A method and device for identifying ice and snow road surface

Abstract

The present invention relates to the technical field of weather data processing, and particularly relates to a method and device for identifying ice and snow roads. The method includes obtaining a visible light video and a thermal infrared video of a road through an in-vehicle camera, and obtaining image data and lane line data based on the visible light video and the thermal infrared video; sending a request to a roadside weather station to obtain weather data of the roadside weather station; calculating the total score of each lane based on the following formula: the total snow accumulation score i = the first image data score i × w1 + the first weather data score i × w2 + the first lane line data score i × w3, where i represents the i-th lane; when the total snow accumulation score i ≥ 0.7, it is determined as an ice and snow road surface; when there is a lane with a total snow accumulation score less than 0.7, recommend this lane to the driver, and when the total snow accumulation scores of all lanes are ≥ 0.7, recommend the lane with the lowest total score to the driver. The present invention does not require a large number of sensors to be arranged along the road, and at the same time improves the accuracy of identification by combining lane line information.

Description

Technical Field

[0001] The present invention relates to the technical field of weather data processing, and in particular to a method and device for identifying an icy and snowy road surface. Background Art

[0002] According to statistics, many traffic accidents are caused by icy and snowy roads. Therefore, it is necessary to carry out early warning and control of road conditions in weather conditions to ensure driving safety.

[0003] There are some technologies for identifying icy and snowy roads. For example, a road image subsystem includes an image sensor and an image collector connected to the image sensor; a road meteorological subsystem includes a meteorological sensor and an A / D converter connected to the meteorological sensor; a road noise subsystem includes a road noise sensor and a high-resolution A / D converter connected to the road noise sensor; a road temperature subsystem includes a road surface temperature telemetry sensor and an image collector connected to the road surface temperature telemetry sensor. This method requires the deployment of a large number of sensors on the road, which is costly. In addition, this method does not combine the specific road conditions for analysis, and the accuracy of the analysis results is insufficient. Summary of the invention

[0004] In view of this, the purpose of the present invention is to propose a method and device for identifying icy and snowy road surfaces to solve the problem that the current identification method requires the deployment of a large number of sensors on the road, which is costly, and this method does not analyze the specific road conditions, and the accuracy of the analysis results is insufficient.

[0005] Based on the above purpose, the present invention provides a method for identifying an icy and snowy road surface, comprising:

[0006] S1: Obtain visible light video and thermal infrared video of the road through the vehicle camera, and obtain image data and lane line data based on the visible light video and thermal infrared video;

[0007] S2: Send a request to the roadside weather station to obtain the weather data of the roadside weather station;

[0008] S3: The snow coverage rate, road surface temperature and texture uniformity of each lane are extracted based on the image data to obtain the first image data score of each lane; the first meteorological data score is calculated based on the temperature, precipitation type, humidity and wind speed in the meteorological data; the first lane line data score of each lane is calculated based on the lane line slope and lane line curvature in the lane line data; and the total score of each lane is calculated based on the following formula:

[0009] Total snow score i = first image data score i × w 1 +First weather data score i×w 2 +First lane line data score i×w 3, i represents the i-th lane;

[0010] w 1 、w 2 、w 3 Indicates weight, when visibility ≥ 1000m: w 1 =0.6, w 2 =0.2, w 3 =0.2; when visibility is between 500-1000m: w 1 =0.4, w 2 =0.4, w 3 =0.2; when visibility <500m: w 1 =0.3, w 2 =0.4, w 3 =0.3;

[0011] S4: When the total snow score i ≥ 0.7, it is determined to be a snowy road;

[0012] S5: When there is a lane with a total snow score less than 0.7, this lane is recommended to the driver. When the total snow scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver and a cautious driving reminder is issued.

[0013] Optionally, the method further includes:

[0014] The second image data score of each lane is calculated based on the temperature gradient mutation extracted from the image data, the mirror reflection area of ​​each lane, and the sharpness of the texture edge of each lane. The second meteorological data score is calculated based on the humidity, temperature, and precipitation type in the meteorological data. The second lane line data score of each lane is calculated based on the lane line slope, lane line obstruction, and lane line curvature in the lane line data. The total score of each lane is calculated based on the following formula:

[0015] Total ice score i = second image data score i × w 1 + Second weather data score i×w 2 +Second lane line data score i×w 3 , i represents the i-th lane;

[0016] When the total ice score i ≥ 0.75, the road is judged to be icy;

[0017] When there is a lane with a total icing score less than 0.75, this lane will be recommended to the driver. When the total icing scores of all lanes are ≥ 0.75, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

[0018] Optionally, the method further includes:

[0019] The third image data score of each lane is obtained by extracting the mirror reflection area, texture volatility, and infrared temperature close to the ambient temperature of each lane based on the image data. The third meteorological data score is obtained based on the temperature and historical precipitation in the meteorological data. The third lane line data score of each lane is obtained based on the lane line width change, lane line obstruction, and lane line slope in the lane line data. The total score of each lane is calculated based on the following formula:

[0020] Total water accumulation score i = third image data score i × w 1 +The third meteorological data score i×w 2 +The third lane line data score i×w 3 , i represents the i-th lane;

[0021] The total water accumulation score i≥0.7 is considered as a flooded road surface;

[0022] When there is a lane with a total water accumulation score less than 0.7, this lane will be recommended to the driver. When the total water accumulation scores of all lanes are ≥ 0.7, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

[0023] Optionally, it is characterized in that

[0024] The first image data score=snow coverage score×0.4+road surface temperature score×0.3+texture uniformity score×0.2;

[0025] The first meteorological data score = temperature score × 0.3 + precipitation type score × 0.3 + humidity score × 0.3 + wind speed score × 0.3;

[0026] The first lane line data score=lane line slope score×0.2+lane line curvature score×0.1.

[0027] Optionally, the second image data score=temperature gradient mutation score×0.4+mirror reflection area score×0.3+texture edge sharpness score×0.2;

[0028] The second meteorological data score = humidity score × 0.3 + temperature score × 0.3 + precipitation type score × 0.2;

[0029] The second lane line data score=lane line slope score×0.2+lane line blocked score×0.1+lane line curvature score×0.1.

[0030] Optional,

[0031] The third image data score = specular reflection area score × 0.4 + texture volatility score × 0.3 + infrared temperature close to ambient temperature score × 0.2;

[0032] The third meteorological data score = temperature score × 0.2 + historical precipitation score × 0.3;

[0033] The third lane line data score=lane line width change score×0.2+lane line obstruction score×0.1+lane line slope score×0.1.

[0034] Based on the same invention, the present invention also provides a device for identifying an icy and snowy road surface, comprising:

[0035] On-board camera: used to obtain visible light video and thermal infrared video of the road;

[0036] Image processing module: used to obtain image data and lane line data based on visible light video and thermal infrared video;

[0037] Request receiving module: used to send a request to the roadside weather station to obtain the weather data of the roadside weather station;

[0038] Data processing and judgment module: extract the snow coverage rate, road surface temperature and texture uniformity of each lane based on the image data to calculate the first image data score of each lane; calculate the first meteorological data score based on the temperature, precipitation type, humidity and wind speed in the meteorological data; calculate the first lane line data score of each lane based on the lane line slope and lane line curvature in the lane line data; and calculate the total score of each lane based on the following formula:

[0039] Total score i = first image data score i×0.5+first meteorological data score i×0.3+first lane line data score i×0.2, where i represents the i-th lane;

[0040] When the total score i ≥ 0.7, it is judged as a snowy road;

[0041] Recommendation prompt module: when there is a lane with a total score less than 0.7, it recommends the lane to the driver. When the total scores of all lanes are ≥ 0.7, it recommends the lane with the lowest total score to the driver and issues a cautious driving prompt.

[0042] When the method is used to identify an icy and snowy road surface, the visible light video (RGB) and thermal infrared video (TIR) ​​of the road are obtained through the vehicle-mounted camera, and the image data and lane line data are obtained based on the visible light video (RGB) and the thermal infrared video (TIR), and then a request is sent to the roadside meteorological station to obtain the meteorological data of the roadside meteorological station. The image data, lane line data and meteorological data obtained in the above manner do not require the deployment of a large number of sensors on the road, thereby reducing the cost, and then the snow coverage rate, road surface temperature and texture uniformity of each lane are extracted based on the image data to calculate the first image data score of each lane, the first meteorological data score is calculated based on the temperature, precipitation type, humidity and wind speed in the meteorological data, and the first lane line data score of each lane is calculated based on the lane line slope and lane line curvature in the lane line data. The total score of each lane is calculated based on the following formula:

[0043] Total snow score i = first image data score i × w 1 +First weather data score i×w 2 +First lane line data score i×w 3 , i represents the i-th lane. When the total score i ≥ 0.7, it is determined to be a snowy road;

[0044] where w 1 、w 2 、w 3 represents the weight, w 1 、w 2 、w 3 The specific value is adjusted according to the visibility information. The visibility information can be obtained through the vehicle camera. When the visibility is very good (≥1000m), the image data can accurately reflect the relevant information. Therefore, w 1 =0.6, w 2 =0.2, w 3 =0.2, increasing the influence of image data on the overall result. When visibility is average, the weight of image data is reduced and the weight of meteorological data is increased. At this time, w 1 =0.4, w 2 =0.4, w 3 =0.2, when visibility is low, such as in foggy or snowy days, when visibility is less than 500m, it is not easy to obtain image data from far away places through the on-board camera. At this time, the weight of image data is further reduced and the weight of lane line data is increased. Lane line data can be obtained from the closer image data. 1 =0.3, w 2 =0.4, w 3 =0.3, reducing over-reliance on visual information.

[0045] When there is a lane with a total score less than 0.7, this lane is recommended to the driver. When the total scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver and a cautious driving reminder is issued.

[0046] It can be seen from the above that the image data, lane line data and meteorological data obtained in the above manner do not require the deployment of a large number of sensors on the road. The meteorological data of the roadside meteorological station can be obtained by using the on-board camera and sending a request to the roadside meteorological station, thereby reducing the cost. At the same time, the first image data score is obtained by calculating the snow coverage rate, road surface temperature and texture uniformity; the first meteorological data score is obtained by calculating the temperature, precipitation type, humidity and wind speed; the first lane line data score of each lane is obtained by calculating the lane line slope and lane line curvature, and the judgment of whether there is snow on the road is made from three dimensions. It not only combines image data and meteorological data, but also judges the specific road conditions by identifying the lane line slope and lane line curvature, such as whether the road condition is a turn or a low-lying area downhill, because when the vehicle is driving on the bend When a vehicle is on the road, centrifugal force will push moisture (such as rainwater and melted snow) toward the outer lane line, resulting in a greater probability of snow accumulation near the outer lane line. There is also a greater probability of snow accumulation in the low-lying areas on the downhill slope. Therefore, in addition to image data and meteorological data, this method also combines relevant data on lane lines to judge whether there is snow on the road from three dimensions, thereby improving the accuracy and robustness of the judgment. When the score of a lane is greater than 0.7, the lane is identified as a lane with snow accumulation. When the score of a lane is less than 0.7, the lane is identified as a lane without snow accumulation. Lanes with scores less than 0.7 are recommended to the driver through voice broadcast. When the total scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver, and a cautious driving reminder is issued, thereby realizing lane-level recognition and prompts. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 Flow chart of the method for identifying icy and snowy road surfaces in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0050] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0051] like Figure 1 As shown, a method for identifying an icy and snowy road surface comprises:

[0052] S1: Obtain visible light video (RGB) and thermal infrared video (TIR) ​​of the road through the vehicle camera, and obtain image data and lane line data based on the visible light video (RGB) and thermal infrared video (TIR);

[0053] S2: Send a request to the roadside weather station to obtain the weather data of the roadside weather station;

[0054] S3: The snow coverage rate, road surface temperature and texture uniformity of each lane are extracted based on the image data to obtain the first image data score of each lane; the first meteorological data score is calculated based on the temperature, precipitation type, humidity and wind speed in the meteorological data; the first lane line data score of each lane is calculated based on the lane line slope and lane line curvature in the lane line data; and the total score of each lane is calculated based on the following formula:

[0055] Total snow score i = first image data score i × w 1 +First weather data score i×w 2 +First lane line data score i×w 3 , i represents the i-th lane;

[0056] w 1 、w 2 、w 3 Indicates weight, when visibility ≥ 1000m: w 1 =0.6, w 2 =0.2, w 3 =0.2; when visibility is between 500-1000m: w 1 =0.4, w 2 =0.4, w 3=0.2; when visibility <500m: w 1 =0.3, w 2 =0.4, w 3 =0.3;

[0057] S4: When the total snow score i ≥ 0.7, it is determined to be a snowy road;

[0058] S5: When there is a lane with a total snow score less than 0.7, this lane is recommended to the driver. When the total snow scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver and a cautious driving reminder is issued.

[0059] When the method is used to identify an icy and snowy road surface, the visible light video (RGB) and thermal infrared video (TIR) ​​of the road are obtained through the vehicle-mounted camera, and the image data and lane line data are obtained based on the visible light video (RGB) and the thermal infrared video (TIR), and then a request is sent to the roadside meteorological station to obtain the meteorological data of the roadside meteorological station. The image data, lane line data and meteorological data obtained in the above manner do not require the deployment of a large number of sensors on the road, thereby reducing the cost, and then the snow coverage rate, road surface temperature and texture uniformity of each lane are extracted based on the image data to calculate the first image data score of each lane, the first meteorological data score is calculated based on the temperature, precipitation type, humidity and wind speed in the meteorological data, and the first lane line data score of each lane is calculated based on the lane line slope and lane line curvature in the lane line data. The total score of each lane is calculated based on the following formula:

[0060] Total snow score i = first image data score i × w 1 +First weather data score i×w 2 +First lane line data score i×w 3 , i represents the i-th lane. When the total score i ≥ 0.7, it is determined to be a snowy road;

[0061] where w 1 、w 2 、w 3 represents the weight, w 1 、w 2 、w 3 The specific value is adjusted according to the visibility information. The visibility information can be obtained through the vehicle camera. When the visibility is very good (≥1000m), the image data can accurately reflect the relevant information. Therefore, w 1 =0.6, w 2 =0.2, w 3 =0.2, increasing the influence of image data on the overall result. When visibility is average, the weight of image data is reduced and the weight of meteorological data is increased. At this time, w 1 =0.4, w2 =0.4, w 3 =0.2, when visibility is low, such as in foggy or snowy days, when visibility is less than 500m, it is not easy to obtain image data from far away places through the on-board camera. At this time, the weight of image data is further reduced and the weight of lane line data is increased. Lane line data can be obtained from the closer image data. 1 =0.3, w 2 =0.4, w 3 =0.3, reducing over-reliance on visual information.

[0062] When there is a lane with a total score less than 0.7, this lane is recommended to the driver. When the total scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver and a cautious driving reminder is issued.

[0063] It can be seen from the above that the image data, lane line data and meteorological data obtained in the above manner do not require the deployment of a large number of sensors on the road. The meteorological data of the roadside meteorological station can be obtained by using the on-board camera and sending a request to the roadside meteorological station, thereby reducing the cost. At the same time, the first image data score is obtained by calculating the snow coverage rate, road surface temperature and texture uniformity; the first meteorological data score is obtained by calculating the temperature, precipitation type, humidity and wind speed; the first lane line data score of each lane is obtained by calculating the lane line slope and lane line curvature, and the judgment of whether there is snow on the road is made from three dimensions. It not only combines image data and meteorological data, but also judges the specific road conditions by identifying the lane line slope and lane line curvature, such as whether the road condition is a turn or a low-lying area downhill. When a vehicle is traveling on a curve, centrifugal force will push water (such as rainwater and melted snow) toward the outer lane line, resulting in a greater probability of snow accumulation near the outer lane line. There is also a greater probability of snow accumulation in low-lying areas on the downhill slope. Therefore, in addition to image data and meteorological data, this method also combines relevant data on lane lines to judge whether there is snow on the road from three dimensions, thereby improving the accuracy and robustness of the judgment. When the score of a lane is greater than 0.7, the lane is identified as a lane with snow accumulation. When the score of a lane is less than 0.7, the lane is identified as a lane without snow accumulation. Lanes with scores less than 0.7 are recommended to the driver through voice broadcast. When the total scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver, and a cautious driving reminder is issued.

[0064] In some embodiments, the method further comprises:

[0065] The second image data score of each lane is calculated based on the temperature gradient mutation extracted from the image data, the mirror reflection area of ​​each lane, and the sharpness of the texture edge of each lane. The second meteorological data score is calculated based on the humidity, temperature, and precipitation type in the meteorological data. The second lane line data score of each lane is calculated based on the lane line slope, lane line obstruction, and lane line curvature in the lane line data. The total score of each lane is calculated based on the following formula:

[0066] Total ice score i = second image data score i × w 1 + Second weather data score i×w 2 +Second lane line data score i×w 3 , i represents the i-th lane;

[0067] When the total ice score i ≥ 0.75, the road is judged to be icy;

[0068] When there is a lane with a total icing score less than 0.75, this lane will be recommended to the driver. When the total icing scores of all lanes are ≥ 0.75, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

[0069] The previous embodiment judges the snow accumulation on the road surface. The present embodiment is used to judge the icing condition of the road surface. The second image data score of each lane is obtained by calculating the temperature gradient mutation, the mirror reflection area of ​​each lane and the sharpness of the texture edge of each lane. The second meteorological data score is obtained by calculating the humidity, temperature and precipitation type. The second lane line data score of each lane is obtained by calculating the lane line slope, lane line obstruction and lane line curvature. Whether the road surface is frozen is judged from three score dimensions. When the total ice score i ≥ 0.75, it is judged to be an icy road surface. When there is a lane with a total ice score less than 0.75, the lane can be recommended to the driver by voice broadcast. When the total scores of all lanes are ≥ 0.75, the lane with the lowest total score is recommended to the driver, and a cautious driving reminder is issued.

[0070] In some embodiments, the method further comprises:

[0071] The third image data score of each lane is obtained by extracting the mirror reflection area, texture volatility, and infrared temperature close to the ambient temperature of each lane based on the image data. The third meteorological data score is obtained based on the temperature and historical precipitation in the meteorological data. The third lane line data score of each lane is obtained based on the lane line width change, lane line obstruction, and lane line slope in the lane line data. The total score of each lane is calculated based on the following formula:

[0072] Total water accumulation score i = third image data score i × w 1 +The third meteorological data score i×w 2+The third lane line data score i×w 3 , i represents the i-th lane;

[0073] The total water accumulation score i≥0.7 is considered as a flooded road surface;

[0074] When there is a lane with a total water accumulation score less than 0.7, this lane will be recommended to the driver. When the total water accumulation scores of all lanes are ≥ 0.7, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

[0075] In this embodiment, water on the road is identified, and a third image data score is obtained by calculating the mirror reflection area, texture volatility, and infrared temperature close to the ambient temperature. The third meteorological data score is obtained by calculating the air temperature and historical precipitation. The third lane line data score is obtained by calculating the lane line width change, lane line obstruction, and lane line slope. The three score dimensions are used to determine whether there is water on the road surface. When the total water accumulation score i ≥ 0.75, it is determined to be a flooded road surface. When there is a lane with a total water accumulation score less than 0.7, the lane can be recommended to the driver by voice broadcast. When the total scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver, and a cautious driving reminder is issued.

[0076] In some embodiments:

[0077] The first image data score=snow coverage score×0.4+road surface temperature score×0.3+texture uniformity score×0.2;

[0078] Optionally, the snow cover score is obtained by a linear mapping, where the coverage percentage is directly converted to a score,

[0079] For example: 70% → 0.7, 90% → 0.9;

[0080] The temperature score can be mapped using intervals, as shown in the following table:

[0081]

[0082] Texture uniformity score = 1 − texture variance / maximum variance threshold texture variance;

[0083] Maximum variance threshold: calibrated through experiments (typical value: the variance of dry road surface is 5000);

[0084] Example: If the current variance is 1000, then the score = 1-1000 / 5000 = 0.8;

[0085] If the variance is close to 0 (perfectly uniform), the score is ≈ 1.0.

[0086] The first meteorological data score = temperature score × 0.3 + precipitation type score × 0.3 + humidity score × 0.3 + wind speed score × 0.3;

[0087] The precipitation type scores are shown in the following table:

[0088]

[0089] Humidity score = humidity / 100;

[0090] Example: 95% → 0.95 (can be rounded to 1.0).

[0091] The wind speed scores are shown in the following table:

[0092]

[0093] The first lane line data score=lane line slope score×0.2+lane line curvature score×0.1.

[0094] The lane slope scores are shown in the following table:

[0095]

[0096] Lane curvature score = min(1.0, curvature / 0.03);

[0097] Example: Curvature 0.02 → 0.02 / 0.03 ≈ 0.67.

[0098] In some embodiments:

[0099] The second image data score=temperature gradient mutation score×0.4+mirror reflection area score×0.3+texture edge sharpness score×0.2;

[0100] Temperature gradient calculation:

[0101] ΔT=(T1-T2) / t, where t is the sampling time interval between temperature T1 and temperature T2;

[0102] Scoring rules:

[0103]

[0104] Specular reflection area score = min(1.0, reflectivity / 20%);

[0105] Example: Reflectivity 25% → 25 / 20 = 1.25 → score 1.0.

[0106] The texture edge sharpness scores are shown in the following table:

[0107]

[0108] The second meteorological data score = humidity score × 0.3 + temperature score × 0.3 + precipitation type score × 0.2;

[0109] The second lane line data score=lane line slope score×0.2+lane line blocked score×0.1+lane line curvature score×0.1.

[0110] The lane line occlusion score can be determined based on the percentage of the occluded area.

[0111] In some embodiments:

[0112] The third image data score = specular reflection area score × 0.4 + texture volatility score × 0.3 + infrared temperature close to ambient temperature score × 0.2;

[0113] texture_volatility_score = min(1.0, volatility / 0.6);

[0114] Example: Volatility 0.75 → 0.75 / 0.6 = 1.25 → Score 1.0.

[0115] The infrared temperature close to the ambient temperature score is obtained as follows:

[0116] ΔT=|Troad surface-Tenvironment|;

[0117] Scoring rules:

[0118]

[0119] Example: road surface temperature 2°C, ambient temperature 3°C → ΔT = 1°C → score 1.0;

[0120] The third meteorological data score = temperature score × 0.2 + historical precipitation score × 0.3;

[0121] The historical precipitation scoring rules are as follows:

[0122]

[0123] Example: 12 mm of precipitation in the past 24 hours → score 1.0.

[0124] The third lane line data score = lane line width change score (water accumulation will affect the width of the lane line) × 0.2 + lane line obstruction score (water accumulation may obstruct the lane line) × 0.1 + lane line slope score × 0.1;

[0125] Based on the same invention, the present invention also provides a device for identifying an icy and snowy road surface, comprising:

[0126] On-board camera: used to obtain visible light video (RGB) and thermal infrared video (TIR) ​​of the road.

[0127] Image processing module: used to obtain image data and lane line data based on visible light video (RGB) and thermal infrared video (TIR);

[0128] Request receiving module: used to send a request to the roadside weather station to obtain the weather data of the roadside weather station;

[0129] Data processing and judgment module: extract the snow coverage rate, road surface temperature and texture uniformity of each lane based on the image data to calculate the first image data score of each lane; calculate the first meteorological data score based on the temperature, precipitation type, humidity and wind speed in the meteorological data; calculate the first lane line data score of each lane based on the lane line slope and lane line curvature in the lane line data; and calculate the total score of each lane based on the following formula:

[0130] Total snow score i = first image data score i × w 1 +First weather data score i×w 2 +First lane line data score i×w 3 , i represents the i-th lane;

[0131] When the total score i ≥ 0.7, it is judged as a snowy road;

[0132] Recommendation prompt module: when there is a lane with a total score less than 0.7, it recommends the lane to the driver. When the total scores of all lanes are ≥ 0.7, it recommends the lane with the lowest total score to the driver and issues a cautious driving prompt.

[0133] The device judges whether there is snow on the road from three dimensions, which improves the accuracy and robustness of the judgment. When the score of a lane is greater than 0.7, the lane is identified as a lane with snow. When the score of a lane is less than 0.7, the lane is identified as a lane without snow. Lanes with scores less than 0.7 are recommended to the driver through voice broadcast. When the total scores of all lanes are ≥0.7, the lane with the lowest total score is recommended to the driver, and a cautious driving reminder is issued.

[0134] It should be understood by those skilled in the art that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the above aspects of the present invention, which are not provided in detail for the sake of simplicity.

[0135] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying icy and snowy roads, characterized in that: include: S1: Obtain visible light video and thermal infrared video of the road through the vehicle camera, and obtain image data and lane line data based on the visible light video and thermal infrared video; S2: Send a request to the roadside weather station to obtain the weather data of the roadside weather station; S3: The snow coverage rate, road surface temperature and texture uniformity of each lane are extracted based on the image data to obtain the first image data score of each lane; the first meteorological data score is calculated based on the temperature, precipitation type, humidity and wind speed in the meteorological data; the first lane line data score of each lane is calculated based on the lane line slope and lane line curvature in the lane line data; and the total score of each lane is calculated based on the following formula: Total snow score i = first image data score i×w1+first meteorological data score i×w2+first lane line data score i×w3, where i represents the i-th lane; w1, w2, w3 represent weights. When visibility is ≥1000m: w1=0.6, w2=0.2, w3=0.2; when visibility is between 500-1000m: w1=0.4, w2=0.4, w3=0.2; when visibility is <500m: w1=0.3, w2=0.4, w3=0.3; S4: When the total snow score i ≥ 0.7, it is determined to be a snowy road; S5: When there is a lane with a total snow score less than 0.7, this lane is recommended to the driver. When the total snow scores of all lanes are ≥ 0.7, the lane with the lowest total score is recommended to the driver and a cautious driving reminder is issued.

2. The method for identifying an icy and snowy road surface according to claim 1, characterized in that: The method further comprises: The second image data score of each lane is calculated based on the temperature gradient mutation extracted from the image data, the mirror reflection area of ​​each lane, and the sharpness of the texture edge of each lane. The second meteorological data score is calculated based on the humidity, temperature, and precipitation type in the meteorological data. The second lane line data score of each lane is calculated based on the lane line slope, lane line obstruction, and lane line curvature in the lane line data. The total score of each lane is calculated based on the following formula: Total ice score i = second image data score i×w1+second meteorological data score i×w2+second lane line data score i×w3, where i represents the i-th lane; When the total ice score i ≥ 0.75, the road is judged to be icy; When there is a lane with a total icing score less than 0.75, this lane will be recommended to the driver. When the total icing scores of all lanes are ≥ 0.75, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

3. The method for identifying an icy and snowy road surface according to claim 1, characterized in that: The method further comprises: The third image data score of each lane is obtained by extracting the mirror reflection area, texture volatility, and infrared temperature close to the ambient temperature of each lane based on the image data. The third meteorological data score is obtained based on the temperature and historical precipitation in the meteorological data. The third lane line data score of each lane is obtained based on the lane line width change, lane line obstruction, and lane line slope in the lane line data. The total score of each lane is calculated based on the following formula: Total water accumulation score i = third image data score i×w1+third meteorological data score i×w2+third lane line data score i×w3, where i represents the i-th lane; The total water accumulation score i≥0.7 is considered as a flooded road surface; When there is a lane with a total water accumulation score less than 0.7, this lane will be recommended to the driver. When the total water accumulation scores of all lanes are ≥ 0.7, the lane with the lowest total score will be recommended to the driver, and a cautious driving reminder will be issued.

4. The method for identifying an icy and snowy road surface according to claim 1, characterized in that: The first image data score=snow coverage score×0.4+road surface temperature score×0.3+texture uniformity score×0.2; The first meteorological data score = temperature score × 0.3 + precipitation type score × 0.3 + humidity score × 0.3 + wind speed score × 0.3; The first lane line data score=lane line slope score×0.2+lane line curvature score×0.

1.

5. The method for identifying an icy and snowy road surface according to claim 2, characterized in that: The second image data score=temperature gradient mutation score×0.4+mirror reflection area score×0.3+texture edge sharpness score×0.2; The second meteorological data score = humidity score × 0.3 + temperature score × 0.3 + precipitation type score × 0.2; The second lane line data score=lane line slope score×0.2+lane line blocked score×0.1+lane line curvature score×0.

1.

6. The method for identifying an icy and snowy road surface according to claim 3, characterized in that: The third image data score = specular reflection area score × 0.4 + texture volatility score × 0.3 + infrared temperature close to ambient temperature score × 0.2; The third meteorological data score = temperature score × 0.2 + historical precipitation score × 0.3; The third lane line data score=lane line width change score×0.2+lane line obstruction score×0.1+lane line slope score×0.

1.

7. A device for executing the method for identifying an icy and snowy road surface as claimed in claim 1, characterized in that: include: On-board camera: used to obtain visible light video and thermal infrared video of the road; Image processing module: used to obtain image data and lane line data based on visible light video and thermal infrared video; Request receiving module: used to send a request to the roadside weather station to obtain the weather data of the roadside weather station; Data processing and judgment module: extract the snow coverage rate, road surface temperature and texture uniformity of each lane based on the image data to calculate the first image data score of each lane; calculate the first meteorological data score based on the temperature, precipitation type, humidity and wind speed in the meteorological data; calculate the first lane line data score of each lane based on the lane line slope and lane line curvature in the lane line data; and calculate the total score of each lane based on the following formula: Total snow score i = first image data score i×w1+first meteorological data score i×w2+first lane line data score i×w3, where i represents the i-th lane; When the total score i ≥ 0.7, it is judged as a snowy road; Recommendation prompt module: when there is a lane with a total score less than 0.7, it recommends the lane to the driver. When the total scores of all lanes are ≥ 0.7, it recommends the lane with the lowest total score to the driver and issues a cautious driving prompt.

Images