braking command.
In a possible implementation manner, the vehicle may be provided with a target time period, and after each warning, if the driver performs a response operation within the target time period after the warning, the vehicle may receive the target instruction, and at this time, the vehicle may acquire the receiving time point t of the target instructionbAnd alarm time twThe time difference between them, i.e. tb-twAnd the receiving time point of the target instruction is also the time point of the response operation of the driver to the alarm.
In addition, since the speed of the vehicle changes if the driver performs a response operation to the present warning after the warning is taken into consideration, the speed information of the vehicle can be acquired in addition to the time difference after the warning. The speed information of the vehicle may be an acceleration of the vehicle, or the speed information of the vehicle includes a first speed of the vehicle at the warning time point and a second speed of the vehicle at the target command reception time point. The process of acquiring the speed information of the vehicle is described in step 201, and is not described herein.
205. And when the target instruction is not received in the target time period after the alarm, assigning the time difference as the time difference corresponding to the target time period, and acquiring the speed information of the vehicle.
In the embodiment of the invention, if the driver considers that no collision danger exists after the vehicle gives an alarm according to the action information, the alarm can be ignored and the response operation is not carried out on the alarm, if the driver does not carry out the response operation on the alarm in the target time period after the alarm, the vehicle does not receive the target instruction, and at the moment, the vehicle can obtain the time difference delta t corresponding to the target time periodrAssigning the time difference to the time difference between the reception time point of the target instruction and the alarm time point, i.e. taking tb-tw＝ΔtrE.g. the Δ trAnd may be 3 seconds.
In addition, after the warning is given, even if the driver does not perform a response operation to the warning, that is, the driver does not perform a braking operation, the driver may perform another operation, such as an acceleration operation, to change the acceleration of the vehicle, and the vehicle may still perform the step of acquiring the speed information of the vehicle.
206. And when the action information indicates that no alarm is given and a target instruction is received in a target time period, assigning the time difference as a target numerical value and acquiring the speed information of the vehicle.
In the embodiment of the invention, whether to alarm is determined according to the weights of the two actions in the step 203, and when the weight of the alarm in the action information is smaller than the weight of the alarm, the alarm is not performed. If the vehicle does not give an alarm, but the driver may think that there is a collision risk at present, a response operation may still be performed so that the vehicle may receive the target instruction, at which point the vehicle may assign a target value to the time difference between the reception time point of the target instruction and the alarm time point, and perform the step of acquiring the speed information of the vehicle. The target value may be an empirical value preset for a case where the target command is received without an alarm, and the target value may be in the same order of magnitude as the time difference between step 204 and step 205.
It should be noted that step 205 and step 206 are optional steps, and after step 202 is executed, if the situations of step 203 and step 204 occur, the step 205 and step 206 do not need to be executed.
207. And obtaining a return value of the alarm according to the time difference and the speed information of the vehicle, wherein the return value is used for evaluating the recognition degree of the driver to the alarm.
In the embodiment of the invention, the alarming time is related to the relative distance and the relative speed between the vehicle and the front target object and the speed and the acceleration of the vehicle. Different drivers have different recognition degrees of alarm time due to the difference of driving experience and driving habits, the early alarm can be regarded as false alarm, and the late alarm can lead the drivers to feel dangerous. Therefore, in order to improve the driving experience, after collision early warning is carried out each time, a return value can be obtained according to the difference between the model output and the actual operation of the driver, and then the warning opportunity is corrected by utilizing the return value. The higher the return value is, the higher the recognition degree of the driver to the alarm is, namely, the closer the opportunity of the alarm is to the expectation of the driver; the smaller the return value is, the smaller the recognition degree of the driver for the alarm is, that is, the more the opportunity of the alarm deviates from the expectation of the driver.
For the case where the speed information of the vehicle is the acceleration of the vehicle, the obtaining process of the reported value may include: and obtaining the return value of the alarm at this time according to the time difference and the acceleration of the vehicle.
For the case where the speed information of the vehicle includes the first speed and the second speed, the obtaining process of the return value may include: the acceleration of the vehicle is obtained from the first speed, the second speed, and the time difference, for example, by dividing the speed difference between the first speed and the second speed by the time difference. And further, obtaining a return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
In the embodiment of the invention, referring to step 204 and step 205, taking the response operation of the driver to the alarm as the brake operation as an example, after collision early warning is performed each time, the vehicle can acquire the alarm time point twAnd twThe latest time t when the driver steps on the brakebDifference (t) ofb-tw) (ii) a At twTime point certain time DeltatrIf there is no brake in the range, then get (t)b-tw)＝Δtr. In addition, since a certain response time is required after the driver hears the warning sound, the response time t of the driver to the warning sound is also required to be considered0. E.g. t0T can be set according to the test result by testing the reaction time of a plurality of drivers in advance0E.g. taking the average of the test results as t0Of course, t0Or may be an empirical value, such as 0.5 seconds, for t0The value of (A) is not limited.
In addition, the vehicle can judge whether to brake and brake amplitude according to the acceleration a for evaluating the feedback of the driver to the collision alarm, wherein a >0 indicates that the vehicle is accelerating, a <0 indicates that the vehicle is decelerating, the alarm needs to be rewarded when the vehicle decelerates, and the alarm is punished if the vehicle continues to accelerate.
After the alarm is sent out, the vehicle can obtain a return value of the alarm at this time according to the time difference and the acceleration of the vehicle, and specifically, the return value of the alarm at this time is obtained according to the time difference, the acceleration of the vehicle and a return function; wherein the return function is represented by the following formula (7):
wherein R istIs the reported value, tbIs the reception time point, t, of the target instructionwAs alarm time point, t0The response time of the driver to the warning is a, the acceleration of the vehicle.
As can be seen from equation (7), the influence factors of the reported value include the time difference and the acceleration between the receiving time point and the alarming time point of the target instruction, after the vehicle gives an alarm, the faster the driver responds to the alarm (such as braking), and the larger the braking amplitude is, the larger t isb-twThe smaller the a is, the larger the return value is, and the larger the recognition degree of the driver on the current alarm is; the slower the driver responds or does not respond to the alarm, tb-twThe larger the return value is, the smaller the acceptance of the driver to the alarm is.
208. And updating the collision early warning model based on the return value.
In the embodiment of the present invention, after the vehicle acquires the return value, the return value may be input to the collision early warning model, an error between the return value and the output result of this time, that is, an error between the return value and the weight value of the alarm in the action information in step 202, is calculated by the collision early warning model, learning is performed in a reinforcement learning manner, and the collision early warning model is updated by an error back propagation algorithm, so that an alarm strategy generated by the model gradually conforms to an expectation, that is, the result output by the model conforms to the expectation of the driver. And the updated collision early warning model is used for performing collision early warning according to the state information acquired next time.
It should be noted that, the above steps 207 to 208 are one possible implementation manner of updating the collision warning model according to the time difference and the speed information of the vehicle. Because the driving habit of the driver is considered, the model is updated according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm time, is more in line with the expectation of the driver.
The technical scheme automatically learns the proper alarming time from the perspective of driving experience based on reinforcement learning. The working principle is that after collision early warning is carried out each time, the time difference between the time point when the driver steps on the brake and the alarming time point is calculated, and the braking amplitude is judged according to the deceleration condition of the vehicle. If the driver does not have braking behavior, the alarm is considered to be too sensitive, and punishment is carried out on the model. If the driver has braking behavior, the alarm is considered to be effective, and the model is rewarded. In addition, for the condition that the driver has brake action but does not give an alarm at this time, the situation that the alarm is missed is considered, punishment is carried out on the model, and early warning can be carried out on the condition that the driver enters the similar condition next time. Compared with the mode of setting the alarm threshold value in the related technology, the technical scheme provided by the embodiment of the invention reduces the logic design and threshold value setting of the model, and better accords with the user experience of a driver; whether the alarm accords with the driving experience or not needs to be confirmed by the operation of the driver after the alarm belongs to the problem of delayed reward, and the method is suitable for reinforcement learning. Compared with the traditional artificial rule alarm strategy, the design of the alarm threshold value is difficult to meet the experience difference of different drivers, and the design of the threshold value is relatively difficult for some complex working conditions. In addition, by designing a return function, the reward or punishment of the alarm can be evaluated according to the current state.
According to the method provided by the embodiment of the invention, the real-time state information of the vehicle is obtained, the corresponding action information is obtained by utilizing the collision early warning model, and after each alarm, the collision early warning model is updated according to the time difference between the alarm time point and the time point of response operation of the driver and the speed information of the vehicle after the alarm. The scheme considers the driving habit of the driver and updates the model according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm opportunity, is more in line with the expectation of the driver, and the accuracy of collision early warning is improved.
Fig. 5 is a schematic structural diagram of a collision warning apparatus according to an embodiment of the present invention. Referring to fig. 5, the apparatus includes:
an obtaining module 501, configured to obtain state information, input the state information into a collision early warning model, and output action information corresponding to the state information, where the state information is used to describe at least one of a relative motion situation between a vehicle and a front target, a driving state of the vehicle, and a current road surface condition, the collision early warning model is used to output action information according to the input state information, and the action information is used to indicate whether to alarm;
an alarm module 502, configured to alarm when the action information indicates an alarm;
the obtaining module 501 is further configured to, when a target instruction is received within a target time period after the warning, obtain a time difference between a receiving time point of the target instruction and a warning time point and speed information of the vehicle, where the target instruction is triggered by a response operation of a driver to the warning;
an updating module 503, configured to update the collision warning model according to the time difference and the speed information of the vehicle.
In one possible implementation, the obtaining module 501 is configured to:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, a target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and acquiring the time required by the collision of the vehicle and the front target object according to the relative distance and the relative speed.
In one possible implementation, the obtaining module 501 is configured to:
the relative distance is obtained using the following equation:
where Δ s is the relative distance, HCY is the ordinate of the target point imaged in the image captured by the image capture apparatus, v is the mounting height of the image capture apparatus0Is thatOrdinate of the center point of the image, fyIs the focal length in the vertical direction, αRIs the tilt angle of the image pickup apparatus.
In one possible implementation, the obtaining module 501 is configured to:
the relative velocity is obtained using the following equation:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, and w is the focal lengthgΔ t is the target time interval, which is the actual scale of the vehicle.
In one possible implementation, the action information includes a weight of the alarm and a weight of the non-alarm;
accordingly, the alarm module 502 is configured to alarm when the weight of the alarm in the action information is greater than the weight of the non-alarm.
In a possible implementation manner, the updating module 503 is configured to obtain a return value of the current warning according to the time difference and the speed information of the vehicle, where the return value is used to evaluate the recognition degree of the driver for the current warning; and updating the collision early warning model based on the return value.
In a possible implementation manner, the speed information of the vehicle is the acceleration of the vehicle, and accordingly, the updating module 503 is configured to obtain the reward value of the present alarm according to the time difference and the acceleration of the vehicle.
In one possible implementation, the speed information of the vehicle includes a first speed of the vehicle at the warning time point, and a second speed of the vehicle at the target command reception time point;
accordingly, the updating module 503 is configured to obtain the acceleration of the vehicle according to the first speed, the second speed, and the time difference; and obtaining the return value of the alarm at this time according to the time difference and the acceleration of the vehicle.
In one possible implementation, the update module 503 is configured to:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the return function is:
wherein R istIs the reported value, tbIs the reception time point, t, of the target instructionwFor the alarm time point, t0The reaction time of the driver to the warning is denoted as a as the acceleration of the vehicle.
In a possible implementation manner, the obtaining module 501 is further configured to assign the time difference to a time difference corresponding to the target time period and execute the step of obtaining the speed information when the target instruction is not received in the target time period after the alarm;
the updating module 503 is further configured to perform the step of updating the collision warning model.
In a possible implementation manner, the obtaining module 501 is further configured to assign the time difference as a target value and execute the step of obtaining the speed information when the action information indicates that no alarm is issued and the target instruction is received within a target time period;
the updating module 503 is further configured to perform the step of updating the collision warning model.
According to the device provided by the embodiment of the invention, the real-time state information of the vehicle is obtained, the corresponding action information is obtained by utilizing the collision early warning model, and after each alarm, the collision early warning model is updated according to the time difference between the alarm time point and the time point of response operation of the driver and the speed information of the vehicle after the alarm. The scheme considers the driving habit of the driver and updates the model according to the feedback of the driver to the alarm, so that the output result of the model, namely the alarm opportunity, is more in line with the expectation of the driver, and the accuracy of collision early warning is improved.
It should be noted that: the collision warning device provided in the above embodiment is exemplified by only dividing the functional modules in collision warning, and in practical applications, the function distribution may be completed by different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above-described functions. In addition, the collision early warning device and the collision early warning method provided by the above embodiments belong to the same concept, and specific implementation processes thereof are detailed in the method embodiments and are not described herein again.
Fig. 6 is a schematic structural diagram of a computer device 600 according to an embodiment of the present invention, where the computer device 600 may be configured in a vehicle for executing the collision warning method provided in the foregoing embodiments. The computer device 600 may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 601 and one or more memories 602, where the memory 602 stores at least one instruction, and the at least one instruction is loaded and executed by the processor 601 to implement the collision warning method provided by the above-mentioned method embodiments. Certainly, the computer device may further have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input and output, and the computer device may further include other components for implementing the functions of the device, which is not described herein again.
In an exemplary embodiment, a computer-readable storage medium, such as a memory, storing at least one instruction, which when executed by a processor, implements the collision warning method in the above embodiments, is also provided. For example, the computer-readable storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a Compact Disc Read-Only Memory (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The above description is only exemplary of the present invention and should not be taken as limiting the invention, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (20)
1. A collision warning method, comprising:
acquiring state information, inputting the state information into a collision early warning model, and outputting action information corresponding to the state information, wherein the state information is time required for a vehicle to collide with a front target object, and the action information comprises a weight value for alarming and a weight value for not alarming;
when the alarm weight is greater than the non-alarm weight, alarming;
when a target instruction is received in a target time period after alarming, acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle, wherein the target instruction is triggered by the response operation of a driver to the alarming;
obtaining a return value of the alarm according to the time difference and the speed information of the vehicle, wherein the return value is used for evaluating the recognition degree of the driver to the alarm;
and updating the collision early warning model based on the return value.
2. The method of claim 1, wherein the obtaining the status information comprises:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, the target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and obtaining the time required by the collision between the vehicle and the front target object according to the relative distance and the relative speed.
3. The method according to claim 2, wherein the obtaining the relative distance according to the internal reference and the external reference of the camera device mounted on the vehicle and the coordinate information of the target point imaged in the image captured by the camera device comprises:
applying the following formula to obtain the relative distance:
wherein Δ s is the relative distance, HCIs the mounting height of the image pickup apparatus, y is the ordinate of the target point imaged in the image taken by the image pickup apparatus, v0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRIs the tilt angle of the camera device.
4. The method of claim 2, wherein the obtaining the relative speed of the vehicle and the front object according to the focal distance, the relative distance, the actual dimension of the front object, a target time interval, and the dimension variation of the front object imaged in the image within the target time interval comprises:
applying the following formula to obtain the relative velocity:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, wgΔ t is the target time interval, which is the actual dimension of the vehicle.
5. The method according to claim 1, characterized in that the speed information of the vehicle is an acceleration of the vehicle,
correspondingly, the obtaining of the return value of the alarm according to the time difference and the speed information of the vehicle includes:
and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
6. The method according to claim 1, wherein the speed information of the vehicle comprises a first speed of the vehicle at the warning time point and a second speed of the vehicle at the target instruction reception time point;
correspondingly, the obtaining of the return value of the alarm according to the time difference and the speed information of the vehicle includes:
obtaining the acceleration of the vehicle according to the first speed, the second speed and the time difference;
and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
7. The method according to claim 5 or 6, wherein the obtaining the return value of the current alarm according to the time difference and the acceleration of the vehicle comprises:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the reward function is:
wherein R istIs the reported value, tbIs the receiving time point, t, of the target instructionwIs the alarm time point, t0The reaction time of the driver to the warning is denoted as a, and the acceleration of the vehicle is denoted as a.
8. The method of claim 1, wherein when the action information indicates an alarm, then after the alarm is raised, the method further comprises:
and when the target instruction is not received in the target time period after alarming, assigning the time difference as the time difference corresponding to the target time period, and executing the steps of acquiring speed information and updating a collision early warning model.
9. The method according to claim 1, wherein after the outputting the action information corresponding to the state information, the method further comprises:
and when the action information indicates that no alarm is given and the target instruction is received in a target time period, assigning the time difference as a target numerical value, and executing the steps of acquiring speed information and updating a collision early warning model.
10. A collision warning apparatus, characterized in that the apparatus comprises:
the system comprises an acquisition module, a collision early warning module and a collision warning module, wherein the acquisition module is used for acquiring state information, inputting the state information into the collision early warning module and outputting action information corresponding to the state information, the state information is time required for a vehicle to collide with a front target object, and the action information comprises a weight value for alarming and a weight value for not alarming;
the alarm module is used for alarming when the weight value of the alarm is greater than the weight value of the non-alarm;
the acquisition module is further used for acquiring the time difference between the receiving time point of the target instruction and the alarming time point and the speed information of the vehicle when the target instruction is received in the target time period after alarming, wherein the target instruction is triggered by the response operation of the driver to the alarming;
the updating module is used for obtaining a return value of the alarm according to the time difference and the speed information of the vehicle, and the return value is used for evaluating the recognition degree of the driver to the alarm;
and updating the collision early warning model based on the return value.
11. The apparatus of claim 10, wherein the obtaining module is configured to:
obtaining a relative distance according to internal parameters and external parameters of camera equipment installed on the vehicle and coordinate information of a target point imaged in an image shot by the camera equipment, wherein the internal parameters comprise a focal length and coordinate information of a central point of the image, and the external parameters comprise an installation height and a pitching angle of the camera equipment;
obtaining the relative speed of the vehicle and the front target object according to the focal length, the relative distance, the actual scale of the front target object, the target time interval and the scale variation of the front target object imaged in the image in the target time interval;
and obtaining the time required by the collision between the vehicle and the front target object according to the relative distance and the relative speed.
12. The apparatus of claim 11, wherein the obtaining module is configured to:
applying the following formula to obtain the relative distance:
wherein Δ s is the relative distance, HCIs the mounting height of the image pickup apparatus, y is the ordinate of the target point imaged in the image taken by the image pickup apparatus, v0Is the ordinate of the center point of the image, fyIs the focal length in the vertical direction, αRFor the purpose of making a video recordingThe pitch angle of the device.
13. The apparatus of claim 11, wherein the obtaining module is configured to:
applying the following formula to obtain the relative velocity:
wherein Δ v is the relative velocity, Δ s is the relative distance, Δ w is the scale variation, f is the focal length, wgΔ t is the target time interval, which is the actual dimension of the vehicle.
14. The apparatus according to claim 10, wherein the speed information of the vehicle is an acceleration of the vehicle, and accordingly the updating module is configured to obtain a return value of the present alarm according to the time difference and the acceleration of the vehicle.
15. The apparatus according to claim 10, wherein the speed information of the vehicle includes a first speed of the vehicle at the warning time point, and a second speed of the vehicle at the reception time point of the target instruction;
correspondingly, the updating module is used for obtaining the acceleration of the vehicle according to the first speed, the second speed and the time difference; and obtaining the return value of the alarm at the time according to the time difference and the acceleration of the vehicle.
16. The apparatus of claim 14 or 15, wherein the update module is configured to:
obtaining a return value of the alarm according to the time difference, the acceleration of the vehicle and a return function; wherein the reward function is:
wherein R istIs the reported value, tbIs the receiving time point, t, of the target instructionwIs the alarm time point, t0The reaction time of the driver to the warning is denoted as a, and the acceleration of the vehicle is denoted as a.
17. The apparatus of claim 10,
the acquisition module is further used for assigning the time difference to be a time difference corresponding to the target time period and executing the step of acquiring the speed information when the target instruction is not received in the target time period after alarming;
the updating module is further configured to perform the step of updating the collision warning model.
18. The apparatus of claim 10,
the obtaining module is further used for assigning the time difference as a target numerical value and executing the step of obtaining the speed information when the action information indicates that no alarm is given and the target instruction is received in a target time period;
the updating module is further configured to execute the step of the collision warning model.
19. A computer device comprising a processor and a memory; the memory is used for storing at least one instruction; the processor, configured to execute at least one instruction stored on the memory to implement the method steps of any of claims 1-9.
20. A computer-readable storage medium, having stored therein at least one instruction, which when executed by a processor, performs the method steps of any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811231736.4A CN111081061B (en) | 2018-10-22 | 2018-10-22 | Collision early warning method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811231736.4A CN111081061B (en) | 2018-10-22 | 2018-10-22 | Collision early warning method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111081061A CN111081061A (en) | 2020-04-28 |
| CN111081061B true CN111081061B (en) | 2021-09-21 |
Family
ID=70309958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811231736.4A Active CN111081061B (en) | 2018-10-22 | 2018-10-22 | Collision early warning method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111081061B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112489425A (en) * | 2020-11-25 | 2021-03-12 | 平安科技（深圳）有限公司 | Vehicle anti-collision early warning method and device, vehicle-mounted terminal equipment and storage medium |
| CN112519790A (en) * | 2020-12-12 | 2021-03-19 | 江西昌河汽车有限责任公司 | Intelligent analysis system for automobile driving habits and control method thereof |
| CN113212449B (en) * | 2021-04-30 | 2022-06-07 | 重庆长安汽车股份有限公司 | Function trigger parameter adjusting method of front collision early warning system |
| CN115410394A (en) * | 2022-08-01 | 2022-11-29 | 江苏航天大为科技股份有限公司 | Internet of vehicles is with dangerous early warning reminding device of big data analysis |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101407199A (en) * | 2008-10-28 | 2009-04-15 | 湖北汽车工业学院 | Vehicle collision prewarning method and apparatus |
| JP2011002869A (en) * | 2009-06-16 | 2011-01-06 | Honda Motor Co Ltd | Collision possibility determination device for vehicle |
| CN102542843A (en) * | 2010-12-07 | 2012-07-04 | 比亚迪股份有限公司 | Early warning method for preventing vehicle collision and device |
| CN102745194A (en) * | 2012-06-19 | 2012-10-24 | 东南大学 | Self-adaption alarming method for preventing tailgating with front car on expressway |
| CN104134371A (en) * | 2014-07-09 | 2014-11-05 | 奇瑞汽车股份有限公司 | Method and device for early warning of vehicle collision |
| JP2016051465A (en) * | 2014-09-01 | 2016-04-11 | ホンダ リサーチ インスティテュート ヨーロッパ ゲーエムベーハーＨｏｎｄａ Ｒｅｓｅａｒｃｈ Ｉｎｓｔｉｔｕｔｅ Ｅｕｒｏｐｅ ＧｍｂＨ | Method and system for performing post-crash maneuvering plan, and vehicle with the system |
| CN105679090A (en) * | 2015-10-21 | 2016-06-15 | 天津师范大学 | Smartphone-based driver nighttime driving assistance system and smartphone-based driver nighttime driving assistance method |
| CN105912998A (en) * | 2016-04-05 | 2016-08-31 | 辽宁工业大学 | Vehicle collision prevention early warning method based on vision |
| CN105930625A (en) * | 2016-06-13 | 2016-09-07 | 天津工业大学 | Design method of Q-learning and neural network combined smart driving behavior decision making system |
| CN106114505A (en) * | 2016-07-04 | 2016-11-16 | 浙江工商大学 | A kind of front truck anti-collision warning method of vehicle DAS (Driver Assistant System) |
| CN106548659A (en) * | 2015-09-23 | 2017-03-29 | 腾讯科技（深圳）有限公司 | To anti-collision warning method and device before a kind of |
| CN106781692A (en) * | 2016-12-01 | 2017-05-31 | 东软集团股份有限公司 | The method of vehicle collision prewarning, apparatus and system |
| CN107145936A (en) * | 2017-04-22 | 2017-09-08 | 大连理工大学 | A kind of vehicle following-model method for building up based on intensified learning |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102009060513B4 (en) * | 2009-12-23 | 2013-11-14 | Siemens Aktiengesellschaft | Method of aligning a towing vehicle, control program and route guidance system |
| US8788176B1 (en) * | 2013-06-19 | 2014-07-22 | Ford Global Technologies, Llc | Adjustable threshold for forward collision warning system |
| US9776632B2 (en) * | 2013-07-31 | 2017-10-03 | Elwha Llc | Systems and methods for adaptive vehicle sensing systems |
| US8935036B1 (en) * | 2013-09-06 | 2015-01-13 | State Farm Mutual Automobile Insurance Company | Systems and methods for updating a driving tip model using telematics data |
| US9656667B2 (en) * | 2014-01-29 | 2017-05-23 | Continental Automotive Systems, Inc. | Method for minimizing automatic braking intrusion based on collision confidence |
| US9108582B1 (en) * | 2014-02-25 | 2015-08-18 | International Business Machines Corporation | System and method for collaborative vehicle crash planning and sequence deployment |
| CN106240458B (en) * | 2016-07-22 | 2017-12-12 | 浙江零跑科技有限公司 | A kind of vehicular frontal impact method for early warning based on vehicle-mounted binocular camera |
-
2018
- 2018-10-22 CN CN201811231736.4A patent/CN111081061B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101407199A (en) * | 2008-10-28 | 2009-04-15 | 湖北汽车工业学院 | Vehicle collision prewarning method and apparatus |
| JP2011002869A (en) * | 2009-06-16 | 2011-01-06 | Honda Motor Co Ltd | Collision possibility determination device for vehicle |
| CN102542843A (en) * | 2010-12-07 | 2012-07-04 | 比亚迪股份有限公司 | Early warning method for preventing vehicle collision and device |
| CN102745194A (en) * | 2012-06-19 | 2012-10-24 | 东南大学 | Self-adaption alarming method for preventing tailgating with front car on expressway |
| CN104134371A (en) * | 2014-07-09 | 2014-11-05 | 奇瑞汽车股份有限公司 | Method and device for early warning of vehicle collision |
| JP2016051465A (en) * | 2014-09-01 | 2016-04-11 | ホンダ リサーチ インスティテュート ヨーロッパ ゲーエムベーハーＨｏｎｄａ Ｒｅｓｅａｒｃｈ Ｉｎｓｔｉｔｕｔｅ Ｅｕｒｏｐｅ ＧｍｂＨ | Method and system for performing post-crash maneuvering plan, and vehicle with the system |
| CN106548659A (en) * | 2015-09-23 | 2017-03-29 | 腾讯科技（深圳）有限公司 | To anti-collision warning method and device before a kind of |
| CN105679090A (en) * | 2015-10-21 | 2016-06-15 | 天津师范大学 | Smartphone-based driver nighttime driving assistance system and smartphone-based driver nighttime driving assistance method |
| CN105912998A (en) * | 2016-04-05 | 2016-08-31 | 辽宁工业大学 | Vehicle collision prevention early warning method based on vision |
| CN105930625A (en) * | 2016-06-13 | 2016-09-07 | 天津工业大学 | Design method of Q-learning and neural network combined smart driving behavior decision making system |
| CN106114505A (en) * | 2016-07-04 | 2016-11-16 | 浙江工商大学 | A kind of front truck anti-collision warning method of vehicle DAS (Driver Assistant System) |
| CN106781692A (en) * | 2016-12-01 | 2017-05-31 | 东软集团股份有限公司 | The method of vehicle collision prewarning, apparatus and system |
| CN107145936A (en) * | 2017-04-22 | 2017-09-08 | 大连理工大学 | A kind of vehicle following-model method for building up based on intensified learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111081061A (en) | 2020-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111081061B (en) | Collision early warning method and device | |
| CN110276985B (en) | Automatic driving safety evaluation method, device and system | |
| CN108062095B (en) | Object tracking using sensor fusion within a probabilistic framework | |
| US9975550B2 (en) | Movement trajectory predicting device and movement trajectory predicting method | |
| US10489664B2 (en) | Image processing device, device control system, and computer-readable storage medium | |
| CN103098110B (en) | Vehicle surroundings monitoring device | |
| JP2017056935A (en) | Classifying objects detected by 3d sensors | |
| US20080084283A1 (en) | Extra-vehicular threat predictor | |
| US20220001858A1 (en) | Dangerous scene prediction device, dangerous scene prediction method, and dangerous scene prediction program | |
| JP2021504218A (en) | State estimator | |
| JP2023085371A (en) | Travel storage system and travel storage method | |
| CN107533133A (en) | For the method and apparatus in the region for monitoring vehicle front | |
| CN112446466A (en) | Measuring confidence in deep neural networks | |
| CN113727064A (en) | Method and device for determining field angle of camera | |
| JP7013722B2 (en) | Driving support device | |
| JP2020034451A (en) | Data structure, storage medium, and storage device | |
| Lee et al. | Real-time associative memory–based rear-end collision warning system | |
| CN113298361A (en) | Dangerous driving behavior evaluation method and device, electronic equipment and system | |
| KR102354734B1 (en) | Method and system for predicting, analyzing and judging accidental hazard of car and object having relative high variations | |
| JP6818948B1 (en) | Driving support devices, driving support systems, driving support methods, and driving support programs | |
| WO2022024947A1 (en) | Path generating device, path generating method, and path generating program | |
| US20180222386A1 (en) | Enhanced lane negotiation | |
| CN114092919B (en) | Vehicle deviation warning method, equipment and medium | |
| CN111767761A (en) | Method and device for determining travel area of mobile equipment and electronic equipment | |
| US20240127694A1 (en) | Method for collision warning, electronic device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |