Method for detecting contamination of a front windshield based on an intelligent sensor

Abstract

This invention relates to the field of vehicle cleaning technology, specifically to a method for detecting windshield contamination based on intelligent sensors. The invention first divides the wiper sweeping stroke, then obtains the frictional resistance coefficient of each spatial sector swept by the wiper to determine the resistance sector. Further, it obtains the clarity change coefficient based on the texture changes of the spatial sectors before and after sweeping to determine the change sector. Then, combining the sector positional relationships, it obtains the correlation factor between the resistance sector and the change sector, thus establishing a correlation. Finally, based on the proportion of uncorrelated and mismatched resistance sectors and the clarity change coefficient of the correlated and matched change sectors, it determines the type and degree of windshield contamination. This invention, by integrating wiper sweeping resistance with the visual changes in the windshield before and after sweeping, and matching and correlating local sectors of mechanical resistance and visual changes, can accurately distinguish the type of windshield contamination, improving the accuracy of windshield contamination type detection.

Description

Technical Field

[0001] This invention relates to the field of vehicle cleaning technology, and specifically to a method for detecting the degree of contamination of a windshield based on a smart sensor. Background Technology

[0002] As the primary window through which drivers obtain their visibility, the cleanliness of the windshield directly impacts driving safety. Besides regular rainfall, glass surfaces often accumulate complex contaminants such as oil films, insect residue, bird droppings, and dried mud. If the wipers operate incorrectly under these non-aqueous conditions, they will not only fail to remove dirt but may also cause oil film buildup leading to blindness, dry wiping damage to the motor and rubber strips, and even malfunction during fogging, interfering with driving. Therefore, accurately identifying the type and degree of contamination on the glass surface is crucial.

[0003] Existing technologies mainly rely on optical sensors or intelligent sensors such as motor current integrated on (electric) vehicles for detection. Optical sensors only determine the presence of water droplets by refractive index and cannot distinguish between oil film, water droplets and fogging. They are prone to misinterpreting changes in the refractive index of oil film as heavy rain, thus exacerbating the smearing. Although motor current detection can sense resistance, it is greatly affected by wind resistance, vehicle acceleration and deceleration and road bumps, making it difficult to extract minute friction features and unable to determine whether the resistance is caused by dry wiping. This leads to frequent false alarms or missed alarms under complex operating conditions, resulting in poor detection of windshield contamination types. Summary of the Invention

[0004] To address the technical problem of poor detection results for windshield contamination types, the present invention aims to provide a windshield contamination detection method based on intelligent sensors. The specific technical solution adopted is as follows: The wiper sweeping stroke is divided to determine all spatial sectors on the windshield; during each wiper sweeping stroke, the wiper sweeping parameters and the windshield image are acquired in real time. During the process of the wipers sweeping each spatial sector, the friction resistance coefficient of the spatial sector is obtained based on the deviation of the sweeping parameters relative to the preset calibration reference at each moment; the clarity change coefficient of the spatial sector is obtained based on the texture change of the windshield image before and after the wipers sweep each spatial sector. The resistance sector is determined based on the friction resistance coefficient, and the change sector is determined based on the sharpness variation coefficient; according to the friction resistance coefficient of each resistance sector and the sharpness variation coefficient of each change sector, and in combination with the sector position relationship, the correlation factor between each resistance sector and each change sector is obtained. Based on the correlation factor, the resistance sector and the change sector are matched. According to the proportion of the resistance sector that is not matched and the clarity change coefficient of the matched change sector, the pollution type and pollution degree of the windshield are determined.

[0005] Furthermore, the sweeping parameters include at least the motor current.

[0006] Furthermore, the method for obtaining the preset calibration reference includes: During factory testing, at each vehicle speed, based on the distribution characteristics of the motor current when the wipers sweep each spatial sector, a preset calibration benchmark for each spatial sector at each vehicle speed is determined.

[0007] Furthermore, the method for obtaining the frictional resistance coefficient includes: During the process of the windshield wipers sweeping each spatial sector, the vehicle speed is collected in real time. Based on the difference between the motor current of the windshield wipers at each moment and the preset calibration benchmark at the corresponding vehicle speed, the current characteristic parameters at each moment are obtained. Based on the distribution characteristics of the current characteristic parameters at all moments, the friction resistance coefficient of the spatial sector is obtained.

[0008] Furthermore, the method for obtaining the sharpness variation coefficient includes: During the process of the wipers sweeping each spatial sector, the spatial sector in the windshield image captured when the wipers are at the starting angle of the spatial sector is taken as the front sweep sector, and the spatial sector in the windshield image captured when the wipers are at the ending angle of the spatial sector is taken as the back sweep sector. Based on the sharpness evaluation algorithm, the sharpness evaluation values ​​of the pre-scan sector and the post-scan sector are obtained respectively. The difference between the sharpness evaluation value of the post-scan sector and the sharpness evaluation value of the pre-scan sector is used as the sharpness variation coefficient of the spatial sector.

[0009] Furthermore, the method for obtaining the resistance sector and the changing sector includes: The spatial sector with a friction resistance coefficient greater than a preset resistance threshold is defined as the resistance sector, and the spatial sector with an absolute value of the sharpness variation coefficient greater than a preset variation coefficient is defined as the variation sector.

[0010] Furthermore, the method for obtaining the correlation factor includes: The spatial distance between each resistance sector and each change sector is negatively correlated to obtain the correlation weight; the friction resistance coefficient of each resistance sector and the sharpness change coefficient of each change sector are fused, and the fusion result is weighted using the correlation weight to obtain the correlation factor between the corresponding resistance sector and the corresponding change sector.

[0011] Furthermore, the association and matching of the resistance sector and the change sector based on the correlation factor includes: Based on the correlation factor, the resistance sector and the change sector are correlated and matched using the bipartite graph maximum weight matching algorithm.

[0012] Further, determining the type and degree of pollution on the windshield includes: The proportion of the resistance sector that is not associated with a match among all resistance sectors is used as an invalid resistance parameter; among the associated matched change sectors, the proportion of the change sector with a sharpness change coefficient less than zero among all associated matched change sectors is used as a smearing feature parameter. When the invalid resistance parameter is greater than the preset dry wiping threshold, the type of pollution on the windshield is determined to be dried stains, and the invalid resistance parameter is used as the degree of pollution. When the smearing feature parameter is greater than the preset smearing threshold and the invalid resistance parameter is less than or equal to the preset dry smearing threshold, the type of contamination on the windshield is determined to be oil film stains, and the smearing feature parameter is used as the degree of contamination. When the invalid resistance parameter is less than or equal to the preset dry shaving threshold and the smearing feature parameter is less than or equal to the preset smearing threshold, the pollution type of the windshield is determined to be a sweepable stain, and the weighted fusion result of the invalid resistance parameter and the smearing feature parameter is taken as the pollution level.

[0013] Furthermore, it also includes: In each sweep stroke of the windshield wiper, the global clarity evaluation value of the windshield is determined based on the distribution characteristics of the clarity evaluation values ​​of the swept sector in all spatial sectors; when there are no resistance sectors and no change sectors, and when the global clarity evaluation value is less than a preset transparency threshold, it is determined that the windshield is fogged up; otherwise, it is considered unfogworthy.

[0014] The present invention has the following beneficial effects: This invention divides the wiper sweeping stroke to determine all spatial sectors on the windshield, enabling precise capture of local texture features within these sectors. During each wiper sweeping stroke, wiper sweeping parameters and a windshield image are acquired in real-time to provide a basis for analysis. As the wiper sweeps each spatial sector, the frictional resistance coefficient of the swept spatial sector is evaluated by comparing the deviation of the sweeping parameters from a preset calibration benchmark at each moment. Furthermore, based on the texture changes in the windshield image before and after the wiper sweeps each spatial sector, the clarity variation coefficient of the spatial sector is obtained. This process identifies resistance sectors and variable sectors. Further, based on the frictional resistance coefficient of each resistance sector and the clarity change coefficient of each variable sector, and considering the sector positional relationships, a correlation factor is obtained to characterize the association matching weight or the probability of the same pollution source between each resistance sector and each variable sector. The resistance sectors and variable sectors are then matched based on this correlation factor. Finally, by combining the resistance performance and visual performance under different pollution conditions, the proportion of unmatched resistance sectors and the clarity change coefficient of matched variable sectors are analyzed to determine the type and degree of pollution in the windshield. This invention, by integrating wiper sweeping resistance with the visual changes in the windshield before and after sweeping, and matching the local sectors that correlate mechanical resistance and visual changes, can accurately distinguish the type of pollution in the windshield, improving the accuracy of windshield pollution type detection. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a windshield contamination detection method based on a smart sensor, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a windshield pollution detection method based on a smart sensor proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details a specific scheme for a windshield contamination detection method based on an intelligent sensor provided by the present invention.

[0020] Please see Figure 1 The diagram illustrates a flowchart of a windshield contamination detection method based on a smart sensor, according to an embodiment of the present invention, specifically including: Step S1: Divide the wiper sweeping stroke to determine all spatial sectors on the windshield; during each wiper sweeping stroke, acquire the wiper sweeping parameters and the windshield image in real time.

[0021] In one embodiment of the present invention, the sweeping stroke of the windshield wiper on the windshield is first divided into a fixed spatial sector area (hereinafter referred to as a spatial sector) so as to accurately capture the local texture features of the spatial sector, thereby achieving accurate capture of local features and improving the sensitivity of identifying minute pollution sources.

[0022] Specifically, the effective upward sweeping range of the windshield wiper (typically 0°-90°) is discretized into K consecutive wiper sweeping sectors, and an integer k∈{1,2,…,K} is used as a unique index. In this embodiment, to adapt to the real-time computing power of the vehicle controller and ensure spatial resolution, K is set to 12, that is, each spatial sector covers a sweeping angle of approximately 7.5°. Implementers can also adjust the value of K according to the actual situation.

[0023] It should be noted that in one embodiment of the present invention, data is collected only during the upward sweeping stroke. The implementer may also adjust it to collect and monitor the downward sweeping stroke. The contamination status on the windshield is analyzed and identified in each upward sweeping stroke. The analysis method for the contamination status on the windshield is consistent in each upward sweeping stroke. Here, only one upward sweeping stroke is used as an example for analysis and description, and will not be repeated.

[0024] Furthermore, during each sweeping stroke of the wipers, the wiper sweeping parameters and the windshield image are acquired in real time. The sweeping parameters help to assess the resistance of the wiper arms during the sweeping process, thereby determining pollution information. The windshield image can provide pollution information directly from a visual perspective.

[0025] In a preferred embodiment of the present invention, considering that when the wiper arm sweeps across the surface of the windshield, the greater the sweeping resistance (such as dry friction or oil film), the greater the load torque that the wiper motor needs to overcome, which in turn leads to a significant increase in the operating current of the wiper motor; the operating current can, to a certain extent, characterize the sweeping resistance of the wiper arm and indirectly characterize the pollution information; therefore, the sweeping parameters include at least the motor current. In other embodiments, the implementer may also use the motor torque instead.

[0026] Specifically, the instantaneous operating current of the wiper motor, i.e., the motor current, is monitored and collected in real time by the on-board wiper motor controller or on-board controller integrated on the (electric) vehicle. This provides a data basis for subsequent evaluation of the friction resistance coefficient when the wiper sweeps each spatial sector. The motor current collection frequency is set to 100Hz, but the implementer can adjust it according to the specific application.

[0027] Since camera capture typically suffers from frame rate limitations and exposure delays, in order to accurately capture the visual changes on the windshield surface during the wiper's sweeping motion, one embodiment of the present invention employs a position-triggered image capture strategy.

[0028] Specifically, through the collaboration of the vehicle controller and the forward-looking camera module, the visual image of the windshield surface, i.e. the windshield image, is triggered. The sweeping angle of the wiper arm can be determined by the wiper arm angle sensor or the motor encoder sensor. When the wiper arm angle is detected to reach the start or end angle of each spatial sector, a trigger signal is immediately sent to the forward-looking camera module or controller to collect the windshield image at each trigger moment.

[0029] It should be noted that the starting angle is the wiper arm angle when it first coincides with the boundary of the spatial sector while sweeping across the sector; similarly, the ending angle is the wiper arm angle when it coincides with the other boundary of the spatial sector for the second (or last) time while sweeping across the sector.

[0030] Step S2: During the process of the wipers sweeping each spatial sector, the friction resistance coefficient of the spatial sector is obtained based on the deviation of the sweeping parameters from the preset calibration reference at each moment; the clarity change coefficient of the spatial sector is obtained based on the texture change of the windshield image before and after the wipers sweep each spatial sector.

[0031] To evaluate the frictional resistance of the wiper sweeping a spatial sector based on the motor current (sweeping parameter), a preset calibration benchmark is first determined. The preset calibration benchmark provides a benchmark for the frictional resistance of the wiper sweeping under the windshield in a clean and uncontaminated state, which can then be used to compare and evaluate the relative frictional resistance during the current sweeping process.

[0032] Preferably, in one embodiment of the present invention, considering that when a vehicle is in motion, the oncoming air (relative wind speed) will exert positive or negative air resistance on the windshield wipers, and the faster the vehicle speed, the greater the oncoming wind resistance; if the vehicle speed is not distinguished and only the reference current at rest is used, the normal wind resistance at high speed will be misjudged as frictional resistance. Based on this, it is necessary to perform test calibration at different vehicle speeds; the method for obtaining the preset calibration reference includes: During factory testing, at each vehicle speed, based on the distribution characteristics of the motor current when the wipers sweep each spatial sector, a preset calibration benchmark for each spatial sector at each vehicle speed is determined.

[0033] Specifically, during the calibration test phase before the (electric) vehicle leaves the factory, the windshield surface is kept clean and moist. The wipers are run at different simulated vehicle speeds v (e.g., 0 km / h, 30 km / h, 60 km / h, 90 km / h, 120 km / h), and the average motor current when the wiper arm sweeps each spatial sector is recorded. The average motor current of spatial sector k at the vehicle speed v is recorded as the preset calibration benchmark Ibase(k,v) and stored in non-volatile memory (Flash) for later retrieval and comparison.

[0034] Furthermore, during the process of the wiper sweeping each spatial sector, the frictional resistance coefficient of the spatial sector can be obtained based on the deviation of the sweeping parameters from the preset calibration benchmark at each moment. The frictional resistance coefficient quantifies the resistance information caused by dirt during the sweeping process of the wiper arm, providing a basis for subsequent judgment of the pollution situation.

[0035] Preferably, in one embodiment of the present invention, considering that during the process of the wiper sweeping each spatial sector, a preset calibration benchmark can be found based on the real-time vehicle speed, thereby eliminating non-frictional resistance components such as air resistance (vehicle speed influence), and determining the additional current caused purely by changes in glass surface friction (such as dry wiping, oil film), i.e., current characteristic parameters; then, all current characteristic parameters during the sweeping of the spatial sector can be comprehensively evaluated to assess the net frictional resistance, avoid misjudgment, and thus accurately determine the frictional resistance coefficient; the method for obtaining the frictional resistance coefficient includes: During the process of the windshield wipers sweeping each spatial sector, the vehicle speed is collected in real time. Based on the difference between the wiper motor current at each moment and the preset calibration benchmark at the corresponding vehicle speed, the current characteristic parameters at each moment are obtained. Based on the distribution characteristics of the current characteristic parameters at all moments, the friction resistance coefficient of the spatial sector is obtained.

[0036] As an example, taking the process of a windshield wiper sweeping across any spatial sector as an example, the vehicle controller is used to monitor and collect the vehicle speed in real time. The sampling frequency of the vehicle speed needs to be aligned with the sampling frequency of the motor current, i.e., 100Hz. In other examples, the implementer can also adjust the speed sampling frequency, but resampling is required to ensure timing alignment. Then, at each (collection) moment, based on the vehicle speed v and the spatial sector index k, a preset calibration benchmark Ibase(k,v) is determined. Then, the wiper motor current at each moment is subtracted from the preset calibration benchmark, and the difference is used as the current characteristic parameter. The current characteristic parameter represents the additional load, i.e., net frictional resistance, that is, the load that exceeds the normal benchmark range at the corresponding moment. When the difference is less than 0, the negative value is truncated, that is, the current characteristic parameter is directly set to 0. Further, the current characteristic parameters at all times during the wiper sweeping the spatial sector are averaged, and the average value is normalized, such as by dividing it by the maximum motor current of the wiper motor. The maximum motor current needs to be determined according to the design parameters of the wiper motor, and in this example, it is taken as 10A. The normalized value is used as the friction resistance coefficient.

[0037] Considering that the visual effect of pollutant smearing on the windshield may differ before and after the wipers sweep each sector, different pollutants may cause different visual effects. For example, when there are oily pollutants, the sector of the windshield may become more blurred due to the oily film before and after sweeping; when there are dried pollutants, there may be no obvious visual change in the sector of the windshield before and after sweeping. Based on this, the embodiments of the present invention further obtain the clarity change coefficient of the spatial sector based on the texture change of the windshield image before and after the wiper sweeps each spatial sector; the clarity change coefficient directly helps to visually evaluate the wiping effect of the wiper, and can be further combined with the friction resistance coefficient to improve the accuracy and robustness of subsequent pollution type assessment.

[0038] Preferably, in one embodiment of the present invention, to facilitate subsequent comparison of texture (clarity) changes before and after sweeping, the pre-sweep sector and post-sweep sector in the windshield image are first determined during each sweep. Considering that the windshield image may contain road background objects, and that the changes in road background objects in the windshield image are relatively small during the short-term scanning of the spatial sector, the main texture changes are affected by the sweeping smearing on the windshield. The clarity evaluation algorithm can help measure texture information, thereby helping to compare texture (clarity) changes before and after sweeping. Therefore, the method for obtaining the clarity change coefficient includes: During the process of the wiper sweeping each spatial sector, the spatial sector in the windshield image captured when the wiper is at the starting angle of the spatial sector is taken as the front sector, and the spatial sector in the windshield image captured when the wiper is at the ending angle of the spatial sector is taken as the back sector. The sharpness evaluation values ​​of the pre-scan and post-scan sectors are obtained based on the sharpness evaluation algorithm. The difference between the sharpness evaluation value of the post-scan sector and the sharpness evaluation value of the pre-scan sector is used as the sharpness variation coefficient of the spatial sector.

[0039] As an example, taking the process of a windshield wiper sweeping any spatial sector as an example, the spatial sector in the windshield image captured when the wiper arm angle reaches the starting angle is taken as the front sector of the sweep, and similarly, the spatial sector in the windshield image captured when the wiper arm angle reaches the starting angle is taken as the back sector of the sweep. The pre-scan and post-scan sectors are converted to grayscale, and then sharpness evaluation values ​​are obtained based on the sharpness evaluation algorithm. Specifically, taking the pre-scan sector as an example, the gradient of each pixel in the pre-scan sector is calculated using the Laplacian operator, and the variance of the gradient magnitude of all pixels is used as the sharpness evaluation value. Similarly, the sharpness evaluation value of the post-scan sector can be calculated. Further, the sharpness evaluation value of the scanned sector is subtracted from the sharpness evaluation value of the pre-scanned sector, and the difference is used as the sharpness variation coefficient of the spatial sector.

[0040] When the clarity variation coefficient is positive (and the larger the absolute value), it indicates that the clarity of the spatial sector after wiping is significantly higher than before wiping, meaning the wipers effectively removed raindrops and other obstructions, which is a normal wiping phenomenon. When the clarity variation coefficient is negative (and the larger the absolute value), it indicates that the clarity of the spatial sector after wiping is lower than before wiping, meaning the wipers spread out the contaminants, causing the windshield to become blurry, which is a typical characteristic of oil film pollution. When the clarity variation coefficient approaches 0, it indicates that the change before and after wiping is small, and there may be dried stains on the windshield causing dry wiping or the windshield itself is very clean.

[0041] Step S3: Determine the resistance sector based on the friction resistance coefficient and the change sector based on the sharpness variation coefficient; based on the friction resistance coefficient of each resistance sector and the sharpness variation coefficient of each change sector, and combined with the sector position relationship, obtain the correlation factor between each resistance sector and each change sector.

[0042] After determining the friction resistance coefficient and sharpness variation coefficient of the spatial sector, the attributes of the spatial sector can be determined from the perspectives of mechanical resistance and visual perception, respectively identifying the resistance sector that may have mechanical resistance and the variation sector that may have visual changes.

[0043] Preferably, in one embodiment of the present invention, the method for obtaining the resistance sector and the changing sector includes: Spatial sectors with a frictional resistance coefficient greater than a preset resistance threshold are defined as resistance sectors, and spatial sectors with an absolute value of a sharpness variation coefficient greater than a preset variation coefficient are defined as variation sectors.

[0044] The preset resistance threshold is set to 0.2, which represents the minimum frictional resistance coefficient when contaminants are present. The preset variation coefficient is set to 15% of the clarity evaluation value of the pre-sweep sector. If the change in the clarity evaluation value of the pre-sweep sector relative to the clarity evaluation value of the post-sweep sector exceeds 15%, the spatial sector is determined to be a variation sector. The preset resistance threshold and preset variation coefficient need to be calibrated based on a large number of contamination sweeping tests. The specific calibration process is a well-known technical method and will not be elaborated here. Implementers may also adjust the values ​​themselves.

[0045] Considering that directly determining the type of pollution based on the frictional resistance coefficient is easily affected by environmental noise such as current fluctuations caused by road bumps, and directly determining the type of pollution based on the sharpness variation coefficient is also easily affected by environmental noise such as changes in ambient light, the reliability of the determination results may be reduced. Furthermore, considering that when a spatial sector has both mechanical resistance and visual changes, it is more likely to contain pollutants, correlation analysis of the frictional resistance coefficient of the resistance sector and the sharpness variation coefficient of the change sector, as well as the positional relationship between the resistance sector and the change sector, can help assess the possibility that mechanical resistance and visual changes originate from the same spatial sector. Based on this, the embodiments of the present invention obtain the correlation factor between each resistance sector and each change sector by combining the frictional resistance coefficient of each resistance sector and the clarity change coefficient of each change sector, and combining the sector position relationship; the correlation factor characterizes the correlation matching weight between the resistance sector and the change sector, quantifies the possibility that the resistance sector and the change sector are the same pollution source, and prepares for subsequent correlation matching.

[0046] Preferably, in one embodiment of the present invention, considering that the smaller the spatial distance between the resistance sector and each changing sector, the greater the probability of corresponding to the same pollution source or pollution type, the association weight can be determined based on spatial distance; when the frictional resistance coefficient is larger or the clarity change coefficient is larger, the degree of pollution is higher; therefore, the method for obtaining the association factor includes: The spatial distance between each resistance sector and each change sector is negatively correlated to obtain the correlation weight; the frictional resistance coefficient of each resistance sector and the clarity change coefficient of each change sector are fused, and the fusion result is weighted using the correlation weight to obtain the correlation factor between the corresponding resistance sector and the corresponding change sector.

[0047] As an example, taking any resistance sector and any change sector as an example, the spatial distance is measured by the sine of the minimum angle between the starting angles of the spatial sectors. Then, the spatial distance is added to a preset positive parameter such as 1 and the reciprocal is calculated to adjust the logic with negative correlation mapping to obtain the association weight. In other examples, the implementer can also use the index difference of the spatial sectors to measure the spatial distance, or other negative correlation mapping methods can be used. Then, the absolute value of the sharpness variation coefficient of the spatial sector is taken and normalized to adjust the value range. This is to prepare for the subsequent integration with the friction resistance coefficient to comprehensively evaluate the related factors, so that it is on the same order of magnitude as the friction resistance coefficient. Specifically, normalization can be performed by dividing by a preset sharpness evaluation value. The preset sharpness evaluation value represents the magnitude of the sharpness variation produced by the standard effective wiping. It needs to be calibrated by experiments, such as the typical sharpness variation value when wiping away moderate raindrops. The calibration process is a well-known technical means and will not be described in detail here. With a weighting of 0.5, the normalized value of the clarity variation coefficient of the variable sector is averaged with the friction resistance coefficient of the resistance sector. The weighted average result is then multiplied by the correlation weight to obtain the correlation factor between the corresponding resistance sector and the corresponding variable sector.

[0048] Step S4: Based on the correlation factor, the resistance sector and the change sector are matched. According to the proportion of the resistance sector that is not matched and the clarity change coefficient of the matched change sector, the pollution type and pollution level of the windshield are determined.

[0049] Once the correlation factors are determined, the resistance sector and the change sector can be further correlated and matched to prepare for subsequent assessment of the pollution type of the windshield.

[0050] Preferably, in one embodiment of the present invention, considering that the correlation factor can characterize the correlation matching weight between the resistance sector and the change sector, and that the bipartite graph maximum weight matching algorithm can find all matching pairs that maximize the sum of weights among different resistance sectors and change sectors, thereby automatically eliminating mismatched resistance sectors and change sectors or those affected by environmental noise; then the correlation matching of resistance sectors and change sectors based on the correlation factor includes: Based on the correlation factor, the bipartite graph maximum weight matching algorithm is used to correlate and match resistance sectors and change sectors.

[0051] It should be noted that the maximum weight matching algorithm for bipartite graphs is the KM algorithm (Kuhn-Munkres algorithm), whose application is a well-known technique and will not be elaborated further. In this algorithm, when the number of resistance sectors and change sectors is not equal, the maximum value L between the number of resistance sectors and the number of change sectors is taken to construct virtual nodes to make up the difference. The correlation factor between the virtual nodes and the remaining sectors is 0, so as to perform subsequent matching.

[0052] Considering the clarity variation coefficient of the variable sectors with associated matching relationships, it can help assess whether effective wiping action leads to a deterioration of the windshield's visibility, thereby helping to assess whether there is oil film contamination; while the resistance sectors without associated matching relationships can help assess whether there is mechanical resistance during wiper sweeping but no visual change, thereby helping to assess whether there are dried stains; based on this, after associating and matching the resistance sectors and variable sectors, this embodiment of the invention further determines the type and degree of contamination of the windshield based on the proportion of resistance sectors without associated matching and the clarity variation coefficient of the variable sectors with associated matching.

[0053] Preferably, in one embodiment of the present invention, determining the type and degree of pollution of the windshield includes: The proportion of unmatched resistance sectors among all resistance sectors is used as the invalid resistance parameter; among matched resistance sectors, the proportion of sectors with a sharpness change coefficient less than zero among all matched resistance sectors is used as the smearing feature parameter. When the ineffective resistance parameter is greater than the preset dry wiping threshold, the type of pollution on the windshield is determined to be dried stains, and the ineffective resistance parameter is used as the degree of pollution. When the smearing feature parameter is greater than the preset smearing threshold and the invalid resistance parameter is less than or equal to the preset dry smearing threshold, the type of pollution on the windshield is determined to be oil film stains, and the smearing feature parameter is used as the degree of pollution. When the invalid resistance parameter is less than or equal to the preset dry swiping threshold and the smearing feature parameter is less than or equal to the preset smearing threshold, the pollution type of the windshield is determined to be a sweepable stain, and the weighted fusion result of the invalid resistance parameter and the smearing feature parameter is used as the pollution level.

[0054] As an example, we first determine the ineffective resistance parameter. The ineffective resistance parameter quantifies the probability that there is mechanical resistance but no visual change during the complete upward sweep of the wiper. The larger the ineffective resistance parameter, the more likely there are dried stains on the windshield. Dried stains cause resistance to the wiper sweep, but because the stains are dried, the wiper sweep cannot effectively remove or spread the stains. Further determination of smearing characteristic parameters quantifies the probability information that there is mechanical resistance and that the windshield visibility is deteriorated during the complete upward sweeping stroke of the wiper. Smearing characteristic parameters indicate that there is a relatively greater possibility of oil film stains on the windshield. Oil film stains have a resistance effect on the wiper sweeping and are further spread and expanded during the wiper sweeping process. The preset application threshold is set to 0.6, and the preset dry scraping threshold is set to 0.8. The implementer can also adjust these values ​​as needed. When the ineffective resistance parameter is greater than 0.8 (regardless of the value of the coating characteristic parameter), it indicates that there are dried stains on the windshield, which affect the resistance of the wiper sweeping but do not bring about visual changes. That is, it is mainly in dry wiping condition. Therefore, the type of pollution on the windshield is determined to be dried stains (such as small hard particles or transparent dried matter, which have resistance to sweeping but low visual change). The ineffective resistance parameter is used as the degree of pollution to characterize the degree of drying. At this time, it is necessary to immediately cut off the power supply to the wiper motor and control the wiper arm to return to the bottom parking area. Then, push the prompt message to the driver through the vehicle human-machine interface (HMI) to check the wiper rubber strips or clean the stubborn stains on the glass. When the smearing characteristic parameter > 0.6 and the invalid resistance parameter ≤ 0.8, it indicates that the possibility of dried stains on the windshield is relatively low, but the possibility of an oil film is relatively high. This oil film affects the wiper sweeping and further spreads, leading to a deterioration of visibility. Therefore, the type of contamination on the windshield is determined to be an oil film stain, and the smearing characteristic parameter is used as the degree of contamination. At this time, the windshield washer pump is turned on to spray a cleaning fluid containing surfactants for t=2s to emulsify the lipid medium on the glass surface. Then, the wiper motor is controlled to run at a low speed constant mode for 3 sweeping strokes (including the up and down strokes; no data collection and monitoring analysis is required during this stroke, only cleaning response) to remove the oil film stains on the windshield. Then, subsequent data collection and monitoring analysis are resumed. The implementer can also adjust the duration of the cleaning fluid spray and the number of sweeping strokes of the cleaning response based on the degree of contamination. When the ineffective resistance parameter is ≤0.8 and the smearing characteristic parameter is ≤0.6, the type of contamination on the windshield is determined to be a sweepable stain, i.e. rainwater or other common stains that are easily swept away. The ineffective resistance parameter and the smearing characteristic parameter are then weighted and summed with a weight of 0.5 to obtain the degree of contamination. In this case, no special treatment is required, and it can be removed by conventional sweeping.

[0055] Considering the possibility that there may be neither a resistance sector nor a change sector, meaning that there is neither obvious resistance nor obvious visual change during the wiper sweeping process, this may be a fogging situation, and further fogging verification is needed.

[0056] Preferably, in one embodiment of the present invention, it further includes: In each sweeping stroke of the windshield wiper, the global clarity evaluation value of the windshield is determined based on the distribution characteristics of the clarity evaluation values ​​of the swept sector after all spatial sectors. When there are no resistance sectors or variable sectors, and when the global clarity evaluation value is less than the preset transparency threshold, it is determined that the windshield is fogged up; otherwise, it is considered uncontaminated.

[0057] As an example, the average of the sharpness evaluation values ​​of the swept sectors of all spatial sectors is used as the global sharpness evaluation value of the windshield. The preset transparency threshold needs to be calibrated under typical conditions where there is no fog inside the windshield and the outside of the vehicle is clean. By creating slight fog on the inside of the windshield until the driver feels a slight obstruction of vision, the critical sharpness is measured, which is the preset transparency threshold. The specific calibration process is a well-known technical means and will not be described in detail here. When there are no resistance sectors or changing sectors, and the global clarity evaluation value is less than the preset transparency threshold, it means that the field of vision changes little before and after the wipers sweep, but is still blurry. This may be because fog is attached to the inside of the glass and cannot be cleared by the external wipers. Therefore, it is determined that there is internal fogging, and a command can be issued to enter the defogging mode. When the global clarity evaluation value is greater than or equal to the preset transparency threshold, it means that the field of vision of the windshield is clear and transparent before and after the wipers sweep, and there may be no pollution.

[0058] In summary, this invention first divides the wiper sweeping stroke, then obtains the frictional resistance coefficient of each spatial sector swept by the wiper to determine the resistance sector. Further, it obtains the sharpness change coefficient based on the texture changes of the spatial sector before and after sweeping to determine the change sector. Then, combining the sector positional relationships, it obtains the correlation factor between the resistance sector and the change sector, thus establishing a correlation. Finally, based on the proportion of uncorrelated and mismatched resistance sectors and the sharpness change coefficient of the correlated and matched change sectors, it determines the type and degree of windshield contamination. This invention, by integrating wiper sweeping resistance with the visual changes in the windshield before and after sweeping, and matching and correlating local sectors of mechanical resistance and visual changes, can accurately distinguish the type of windshield contamination, improving the accuracy of windshield contamination type detection.

[0059] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0060] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for detecting windshield contamination based on intelligent sensors, characterized in that, The method includes: The wiper sweeping stroke is divided to determine all spatial sectors on the windshield; during each wiper sweeping stroke, the wiper sweeping parameters and the windshield image are acquired in real time. During the process of the wipers sweeping each spatial sector, the friction resistance coefficient of the spatial sector is obtained based on the deviation of the sweeping parameters relative to the preset calibration reference at each moment; the clarity change coefficient of the spatial sector is obtained based on the texture change of the windshield image before and after the wipers sweep each spatial sector. The resistance sector is determined based on the friction resistance coefficient, and the change sector is determined based on the sharpness variation coefficient; according to the friction resistance coefficient of each resistance sector and the sharpness variation coefficient of each change sector, and in combination with the sector position relationship, the correlation factor between each resistance sector and each change sector is obtained. Based on the correlation factor, the resistance sector and the change sector are matched. According to the proportion of the resistance sector that is not matched and the clarity change coefficient of the matched change sector, the pollution type and pollution degree of the windshield are determined.

2. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, The sweeping parameters include at least the motor current.

3. The method for detecting windshield contamination based on intelligent sensors according to claim 2, characterized in that, The method for obtaining the preset calibration benchmark includes: During factory testing, at each vehicle speed, based on the distribution characteristics of the motor current when the wipers sweep each spatial sector, a preset calibration benchmark for each spatial sector at each vehicle speed is determined.

4. The method for detecting windshield contamination based on intelligent sensors according to claim 3, characterized in that, The method for obtaining the friction resistance coefficient includes: During the process of the windshield wipers sweeping each spatial sector, the vehicle speed is collected in real time. Based on the difference between the motor current of the windshield wipers at each moment and the preset calibration benchmark at the corresponding vehicle speed, the current characteristic parameters at each moment are obtained. Based on the distribution characteristics of the current characteristic parameters at all moments, the friction resistance coefficient of the spatial sector is obtained.

5. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, The method for obtaining the sharpness variation coefficient includes: During the process of the wipers sweeping each spatial sector, the spatial sector in the windshield image captured when the wipers are at the starting angle of the spatial sector is taken as the front sweep sector, and the spatial sector in the windshield image captured when the wipers are at the ending angle of the spatial sector is taken as the back sweep sector. Based on the sharpness evaluation algorithm, the sharpness evaluation values ​​of the pre-scan sector and the post-scan sector are obtained respectively. The difference between the sharpness evaluation value of the post-scan sector and the sharpness evaluation value of the pre-scan sector is used as the sharpness variation coefficient of the spatial sector.

6. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, The method for obtaining the resistance sector and the change sector includes: The spatial sector with a friction resistance coefficient greater than a preset resistance threshold is defined as the resistance sector, and the spatial sector with an absolute value of the sharpness variation coefficient greater than a preset variation coefficient is defined as the variation sector.

7. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, The method for obtaining the correlation factor includes: The spatial distance between each resistance sector and each change sector is negatively correlated to obtain the correlation weight; the friction resistance coefficient of each resistance sector and the sharpness change coefficient of each change sector are fused, and the fusion result is weighted using the correlation weight to obtain the correlation factor between the corresponding resistance sector and the corresponding change sector.

8. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, The association and matching of the resistance sector and the change sector based on the correlation factor includes: Based on the correlation factor, the resistance sector and the change sector are correlated and matched using the bipartite graph maximum weight matching algorithm.

9. The method for detecting windshield contamination based on intelligent sensors according to claim 1, characterized in that, Determining the type and degree of contamination on the windshield includes: The proportion of the resistance sector that is not associated with a match among all resistance sectors is used as an invalid resistance parameter; among the associated matched change sectors, the proportion of the change sector with a sharpness change coefficient less than zero among all associated matched change sectors is used as a smearing feature parameter. When the invalid resistance parameter is greater than the preset dry wiping threshold, the type of pollution on the windshield is determined to be dried stains, and the invalid resistance parameter is used as the degree of pollution. When the smearing feature parameter is greater than the preset smearing threshold and the invalid resistance parameter is less than or equal to the preset dry smearing threshold, the type of contamination on the windshield is determined to be oil film stains, and the smearing feature parameter is used as the degree of contamination. When the invalid resistance parameter is less than or equal to the preset dry shaving threshold and the smearing feature parameter is less than or equal to the preset smearing threshold, the pollution type of the windshield is determined to be a sweepable stain, and the weighted fusion result of the invalid resistance parameter and the smearing feature parameter is taken as the pollution level.

10. The method for detecting windshield contamination based on intelligent sensors according to claim 5, characterized in that, Also includes: In each sweep stroke of the windshield wiper, the global sharpness rating of the windshield is determined based on the distribution characteristics of the sharpness rating values ​​of the swept sector in all spatial sectors. When neither the resistance sector nor the change sector exists, and when the global clarity evaluation value is less than the preset transparency threshold, internal fogging is determined; otherwise, there is no pollution.

Images