Inspection methods and devices for anti-glare failure zones in highway median strips

Abstract

This invention provides a method for inspecting anti-glare failure zones in highway median strips, comprising: adjusting the horizontal positions of two image sensors so that the angle between the horizontal projection of the central optical axis of the image sensors and the driving direction is the lateral half-field of view angle of the driver at different driving speeds; scanning the median strip area to acquire images of anti-glare facilities; extracting all pixels between the m-th and n-th rows from top to bottom of the original image to generate a key area image; calculating the average value of each primary color component in the RGB three primary colors for all pixels in the key area image; determining the type of anti-glare facility; for anti-glare panels or anti-glare nets, using a contour extraction-based color analysis method to detect anti-glare failure zones in the key area image; for anti-glare plants, using an information entropy-based contour size method to detect anti-glare failure zones in the key area image. This invention also provides a device for inspecting anti-glare failure zones in highway median strips.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of highway traffic safety facility detection, in particular to a highway central median strip anti-dazzle failure zone inspection method and device. BACKGROUND

[0002] When driving at night, drivers often experience visual function decline and emotional tension fatigue and other discomforts when affected by strong light from oncoming vehicles, forming potential traffic safety hazards. It is generally required in China to set anti-dazzle facilities in the central median strip of expressways and first-class highways, which mainly include anti-dazzle plates, anti-dazzle nets and anti-dazzle plants. The visual sense after installation of anti-dazzle plates and anti-dazzle nets is a continuous non-light-transmitting pure-color homogeneous facade, and there may be a weak vertical boundary between the plates or between the net frames. The visual effect after planting of anti-dazzle plants is a non-pure-color rough facade with complex branches and leaves, and there is usually no regularly distributed profile in the facade.

[0003] During highway operation, the bottom installation of anti-dazzle facilities may fall off, the plate and net body may be damaged, or the plants may be excessively pruned due to factors such as impact of accident vehicles, extreme wind blowing, crown shape pruning and human damage, thereby causing failure of the anti-dazzle function in some areas of the central median strip.

[0004] Currently, the investigation of anti-dazzle failure zones is mainly based on manual walking and subjective judgment, which has high safety risks and poor accuracy. How to realize normalized, intelligent and fine inspection and determination of anti-dazzle failure zones without disturbing normal traffic is a technical problem that has not been solved in the highway transportation industry. SUMMARY

[0005] In view of the deficiencies in the prior art, the present application provides a highway central median strip anti-dazzle failure zone inspection method and device to improve the technical problems in the prior art that the investigation of anti-dazzle failure zones in the central median strip of highways is mainly based on manual walking and subjective judgment, has high safety risks and poor accuracy.

[0006] The technical solution adopted by the present application is as follows:

[0007] In a first aspect, a highway central median strip anti-dazzle failure zone inspection method is provided, comprising the following steps:

[0008] Adjusting the horizontal position of the color line array image sensor so that the included angle between the horizontal projection of the center optical axis of the image sensor and the driving direction is the lateral half field of view angle of the driver at different driving speeds;

[0009] Driving the inspection vehicle along the center line of the lane, and scanning the central median strip area through the image sensor to collect images of the anti-dazzle facilities;

[0010] The original image collected by the image sensor is extracted from the mth row to the nth row to generate a key region image;

[0011] The average value of each color component in RGB three primary colors is calculated for all pixels in the key region image; and the average value is used to determine the type of anti-glare facility;

[0012] For anti-glare plates or anti-glare nets, a color analysis method based on contour extraction is used to detect the anti-glare failure area in the key region image; for anti-glare plants, a contour size method based on information entropy is used to detect the anti-glare failure area in the key region image.

[0013] Further, the different line speeds are the maximum speed limit of the road section and the minimum speed limit of the road section.

[0014] Further, the pixels from the mth row to the nth row are extracted, and m and n are calculated by the following formula:

[0015]

[0016] In the above formula, NUM row represents the total number of rows of the original image, L represents the straight line distance from the image sensor to the intersection point of the central optical axis and the central median strip, represents the horizontal half field angle of the image sensor, represents the height of the anti-glare region given by the design file, represents the height of the anti-glare region given by the design file, represents the height of the image sensor from the road surface, According to , L and determined by the trigonometric function relationship.

[0017] Further, if is less than the height of the central median strip safety guardrail from the road surface , then takes the value of .

[0018] Further, the anti-glare facility includes anti-glare plates or anti-glare nets, and the surface colors of the anti-glare plates and the anti-glare nets include pure green and pure blue;

[0019] When the surface of the anti-glare plate and the anti-glare net is pure green, if one of the following conditions is met, the anti-glare facility can be determined as an anti-glare plate or an anti-glare net:

[0020] G avg >Thed G , R avg <Thed R , B avgThe B Three conditions are met simultaneously; or

[0021] G avg -R avg The G-R and G avg -B avg The G-B are met simultaneously;

[0022] When the anti-glare board and the anti-glare net surface are both pure blue, one of the following conditions is met, which can determine whether the anti-glare facility is an anti-glare board or an anti-glare net:

[0023] B avg The B , R avg The R , G avg The G Three conditions are met simultaneously; or

[0024] B avg -R avg The B-R and B avg -G avg The B-G are met simultaneously;

[0025] Wherein, R avg is the average value of R component, G avg is the average value of G component, and B avg is the average value of B component; Thed G , Thed R , Thed B , Thed G-R , Thed G-B are all set threshold values.

[0026] Further, the color analysis method based on contour extraction includes:

[0027] Selecting a primary color in RGB three primary colors as a main color, extracting the main color component of each pixel in the key area image to generate a gray scale image, and sequentially performing edge detection, binarization, setting the first and last rows and the first and last columns of the image as edge pixels, dilation operation and erosion operation on the gray scale image to obtain a first edge image; and identifying all closed contours in the first edge image;

[0028] For a single closed contour, the total number of pixels in the enclosed area is counted, and the average value of the main color component of all pixels in the closed contour-enclosed area and the average value of the two non-main color components are calculated in the corresponding key area image;

[0029] The value of the total number of pixels, the average value of the primary color component, and the average value of the two non-primary color components are used to determine whether the inside of the closed contour is an anti-glare panel or anti-glare mesh surface; if not, the closed contour is an anti-glare failure zone.

[0030] Furthermore, the interior of the closed contour can be determined not to be an anti-glare panel or anti-glare mesh surface when one of the following conditions is met:

[0031] NUM pixel >Thed num AS local <Thed AS Simultaneously established; or

[0032] NUM pixel >Thed num , max(AS local BS local CS local )≠AS local Established simultaneously;

[0033] Among them, NUM pixel This represents the total number of pixels within the area enclosed by a single closed contour; AS local AS represents the average value of the primary color components of all pixels within the area enclosed by the closed-loop outline. local BS local CS loca Thed represents the average value of all non-primary color components of all pixels within the area enclosed by the closed-loop outline; num and Thed AS All of these are set thresholds.

[0034] Furthermore, contour sizing methods based on information entropy include:

[0035] The key region image is converted to grayscale and then divided into several rectangular grayscale sub-images of the same size;

[0036] Histogram transformation and normalization are performed on a single rectangular grayscale sub-image to obtain the probability function of pixel values;

[0037] Calculate the information entropy of the rectangular grayscale sub-image based on the probability function;

[0038] When the information entropy is less than the third set threshold, the rectangular grayscale sub-image is sequentially subjected to edge detection, binarization, setting the first and last row and first and last column pixels of the image as edge pixels, dilation operation and erosion operation to obtain the second edge image, and all closed contours in the second edge image are identified.

[0039] The number of pixels within the area enclosed by each closed contour is counted. When the maximum number of pixels exceeds the fourth set threshold, the closed contour is determined to be an anti-glare failure area.

[0040] Further, after detecting the anti-dazzle failure area, marking processing is performed, and all position post number information of the detected road section marked as the anti-dazzle failure area is counted and output.

[0041] From the above technical solution, the beneficial technical effects of the present application are as follows:

[0042] The failure area detection is suitable for various types of anti-dazzle facilities, and the inspection method has high automation degree and objective and safe determination of the anti-dazzle failure area of the central separation belt.

[0043] In the second aspect, an inspection device for a highway central separation belt anti-dazzle failure area inspection method is provided, which comprises an image acquisition module, a rotary encoder and a processing module.

[0044] The image acquisition module is installed at the front of the inspection car through an adjustable support, and after installation, the center optical axis is horizontally forward and the scanning line is perpendicular to the road surface, which is used for acquiring images of the anti-dazzle facility.

[0045] The rotary encoder is coaxially installed with the rear wheel of the inspection car and is also electrically connected with the image sensor in the image acquisition module; the rotary encoder is used for sensing the position movement of the car and triggering the image sensor to acquire images.

[0046] The processing module is electrically connected with the image acquisition module and is used for data storage, analysis calculation and man-machine interaction in the inspection process.

[0047] Further, the image acquisition module comprises an adjustable support, a tray, a rotatable platform and an image sensor.

[0048] The adjustable support comprises a suction cup and a support, the suction cup is connected with the inspection car, and the support is connected with the tray; the suction cup and the support are connected through an adjustable universal ball, and a first fixed knob is arranged beside the adjustable universal ball.

[0049] A level bubble is arranged at the center position of the side edge of the tray, which is used for calibrating whether the tray is in a horizontal position.

[0050] The rotatable platform is arranged on the top of the tray, and a second fixed knob is arranged beside the rotatable platform.

[0051] The image sensor is fixedly connected with the top of the rotatable platform and is used for image acquisition.

[0052] From the above technical solution, the beneficial technical effects of the present application are as follows:

[0053] The inspection device is portable, easy to install and disassemble, and does not need to be limited to the type of the inspection car during use, and has strong adaptability. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the description of the embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference signs. In the drawings, the elements or parts are not necessarily drawn according to the actual proportions.

[0055] Figure 1 Structure schematic diagram of the patrol device of the embodiment of the present application;

[0056] Figure 2 Flow chart of the patrol method of the embodiment of the present application;

[0057] Figure 3 Schematic diagram of the spatial position relationship of each part when generating the key area image of the embodiment of the present application;

[0058] Reference signs:

[0059] 1-image sensor, 2-rotatable platform, 3-tray, 4-level bubble, 5-fixed knob, 6-adjustable universal ball, 7-suction cup. DETAILED DESCRIPTION

[0060] The embodiments of the technical solutions of the present application will be described in detail below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0061] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in the present application should be understood as the usual meanings understood by the skilled person in the field to which the present application belongs.

[0062] EMBODIMENT

[0063] The present embodiment provides a patrol device for use in the patrol of the anti-dazzle failure zone of the central median strip of the highway, as shown in the figure, the patrol device comprises: Figure 1

[0064] An image acquisition module, a rotary encoder and a processing module.

[0065] The image acquisition module is installed at the front of the patrol car through an adjustable support, and after installation, the center optical axis is horizontally forward and the scanning line is perpendicular to the road surface, which is used to acquire real-time images of the anti-dazzle facility.

[0066] The rotary encoder is coaxially installed with the rear wheel of the patrol car, and is also electrically connected with the image sensor in the image acquisition module; the rotary encoder is used to sense the position movement of the car and trigger the image sensor to acquire images. In a specific embodiment, the implementation form of the rotary encoder is not limited, and any implementation form in the prior art can be used. ​

[0067] The processing module is electrically connected with the image acquisition module, and is used for data storage, analysis calculation and man-machine interaction in the inspection process. In the specific embodiment, the implementation form of the processing module is not limited, and any implementation form in the prior art can be used, such as an industrial computer.

[0068] The image acquisition module comprises an adjustable support, a tray, a rotatable platform and an image sensor.

[0069] The adjustable support comprises a suction cup and a support, the suction cup is connected with the inspection vehicle, and the support is connected with the tray; the suction cup and the support are connected through an adjustable universal ball, and a first fixing knob is arranged beside the adjustable universal ball. In the specific embodiment, the implementation form of the suction cup, the support and the adjustable universal ball is not limited, and any implementation form in the prior art can be used; by adjusting the adjustable universal ball, the spatial position and the angle of the tray can be changed.

[0070] The tray is provided with a bubble level at the center of the side edge, and the bubble level is used for calibrating whether the tray is in a horizontal position.

[0071] The rotatable platform is arranged on the top of the tray; a second fixing knob is arranged beside the rotatable platform. In the specific embodiment, the rotatable platform is divided into two left and right parts; the implementation form of the rotatable platform and the second fixing knob is not limited, and any implementation form in the prior art can be used; by tightening the second fixing knob, the rotation angle of the rotatable platform can be fixed at a required angle and kept unchanged.

[0072] The image sensor is fixedly connected with the top of the rotatable platform and is used for image acquisition. In order to facilitate subsequent image processing, in the specific embodiment, a color linear array CCD image sensor is selected for the image sensor. The reason for selecting the linear array image sensor is that the linear array scanning of the highway median strip can be realized by combining the rotation of the rotary encoder during image acquisition. The reason for selecting the color image sensor is that RGB three primary colors are needed in subsequent image analysis. The number of image sensors is at least two, which are arranged on the left and right sides. The reason for using at least two image sensors is that the driver needs to analyze different situations of the highway median strip at the highest speed limit and the lowest speed limit of the highway. The number of image sensors can also be multiple, corresponding to different driving speeds.

[0073] The use instruction of the inspection device is as follows: the suction cup is adsorbed on the front windshield of the inspection vehicle, the adjustable universal ball is adjusted to change the height of the tray, the height is between one meter and one meter and five, the driver's line of sight is used as the reference, the air bubbles in the level bubble are in the middle, and the tray is kept horizontal. After the adjustment is completed, the adjustable universal ball is locked and fixed by using the fastening bolt. The rotatable platform is rotated to the required angle θ, and then the fixing knob is tightened to keep the angle of the rotatable platform consistent during driving, so that the image sensor can collect images in a fixed simulation driver's view angle, and the effect of the anti-dazzle facility is verified.

[0074] The embodiment also provides a method for inspecting the anti-dazzle invalid area of a highway central median strip by using the inspection device described above, as shown in the figure, comprising the following steps: Figure 2

[0075] S1, the horizontal position of the color line array image sensor is adjusted, so that the horizontal projection of the center optical axis of the image sensor and the driving direction is respectively θ1 and θ2.

[0076] The inspection device is fixed on the front windshield of the inspection vehicle by using the suction cup, the adjustable universal ball is rotated to keep the bearing tray in a horizontal state, and the rotatable platform is adjusted so that the horizontal projection of the center optical axis of the two color line array image sensors and the driving direction is respectively θ1 and θ2.

[0077] The driving direction refers to the driving direction of the inspection vehicle. θ1 and θ2 represent the lateral half-view angle of the driver at different driving speeds; the value of θ1 is determined according to the highest speed limit of the road section, the higher the speed limit, the smaller the value of θ1, typically, when the speed limit is 40 km / h, θ1 = 50°, when the speed limit is 70 km / h, θ1 = 33°, and when the speed limit is 100 km / h, θ1 = 20°; the value of θ2 is determined according to the lowest speed limit of the road section, typically, when the speed limit of the highway is 60 km / h, θ2 = 37°, and when the speed limit of the first-class highway is 50 km / h, θ2 = 40°.

[0078] S2, the inspection vehicle drives along the center line of the lane in the lane, and the image sensor scans the central median strip area to collect images of the anti-dazzle facility

[0079] In some embodiments, the inspection vehicle can drive along the center line of the lane in the leftmost lane in the driving direction, the position movement of the car is sensed by the rotary encoder and the image sensor is triggered to collect images, and the images are sent to the processing module.

[0080] S3, the original images collected by the image sensor are extracted to generate key area images

[0081] The original images collected are Img org ​, the generated key area image is Img roi , m, n are calculated by the following formula:

[0082]

[0083] In the above formula, NUM row represents the total number of rows of Img org , L represents the distance from the image sensor to the straight line where the central optical axis intersects the central dividing strip, represents the horizontal half field angle of the image sensor, represents the height of the anti-glare area given by the design file, represents the height of the bottom of the anti-glare area from the road surface given by the design file, represents the height of the image sensor from the road surface, According to , L and , the specific spatial position relationship can be referred to Figure 3 .

[0084] In some embodiments, if is less than the height of the top of the central dividing strip safety guardrail (such as a corrugated beam steel guardrail, a concrete guardrail, etc.) from the road surface , the value of should be taken as .

[0085] S4, for all pixels in the key area image, the average values of R component, G component and B component of RGB three primary colors are calculated respectively; according to the average values, it is judged which one of anti-glare board, anti-glare net or anti-glare plant the anti-glare facility is

[0086] For all pixels in the key area image Img roi , the average value R avg of R component, the average value G avg of G component and the average value B avg of B component of RGB three primary colors are calculated, and according to the average values R avg , G avg , B avg , it is judged which one of anti-glare board, anti-glare net or anti-glare plant the anti-glare facility is.

[0087] The surfaces of anti-glare board and anti-glare net are generally pure green or pure blue; in a specific embodiment, when the surfaces of anti-glare board and anti-glare net are both pure green, the value of G avg will be obviously greater than the values of R avg and B avg , so if one of the following conditions is met, it can be determined that the anti-glare facility is anti-glare board or anti-glare net:

[0088] ① G avg >Thed G 、R avg <Thed R 、B avg <Thed B three conditions are met simultaneously;

[0089] ② G avg -R avg >Thed G-R and G avg -B avg >Thed G-B are met simultaneously;

[0090] wherein Thed G , Thed R , Thed B , Thed G-R , Thed G-B are set threshold values. In a specific embodiment, the above threshold values can be taken as 180, 80, 80, 90, 90, respectively.

[0091] When the surface of the anti-glare board and the anti-glare net are both pure blue, the B avg value will be obviously greater than the R avg value and the G avg value among the three average values, and therefore, if one of the following conditions is met, it can be determined that the anti-glare facility is an anti-glare board or an anti-glare net:

[0092] ① B avg >Thed B , R avg <Thed R , and G avg <Thed G three conditions are met simultaneously;

[0093] ② B avg -R avg >Thed B-R and B avg -G avg >Thed B-G are met simultaneously;

[0094] wherein Thed B , Thed R , Thed G , Thed B-R , Thed B-G are set threshold values. In a specific embodiment, the above threshold values can be taken as 180, 80, 80, 90, 90, respectively. When the anti-glare facility is an anti-glare board or an anti-glare net, the following step 5 is executed;

[0095] If the anti-glare facility does not meet the aforementioned conditions ① or ② according to the average value, it is considered that the anti-glare facility is an anti-glare plant; when the anti-glare facility is an anti-glare plant, step 6 below is executed.

[0096] S5, for the anti-glare plate or the anti-glare net, a color analysis method based on contour extraction is used to detect the anti-glare failure area in the key area image

[0097] The color analysis method based on contour extraction is as follows:

[0098] S51, select a primary color in RGB three primary colors as the main color, and extract the key area image Img roi The main color component of each pixel in the image generates a gray image Img gray , for the gray image Img gray , edge detection, binarization, setting the first and last rows and the first and last columns of the image as edge pixels, dilation operation and erosion operation are sequentially performed to obtain the first edge image Img edge , and all closed contours in the first edge image Img edge are identified.

[0099] S52, for a single closed contour, count the total number of pixels in the area surrounded by the contour and record it as NUM pixel , and calculate the average value AS roi of the main color component of all pixels in the area surrounded by the closed contour in the corresponding key area image Img local , and the average values BS local and CS local of the two non-main color components.

[0100] S53, according to the values of NUM pixel , AS local , BS local , and CS local , judge whether the inside of the closed contour is the surface of the anti-glare plate or the anti-glare net. If not, it is considered that the closed contour is an anti-glare failure area; if yes, it is considered that the closed contour is not an anti-glare failure area.

[0101] In the specific embodiment, the surfaces of the anti-glare plate and the anti-glare net are pure green, and G in RGB three primary colors should be selected as the main color component, and R and B as the non-main color components. If the values of NUM pixel , AS local , BS local , and CS local satisfy one of the following conditions, it is judged that the inside of the closed contour is not the surface of the anti-glare plate or the anti-glare net:

[0102] ① NUM pixel >Thed num , AS localThe AS Simultaneously

[0103] ② NUM pixel The num , max(AS local , BS local , CS local )≠ AS local Simultaneously

[0104] Wherein, Thed num and Thed AS are set threshold values. In a specific embodiment, the above threshold values can be taken as 100, 128, respectively.

[0105] S54, judging whether there is still an unanalyzed closed contour in the current first edge image Img edge ? If yes, go to step S52 to analyze the next closed contour; if no, end the detection process.

[0106] This step uses a color analysis method based on contour extraction to detect the anti-dazzle failure zone in the key region images Img roi corresponding to the two image sensors, respectively.

[0107] S6, for the anti-dazzle plant, using a contour size method based on information entropy to detect the anti-dazzle failure zone in the key region image

[0108] The contour size method based on information entropy is as follows:

[0109] S61, after the key region image Img roi is grayed, it is segmented into a plurality of rectangular gray sub-images Img sub of the same size.

[0110] S62, histogram transformation and normalization processing are performed on a single rectangular gray sub-image Img sub to obtain a probability function p(x) of pixel value x:

[0111]

[0112] Wherein, K represents the total number of pixels of Img sub , k(x) represents the number of pixels of Img sub with value x, NUM grade represents the maximum gray level of the image, and x can take any integer value between 0 and NUM grade .

[0113] S63, the information entropy Ety of the rectangular gray sub-image Img sub is calculated according to the probability function p(x):

[0114]

[0115] wherein λ represents a proportional coefficient, and λ>0; NUM grade represents the maximum level of image gray scale.

[0116] S64, judging whether the information entropy Ety is less than a third set threshold Thed ety If yes, the rectangular gray scale sub-image Img sub is subjected to edge detection, binarization, setting the first and last rows and the first and last columns of pixels as edge pixels, dilation operation and erosion operation in sequence to obtain a second edge image Img bian , and all the closed contours in the second edge image Img bian are identified.

[0117] In a specific embodiment, the third set threshold can be 100.

[0118] S65, counting the number of pixels in the region surrounded by each closed contour, and recording the maximum value of the number of pixels as NUM max ; judging whether the maximum value of the number of pixels NUM max is greater than a fourth set threshold Thed max If yes, the closed contour is considered as an anti-glare failure zone; if not, the closed contour is not considered as an anti-glare failure zone.

[0119] In a specific embodiment, the fourth set threshold is set according to the size and segmentation of the image.

[0120] S66, judging whether there is still an unanalyzed rectangular gray scale sub-image Img roi in the current key region image Img sub ; if yes, the next rectangular gray scale sub-image Img sub is analyzed in step S62; if not, the detection process is ended.

[0121] In this step, the anti-glare failure zone is detected in the key region image Img roi corresponding to each image sensor by using the contour size method based on information entropy.

[0122] In a specific embodiment, the execution order of steps S5 and S6 is not fixed.

[0123] In some embodiments, for the convenience of subsequent work, the anti-glare failure zones detected in steps S5 and S6 can be marked, and then all the location post number information of the detected road section marked as the anti-glare failure zone of the central divider is counted and output.

[0124] The technical scheme provided by the embodiment is suitable for failure area detection of various types of anti-dazzle facilities; the inspection device is portable, easy to install and disassemble, and does not need to be limited to the type of the inspection vehicle when in use, and is highly adaptable. The inspection method has a high degree of automation, and the determination of the anti-dazzle failure area of the central separation belt is objective and safe.

[0125] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the description of the present application.

Claims

1. A highway median anti-glare failure zone inspection method, characterized in that, The method comprises the following steps: Adjusting the horizontal position of the color line array image sensor so that the horizontal projection of the center optical axis of the image sensor and the driving direction forms an angle equal to the lateral half field of view of the driver at different driving speeds; Driving the inspection vehicle along the center line of the lane and scanning the central divider area through the image sensor to collect images of the anti-glare facility; Extracting all pixels between the mth row and the nth row from top to bottom in the original image collected by the image sensor to generate a key area image; m and n are calculated by the following formula: In the above formula, NUM row represents the total number of rows of the original image, L represents the straight-line distance from the image sensor to the point where the central optical axis intersects the central dividing strip, represents the horizontal half field angle of the image sensor, represents the height of the anti-dazzle area given by the design file, represents the height of the anti-dazzle area bottom from the road surface given by the design file, represents the height of the image sensor from the road surface, According to , L and determined by the trigonometric relationship; Calculating the average value of each component of the RGB three primary colors for all pixels in the key area image; and judging the type of the anti-glare facility according to the average value; For the anti-glare board or the anti-glare net, a color analysis method based on contour extraction is used to detect the anti-glare failure area in the key area image; for the anti-glare plant, a contour size method based on information entropy is used to detect the anti-glare failure area in the key area image.

2. The method of claim 1, wherein, The different driving speeds are the maximum speed limit of the road section and the minimum speed limit of the road section.

3. The method of claim 1, wherein, If Height of guardrail top above road surface Then The value of is adopted.

4. The method of claim 1, wherein, The anti-glare facility includes an anti-glare board or an anti-glare net, and the surface color of the anti-glare board and the anti-glare net includes pure green and pure blue; When the surface of the anti-glare board and the anti-glare net is pure green, if one of the following conditions is met, the anti-glare facility can be determined as an anti-glare board or an anti-glare net: G avg >Thed G 、R avg <Thed R 、B avg <Thed B three conditions are met simultaneously; or G avg - R avg > The d G-R and G avg - B avg > The d G-B Simultaneously; When the surface of the anti-glare board and the anti-glare net is pure blue, if one of the following conditions is met, the anti-glare facility can be determined as an anti-glare board or an anti-glare net: B avg >Thed B 、R avg <Thed R 、G avg <Thed G three conditions are met simultaneously; or B avg - R avg > The d B-R and B avg - G avg > The d B-G Simultaneously wherein R avg is the average value of the R component, G avg is the average value of the G component, and B avg is the average value of the B component; Thed G , Thed R , Thed B , Thed G-R , Thed G-B are set threshold values.

5. The method of claim 1, wherein, The color analysis method based on contour extraction comprises: Selecting a primary color in the RGB three primary colors as the main color, extracting the main color component of each pixel in the key area image to generate a gray-scale image, and sequentially performing edge detection, binarization, setting the first and last rows and the first and last columns of the image as edge pixels, dilation operation and erosion operation on the gray-scale image to obtain a first edge image; and identifying all closed contours in the first edge image; For a single closed contour, the total number of pixels in the enclosed area is counted, and the average value of the main color component and the average value of the two non-main color components of all pixels in the enclosed area of the closed contour are calculated in the corresponding key area image; According to the total number of pixels, the average value of the main color component and the average value of the two non-main color components, it is judged whether the inside of the closed contour is the surface of the anti-glare board or the anti-glare net; if not, the closed contour is an anti-glare failure area.

6. The method of claim 5, wherein, When one of the following conditions is met, it can be determined that the inside of the closed contour is not the surface of the anti-glare board or the anti-glare net: NUM pixel >Thed num 、AS local <Thed AS simultaneously; or NUM pixel >Thed num 、max(AS local 、BS local 、CS local )≠AS local simultaneously Among them, NUM pixel This represents the total number of pixels within the area enclosed by a single closed contour; AS local AS represents the average value of the primary color components of all pixels within the area enclosed by the closed-loop outline. local BS local CS loca Thed represents the average value of all non-primary color components of all pixels within the area enclosed by the closed-loop outline; num and Thed AS All of these are set thresholds.

7. The method of claim 1, wherein, The contour size method based on information entropy comprises: The key area image is grayed and divided into a plurality of rectangular gray-scale sub-images with the same size; For a single rectangular gray-scale sub-image, a histogram transformation and a normalization processing are performed to obtain a probability function of pixel value; The information entropy of the rectangular gray-scale sub-image is calculated according to the probability function; When the information entropy is less than a third set threshold, the rectangular gray-scale sub-image is sequentially subjected to edge detection, binarization, setting the first and last rows and the first and last columns of the image as edge pixels, dilation operation and erosion operation to obtain a second edge image, and all closed contours in the second edge image are identified; Count the number of pixels in the area enclosed by each closed contour, and when the maximum pixel number is greater than a fourth set threshold, determine that the closed contour is a glare failure zone.

8. The method of claim 1, wherein, Also includes: After detecting the glare failure zone, mark the processing, statistics and output the marked as glare failure zone of the test section of the entire location post information.

9. An inspection device for use in the inspection method of any one of claims 1 to 8, characterized in that Including: Image acquisition module, rotary encoder and processing module; The image acquisition module is installed in the front of the inspection car through an adjustable support, and the center optical axis is horizontal and forward after installation, and the scanning line is perpendicular to the road surface, which is used for collecting images of glare facilities; The rotary encoder is coaxially installed with the rear wheel of the inspection car and is also electrically connected with the image sensor in the image acquisition module; the rotary encoder is used for sensing the position movement of the car and triggering the image sensor to collect images; The processing module is electrically connected with the image acquisition module, and is used for data storage, analysis and calculation and man-machine interaction in the inspection process.

10. The patrol device according to claim 9, wherein The image acquisition module comprises an adjustable support, a tray, a rotatable platform and an image sensor. The adjustable support comprises a suction cup and a support, the suction cup is connected with the inspection car, and the support is connected with the tray; the suction cup and the support are connected through an adjustable universal ball, and a first fixed knob is arranged beside the adjustable universal ball; A level bubble is arranged at the center position of the side edge of the tray, and the level bubble is used for calibrating whether the tray is in a horizontal position; The rotatable platform is arranged on the top of the tray, and a second fixed knob is arranged beside the rotatable platform; The image sensor is fixedly connected with the top of the rotatable platform and is used for image acquisition.

Images