CN-122009025-A - Display method and system of electronic rearview mirror based on multi-mode risk field coupling

CN122009025ACN 122009025 ACN122009025 ACN 122009025ACN-122009025-A

Abstract

The invention discloses a display method and a system of an electronic rearview mirror based on multi-mode risk field coupling, comprising the steps of obtaining running state data of a current vehicle and one or more target vehicles and observation point data of a driver of the current vehicle; the method comprises the steps of constructing a coupled dynamic risk field model based on driving state data to generate a comprehensive risk value, generating attention entropy and visual search efficiency value based on observation point data, generating a display control signal through a visual information flow optimizer based on an optimal control theory, simplifying a background area according to the display control signal by adopting an anisotropic diffusion equation, enhancing a foreground object by utilizing a nonlinear rendering pipeline to generate an electronic rearview mirror synthetic picture, and displaying the electronic rearview mirror synthetic picture on an in-vehicle display screen. The invention realizes the highlighting of key risk information and the effective compression of redundant information, reduces the cognitive load of a driver and improves the driving decision efficiency and the safety.

Inventors

HUANG CHANGQING
WEI XIAOLAN
YANG CAISHENG
LAN JINYAO
WAN GUOYI

Assignees

上汽通用五菱汽车股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251224

Claims (10)

1. The display method of the electronic rearview mirror based on the multi-mode risk field coupling is characterized by comprising the following steps of: Acquiring running state data of one or more target vehicles in the current vehicle and the electronic rearview mirror and observation point data of a driver of the current vehicle; Constructing a coupled dynamic risk field model formed by linear weighted fusion of a kinematic risk field, a collision time risk field and a road structure risk field based on the driving state data, and generating a comprehensive risk value R_i (t) of a target vehicle i at time t; Generating an attention entropy H_ attention and a visual search efficiency value eta_search on a display screen based on the gaze point of the current vehicle driver in a preset time window based on the observation point data of the current vehicle driver; generating display control signals through a visual information flow optimizer based on an optimal control theory by the comprehensive risk value R_i (t), the attention entropy H_ attention and the visual search efficiency value eta_search; and simplifying a background area by adopting an anisotropic diffusion equation according to the display control signal, enhancing a foreground object by utilizing a nonlinear rendering pipeline, generating an electronic rearview mirror synthesized picture, and displaying the picture on an in-vehicle display screen.
2. The method for displaying the electronic rearview mirror based on the multi-modal risk field coupling according to claim 1, wherein the driving state data comprises relative acceleration, relative speed and relative distance between the current vehicle and one or more target vehicles, and the observation point data is specifically current vehicle driver gaze point distribution data.
3. The method for displaying an electronic rearview mirror based on multi-modal risk field coupling according to claim 2, wherein the step of constructing a coupled dynamic risk field model based on linear weighted fusion of a kinematic risk field, a collision time risk field, an uncertainty risk field and a road structure risk field based on the driving state data, and generating a comprehensive risk value r_i (t) of the target vehicle i at time t specifically comprises: The kinematic risk field r_kinemic_i (t) is constructed, expressed as: Constructing a collision time risk field R_ttc_i (t), expressed as: building a road structure risk field r_road_i (t), expressed as: Building an uncertainty risk field r_ uncertainty _i (t), expressed as: Constructing a coupled dynamic risk field model formed by linear weighted fusion of a kinematic risk field, a collision time risk field, an uncertainty risk field and a road structure risk field, and generating a comprehensive risk value R_i (t) of a target vehicle i at time t, wherein the comprehensive risk value R_i (t) is expressed as: R_i(t)=w_k R_kinematic_i(t)+w_ttc R_ttc_i(t)+w_r R_road_i(t)+w_u R_uncertainty_i(t) Wherein m_i represents a mass equivalent coefficient of the target vehicle i, |·|2 represents an L2 norm of an acceleration vector, a_relative_i (t) represents a relative acceleration of the target vehicle i and the current vehicle, v_relative_i (t) represents a relative speed of the target vehicle i and the current vehicle, d_i (t) represents a relative distance of the target vehicle i and the current vehicle, b is a constant term, c is a speed weight, γ is a distance attenuation index, and ε is a control 0 decimal; ttc—min represents the minimum safe collision time threshold, τ represents the time decay constant, Σ—i (t) represents the covariance matrix of the target vehicle i state estimate, det (·) represents the determinant, Representing the shape adjustment coefficients, l_i (t) and l_ ego are the positions of the target vehicle i and the current vehicle in the lane lateral coordinates, respectively, l_safe is the safe lateral distance threshold, The scaling coefficients, w_k, w_ttc, w_r, and w_u, are represented, and the sum is 1.
4. The method for displaying an electronic rearview mirror based on multi-modal risk field coupling according to claim 3, wherein the process of generating the attention entropy h_ attention and the visual search efficiency value η_search based on the gaze point of the current vehicle driver on the display screen within the preset time window based on the observation point data of the current vehicle driver specifically comprises: Extracting the distribution probability p (x, y) of the gaze point of the current vehicle driver in a preset time window on the display screen based on the observation point data of the current vehicle driver, and generating attention entropy H_ attention, which is expressed as: The visual search efficiency value η_search is expressed as: Wherein T is an observation time window, g (T) represents actual gaze point coordinates, g_optimal (T) is a theoretical optimal gaze point calculated based on risk distribution, and sigma is a tolerance parameter.
5. The method for displaying an electronic rearview mirror based on multi-modal risk field coupling according to claim 4, wherein the process of generating display control signals by a visual information flow optimizer based on optimal control theory through the integrated risk value r_i (t), the attention entropy h_ attention and the visual search efficiency value η_search specifically comprises: calculating a priority visual intensity e_i (t) based on the integrated risk value r_i (t) and the attention entropy h_ attention, expressed as: E_i(t)=G [R_i(t)/(1+aH_attention)] using the preferential visual intensity e_i (t) for regulating the brightness l_halo_i and the pulsation frequency f_pulse_i of the target vehicle contour halo; The brightness l_halo_i of the contour halo is expressed as: L_halo_i= Wherein the pulsation frequency f_pulse_i is expressed as: f_pulse_i= constructing an objective function of a display control signal And constraints, wherein the objective function Expressed as: the constraint is expressed as: E_min≤E(t)≤E_max where G is the global gain coefficient, a is the suppression coefficient for attention entropy, Representing the minimum value of the target vehicle contour halation brightness, Represents the maximum value of the halation brightness of the contour of the target vehicle, As a function of the growth of the S-type, In order to enhance the degree of activation threshold, Representing the minimum value of the pulsation frequency of the target vehicle, Representing the maximum value of the target vehicle pulsation frequency, E (t) is the system enhancement output, E_desired (t) is the theoretical expected enhancement based on risk calculations, , , For the weight coefficients, E_min and E_max are the minimum and maximum limits of the enhancement.
6. The method for displaying an electronic rearview mirror based on multi-modal risk field coupling according to claim 5, wherein the step of simplifying a background area by using an anisotropic diffusion equation and enhancing a foreground object by using a nonlinear rendering pipeline according to the display control signal to generate an electronic rearview mirror composite picture and displaying the electronic rearview mirror composite picture on an in-vehicle display screen comprises: Acquiring a background area image of the electronic rearview mirror based on the display control signal, dividing the background area image into a plurality of blocks, and calculating an image information entropy H_image_j of each block j, wherein the image information entropy H_image_j is expressed as: p_k represents the probability that the gray value or the color value of the pixel in the block j falls in the kth interval, and M is the total number of intervals; If the image information entropy h_image_j of the block j is lower than the simplification threshold h_th and the block j corresponds to a static background, simplifying the block j, simplifying by adopting a Perona-Malik anisotropic diffusion equation, and expressing a pixel value I_simple of a simplified background image as follows: I_simple=ω I_blur+(1-ω) I_mean I_blast is the Gaussian blur result of block j, I_mean is the average color value of block j, and ω is the blur weight; the nonlinear rendering pipeline is utilized to enhance the high-risk target, and the high-risk target is highlighted through dynamic changes of contour halation brightness and pulsation frequency; And synthesizing the simplified background image, the enhanced high-risk target and the prospective risk probability cloud picture, generating an electronic rearview mirror synthesized picture, and displaying the picture on an in-vehicle display screen.
7. The method for displaying an electronic rearview mirror based on multi-modal risk field coupling according to claim 6, wherein the prospective risk probability cloud map is superimposed and generated in the form of a semitransparent cloud map based on a random differential equation and a Fokker-plane equation.
8. A display system based on a multimodal risk field coupling for electronic rear view mirrors implementing the method of any of claims 1-7, comprising: The information acquisition module is used for acquiring running state data of one or more target vehicles in the current vehicle and the electronic rearview mirror and observation point data of a driver of the current vehicle; The multi-mode risk field coupling processing module is used for constructing a coupling dynamic risk field model formed by linearly weighting and fusing a kinematic risk field, a collision time risk field and a road structure risk field based on the driving state data, and generating a comprehensive risk value R_i (t) of the target vehicle i at time t; The attention state analysis module is used for generating an attention entropy H_ attention and a visual search efficiency value eta_search on a display screen based on the point of regard of the current vehicle driver in a preset time window based on the observation point data of the current vehicle driver; The visual information flow optimization module is used for generating display control signals through a visual information flow optimizer based on an optimal control theory by using the comprehensive risk value R_i (t), the attention entropy H_ attention and the visual search efficiency value eta_search; and the image rendering display module is used for simplifying a background area by adopting an anisotropic diffusion equation according to the display control signal, enhancing a foreground object by utilizing a nonlinear rendering pipeline, generating an electronic rearview mirror synthesized picture and displaying the electronic rearview mirror synthesized picture on an in-vehicle display screen.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the method for displaying an electronic rearview mirror based on multi-modal risk field coupling of any one of claims 1& # x2011, 7.
10. A computer readable storage medium having stored thereon a computer program or instructions, which when executed by a processor, is adapted to implement the method of displaying an electronic rear view mirror based on multimodal risk field coupling as claimed in any of claims 1& #x2011, 7.

Description

Display method and system of electronic rearview mirror based on multi-mode risk field coupling Technical Field The invention belongs to the technical field of automobiles, and particularly relates to a display method and a display system of an electronic rearview mirror based on multi-mode risk field coupling. Background An electronic rear view mirror (CMS) captures an image through a camera and displays it on an in-vehicle display screen to replace a conventional optical outside rear view mirror. Compared with the traditional rearview mirror, the electronic rearview mirror CMS has the advantages of wider visual field, small influence by weather, low wind resistance and the like. However, the existing electronic rearview mirror solutions still have a number of disadvantages: (1) The visual field is fixed, most CMSs provide images with fixed visual fields, and the display content cannot be dynamically optimized according to driving situations such as vehicle speed, steering and the like. Too wide a field of view at high speeds may result in the driver not being able to quickly focus on the distant critical objects, and may not provide adequate lateral vision at low speeds. (2) Information overload and interference, that is, a simple wide-angle view can present a large amount of irrelevant information (such as sky and roadside vegetation) to a driver, and in severe weather (such as rain and fog), the image quality is reduced, and key targets are difficult to identify. (3) Lack of intelligent interaction-existing systems fail to provide active, personalized display services in combination with the driver's real-time status (e.g., gaze focus). Disclosure of Invention The invention aims to overcome the technical defects of fixed visual field, information overload and lack of intelligent interaction of the traditional electronic rearview mirror, and provides a display method and a display system of the electronic rearview mirror based on multi-mode risk field coupling. According to the method, through accurately quantifying the rear traffic environment risk and perceiving the attention state of the driver in real time, the information density and the rendering mode of the display picture are dynamically optimized, the highlighting of key risk information and the effective compression of redundant information are realized, the cognitive load of the driver is reduced, and the driving decision efficiency and the safety are improved. In order to achieve the above purpose, the present invention proposes the following technical solutions: in a first aspect, the present invention provides a display method of an electronic rearview mirror based on multi-modal risk field coupling, the display method comprising the steps of: Acquiring running state data of one or more target vehicles in the current vehicle and the electronic rearview mirror and observation point data of a driver of the current vehicle; Constructing a coupled dynamic risk field model formed by linear weighted fusion of a kinematic risk field, a collision time risk field and a road structure risk field based on the driving state data, and generating a comprehensive risk value R_i (t) of a target vehicle i at time t; Generating an attention entropy H_ attention and a visual search efficiency value eta_search on a display screen based on the gaze point of the current vehicle driver in a preset time window based on the observation point data of the current vehicle driver; generating display control signals through a visual information flow optimizer based on an optimal control theory by the comprehensive risk value R_i (t), the attention entropy H_ attention and the visual search efficiency value eta_search; and simplifying a background area by adopting an anisotropic diffusion equation according to the display control signal, enhancing a foreground object by utilizing a nonlinear rendering pipeline, generating an electronic rearview mirror synthesized picture, and displaying the picture on an in-vehicle display screen. Further, the driving state data comprise relative acceleration, relative speed and relative distance between the current vehicle and one or more target vehicles, and the observation point data are specifically gaze point distribution data of the current vehicle driver. Further, based on the driving state data, constructing a coupled dynamic risk field model formed by linearly weighting and fusing a kinematic risk field, a collision time risk field, an uncertainty risk field and a road structure risk field, and the step of generating the comprehensive risk value r_i (t) of the target vehicle i at time t specifically includes: The kinematic risk field r_kinemic_i (t) is constructed, expressed as: Constructing a collision time risk field R_ttc_i (t), expressed as: building a road structure risk field r_road_i (t), expressed as: Building an uncertainty risk field r_ uncertainty _i (t), expressed as: Constructing a coupled dynamic risk field model f