A method and device for classifying, identifying and measuring the brightness of tunnel luminous signs

Abstract

The present invention relates to the field of electromechanical detection technology for highway tunnels, and specifically to a method and device for classifying, identifying, and measuring the brightness of tunnel luminous signs. The method comprises: acquiring images inside and outside the tunnel to obtain a plurality of tunnel image frames; the tunnel images include tunnel linear array images and tunnel area array images; acquiring a starting point outside the tunnel based on the tunnel linear array images; determining a key image based on the tunnel area array images and the starting point outside the tunnel; acquiring the type of tunnel sign based on the tunnel linear array images and the key image; the types of tunnel signs include variable signs and fire signs; and determining the brightness of the variable signs and fire signs, respectively, based on the key image. The acquired tunnel images are used to acquire the type of tunnel sign, and then the brightness of the variable signs and fire signs is acquired, thereby achieving full coverage testing of the luminous signs in the tunnel. The brightness of the luminous signs in the tunnel does not need to be measured one by one, thereby improving detection efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of electromechanical detection of highway tunnels, and in particular to a method and device for classifying, identifying and measuring the brightness of tunnel luminous signs. Background Art

[0002] Tunnels are prone to accidents due to dim lighting and poor visibility. Therefore, they are often equipped with numerous signboards, including variable tunnel signs for information and fire signs for fire safety. To better alert drivers, these signs are designed to be luminous. Tunnel maintenance personnel and inspectors also test the brightness of these signs when inspecting tunnel electromechanical equipment to ensure they meet applicable standards.

[0003] The brightness of illuminated signs in tunnels (except for evacuation signs) should be measured at a stop sight distance in front of them; otherwise, the measured data is meaningless. The current common practice is to manually identify each type of illuminated sign on foot, while ensuring proper tunnel traffic management, and then measure each sign individually using a spot-tracing luminance meter installed on the road surface at a stop sight distance from the sign. This is inefficient and fails to fully measure the illuminated area of ​​each illuminated sign. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention proposes a method and device for classifying, identifying and measuring the brightness of tunnel luminous signs to solve the technical problems in the existing technology of low brightness measurement efficiency and inability to fully cover the luminous areas of each luminous sign.

[0005] In a first aspect, the present invention provides a method for classifying, identifying and measuring the brightness of tunnel luminous signs.

[0006] In a first possible implementation, a method for classifying, identifying, and measuring the brightness of tunnel luminous signs includes:

[0007] Capturing images inside and outside the tunnel to obtain a plurality of frames of tunnel images; the tunnel images include tunnel linear array images and tunnel area array images;

[0008] Obtain the starting point outside the tunnel according to the tunnel linear array image;

[0009] Determine the key image based on the tunnel array image and the starting point position outside the tunnel;

[0010] Acquire the type of tunnel sign according to the tunnel linear array image and the key image; the type of tunnel sign includes a variable sign and a fire sign;

[0011] The brightness of the variable sign and the fire sign are determined separately according to the key image.

[0012] In combination with the first possible implementation, in the second possible implementation, collecting images inside and outside the tunnel to obtain a tunnel image includes:

[0013] A forward color area array image sensor and a black-and-white linear array image sensor are installed on the roof of the tunnel inspection vehicle, and a side color area array image sensor and a color linear array image sensor are installed on the side of the vehicle. The inspection vehicle starts collecting data from the tunnel entrance until it leaves the tunnel, obtaining several frames of black-and-white linear array images, color linear array images, forward color area array images, and side color area array images.

[0014] In combination with the first implementable manner, in a third implementable manner, obtaining the starting point position outside the tunnel according to the tunnel linear array image includes:

[0015] Performing feature processing on the tunnel linear array image to obtain a linear array feature image;

[0016] Detect straight line segments and circular arc segments in linear array feature images;

[0017] Traversing all straight line segments and circular arc segments, if there is a straight line segment or circular arc segment that passes through all columns of the linear array feature image and whose row span is less than a first preset threshold, then determining the straight line segment or circular arc segment that passes through all columns of the linear array feature image and whose row span is less than the first preset threshold as a tunnel entrance, and obtaining the row label of the tunnel entrance in the linear array feature image;

[0018] The starting point outside the tunnel is obtained based on the displacement of the detection vehicle between the tunnel entrance and two adjacent rows of pixels.

[0019] In combination with the second implementable manner, in a fourth implementable manner, determining a key image based on the tunnel array image and the starting position outside the tunnel includes:

[0020] Identify the tunnel sign rectangular area in the tunnel linear array image;

[0021] Obtaining a first distance from the tunnel sign to the starting position outside the tunnel according to the rectangular area of ​​the tunnel sign; the tunnel sign includes a variable sign and a fire sign;

[0022] For the variable sign, a first key image is selected based on the first distance and the displacement of the detection vehicle between two adjacent frames of the forward color area array image; the first key image is a forward color area array image captured at a stopping sight distance behind the driving direction of the sign;

[0023] For the fire sign, the second key image is selected based on the first distance and the displacement of the detection vehicle between two adjacent frames of the lateral color area array image; the second key image is a lateral color area array image collected at a stopping sight distance behind the driving direction of the sign.

[0024] In combination with the fourth possible implementation, in a fifth possible implementation, identifying a tunnel sign rectangular area in a tunnel linear array image includes:

[0025] Identify all corner points in the tunnel linear array image;

[0026] Get all rectangles formed by corner points;

[0027] Determine a rectangle that meets the first preset condition as a tunnel sign rectangular area;

[0028] The first precondition is:

[0029] Among them, H is the actual thickness of the sign board, S 行间 To detect vehicle displacement between adjacent pixels, x1 and x3 are the row and third corner numbers of the rectangle, y1 and y2 are the column and first corner numbers of the rectangle, and T1, T2, and T3 are all constants.

[0030] In combination with the fifth possible implementation, in a sixth possible implementation, obtaining the type of tunnel sign according to the tunnel linear array image and the key image includes:

[0031] Acquire a first tunnel sign rectangle in the black-and-white tunnel linear image, determine the first tunnel sign rectangle that meets a first preset condition as a variable tunnel sign, and determine the type of the variable sign based on the actual thickness of the sign panel and the range of spacing between adjacent column numbers of the rectangle;

[0032] Acquire a second tunnel sign rectangle in the tunnel color line array image, and determine the second tunnel sign rectangle that satisfies both the first preset condition and the second preset condition as a tunnel fire sign;

[0033] The second preset condition is: T7≤|y1+y2| / 2≤T8; where T7 and T8 are both constants;

[0034] Determine the fire characteristics of the tunnel fire signs, including whether there is a horizontal arrow, whether there are two figures, whether there is a vertical long arc, and whether there are two vertical straight line segments that meet preset conditions;

[0035] The type of tunnel fire sign is identified based on the fire feature judgment results.

[0036] In combination with the first implementable manner, in a seventh implementable manner, respectively determining the brightness of the variable sign and the fire sign according to the key image includes:

[0037] In the case where the tunnel sign is a variable sign, a variable sign surface is determined from the first key image; a brightness mapping model and a chromaticity mapping model are constructed, and based on the brightness mapping model and the chromaticity mapping model, the brightness and chromaticity of the variable sign are obtained using the variable sign surface;

[0038] In the case that the tunnel sign is a fire sign, the fire sign surface is determined from the second key image; the white area binary template and the green area binary template of the fire sign are determined; the white area binary template and the green area binary template are scaled to the same size as the fire sign surface; the white area binary template and the green area binary template of the same size are respectively multiplied by the corresponding elements of the fire sign surface to obtain a white fire product image and a green fire product image; based on the brightness mapping model, the average brightness of all non-zero pixels in the white fire image and the green fire image is obtained, and the average brightness of the white fire image and the green fire image is used as the brightness of the white area and the green area in the fire sign, respectively.

[0039] In combination with the seventh possible implementation, in an eighth possible implementation, based on the brightness mapping model and the chromaticity mapping model, obtaining the brightness and chromaticity of the variable sign using the variable sign surface includes:

[0040] For information board signs, the brightness and chromaticity are directly obtained using the variable sign board area based on the brightness mapping model and chromaticity mapping module;

[0041] For lane indicator signs, a two-dimensional template of the lane indicator panel pattern is pre-built and scaled to the same size as the variable sign panel. The two-dimensional template of the lane indicator panel pattern of the same size is multiplied with the pixels at corresponding positions on the variable sign panel to obtain a variable product image. Based on the brightness mapping model and the chromaticity mapping module, the variable product image is used to obtain the brightness and chromaticity.

[0042] In combination with the third possible implementation, the ninth possible implementation further includes:

[0043] The tunnel test length is obtained based on the tunnel entrance and exit in the linear array feature image and the displacement of the detection vehicle between adjacent pixels;

[0044] Obtaining a position correction factor based on the tunnel test length and the actual tunnel length;

[0045] Correct the position of each tunnel sign according to the position correction coefficient;

[0046] The corrected positions, brightness and chromaticity of each tunnel sign are plotted into a luminous sign information layout diagram in the tunnel.

[0047] In a second aspect, the present invention provides a device for classifying, identifying and measuring the brightness of tunnel luminous signs.

[0048] In a tenth possible implementation, a device for classifying, identifying, and measuring the brightness of tunnel luminous signs includes:

[0049] Inspection vehicle,

[0050] A forward-facing color area array image sensor is mounted on the roof of the inspection vehicle. After installation, the central optical axis of the forward-facing color area array image sensor is projected onto the road surface parallel to the road surface centerline, and the horizontal distance from the intersection of the central optical axis and the tunnel vault to the inspection vehicle is one stopping sight distance. The forward-facing color area array image sensor is used to capture images of the suspended information board and lane indicator.

[0051] A lateral color area array image sensor is mounted on the side of the inspection vehicle. After installation, the central optical axis of the lateral color area array image sensor points toward the tunnel side wall at a height of 2.5 meters from the ground, and the horizontal distance from the intersection of the central optical axis and the tunnel side wall to the inspection vehicle is one stopping sight distance. The lateral color area array image sensor is used to capture images of various fire electro-optical signs with panels perpendicular to the side wall.

[0052] A color line array image sensor is installed on the side of the inspection vehicle. After installation, the central optical axis of the color line array image sensor is projected on the road surface perpendicular to the center line of the road surface. The color line array image sensor is used to collect images of the tunnel side walls.

[0053] A black and white linear array image sensor is installed on the roof of the inspection vehicle, with the central optical axis of the black and white linear array image sensor pointing vertically upward after installation, and the black and white linear array image sensor is used to capture images directly above the inspection vehicle;

[0054] A rotary encoder is coaxially mounted with the rear wheel of the inspection vehicle. The rotary encoder is used to sense the displacement of the inspection vehicle and send a pulse signal. The rotary encoder is used to directly trigger the color linear array image sensor and the black and white linear array image sensor to capture images.

[0055] a pulse modulator, configured to receive the pulse signal emitted by the rotary encoder and perform frequency reduction processing on the pulse signal to obtain a frequency-reduced pulse signal, wherein the frequency-reduced pulse signal is used to trigger the forward color area array image sensor and the side color area array image sensor to capture an image;

[0056] A control processing unit is electrically connected to each image sensor, and is used to execute the above-mentioned method for classifying, identifying and measuring the brightness of tunnel luminous signs.

[0057] It can be seen from the above technical solution that the beneficial technical effects of the present invention are as follows:

[0058] 1. The inspection vehicle collects images inside and outside the tunnel, uses the collected tunnel images to determine the type of tunnel signs, and then obtains the brightness of variable signs and fire signs. This can achieve full coverage testing of luminous signs in the tunnel, eliminating the need to measure the brightness of each luminous sign in the tunnel one by one, improving inspection efficiency and solving the technical problems of low brightness measurement efficiency and the inability to fully cover the luminous area of ​​each luminous sign in the existing technology.

[0059] 2. Using tunnel images, a color image captured at one stop sight distance behind the sign in the direction of travel is used as the key image. This allows accurate luminance and chromaticity measurement to be performed at one stop sight distance in front of the illuminated sign (excluding evacuation signs), improving measurement accuracy. This solution also enables automated measurement, making it efficient, objective, and safe. The inspection process does not require road closures or traffic interruptions, and does not disrupt normal tunnel traffic. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0061] Figure 1 A schematic diagram of a method for classifying, identifying, and measuring the brightness of tunnel luminous signs provided in this embodiment;

[0062] Figure 2 A schematic diagram of the structure of a device for classifying, identifying, and measuring the brightness of tunnel luminous signs provided in this embodiment;

[0063] Figure 3-A A binary template image of a green arrow of a two-dimensional template of a lane indicator panel pattern provided in this embodiment;

[0064] Figure 3-B A binary template image of a red cross of a two-dimensional template of a lane indicator plate pattern provided in this embodiment;

[0065] Figure 4 A flow chart for identifying the type of tunnel fire signs provided in this embodiment;

[0066] Figure 5 A layout diagram of a luminous sign provided in this embodiment;

[0067] Reference numerals:

[0068] 1- Forward color area array image sensor, 2- Lateral color area array image sensor, 3- Color linear array image sensor, 4- Black and white linear array image sensor, 5- Rotary encoder, 6- Pulse modulator, 7- Control processing unit. DETAILED DESCRIPTION

[0069] The following embodiments of the technical solution of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore only examples and are not intended to limit the scope of protection of the present invention.

[0070] It should be noted that, unless otherwise specified, the technical or scientific terms used in this application should have the common meanings understood by those skilled in the art to which this invention belongs. The terms "first," "second," and so on, in the description and claims of the embodiments of the present disclosure, and in the accompanying drawings, are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate to facilitate the implementation of the embodiments of the present disclosure described herein. In addition, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. Unless otherwise specified, the term "plurality" means two or more. In the embodiments of the present disclosure, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B means: A or B. The term "and / or" describes an association relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or, A and B. The term "corresponding" can refer to an association relationship or a binding relationship. A and B corresponding means that there is an association relationship or a binding relationship between A and B.

[0071] Combine Figure 1 As shown, this embodiment provides a method for classifying and identifying tunnel luminous signs and measuring brightness, including:

[0072] Step S01: Capture images inside and outside the tunnel to obtain a plurality of frames of tunnel images; the tunnel images include tunnel linear array images and tunnel area array images;

[0073] Step S02: obtaining the starting point outside the tunnel according to the tunnel linear array image;

[0074] Step S03: determining a key image based on the tunnel array image and the starting position outside the tunnel;

[0075] Step S04: acquiring the type of tunnel sign according to the tunnel linear array image and the key image; the type of tunnel sign includes a variable sign and a fire sign;

[0076] Step S05: Determine the brightness of the variable sign and the fire sign respectively according to the key image.

[0077] Optionally, collecting images inside and outside the tunnel to obtain tunnel images includes:

[0078] A forward color area array image sensor and a black-and-white linear array image sensor are installed on the roof of the tunnel inspection vehicle, and a side color area array image sensor and a color linear array image sensor are installed on the side of the vehicle. The inspection vehicle starts collecting data from the tunnel entrance until it leaves the tunnel, obtaining several frames of black-and-white linear array images, color linear array images, forward color area array images, and side color area array images.

[0079] In some embodiments, the detection vehicle starts collecting images when it is hundreds of meters away from the tunnel entrance. While driving in the tunnel, the forward color area array image sensor, the side color area array image sensor, the color line array image sensor, and the black and white line array image sensor are triggered by pulse signals to collect images. The detection vehicle stops collecting after it leaves the tunnel, and finally obtains several frames of black and white line array images, forward color area array images, side color area array images, and color line array images.

[0080] This embodiment provides a device for classifying, identifying, and measuring the brightness of tunnel luminous signs, including:

[0081] Inspection vehicle,

[0082] A forward-facing color area array image sensor is mounted on the roof of the inspection vehicle. After installation, the central optical axis of the sensor is projected onto the road surface parallel to the road centerline. The horizontal distance from the intersection of the central optical axis and the tunnel vault to the inspection vehicle is one stopping sight distance. The forward-facing color area array image sensor is used to capture images of the suspended information board and lane indicator.

[0083] The lateral color area array image sensor is installed on the side of the inspection vehicle. After installation, the central optical axis of the lateral color area array image sensor points to the tunnel side wall at a height of 2.5 meters from the ground. The horizontal distance from the intersection of the central optical axis and the tunnel side wall to the inspection vehicle is one stopping sight distance. The lateral color area array image sensor is used to collect images of various fire electro-optical signs installed perpendicular to the side wall.

[0084] A color linear array image sensor is installed on the side of the inspection vehicle. After installation, the central optical axis of the color linear array image sensor is projected perpendicular to the road centerline. The color linear array image sensor is used to capture images of the tunnel side walls.

[0085] A black and white linear array image sensor is installed on the roof of the inspection vehicle. After installation, the central optical axis of the black and white linear array image sensor is vertically upward. The black and white linear array image sensor is used to capture images directly above the inspection vehicle.

[0086] A rotary encoder is coaxially mounted with the rear wheel of the inspection vehicle. The rotary encoder senses the displacement of the inspection vehicle and sends a pulse signal. The rotary encoder is used to directly trigger the color linear array image sensor and the black and white linear array image sensor to capture images.

[0087] a pulse modulator for receiving a pulse signal from the rotary encoder and performing frequency reduction processing on the pulse signal to obtain a frequency-reduced pulse signal, wherein the frequency-reduced pulse signal is used to trigger the forward color area array image sensor and the side color area array image sensor to acquire an image;

[0088] a control processing unit electrically connected to each image sensor, the control processing unit being configured to execute any of the methods for classifying, identifying, and measuring brightness of tunnel luminous signs as described above and below;

[0089] In some embodiments, combined Figure 2 As shown, a device for classifying, identifying, and measuring the brightness of tunnel luminous signs comprises a forward color area array image sensor 1, a lateral color area array image sensor 2, a color line array image sensor 3, a black-and-white line array image sensor 4, a rotary encoder 5, a pulse modulator 6, and a control processing unit 7. The rotary encoder 5 is connected to one end of the color line array image sensor 3, one end of the black-and-white line array image sensor 4, and one end of the pulse modulator 6, respectively. The other end of the pulse modulator 6 is connected to one end of the forward color area array image sensor 1 and one end of the lateral color area array image sensor 2, respectively. The other ends of the color line array image sensor 3, the other ends of the black-and-white line array image sensor 4, the other ends of the forward color area array image sensor 1, and the other ends of the lateral color area array image sensor 2 are respectively connected to the control processing unit 7.

[0090] Optionally, the control processing unit is used for image storage and analysis, luminous sign recognition and positioning, brightness and chromaticity calculation, drawing of luminous sign layout in the tunnel, and human-computer interaction.

[0091] Optionally, before obtaining the starting position outside the tunnel based on the black and white linear array image of the tunnel, it also includes: calibrating the vehicle displacement detection between two adjacent frames of the forward color area array image sensor, the vehicle displacement detection between two adjacent frames of the lateral color area array image sensor, the vehicle displacement detection between two adjacent rows of pixels of the color linear array image sensor, and the vehicle displacement detection between two adjacent rows of pixels of the black and white linear array image sensor.

[0092] Optionally, the displacement of the detected vehicle between two adjacent frames of the area array image sensor is calculated using the following formula:

[0093] S 帧间 =L 车轮 *N 帧间 / N 调制 ; Among them, S 帧间 The displacement of the vehicle detected between two adjacent frames of the area array image sensor, L 车轮The circumference of the wheel on which the encoder is installed, N 帧间 Set the number of trigger pulses between two adjacent frames of the area array image sensor, N 调制 The number of pulses output by the pulse modulator when the wheel rotates one circle.

[0094] Optionally, the displacement of the detected vehicle between two adjacent rows of pixels of the linear array image sensor is calculated using the following formula:

[0095] S 行间 =L 车轮 / N 编码 ; Among them, S 行间 The displacement of the vehicle is detected between two adjacent rows of pixels of the linear array image sensor, L 车轮 The circumference of the wheel on which the encoder is installed, N 编码 It is the number of pulses output by the encoder when the wheel rotates one circle.

[0096] Optionally, obtaining the starting point position outside the tunnel based on the tunnel linear array image includes: performing feature processing on the tunnel linear array image to obtain a linear array feature image; detecting straight line segments and circular arc segments in the linear array feature image; traversing all straight line segments and circular arc segments, and if there is a straight line segment or circular arc segment that runs through all columns of the linear array feature image and has a row span less than a first preset threshold, then determining the straight line segment or circular arc segment that runs through all columns of the linear array feature image and has a row span less than the first preset threshold as the tunnel entrance, and obtaining the row label of the tunnel entrance in the linear array feature image; obtaining the starting point position outside the tunnel based on the displacement of the detection vehicle between the tunnel entrance and two adjacent rows of pixels.

[0097] Optionally, performing feature processing on the tunnel linear array image to obtain the linear array feature image includes: performing edge detection, dilation and corrosion, binarization and other processing on the linear array image to obtain the linear array feature image.

[0098] Optionally, performing feature processing on the black and white line array image of the tunnel to obtain the line array feature image includes: performing edge detection, expansion and corrosion, binarization and other processing on the black and white line array image to obtain the black and white line array feature image Edge 黑阵 .

[0099] Optionally, performing feature processing on the tunnel color line array image to obtain the line array feature image includes: performing edge detection, expansion and corrosion, binarization and other processing on the color line array image to obtain the color line array feature image Edge 彩阵 .

[0100] In some embodiments, the Hough transform algorithm is used to detect straight line segments and circular arc segments in the linear array feature image, and the first straight line segment or circular arc segment that runs through all columns and whose row span is less than a set threshold is searched from the first row. The found straight line segment or circular arc segment is determined as the tunnel entrance and its center row is labeled as k. 入口, use the formula to calculate the distance value outside the tunnel, and determine the position of the distance value from the tunnel entrance to the outside of the tunnel as the starting point outside the tunnel.

[0101] Alternatively, using the formula L 洞外 =k 入口 ·S 行间 Calculate the distance outside the cave; where L 洞外 is the distance outside the cave, k 入口 is the center row number of the tunnel entrance, S 行间 Detect the displacement of the car between two adjacent rows of pixels.

[0102] Optionally, a key image is determined based on the tunnel area array image and the starting position outside the tunnel, including: identifying a rectangular area of ​​a tunnel sign in the tunnel linear array image; obtaining a first distance from the tunnel sign to the starting position outside the tunnel based on the rectangular area of ​​the tunnel sign; the tunnel sign includes a variable sign and a fire sign; for the variable sign, a first key image is filtered based on the first distance and the displacement of the detection vehicle between two adjacent frames of the forward color area array image; the first key image is a forward color area array image captured at a stopping sight distance behind the driving direction of the sign; for the fire sign, a second key image is filtered based on the first distance and the displacement of the detection vehicle between two adjacent frames of the lateral color area array image; the second key image is a lateral color area array image captured at a stopping sight distance behind the driving direction of the sign.

[0103] Optionally, identifying the tunnel marking rectangular area in the tunnel linear array image includes: identifying all corner points in the tunnel linear array image; obtaining all rectangles formed by the corner points; and determining a rectangle that meets a first preset condition as the tunnel marking rectangular area;

[0104] The first precondition is:

[0105] Among them, H is the actual thickness of the sign board, S 行间 To detect vehicle displacement between adjacent pixels, x1 and x3 are the row and third corner numbers of the rectangle, y1 and y2 are the column and first corner numbers of the rectangle, and T1, T2, and T3 are all constants.

[0106] In some embodiments, a corner detection algorithm is used to identify all corners in the line array image.

[0107] Optionally, obtaining all rectangles formed by corner points includes: selecting any four corner points cp1, cp2, cp3, cp4, whose row labels in the image are x cp1 、x cp2 、x cp3 、x cp4 , the column labels are y cp1 、y cp2 、ycp3 、y cp4 , the four corner points are considered to form a rectangle if and only if they simultaneously satisfy the following two conditions, and all rectangles can be found from this:

[0108] ①|x cp1 -x cp2 |≤Thed0 and|x cp3 -x cp4 |≤Thed0

[0109] ②|y cp1 -y cp3 |≤Thed0 and|y cp2 -y cp4 |≤Thed0

[0110] Wherein, Thed0 is a constant obtained by experiment. In some embodiments, Thed0 is 3.

[0111] Optionally, obtaining the first distance from the tunnel sign to the starting point outside the tunnel according to the tunnel sign rectangular area includes: using the formula L=0.5*(x cp1 +x cp3 )·S 行间 Get the first distance from the tunnel sign to the starting point outside the tunnel; where L is the first distance, x cp1 and x cp3 are the row numbers of the first and third corner points of the tunnel marking rectangle, S 行间 Detect vehicle displacement between adjacent pixels.

[0112] In some embodiments, tunnel luminous signs are divided into two categories: variable information signs and fire electro-optical signs. Variable information signs include information boards and lane indicators. Both are suspended from the tunnel vault with the board facing the direction of oncoming vehicles, and the board thickness is approximately 15cm to 30cm. Fire electro-optical signs include evacuation signs, fire extinguisher signs, emergency telephone signs, emergency parking lane signs, vehicle crosswalk signs, and pedestrian crosswalk signs. Except for the evacuation signs, which are installed close to the tunnel side walls, the rest are installed perpendicular to the side walls. Typically, a single tunnel has approximately 40 evacuation signs, approximately 20 fire extinguisher signs, and approximately 15 other luminous signs per kilometer.

[0113] Optionally, a first tunnel sign rectangle is obtained in the black and white linear array image of the tunnel, and the first tunnel sign rectangle that meets the first preset condition is determined as a tunnel variable sign; and the type of the variable sign is determined according to the actual thickness of the sign plate and the range of the spacing between adjacent column labels of the rectangle; a second tunnel sign rectangle is obtained in the color linear array image of the tunnel, and the second tunnel sign rectangle that meets both the first preset condition and the second preset condition is determined as a tunnel fire sign; the second preset condition is: T7≤|y1+y2| / 2≤T8; wherein T7 and T8 are both constants; a fire feature judgment is performed on the tunnel fire sign, and the fire feature includes whether there is a horizontal arrow, whether there are two figures, whether there is a vertical long arc, and whether there are two vertical straight line segments that meet the preset conditions; and the type of the tunnel fire sign is identified according to the fire feature judgment result.

[0114] In some embodiments, a first distance from each variable mark to a starting point outside the tunnel is identified based on the black-and-white line array image, and then, based on the first distance from each variable mark to the starting point outside the tunnel and the vehicle displacement detected between two adjacent frames of the forward color area array image sensor, a key image Img captured by the forward color area array image sensor at a stopping sight distance behind the variable mark in the driving direction is found. 可变 .

[0115] In some embodiments, a first distance from each fire sign to a starting point outside the tunnel is identified based on the color line array image, and then, based on the first distance from each fire sign to the starting point outside the tunnel and the vehicle displacement detected between two adjacent frames of the lateral color area array image sensor, an image Img captured by the lateral color area array image sensor at a stopping sight distance behind the fire sign in the driving direction is searched. 消防 .

[0116] Optionally, the variable sign types include lane indicators and information boards. If the actual thickness of the sign board and the range between adjacent column numbers of the rectangle meet a preset lane indicator condition equation, the rectangle is determined to be a rectangle in the bottom area of ​​a lane indicator, and the rectangle is determined to be a lane indicator sign. If the preset lane indicator condition equation is not met, the determination is continued as to whether the rectangle meets a preset information board condition equation. If so, the rectangle is determined to be in the bottom area of ​​an information board, and is therefore a information board sign.

[0117] Optionally, the preset lane indicator condition equation is as follows:

[0118] Among them, H 车指器 is the actual thickness of the lane indicator, T 21 and T 31 is a constant obtained by experiment. In some embodiments, T1 is 5.

[0119] Optionally, the preset information board condition equation is as follows:

[0120] Among them, H 情报板 is the actual thickness of the information board, T1, T 22 and T 32 is a constant obtained through experiments.

[0121] Optionally, it is determined whether the four corner points of the rectangle satisfy the following conditions at the same time. If so, the rectangle is determined to be a side rectangular area of ​​the fire sign, and the side rectangular area contains the fire sign:

[0122] Among them, H 消防标 is the actual thickness of the fire sign, T4, T5, T6, T7 and T8 are constants obtained through experiments, and in some embodiments, T4 is 5.

[0123] Optionally, the brightness of the variable sign and the fire sign is determined separately according to the key image, including: when the tunnel sign is a variable sign, determining the variable sign board from the first key image; constructing a brightness mapping model and a chromaticity mapping model, and based on the brightness mapping model and the chromaticity mapping model, using the variable sign board to obtain the brightness and chromaticity of the variable sign.

[0124] Optionally, determining the variable sign surface from the first key image includes: gray-scaling the first key image corresponding to the variable sign, performing edge detection, dilation and corrosion, binarization, etc. on the gray-scaled image to obtain a variable key feature image; determining a rectangular area in the variable key feature image, and determining the largest rectangular area in the variable key feature image as the variable sign surface, where the size of the variable sign surface is m×n pixels.

[0125] Optionally, based on the brightness mapping model and the chromaticity mapping model, obtaining the brightness and chromaticity of the variable sign using the variable sign surface includes: for the information board sign, directly obtaining the brightness and chromaticity using the variable sign surface according to the brightness mapping model and the chromaticity mapping module; for the lane indicator sign, pre-constructing a two-dimensional template of the lane indicator surface pattern, scaling the two-dimensional template of the lane indicator surface pattern to the same size as the variable sign surface area, multiplying the two-dimensional template of the lane indicator surface pattern of the same size with the pixels at corresponding positions in the variable sign surface area to obtain a variable product image; according to the brightness mapping model and the chromaticity mapping module, obtaining the brightness and chromaticity using the variable product image.

[0126] In some embodiments, a brightness mapping model between the grayscale of the image of the forward color area array image sensor and the actual brightness, as well as a chromaticity mapping model between the three primary color components (R, G, B) of the image pixel and the chromaticity coordinates (p, q) are pre-built. For the information board sign, the variable key feature image Img is calculated based on the brightness mapping model.可变 The actual average brightness of the variable sign board is used as the brightness measurement result of the information board; Img is calculated based on the chromaticity mapping model 可变 The average value of the chromaticity coordinates of the variable sign board surface in the image is used as the chromaticity measurement result of the information board.

[0127] In some embodiments, for the lane indicator sign, the binary template of the green arrow of the lane indicator plate pattern two-dimensional template is as follows: Figure 3-A As shown, the binary template of the red cross is as follows Figure 3-B As shown. Calculate the variable product image Img according to the brightness mapping model 可变乘积 The actual average brightness of all non-zero pixels in is used as the lane indicator brightness measurement result; the variable image Img is calculated according to the chromaticity mapping model 可变 Img 可变乘积 The average value of the chromaticity coordinates of all pixels corresponding to the non-zero pixels is taken as the lane indicator chromaticity measurement result.

[0128] Optionally, the brightness of the variable sign and the fire sign is determined separately according to the key image, including: when the tunnel sign is a fire sign, determining the fire sign board from the second key image; determining the white area binary template and the green area binary template of the fire sign; scaling the white area binary template and the green area binary template to the same size as the fire sign board; performing corresponding element multiplication operations on the white area binary template and the green area binary template of the same size with the fire sign board, respectively, to obtain a white fire product image and a green fire product image; based on a brightness mapping model, obtaining the average brightness of all non-zero pixels in the white fire image and the green fire image, and using the average brightness of the white fire image and the green fire image as the brightness of the white area and the green area in the fire sign, respectively.

[0129] Optionally, determining the fire sign panel from the second key image includes: graying the second key image corresponding to the fire sign, performing edge detection, dilation and corrosion, binarization, etc. on the grayed image to obtain a fire key feature image; determining a rectangular area in the fire key feature image, determining a rectangular area in the fire key feature image that meets a preset pixel size and is closest to the image center as the fire sign panel, and segmenting the fire sign panel grayscale image Edge that is the same as the fire sign panel area from the grayed image. 板面灰度 .

[0130] Combine Figure 4 As shown, the process of identifying the type of tunnel fire sign is as follows:

[0131] Step S11: Analyze and process the second key image corresponding to the tunnel fire sign to obtain the fire sign board Edge 板面 ;

[0132] Step S12: Determine Edge 板面 Is there a long horizontal arrow? If yes, go to step S13; if not, go to step S16;

[0133] Step S13: Determine Edge 板面 Whether there are two graphics, if yes, proceed to step S14, if no, proceed to step S15; in some embodiments, the two graphics are of the same size and are symmetrically distributed;

[0134] Step S14: Identify the vehicle as a cross-passage sign;

[0135] Step S15: Identify the pedestrian crossing sign;

[0136] Step S16: Determine Edge 板面 Check whether there is a vertical long arc that is symmetrical up and down. If yes, go to step S17; if not, go to step S18;

[0137] Step S17: Identify the emergency call sign;

[0138] Step S18: Determine Edge 板面 Whether there is a vertical straight line segment that meets the preset conditions, if so, go to step S19, if not, go to step S20; the preset conditions are two straight line segments of equal length, symmetrical up and down, and at a 45-degree angle to the horizontal direction;

[0139] Step S19: Identify as an emergency stop sign;

[0140] Step S20: Identify the fire extinguisher symbol.

[0141] In some embodiments, the brightness of the fire sign is obtained by the following steps:

[0142] Step 21: Preset and construct white area binary templates and green area binary templates of various fire signs, and construct a brightness mapping model between the grayscale of the image of the lateral color area array image sensor and the actual brightness;

[0143] Step 22: Analyze and process the key image corresponding to the tunnel fire sign to obtain the grayscale image Edge of the fire sign board. 板面灰度 and Fire Sign Panel Area Image Edge 板面 ;

[0144] Step 23: Use the Edge of the Fire Sign Panel Area Image 板面 Identify specific types of fire signs;

[0145] Step 24: Select a white area binary template and a green area binary template corresponding to the fire sign according to the specific type of the fire sign;

[0146] Step 25: Scale the selected white area binary template and green area binary template to the edge of the grayscale image of the fire sign. 板面灰度 Same size;

[0147] Step 26: Combine the white area binary template and green area binary template of the same size with Edge 板面灰度 Perform corresponding element multiplication operation to obtain the white fire product image Img 消防乘积_白色 and the green fire product image Img 消防乘积_绿色 ;

[0148] Step 27: Calculate Img according to the brightness mapping model 消防乘积_白色 and Img 消防乘积_绿色 The actual average brightness of all non-zero pixels in the image is used as the brightness measurement result of the white area and the green area of ​​the fire sign.

[0149] Optionally, the method for classifying, identifying and measuring the brightness of tunnel luminous signs also includes: obtaining all rectangles in the color linear array image, determining the rectangular area that is close to the tunnel side wall and is on the front as an evacuation sign, and determining the distance from the evacuation sign to the starting point outside the tunnel according to the distance from the center pixel row number of the rectangular area of ​​the evacuation sign to the starting position outside the tunnel, that is, the location of the evacuation sign; performing brightness analysis on the rectangular area of ​​the evacuation sign to obtain the brightness of the evacuation sign.

[0150] In some embodiments, in the color line array image, the front rectangular area of ​​the evacuation sign set close to the tunnel side wall is automatically identified, the distance from each evacuation sign to the starting point outside the tunnel is calculated based on the row label of the center pixel of the rectangular area, the rectangular area image is analyzed and the brightness of the evacuation sign is calculated.

[0151] Optionally, the method for classifying, identifying and measuring the brightness of tunnel luminous signs also includes: obtaining the tunnel test length based on the tunnel entrance and tunnel exit in the linear array feature image, and the displacement of the detection vehicle between adjacent pixels; obtaining a position correction coefficient based on the tunnel test length and the actual length of the tunnel; correcting the position of each tunnel sign based on the position correction coefficient; and plotting the corrected position, brightness and chromaticity of each tunnel sign into a luminous sign information layout diagram in the tunnel.

[0152] Optionally, the tunnel measurement length L 测量 =(k 出口 -k 入口 )·S 行间 ; where k 出口 is the center row number of the tunnel exit, k 入口 is the center row number of the tunnel entrance, S 行间 To detect the vehicle displacement between adjacent pixels, L 测量Measure the length of the tunnel.

[0153] Optionally, the position correction coefficient λ = L 实际 ÷L 测量 , where L 实际 The actual length of the tunnel given in the tunnel as-built drawings.

[0154] In some embodiments, the position of each tunnel sign is corrected according to the position correction coefficient, and the corrected tunnel luminous sign and brightness / chromaticity information are drawn to the corresponding position in the layout diagram. The final luminous sign layout diagram is as follows: Figure 5 As shown. Combined Figure 5 As shown, the layout diagram of luminous signs in the tunnel is marked with the numbers of the luminous signs in the tunnel, as well as the distance to the tunnel, equipment classification, average brightness and chromaticity coordinates of each sign. Equipment classification includes evacuation signs, fire extinguisher signs, emergency telephone signs, emergency parking lane signs, vehicle crossing signs, pedestrian crossing signs, information boards and lane indicators. For example, the number E represents a lane indicator, and E1 means that the distance to the tunnel of one lane indicator is XXXm, the equipment classification is lane indicator, and the average brightness is XX cd / m 2 , chromaticity coordinates are (x, y); for example, label B represents an emergency phone sign, B2 represents the distance of one of the emergency phone signs into the hole is xxx m; the device classification is an emergency phone sign; the average luminance is xx cd / m 2 .

[0155] In some embodiments, whether it is the inspection of tunnel electromechanical quality or the evaluation of tunnel electromechanical maintenance effect, there is a certain demand for the classification and identification of luminous signs and the measurement of brightness. The method and device provided by this solution can well meet the following four requirements for the classification and identification of luminous signs and the measurement of brightness, providing a more convenient and efficient option for the inspection of tunnel electromechanical quality or the evaluation of tunnel electromechanical maintenance effect. This solution can achieve: (1) Identify the information board and determine its distance to the tunnel entrance, and measure the brightness / chromaticity value of the rectangular luminous panel of the information board when the screen is full of red, green, and yellow. (2) Identify the lane indicator and determine its distance to the tunnel entrance, and measure the brightness / chromaticity value of the green "↓" luminous pattern and the red "×" luminous pattern of the lane indicator. (3) Identify the specific type of fire electro-optical sign and determine its distance to the tunnel entrance, and measure the brightness value of the white area and the green area in the luminous panel. (4) Since the brightness / chromaticity measurement of the luminous sign (except the evacuation sign) should be measured at a parking sight distance in front of the sign, otherwise the measurement data is meaningless. In this solution, the brightness / chromaticity measurements of luminous signs (except evacuation signs) are all measured at a stopping sight distance in front of the sign, thereby ensuring that the measurement data is meaningful.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.

Claims

1. A method for classifying, identifying and measuring the brightness of tunnel luminous signs, characterized in that: include: Capture images inside and outside the tunnel to obtain several frames of tunnel images; The tunnel image includes a tunnel line array image and a tunnel area array image; Obtaining a starting point position outside the tunnel based on a tunnel linear array image; comprising: performing feature processing on the tunnel linear array image to obtain a linear array feature image; detecting straight line segments and circular arc segments in the linear array feature image; traversing all straight line segments and circular arc segments, and if there is a straight line segment or circular arc segment that passes through all columns of the linear array feature image and has a row span less than a first preset threshold, determining the straight line segment or circular arc segment that passes through all columns of the linear array feature image and has a row span less than the first preset threshold as a tunnel entrance, and obtaining a row label of the tunnel entrance in the linear array feature image; obtaining a starting point position outside the tunnel based on the displacement of the detection vehicle between the tunnel entrance and two adjacent rows of pixels; Determine the key image based on the tunnel array image and the starting point position outside the tunnel; Acquire the type of tunnel sign according to the tunnel linear array image and the key image; the type of tunnel sign includes a variable sign and a fire sign; The brightness of the variable sign and the fire sign are determined separately according to the key image.

2. The method according to claim 1, characterized in that Capture images inside and outside the tunnel to obtain tunnel images, including: A forward color area array image sensor and a black-and-white linear array image sensor are installed on the roof of the tunnel inspection vehicle, and a side color area array image sensor and a color linear array image sensor are installed on the side of the vehicle. The inspection vehicle starts collecting data from the tunnel entrance until it leaves the tunnel, obtaining several frames of black-and-white linear array images, color linear array images, forward color area array images, and side color area array images.

3. The method according to claim 2, characterized in that Determine the key images based on the tunnel array image and the starting position outside the tunnel, including: Identify the tunnel sign rectangular area in the tunnel linear array image; Obtaining a first distance from the tunnel sign to the starting position outside the tunnel according to the rectangular area of ​​the tunnel sign; the tunnel sign includes a variable sign and a fire sign; For the variable sign, a first key image is selected based on the first distance and the displacement of the detection vehicle between two adjacent frames of the forward color area array image; the first key image is a forward color area array image captured at a stopping sight distance behind the driving direction of the sign; For the fire sign, the second key image is selected based on the first distance and the displacement of the detection vehicle between two adjacent frames of the lateral color area array image; the second key image is a lateral color area array image collected at a stopping sight distance behind the driving direction of the sign.

4. The method according to claim 3, characterized in that Identify the tunnel sign rectangular area in the tunnel linear array image, including: Identify all corner points in the tunnel linear array image; Get all rectangles formed by corner points; Determine a rectangle that meets the first preset condition as a tunnel sign rectangular area; The first precondition is: ; Among them, H is the actual thickness of the sign board, To detect vehicle displacement between adjacent pixels, and is the row number of the first corner point and the row number of the third corner point of the rectangle, and is the column index of the first corner point and the column index of the second corner point of the rectangle, 、 and are all constants.

5. The method according to claim 4, characterized in that The types of tunnel signs obtained based on tunnel linear images and key images include: Acquire a first tunnel sign rectangle in the black-and-white tunnel linear image, determine the first tunnel sign rectangle that meets a first preset condition as a variable tunnel sign, and determine the type of the variable sign based on the actual thickness of the sign panel and the range of spacing between adjacent column numbers of the rectangle; Acquire a second tunnel sign rectangle in the tunnel color line array image, and determine the second tunnel sign rectangle that satisfies both the first preset condition and the second preset condition as a tunnel fire sign; The second precondition is: ;in, and are all constants; Determine the fire characteristics of the tunnel fire signs, including whether there is a horizontal arrow, whether there are two figures, whether there is a vertical long arc, and whether there are two vertical straight line segments that meet preset conditions; The type of tunnel fire sign is identified based on the fire feature judgment results.

6. The method according to claim 1, wherein Determine the brightness of variable signs and fire signs based on the key image, including: In the case where the tunnel sign is a variable sign, a variable sign surface is determined from the first key image; a brightness mapping model and a chromaticity mapping model are constructed, and based on the brightness mapping model and the chromaticity mapping model, the brightness and chromaticity of the variable sign are obtained using the variable sign surface; In the case that the tunnel sign is a fire sign, the fire sign surface is determined from the second key image; the white area binary template and the green area binary template of the fire sign are determined; the white area binary template and the green area binary template are scaled to the same size as the fire sign surface; the white area binary template and the green area binary template of the same size are respectively multiplied by the corresponding elements of the fire sign surface to obtain a white fire product image and a green fire product image; based on the brightness mapping model, the average brightness of all non-zero pixels in the white fire image and the green fire image is obtained, and the average brightness of the white fire image and the green fire image is used as the brightness of the white area and the green area in the fire sign, respectively.

7. The method according to claim 6, characterized in that Based on the brightness mapping model and the chromaticity mapping model, the brightness and chromaticity of the variable sign are obtained by using the variable sign board, including: For information board signs, the brightness and chromaticity are directly obtained using the variable sign board area based on the brightness mapping model and chromaticity mapping module; For lane indicator signs, a two-dimensional template of the lane indicator panel pattern is pre-built and scaled to the same size as the variable sign panel. The two-dimensional template of the lane indicator panel pattern of the same size is multiplied with the pixels at corresponding positions on the variable sign panel to obtain a variable product image. Based on the brightness mapping model and the chromaticity mapping module, the variable product image is used to obtain the brightness and chromaticity.

8. The method according to claim 1, characterized in that Also includes: The tunnel test length is obtained based on the tunnel entrance and exit in the linear array feature image and the displacement of the detection vehicle between adjacent pixels; Obtain position correction coefficient based on tunnel test length and actual tunnel length; Correct the position of each tunnel sign according to the position correction coefficient; The corrected positions, brightness and chromaticity of each tunnel sign are plotted into a luminous sign information layout diagram in the tunnel.

9. A device for classifying, identifying and measuring the brightness of tunnel luminous signs, characterized in that: include: Inspection vehicle, A forward-facing color area array image sensor is mounted on the roof of the inspection vehicle. After installation, the central optical axis of the forward-facing color area array image sensor is projected onto the road surface parallel to the road surface centerline, and the horizontal distance from the intersection of the central optical axis and the tunnel vault to the inspection vehicle is one stopping sight distance. The forward-facing color area array image sensor is used to capture images of the suspended information board and lane indicator. A lateral color area array image sensor is mounted on the side of the inspection vehicle. After installation, the central optical axis of the lateral color area array image sensor points toward the tunnel side wall at a height of 2.5 meters from the ground, and the horizontal distance from the intersection of the central optical axis and the tunnel side wall to the inspection vehicle is one stopping sight distance. The lateral color area array image sensor is used to capture images of various fire electro-optical signs with panels perpendicular to the side wall. A color line array image sensor is installed on the side of the inspection vehicle. After installation, the central optical axis of the color line array image sensor is projected on the road surface perpendicular to the center line of the road surface. The color line array image sensor is used to collect images of the tunnel side walls. A black and white linear array image sensor is installed on the roof of the inspection vehicle, with the central optical axis of the black and white linear array image sensor pointing vertically upward after installation, and the black and white linear array image sensor is used to capture images directly above the inspection vehicle; A rotary encoder is coaxially mounted with the rear wheel of the inspection vehicle. The rotary encoder is used to sense the displacement of the inspection vehicle and send a pulse signal. The rotary encoder is used to directly trigger the color linear array image sensor and the black and white linear array image sensor to capture images. a pulse modulator, configured to receive the pulse signal emitted by the rotary encoder and perform frequency reduction processing on the pulse signal to obtain a frequency-reduced pulse signal, wherein the frequency-reduced pulse signal is used to trigger the forward color area array image sensor and the side color area array image sensor to capture an image; A control processing unit is electrically connected to each image sensor, and is used to execute the method for classifying, identifying and measuring the brightness of tunnel luminous signs according to any one of claims 1 to 8.

Images