Illumination glare suppression control method and system fused with visual model

Abstract

The invention is suitable for the technical field of urban intelligent illumination and traffic vision safety control, and provides an illumination glare suppression control method and system fused with a vision model, and the method comprises the steps: judging whether a current road surface wet reflection coefficient is in a road surface wet reflection coefficient range corresponding to an optimal interval or not if the optimal interval exists; if yes, the illumination control parameters of the high-curvature road section are kept unchanged; and if not, calling a predefined visual response prediction model, and performing prediction calculation by iteratively adjusting the illumination control parameters under the current road surface wet reflection coefficient, so that the driving visual response comprehensive index output by the model falls into the index range corresponding to the optimal interval. According to the method, reverse optimization of the visual effect under the condition that the humidity cannot be changed is achieved, the phenomena of dazzle light and elevated fence boundary blurring caused by high-curvature road section fence reflection are effectively restrained, and night driving safety and visual comfort are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of urban intelligent lighting and traffic visual safety control, and particularly relates to a lighting glare suppression control method and system fusing a visual model. BACKGROUND

[0002] In the urban road lighting system, high curvature sections such as viaducts and ramp interchanges are the key areas of night traffic safety control. The existing urban lighting control system usually dynamically adjusts the lighting equipment according to environmental brightness, traffic volume or time period, etc. to achieve the balance between energy saving and safety. However, in areas with complex curvature and fence structure such as viaduct curves and ramps, the lighting light will produce multiple reflections and scattering between the road surface and the fence surface, resulting in uneven spatial brightness distribution, forming local glare or dark areas, which brings significant interference to the driver's visual recognition. Especially in rainy and slippery environment, the reflection characteristics of light on the wet road surface change, which easily causes driver's visual illusion, fence blur and halo effect in high curvature sections (especially when the traffic volume is small and the driver has no motion reference object), thereby affecting the driving decision and increasing the risk of abnormal deceleration or misoperation.

[0003] Although there are some lighting control schemes based on visual perception optimization in the prior art, most of them only perform static optimization for brightness adjustment or glare threshold suppression, and do not fully consider the dynamic influence of the wet-state reflection characteristics of the road surface on the driver's visual perception. Specifically, the reflection mode of the wet road surface will change from diffuse reflection to specular reflection as the humidity changes, and the reflection light distribution characteristics under different wet states directly determine the driver's visual clarity and spatial perception ability. However, the current lighting control technology generally lacks modeling and real-time response mechanism for this reflection change rule, and it is difficult to maintain the visual stability of the driver when the environmental humidity fluctuates. In addition, the existing system mostly uses single light intensity or brightness feedback control, and cannot comprehensively optimize the traffic behavior data and visual response data, resulting in lagging or excessive lighting adjustment.

[0004] Therefore, there is an urgent need for a lighting control method that can fuse the road surface wet-state reflection characteristics and the driving visual response characteristics, to dynamically identify the visual optimal response interval under different humidity and environmental conditions, and to realize visual equivalent optimization through intelligent adjustment of lighting parameters when the current wet-state reflection condition deviates from the optimal state. SUMMARY

[0005] The purpose of the present application is to provide a lighting glare suppression control method and system fusing a visual model, which aims to solve the problems raised in the background art.

[0006] The present application is implemented as follows: a lighting glare suppression control method fusing a visual model, the method comprising:

[0007] Obtain multi-source perception data of the elevated monitoring section, the monitoring section including a high-curvature road section and a specific road section related to visual interference thereof;

[0008] Screen a plurality of samples with different road surface wet-state reflectance coefficients of the high-curvature road section from the multi-source perception data consistent with the current environmental conditions, and determine the driving visual response comprehensive index and the abnormal deceleration event occurrence rate of the driving group corresponding to the samples located in the specific road section;

[0009] Analyze whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate obtained from the change of the road surface wet-state reflectance coefficient in different samples exist fluctuations and a coupling relationship, and if so, determine whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold;

[0010] If the optimal interval exists, determine whether the current road surface wet-state reflectance coefficient is within the range of the road surface wet-state reflectance coefficient corresponding to the optimal interval; if yes, keep the lighting control parameters of the high-curvature road section unchanged; if no, call a predefined visual response prediction model to perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflectance coefficient, so that the driving visual response comprehensive index output by the model falls within the index range corresponding to the optimal interval.

[0011] As a further limitation of the technical scheme of the embodiment of the present application, the high-curvature road section refers to a road section with a curvature radius less than a preset radius threshold and provided with a high fence structure for protection; and the specific road section related to visual interference thereof refers to a driving section located downstream of the high-curvature road section and affected by fence reflection light or glare of the high-curvature road section within the field of view of the driver.

[0012] As a further limitation of the technical scheme of the embodiment of the present application, the consistency with the current environmental conditions refers to that the differences between the external parameters such as environmental brightness, meteorological state, visibility, traffic flow and average vehicle speed at the time of sample collection and the corresponding parameters of the current monitoring section are within a preset tolerance range.

[0013] As a further limitation of the technical scheme of the embodiment of the present application, the road surface wet-state reflectance coefficient refers to the ratio of unit incident light intensity to light intensity reflected by the road surface when there is a water film or wet state on the road surface, which is used to represent the reflection characteristics and degree of the high-curvature road section to the lighting light.

[0014] As a further limitation of the technical scheme of the embodiment of the present application, the driving visual response comprehensive index refers to a comprehensive evaluation quantity calculated based on the objective parameters such as the environmental brightness distribution, lighting uniformity and reflected light intensity gradient of the high-curvature road section observed in the field of view of the driver entering the specific road section, which is used to represent the visual perception adaptability of the driver.

[0015] As a further limitation of the technical solutions of the embodiments of the present application, the abnormal deceleration event occurrence rate of the driving group refers to the proportion of the number of events in which the deceleration exceeds the preset deceleration threshold to the total number of vehicles in the range of the specific road section, based on the video monitoring data in the multi-source perception data, excluding the deceleration caused by traffic interference factors such as deceleration, lane change or cutting in front of the vehicle.

[0016] As a further limitation of the technical solutions of the embodiments of the present application, the change trend of the driving visual response comprehensive index obtained by analyzing different samples as the road surface wet-state reflection coefficient changes and the change trend of the abnormal deceleration event occurrence rate are both fluctuating and have a coupling relationship, if so, the step of determining whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously below the preset threshold value comprises:

[0017] Sort a plurality of samples according to the road surface wet-state reflection coefficient from small to large to obtain a sample sequence;

[0018] Obtain the change trend of the driving visual response comprehensive index and the change trend of the abnormal deceleration event occurrence rate corresponding to the sample sequence, and when both show consistency in overall change direction, fluctuation amplitude and stage turning characteristics, it is determined that both have a coupling relationship;

[0019] After confirming the existence of the coupling relationship, the change trend of the abnormal deceleration event occurrence rate is analyzed, and if a continuous interval below the preset threshold value is detected, the interval is determined as the optimal interval.

[0020] As a further limitation of the technical solutions of the embodiments of the present application, if there is an optimal interval, it is determined whether the current road surface wet-state reflection coefficient is within the range of the road surface wet-state reflection coefficient corresponding to the optimal interval; if so, the lighting control parameters of the high-curvature road section are kept unchanged; if not, a predefined visual response prediction model is called to perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflection coefficient, so that the driving visual response comprehensive index output by the model falls within the index range corresponding to the optimal interval, the step comprising:

[0021] After determining that there is an optimal interval in the change trend of the abnormal deceleration event occurrence rate, the current road surface wet-state reflection coefficient is obtained based on the multi-source perception data, and the sample set corresponding to the optimal interval is extracted;

[0022] The road surface wet-state reflection coefficient included in the sample set is set as the reflection coefficient range corresponding to the optimal interval, and the driving visual response comprehensive index corresponding to the sample set is set as the index range corresponding to the optimal interval;

[0023] determining whether the current road surface wet-state reflection coefficient falls within the reflection coefficient range, if yes, keeping the lighting control parameters of the high-curvature road section lighting device unchanged;

[0024] if no, calling a predefined visual response prediction model, inputting the current road surface wet-state reflection coefficient into the model, and iteratively adjusting the lighting control parameters under the current environmental conditions to make the model output driving visual response comprehensive index fall within the index range;

[0025] the lighting control parameters obtained at this time are taken as the target control parameters of the high-curvature road section lighting device under the current environmental conditions and the road surface wet-state reflection coefficient, and the target control parameters are sent to the corresponding lighting device to perform lighting adjustment.

[0026] As a further limitation of the technical scheme of the embodiment of the application, the predefined visual response prediction model refers to a visual response adjustment model, which models the control parameters of the lighting device according to the spatial structure parameters, curvature characteristics and fence reflection characteristics of different specific road sections, combines different road surface wet-state reflection coefficients and environmental illumination conditions, and predicts the corresponding driving visual response comprehensive index in the specific road section driver's field of view under different lighting control parameter combinations; the control parameters include irradiation angle, luminous flux, color temperature and illumination distribution.

[0027] A lighting glare suppression control system fusing a visual model, the system comprising:

[0028] a multi-source perception acquisition module for acquiring multi-source perception data of the elevated monitoring section, the monitoring section including a high-curvature road section and a specific road section related to visual interference thereof;

[0029] a sample screening and feature extraction module for screening a plurality of samples with consistent current environmental conditions but different road surface wet-state reflection coefficients of the high-curvature road section from the multi-source perception data, and determining the corresponding driving visual response comprehensive index of the specific road section and the abnormal deceleration event occurrence rate of the driving group of the samples;

[0030] a trend analysis and coupling identification module for analyzing whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate with the change of the road surface wet-state reflection coefficient in different samples exist fluctuations and have a coupling relationship, if yes, determining whether there is an optimal interval in which the abnormal deceleration event occurrence rate continuously falls below a preset threshold;

[0031] The visual model regulation and lighting control module is used to determine whether the current road surface wet-state reflection coefficient is in the road surface wet-state reflection coefficient range corresponding to the optimal interval if the optimal interval exists; if yes, the lighting control parameters of the high-curvature road section are kept unchanged; if no, a predefined visual response prediction model is called to perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflection coefficient, so that the driving visual response comprehensive index output by the model falls into the index range corresponding to the optimal interval.

[0032] Compared with the prior art, the present application has the following beneficial effects:

[0033] The present application proposes a lighting glare suppression control method fusing visual model for the glare and visual blur problems easily caused by high-curvature road sections such as urban elevated ramps under night wet-slip environment. The method fully utilizes the physical law that the road surface wet-state reflection characteristics transition from diffuse reflection to specular reflection under different humidity conditions, constructs a coupling model of "road surface wet-state reflection-visual response-driving behavior", and identifies an optimal wet-state reflection interval with optimal driving visual response and lowest abnormal deceleration event. When the real-time road surface wet-state reflection coefficient deviates from the interval, the lighting parameters (illumination angle, luminous flux, color temperature, and illumination distribution) are adaptively and iteratively adjusted through the visual response prediction model, so that the driver obtains equivalent visual clarity to the optimal wet-state under the current humidity.

[0034] The present application realizes reverse optimization of visual effect under the condition that humidity cannot be changed, effectively suppresses the glare and elevated fence boundary blur phenomenon caused by fence reflection of high-curvature road sections, and significantly improves the night driving safety and visual comfort. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 The flowchart of the method provided for the embodiments of the present application;

[0036] Figure 2 The flowchart of analyzing the coupling relationship of different samples and identifying the optimal interval in the method provided for the embodiments of the present application;

[0037] Figure 3 The flowchart of adjusting the lighting control parameters based on the visual response prediction model in the method provided for the embodiments of the present application;

[0038] Figure 4 The application architecture diagram of the system provided for the embodiments of the present application. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application is further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0040] Figure 1 A flow chart of the method provided by the embodiment of the application is shown.

[0041] Specifically, the method for suppressing glare of lighting by fusing visual models specifically comprises the following steps:

[0042] In step S100, multi-source perception data of an elevated monitoring section is acquired, the monitoring section including a high-curvature road section and a specific road section related to visual interference thereof.

[0043] The high-curvature road section refers to a road section with a curvature radius less than a preset radius threshold and provided with an elevated fence structure for protection; and the specific road section related to visual interference thereof refers to a driving section located downstream of the high-curvature road section and affected by fence reflection light or glare of the high-curvature road section within a field of view of a driver.

[0044] In the embodiment of the application, the application scenario to which the application is applied is intelligent lighting safety control of urban elevated roads, especially for an elevated monitoring section including a high-curvature road section and a specific road section related to visual interference thereof.

[0045] In a night driving environment, the elevated monitoring section usually includes a ramp or a curved road section with a large curvature, and such a high-curvature road section is usually provided with a fence structure made of concrete (mainly), metal, glass or composite material for protection and isolation. Due to the large spatial curvature and the reflection characteristics of the fence surface, when light from a lighting device is irradiated to the area, the light will produce complex multiple reflections and scattering between the road surface and the fence surface, resulting in unevenness of spatial brightness distribution.

[0046] For a specific road section located downstream of the high-curvature road section but still within the field of view of the driver in front, the driver will be affected by multiple factors such as light intensity attenuation, fence reflection brightness change and background fusion when observing the distant high-curvature road section at this position, resulting in a visual illusion that the high-curvature road section and the road surface background have reduced light and shade levels, blurred contours, and even "merge into the ground". Such visual distortion makes it difficult for the driver to accurately identify the boundary line, solid line marking and road space form of the remote road, and it is easy to produce slight deceleration or instantaneous braking operation in the case of uncertain perception.

[0047] Especially in rainy or wet conditions, the visual blur problem is more likely to occur and more complex. In practical research, the skilled person in the art found that during the rainfall process, a thick and uneven water film layer is formed on the road surface, and its surface wet-state reflection coefficient fluctuates significantly with the change of wetness, causing the reflection mode of the lighting light to change dynamically: when the humidity is low, the main reflection is diffuse reflection, the road surface brightness is insufficient, the contrast is reduced, and the boundary recognition of the far high-curvature road section is reduced; when the humidity increases to a moderate level, the diffuse reflection and the specular reflection components are superimposed, the light distribution tends to be balanced, and the road surface profile and the fence structure reflection form a complementary brightness, thereby temporarily forming an "optimal reflection state" with higher visual clarity.

[0048] Therefore, in the overall wet and slippery environment on rainy days, although the visual interference risk of high-curvature road sections still exists, in a certain wet-state reflection coefficient range, there is still a state interval in which the driver's visual response is optimal, i.e., a visual optimization interval.

[0049] However, in further research, the skilled person in the art found that although the optimal visual state corresponding to a certain wet state can be identified statistically, in actual road operation, influenced by many factors such as environmental light changes, traffic flow, wind speed, rainfall intensity, and road drainage conditions, the real-time road wet-state reflection coefficient often deviates from the optimal humidity interval, so that the lighting conditions no longer match the visual perception. The prior art lacks a technical means that can actively and dynamically adjust the lighting conditions according to the current wet-state reflection conditions, so that the visual effect approaches the optimal visual response state.

[0050] The multi-source perception data includes multi-dimensional information data for characterizing road environment state, lighting conditions, and traffic operation characteristics. Specifically, it includes:

[0051] Environmental perception data: acquired in real time by road illumination sensors, meteorological monitoring nodes, humidity and rainfall sensors, temperature and humidity detection modules, etc., for determining external illumination, meteorological state, visibility, and rainfall intensity parameters;

[0052] Road surface state data: collected by infrared imaging sensors, spectral reflection detectors, or laser reflection ranging devices arranged on high-curvature road sections and specific road sections, for calculating the road wet-state reflection coefficient and its spatial distribution;

[0053] Traffic operation data: collected by video monitoring systems, geomagnetic detection units, or millimeter wave radars, for identifying vehicle density, driving speed, and deceleration event information;

[0054] Lighting device state data: real-time lighting power, irradiation angle, color temperature, and luminous flux parameters fed back by the lighting control module.

[0055] The aforementioned multi-source sensing data can all be obtained through existing sensing equipment such as roadside units (RSUs), smart light pole nodes, or traffic monitoring centers. They fall within the scope of existing sensing technologies in this field and do not require special modifications to the sensing devices. The functional objectives of this invention can be achieved simply by optimizing the algorithms and reconstructing the logic at the data fusion and visual model analysis levels.

[0056] Furthermore, the illumination glare suppression control method based on the fused visual model also includes the following steps:

[0057] Step S200: Select several samples from the multi-source perception data that are consistent with the current environmental conditions but have different road surface wet reflectance coefficients in high-curvature road sections, and determine the comprehensive driving visual response index and the abnormal deceleration event rate of the driving group corresponding to the sample in a specific road section.

[0058] The phrase "consistent with current environmental conditions" means that the differences between external parameters such as ambient brightness, weather conditions, visibility, traffic flow, and average vehicle speed at the time of sample collection and the corresponding parameters of the current monitoring section are all within the preset tolerance range.

[0059] The road surface wet reflectance coefficient refers to the ratio of the unit incident light intensity to the light intensity reflected from the road surface when there is a water film or wet state on the road surface. It is used to characterize the reflection characteristics and degree of reflection of lighting light on road surfaces with high curvature.

[0060] The comprehensive driving visual response index refers to a comprehensive evaluation quantity used to characterize the driver's visual perception adaptability, calculated based on objective parameters such as the ambient brightness distribution, illumination uniformity, and reflected light intensity gradient of the high-curvature road section observed in the driver's field of vision when entering a specific road section.

[0061] The abnormal deceleration event rate of the driving group refers to the proportion of the number of events in which the deceleration exceeds a preset deceleration threshold among vehicles entering a specific road segment and traveling a preset distance, after excluding deceleration caused by traffic interference factors such as deceleration, lane change, or cutting in front of the vehicle. This is based on video monitoring data identified from multi-source perception data.

[0062] In this embodiment of the invention, to achieve accurate analysis of visual interference and lighting adaptability on high-curvature road sections, it is necessary to screen out comparative samples from a large amount of historical monitoring data that can truly reflect the differences in visual response under different wet reflection conditions. Therefore, in step S200, the invention first performs environmental condition consistency screening on the multi-source sensing data. The purpose is to eliminate interference from non-wet factors (such as ambient brightness, weather changes, traffic flow fluctuations, etc.) on visual response and driving behavior, ensuring that the comparison between samples stems solely from differences in the wet reflection characteristics of the road surface.

[0063] Specifically, the "consistent with the current environmental conditions" refers to that the key parameters such as the external light intensity, weather state, visibility, traffic flow and vehicle average speed at the time of sample collection are within the preset tolerance range compared with the real-time environmental conditions of the current monitoring section. This screening process enables the comparative analysis between samples to be carried out under similar external light and traffic background, thereby ensuring the comparability and accuracy of the analysis results. Thus, each sample data can be regarded as the actual response performance of the elevated high-curvature road section-specific road section system under the same external environment but different road surface wet-state reflection conditions, reflecting the objective influence of different humidity levels on visual clarity and driving behavior.

[0064] When analyzing the samples, the road surface wet-state reflection coefficient and the driving visual response comprehensive index are two key evaluation variables.

[0065] The road surface wet-state reflection coefficient is used to quantitatively represent the light reflection ability of the road surface under different wet states, and its calculation can be based on the incident light intensity I0 and I r obtained by the light sensor or the spectral reflection measuring device. The reflection coefficient can be obtained by the formula P = I r / I0. The reflection coefficient can reflect the changes in the thickness of the water film and the surface optical properties, and is an input parameter for subsequent visual model calculation. This measurement method has been widely used in the field of road optical detection and belongs to the mature prior art.

[0066] The driving visual response comprehensive index is used to quantitatively reflect the perception and adaptation of the driver to the visual clarity of the remote high-curvature road section when entering the specific road section. This index can be obtained by setting up a high-definition video camera or an image perception module on the specific road section to calculate the brightness distribution, lighting uniformity and reflected light intensity gradient of the high-curvature road section in the driver's field of view. Specifically, the visual contrast can be calculated using the image gray histogram and brightness uniformity function, and a normalized evaluation value can be formed by combining the reflected light intensity gradient change rate and the spatial light distribution uniformity to represent the visual adaptation degree of the road section. This visual parameter calculation method is based on mature computer vision brightness evaluation and road lighting detection algorithms and belongs to the prior art.

[0067] In addition, in order to reflect the behavior response characteristics of the driving group under specific visual conditions, the present application defines an "abnormal deceleration event occurrence rate" index. This index identifies the vehicle trajectory within the preset distance range entering the specific road section through video monitoring or geomagnetic detection unit, and after excluding the deceleration events caused by external traffic interference factors such as deceleration, lane change, and cutting of the vehicle in front, counts the number of times the deceleration exceeds the preset deceleration threshold (e.g. 0.5 m / s² or adaptively set according to the road grade), and calculates the proportion of the total number of vehicles in the time period.

[0068] The parameter can effectively reflect the instinctive deceleration behavior of the driver caused by visual blur, glare or spatial sense illusion under the condition of no external interference, and is a behavior level index for evaluating driving safety and visual adaptability. Through the change trend of the index, the change of driving load and the potential risk level under different wet-state reflection conditions can be directly judged.

[0069] In summary, the road surface wet-state reflection coefficient reflects the physical characteristics of the interaction of light and road surface, the driving visual response comprehensive index reflects the visual perception effect of the driver, and the abnormal deceleration event occurrence rate reflects the safety feedback of the driving behavior level. The combination of the three constitutes a multi-dimensional evaluation system for the visual safety state of the elevated high-curvature section, which provides a reliable data basis for subsequent determination of the visual optimization interval and adaptive lighting control.

[0070] Further, the lighting glare suppression control method of the fusion visual model further comprises the following steps:

[0071] Step S300, analyze whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate obtained as the road surface wet-state reflection coefficient changes in different samples exist fluctuations and have a coupling relationship, and if so, determine whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold.

[0072] Specifically, Figure 2 A flowchart for analyzing the coupling relationship and optimal interval identification of different samples is shown.

[0073] Among them, analyzing whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate obtained as the road surface wet-state reflection coefficient changes in different samples exist fluctuations and have a coupling relationship, and if so, determining whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold specifically comprises the following steps:

[0074] Step S301, sort a plurality of samples according to the road surface wet-state reflection coefficient from small to large to obtain a sample sequence;

[0075] Step S302, obtain the corresponding driving visual response comprehensive index change trend and abnormal deceleration event occurrence rate change trend in the sample sequence, and when the two show consistency in overall change direction, fluctuation amplitude and stage transition characteristics, it is determined that the two have a coupling relationship;

[0076] Step S303, after confirming the existence of the coupling relationship, interval analysis is performed on the change trend of the abnormal deceleration event occurrence rate, and if a continuously lower than the preset threshold interval is detected, the interval is determined as the optimal interval.

[0077] In the embodiment of the present application, the core of step S300 is to perform trend analysis on the sample data under different wet-state reflection conditions to reveal the dynamic relationship between the road wet-state reflection coefficient and the driver's visual response comprehensive index and the abnormal deceleration event occurrence rate. This process is completed through three parts of trend fluctuation, coupling determination and interval identification, which is the key logical link from the perception level to the control level.

[0078] Firstly, the "fluctuation of change trend" means that in the process of gradual change of the road wet-state reflection coefficient corresponding to different samples, the values of the driver's visual response comprehensive index and the abnormal deceleration event occurrence rate do not change monotonously, but show a dynamic fluctuation form of first rising and then falling or first falling and then rising. This fluctuation phenomenon reflects that under different wet conditions, the driver's visual clarity and driving safety behavior are not linearly corresponding, but there is a nonlinear response rule. Specifically, when the road humidity is low, the visual clarity is poor and the deceleration events are more; when the humidity increases to a certain range, the visual clarity improves and the deceleration events decrease; when the humidity is too high, the reflection is enhanced and the glare rises, and the deceleration events increase again. As can be seen, this fluctuation rule is consistent with the generation mechanism of the visual blur phenomenon in the high-curvature road segment in step S100, and embodies the dynamic corresponding relationship between visual response and reflection characteristics.

[0079] Secondly, the process of determining the "coupling relationship" is a necessary step. The reason is that observing the change trend of the driver's visual response comprehensive index or the abnormal deceleration event occurrence rate alone cannot accurately explain whether there is a causal relationship between the two. By analyzing the direction consistency, fluctuation amplitude similarity and phase transition point synchronization of the two change curves, it can be judged whether there is a cooperative response relationship between the two in the physical and behavioral levels. When the two show similar periodicity or turning correspondence in the change trend, it means that there is a coupling between the driver's visual perception change and the behavior feedback, that is, the change of visual clarity can directly affect the safety of driving behavior. The coupling relationship at this time is a prerequisite for determining the optimal interval, which ensures that the extracted safe interval is indeed caused by visual improvement rather than other random factors.

[0080] After confirming the existence of the coupling relationship, the interval analysis of the abnormal deceleration event occurrence rate is to identify the range in which the driving safety is optimal in the overall trend fluctuation. The interval "continuously below the preset threshold" is selected as the optimal interval because the driving behavior has the characteristics of time continuity and group consistency. Only when the abnormal deceleration events remain below the safety level in a continuous wet-state reflection range, it can be concluded that the visual perception state corresponding to this interval is stable and universally effective. If only individual point values are below the threshold, it may be random fluctuation rather than real optimal state.

[0081] The "preset threshold value" can be set according to the road grade, speed limit standard and historical traffic safety data. For example, for an urban expressway, a safe upper limit value (such as 2%-5%) of the abnormal deceleration event occurrence rate can be obtained according to the monitoring data of multiple periods, as a standard for judging the safe driving state. The threshold value can be a fixed parameter, or can be self-adaptively adjusted according to the real-time traffic density and meteorological conditions.

[0082] Finally, the determined "optimal interval" represents a range of road surface wet-state reflection coefficients, in which the comprehensive index of driver visual response is at a relatively optimal value, the abnormal deceleration event occurrence rate is at a lowest stable level, and the change trends of the two are well coupled. The interval defines a "optimal balance state of vision and safety" in a statistical sense, which corresponds to the visual blur phenomenon described in step S100: the former reveals the physical and visual nature of the problem, and the latter provides a quantitative reference basis for subsequent lighting control through data coupling and safety threshold determination, so as to realize "lighting parameter optimization with visual response as the core" in the control logic.

[0083] Further, the lighting glare suppression control method of the fused visual model further includes the following steps:

[0084] Step S400, if the optimal interval exists, it is judged whether the current road surface wet-state reflection coefficient is in the range of road surface wet-state reflection coefficients corresponding to the optimal interval; if yes, the lighting control parameters of the high-curvature road section are kept unchanged; if no, a predefined visual response prediction model is called, and the lighting control parameters are iteratively adjusted under the current road surface wet-state reflection coefficient for prediction calculation, so that the comprehensive index of driving visual response output by the model falls into the index range corresponding to the optimal interval.

[0085] Specifically, Figure 3 A flowchart of adjusting the lighting control parameters based on the visual response prediction model is shown.

[0086] If the optimal interval exists, it is judged whether the current road surface wet-state reflection coefficient is in the range of road surface wet-state reflection coefficients corresponding to the optimal interval; if yes, the lighting control parameters of the high-curvature road section are kept unchanged; if no, a predefined visual response prediction model is called, and the lighting control parameters are iteratively adjusted under the current road surface wet-state reflection coefficient for prediction calculation, so that the comprehensive index of driving visual response output by the model falls into the index range corresponding to the optimal interval. Specifically, the following steps are included:

[0087] Step S401, after determining that the change trend of the abnormal deceleration event occurrence rate has the optimal interval, the current road surface wet-state reflection coefficient is obtained based on the multi-source perception data, and a sample set corresponding to the optimal interval is extracted;

[0088] Step S402, set the wet-state road surface reflectance included in the sample set to the reflectance range corresponding to the optimal interval, and set the driving visual response comprehensive index corresponding to the sample set to the index range corresponding to the optimal interval;

[0089] Step S403, determine whether the current wet-state road surface reflectance falls within the reflectance range, if yes, keep the lighting control parameters of the high-curvature road segment lighting device unchanged;

[0090] Step S404, if not, call the pre-defined visual response prediction model, input the current wet-state road surface reflectance into the model, and iteratively adjust the lighting control parameters under the current environmental conditions to make the driving visual response comprehensive index output by the model fall within the index range;

[0091] Step S405, take the lighting control parameters obtained at this time as the target control parameters of the high-curvature road segment lighting device under the current environmental conditions and the wet-state road surface reflectance, and send the target control parameters to the corresponding lighting device to perform lighting adjustment.

[0092] The pre-defined visual response prediction model refers to a visual response adjustment model, which models the control parameters of the lighting device according to the spatial structure parameters, curvature characteristics and fence reflection characteristics of different specific road segments, combines different road surface wet-state reflectances and environmental illumination conditions, to predict the corresponding driving visual response comprehensive index in the driver's field of view on a specific road segment under different lighting control parameter combinations; the control parameters include irradiation angle, luminous flux, color temperature and illumination distribution.

[0093] In the embodiments of the present application, the purpose of steps S100 to S300 is to determine the optimal wet-state reflectance interval corresponding to the current elevated monitoring section under the existing environmental conditions, i.e. the road surface humidity state with the best driving visual response comprehensive index and the lowest abnormal deceleration event occurrence rate, through analysis of historical samples and real-time monitoring data. This optimal interval represents the balance state between driver visual perception and safe driving behavior. However, in actual application, road surface humidity is greatly affected by environmental and meteorological conditions and is an uncontrollable variable, so the current wet-state reflectance of the road surface does not fall within the optimal interval in most cases. In view of this situation, the core idea of the present application is to perform "visual reverse compensation" through the lighting control system, i.e. in the case where the road surface humidity cannot be directly changed, by adjusting the lighting control parameters, the visual perception effect of the driver under the current environment is equivalent to the visual performance under the optimal humidity condition, so as to realize dynamic optimization of visual response and continuous guarantee of safe driving.

[0094] To this end, the application introduces a predefined visual response prediction model in step S400. This model combines the spatial geometric features of high-curvature road sections (including the radius of curvature, road surface slope, direction angle), the reflective characteristics of the fence (including the reflection coefficient, surface roughness, material optical properties), and the environmental illumination conditions (including the ambient light brightness and color temperature distribution), while inputting the currently detected road surface wet-state reflection coefficient, to realize the prediction of the comprehensive index of driving visual response through light simulation and visual response modeling. The construction of this model can adopt a combination of physical modeling and data-driven methods, i.e., using a light reflection calculation model based on the physical rendering equation to describe the light propagation process, and then establishing a nonlinear mapping relationship between the lighting parameters and the comprehensive index of visual response through a machine learning module trained based on historical sample data. Such modeling methods have mature applications in adaptive vehicle light control systems (ADB), tunnel gradual lighting design, and visual glare prediction of intelligent lighting systems, and belong to the range that can be realized by existing technologies.

[0095] In actual application process, when judging that the current road surface wet-state reflection coefficient is not in the optimal interval, the system automatically executes dynamic adjustment of the lighting parameters through the visual response prediction model. This adjustment process includes joint optimization of multiple parameters such as illumination intensity, illumination angle, color temperature, and illumination distribution. Specifically, when the road surface wet-state reflection coefficient is too high, causing enhanced specular reflection and rising glare, the system reduces the luminous flux of the lighting device (e.g., by 10% to 20%), fine-tunes the illumination angle (upward adjustment of 2° to 3° to reduce the reflection incidence angle), and reduces the color temperature of the light source (e.g., from 4500K to 3500K), to reduce reflection interference and restore the visual hierarchy; when the road surface wet-state reflection coefficient is too low, causing the road surface to be dull and the contrast to be insufficient, the system appropriately increases the luminous flux (by 5% to 15%) and raises the color temperature (e.g., from 3500K to 4000K to 4500K), to enhance the recognizability of the road surface profile and enable the driver to obtain a clearer visual sense of spatial depth. After each round of parameter adjustment, the system inputs the new control parameters into the visual response prediction model for simulation and prediction until the output comprehensive index of driving visual response falls within the index range corresponding to the optimal interval.

[0096] At the lighting implementation level, the target control parameters are issued by the system to the intelligent lighting devices corresponding to the high-curvature section in real time. For LED lighting systems controlled by DALI or 0-10V bus, fine adjustment of luminous flux and color temperature can be realized directly through digital dimming signals; for intelligent projection lamps with adjustable optical modules, spatial directional lighting optimization of the high-curvature section area can be realized by adjusting the irradiation angle through motor drive; for integrated road lighting control terminals, automatic closed-loop updating and remote control of lighting parameters can be realized through the central control platform. For example, in a wet and slippery environment after rain, the system can reduce the luminous flux of the first group of lighting lamps outside the high-curvature section by 10%, increase the irradiation angle by 3°, and adjust the color temperature to 3600K, so as to reduce the intensity of the fence reflection in the driver's field of view and enhance the clarity of the road boundary, thereby significantly reducing the deceleration behavior caused by visual blur.

[0097] The present application realizes visual equivalent optimization under the condition that the road surface humidity cannot be directly changed, and constructs a dynamic lighting regulation and control system based on visual feedback, which has significant technical advantages and application value. This method not only effectively alleviates the visual interference caused by fence reflection on high-curvature sections of elevated roads at night, but also significantly reduces the visual load and abnormal deceleration events of drivers, thereby improving the overall safety and comfort of driving. The technical path of the present application exactly responds to the core research problem proposed in step S100, that is, to realize dynamic optimization of visual blur under different wet conditions. Through the combination of visual response prediction model and lighting adaptive control, the present application realizes the transformation from "passive response to reflection interference" to "active remodeling of visual environment".

[0098] In addition, the present application also has a key technical advantage, that is, to provide an accurate lighting adjustment range based on visual response characteristics in a wet and slippery environment. By identifying the optimal wet-state reflection interval corresponding to "abnormal deceleration event occurrence rate continuously below the preset threshold", the present application can avoid the problem of excessive or insufficient illumination adjustment under the combined action of high-curvature fence reflection and wet and slippery road reflection, ensure that the lighting adjustment is neither too bright to cause glare nor too dark to cause recognition decline, and thus realize the optimal balance between light control and driving visual safety.

[0099] This technical solution has good engineering feasibility and promotion prospect, and can be widely applied to night lighting control of multi-reflection and high-risk sections such as urban elevated roads, ramps, ring interchanges and mountain roads. Combined with existing intelligent lighting systems and vehicle-road cooperation platforms, it can be further expanded into a visual safety adaptive lighting system based on real-time traffic and environmental data, providing key safety vision support for future intelligent transportation and automatic driving infrastructure.

[0100] Further, Figure 4 The application architecture diagram of the system provided by the embodiment of the present application is shown.

[0101] In a further preferred embodiment provided by the present application, a lighting glare suppression control system based on a fusion visual model comprises:

[0102] The multi-source perception acquisition module 100 is configured to acquire multi-source perception data of the elevated monitoring section, and the monitoring section includes a high-curvature road section and a specific road section related to visual interference.

[0103] Further, the lighting glare suppression control system based on the fusion visual model further comprises:

[0104] The sample screening and feature extraction module 200 is configured to screen a plurality of samples with different road surface wet-state reflection coefficients of the high-curvature road section consistent with the current environmental conditions from the multi-source perception data, and determine a driving visual response comprehensive index corresponding to the samples and located in the specific road section and an abnormal deceleration event occurrence rate of a driving group.

[0105] Further, the lighting glare suppression control system based on the fusion visual model further comprises:

[0106] The trend analysis and coupling identification module 300 is configured to analyze whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate obtained by changing the road surface wet-state reflection coefficient in different samples exist fluctuations and have a coupling relationship, and if so, determine whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold.

[0107] Further, the lighting glare suppression control system based on the fusion visual model further comprises:

[0108] The visual model regulation and lighting control module 400 is configured to, if the optimal interval exists, determine whether the current road surface wet-state reflection coefficient is within a range of road surface wet-state reflection coefficients corresponding to the optimal interval; if so, keep the lighting control parameters of the high-curvature road section unchanged; and if not, call a predefined visual response prediction model to perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflection coefficient, so that the driving visual response comprehensive index output by the model falls within an index range corresponding to the optimal interval.

[0109] It should be understood that, although the steps in the flowcharts of the embodiments of the present application are shown in a certain order according to the arrows, the steps are not necessarily executed in the order of the arrows. Unless otherwise specified herein, the execution of the steps is not strictly limited in order, and the steps can be executed in other orders. Moreover, at least some of the steps in the embodiments can include a plurality of sub-steps or a plurality of stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of the sub-steps or stages is not necessarily sequential, but can be round-robin or alternately executed with at least some of the other steps or sub-steps or stages of the other steps.

[0110] It can be understood by those skilled in the art that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a non-volatile computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments of the methods. Any reference to memory, storage, database or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0111] The technical features of the above-mentioned embodiments can be combined in any way. In order to make the description concise, all possible combinations of the technical features in the above-mentioned embodiments are not described, but as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0112] The above embodiments only express several implementation manners of the present application, which are described in a more specific and detailed manner, but should not be understood as a limitation on the patent scope of the present application. It should be noted that, for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

[0113] The above merely describes the preferred embodiments of the present application and should not be used to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method of controlling glare suppression of lighting by fusing vision models, characterized by, The method comprises: obtaining multi-source perception data of an elevated monitoring section, the monitoring section comprising a high-curvature road section and a specific road section related to visual interference thereof; screening, from the multi-source perception data, a plurality of samples consistent with current environmental conditions but having different road surface wet-state reflectance coefficients of the high-curvature road section, and determining driving visual response comprehensive indexes corresponding to the samples and an abnormal deceleration event occurrence rate of a driving group located in the specific road section; analyzing whether fluctuations and coupling relationships exist between the driving visual response comprehensive indexes and the abnormal deceleration event occurrence rate obtained by changing the road surface wet-state reflectance coefficients in different samples, and if so, determining whether an optimal interval exists in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold value; if the optimal interval exists, determining whether the current road surface wet-state reflectance coefficient is within a range of road surface wet-state reflectance coefficients corresponding to the optimal interval; if so, keeping the lighting control parameters of the high-curvature road section unchanged; and if not, calling a predefined visual response prediction model to perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflectance coefficient, so that the driving visual response comprehensive index output by the model falls within an index range corresponding to the optimal interval.

2. The glare suppression control method of the fused vision model according to claim 1, characterized in that, The high-curvature road section refers to a road section having a curvature radius smaller than a preset radius threshold value and being provided with a high fence structure for protection; and the specific road section related to visual interference thereof refers to a driving section located downstream of the high-curvature road section and being affected by fence reflected light or glare of the high-curvature road section within a field of view of a driver.

3. The illumination glare suppression control method of fusing vision models according to claim 1, wherein, The consistency with current environmental conditions refers to that differences between external parameters such as environmental brightness, meteorological state, visibility, traffic flow and average vehicle speed at the time of sample collection and corresponding parameters of the current monitoring section are within a preset tolerance range.

4. The illumination glare suppression control method of fusing vision models according to claim 1, wherein, The road surface wet-state reflectance coefficient refers to a parameter representing a ratio of unit incident light intensity to light intensity reflected by a road surface in a water film or wet state, used to represent the reflection characteristics and degree of the high-curvature road section.

5. The method of claim 1, wherein the method of suppressing glare of illumination light based on a fusion visual model is characterized by, The driving visual response comprehensive index refers to a comprehensive evaluation quantity calculated based on objective parameters such as environmental brightness distribution, lighting uniformity and reflected light intensity gradient of the high-curvature road section observed in a field of view of a driver entering the specific road section, used to represent visual perception adaptability of the driver.

6. The illumination glare suppression control method of fusing vision models according to claim 1, wherein, The abnormal deceleration event occurrence rate of the driving group refers to a proportion of the number of events in which a vehicle has a deceleration exceeding a preset deceleration threshold value to a total number of vehicles within a preset distance range after driving into the specific road section, which is identified based on video monitoring data in the multi-source perception data and excludes deceleration situations caused by traffic interference factors such as deceleration, lane changing or cutting in of a front vehicle.

7. The method of claim 1, wherein the method of suppressing glare of illumination light in a fusion vision model is characterized by, The step of analyzing whether fluctuations and coupling relationships exist between the driving visual response comprehensive indexes and the abnormal deceleration event occurrence rate obtained by changing the road surface wet-state reflectance coefficients in different samples, and if so, determining whether an optimal interval exists in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold value, comprises: sorting the plurality of samples according to the road surface wet-state reflectance coefficients from small to large to obtain a sample sequence; obtain the change trend of the corresponding driving visual response comprehensive index and the change trend of the abnormal deceleration event occurrence rate in the sample sequence, and determine that the two have a coupling relationship when the two show consistency in overall change direction, fluctuation amplitude, and stage transition characteristics; after confirming the existence of the coupling relationship, perform interval analysis on the change trend of the abnormal deceleration event occurrence rate, and determine the interval as an optimal interval if a continuous interval below a preset threshold is detected.

8. The method of claim 7, wherein the method of suppressing glare of illumination light for a fusion vision model is characterized by, if the optimal interval exists, determine whether the current road surface wet-state reflection coefficient is within the range of the road surface wet-state reflection coefficient corresponding to the optimal interval; if yes, keep the lighting control parameters of the high-curvature road section unchanged; if no, call a predefined visual response prediction model, and perform prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflection coefficient, so that the driving visual response comprehensive index output by the model falls within the index range corresponding to the optimal interval, which includes the following steps: after determining that the change trend of the abnormal deceleration event occurrence rate has the optimal interval, obtain the current road surface wet-state reflection coefficient based on the multi-source perception data, and extract a sample set corresponding to the optimal interval; set the road surface wet-state reflection coefficient included in the sample set as the reflection coefficient range corresponding to the optimal interval, and set the driving visual response comprehensive index corresponding to the sample set as the index range corresponding to the optimal interval; determine whether the current road surface wet-state reflection coefficient falls within the reflection coefficient range; if yes, keep the lighting control parameters of the lighting equipment of the high-curvature road section unchanged; if no, call the predefined visual response prediction model, input the current road surface wet-state reflection coefficient into the model, and iteratively adjust the lighting control parameters under the current environmental condition, so that the driving visual response comprehensive index output by the model falls within the index range; use the lighting control parameters obtained at this time as the target control parameters of the lighting equipment of the high-curvature road section under the current environmental condition and the road surface wet-state reflection coefficient, and send the target control parameters to the corresponding lighting equipment to perform lighting adjustment.

9. The fused visual model's lighting glare suppression control method according to claim 8, wherein, The predefined visual response prediction model refers to a visual response adjustment model that models the control parameters of the lighting equipment according to the spatial structure parameters, curvature characteristics, and fence reflection characteristics of different specific road sections, in combination with different road surface wet-state reflection coefficients and environmental illumination conditions, to predict the driving visual response comprehensive index in the field of view of the driver on a specific road section under different combinations of lighting control parameters; the control parameters include irradiation angle, luminous flux, color temperature, and illumination distribution.

10. A lighting glare suppression control system that integrates a visual model, characterized in that, The system includes: a multi-source perception acquisition module for obtaining multi-source perception data of a high-level monitoring section, the monitoring section including a high-curvature road section and a specific road section related to visual interference thereof; a sample screening and feature extraction module for screening a plurality of samples with consistent current environmental conditions but different road surface wet-state reflection coefficients of the high-curvature road section from the multi-source perception data, and determining the driving visual response comprehensive index corresponding to the samples located in the specific road section and the abnormal deceleration event occurrence rate of the driving group; The trend analysis and coupling identification module is configured to analyze whether the change trends of the driving visual response comprehensive index and the abnormal deceleration event occurrence rate obtained by changing the road surface wet-state reflection coefficient in different samples exist fluctuations and have a coupling relationship, and if so, determine whether there is an optimal interval in which the abnormal deceleration event occurrence rate is continuously lower than a preset threshold. The visual model regulation and lighting control module is configured to, if there is the optimal interval, determine whether the current road surface wet-state reflection coefficient is within a range of road surface wet-state reflection coefficients corresponding to the optimal interval; if so, keep the lighting control parameters of the high-curvature road section unchanged; and if not, call a predefined visual response prediction model to perform a prediction calculation by iteratively adjusting the lighting control parameters under the current road surface wet-state reflection coefficient, so that the driving visual response comprehensive index output by the model falls within an index range corresponding to the optimal interval.

Images