CN117314879A

CN117314879A - Self-adaptive operation and maintenance judging method and monitoring device for road indication board

Info

Publication number: CN117314879A
Application number: CN202311354293.9A
Authority: CN
Inventors: 曹国雄; 吴花锋; 岳生晓; 李栋; 张立惟
Original assignee: Gansu Luqiao Feiyu Transportation Facilities Co ltd
Current assignee: Gansu Luqiao Feiyu Transportation Facilities Co ltd
Priority date: 2023-10-19
Filing date: 2023-10-19
Publication date: 2023-12-29

Abstract

The invention discloses a road sign self-adaptive operation and maintenance judging method and a monitoring device; s1.2, crack identification: calculating crack information EM (x, y) by using a Canny operator; s1.3, fading identification: transferring the RGB color space to the HSV color space, comprising the steps of: generating an information vector DV: dv= [ EM (x, y), HSV ]; the base is fixedly connected to the road sign support column, and an adjusting mechanism is mounted on the base; the adjusting mechanism comprises a rotational degree of freedom and a first linear degree of freedom; the invention realizes the rapid and efficient detection of the surface defects of the road indication board through the machine vision detection and the automatic device, and greatly improves the detection efficiency compared with the traditional manual detection. By adopting the self-adaptive device structure, the detection unit can be ensured to cover each position of the road indication board, and the information such as cracks, fading and the like can be accurately detected, so that omission is avoided.

Description

Self-adaptive operation and maintenance judging method and monitoring device for road indication board

Technical Field

The invention relates to the technical field of guideboards, in particular to a self-adaptive operation and maintenance judging method and a monitoring device for a road sign.

Background

The operation and maintenance of the guideboard refers to regular inspection, maintenance and repair of the guideboard on the road, ensuring normal operation and clear visibility of the guideboard so as to provide accurate traffic information. The guideboard is inspected regularly, so that the problems are found out to be timely maintained and repaired, and the good state of the guideboard is maintained. Damaged, discolored, cracked or destroyed guideboards are repaired or replaced to ensure accurate communication of traffic information. The staff can record the maintenance conditions of each time, including maintenance time, maintenance content and maintenance staff, so as to track and manage the maintenance conditions of the guideboard. Materials, manufacturing processes and maintenance methods of the guideboard are continuously improved and upgraded according to actual conditions and technical development so as to improve the durability and reliability of the guideboard.

Through effective operation and maintenance, the long-term effective operation of the guideboard can be ensured, and the safety and efficiency of road traffic are improved. In the conventional technology, the surface of a guideboard is usually required to be directly observed by operation and maintenance personnel by naked eyes, so that the problems of cracks, fading, damage, dirt and the like are found. They will walk through each guideboard, scrutinize and record any anomalies. Photographs or videos are then taken to record the actual status of the guideboard and used in subsequent analysis. The traditional detection modes require operation and maintenance personnel to have abundant experience and expertise so as to accurately identify and record various defects of the guideboard and provide basis for formulating maintenance operation strategies.

However, the conventional technology restricts large-scale guideboard operation and maintenance implementation due to high labor intensity and low detection efficiency, and meanwhile, the manual mode relies on subjectivity, so that predictive guidance information can not be provided more effectively, and the establishment of maintenance operation strategies is restricted in a variable way.

Therefore, a road sign self-adaptive operation and maintenance judging method and a monitoring device are provided.

Disclosure of Invention

In view of this, an embodiment of the present invention is expected to provide a method and a device for determining and monitoring an adaptive operation and maintenance of a road sign, so as to solve or alleviate the technical problem existing in the prior art, namely (), and at least provide a beneficial choice for the method and the device;

the technical scheme of the embodiment of the invention is realized as follows:

first aspect

Self-adaptive operation and maintenance judging method for road indication board

Summary (one) overview

The method disclosed by the invention utilizes a machine vision technology and a cellular automaton model to realize the operation maintenance judgment of the road sign by combining with a D-S evidence theory. The method can detect the surface crack and fading condition of the road sign, forecast the state of the next time step, and provide the basis for maintaining the operation strategy.

(II) technical content

S1, visual inspection

And performing machine vision detection on the surface of the road sign to generate an information vector DVDV which comprises crack information and fading information.

S1.1, the machine vision detection: shooting the surface of the road indication board to generate an image matrix;

s1.2, crack identification: calculating crack information EM (x, y) by using a Canny operator;

s1.3, fading identification: transferring the RGB color space to the HSV color space, comprising the steps of:

generating an information vector DV:

DV＝[EM(x,y)，HSV]

where symbol "is a vector concatenation operation.

S2, executing cellular automata

Drawing the surface of the road signLike a two-dimensional grid N, divided into a number of equally sized small areas, each of which is considered a cell N. Defining a cell n to execute Moore neighborhood, predicting and outputting an information vector DV of the next time step through a transfer function f ₊₁ 。

In the S2, it includes:

s2.1, the Moore neighborhood: the Moore neighborhood around the cell n is formed by 8 adjacent cells;

s2.2, the transfer function f:

s2.2.1 and calculating the weight w _i : the distance from the central cell is used for determining:

d _i representing adjacent cells n _i A distance to the central cell n;

p is an adjustable parameter, and takes a value greater than or equal to 1;

S2.2.2, calculating a weighted sum of states in Moore neighborhood N:

k is the total number of adjacent cells, each cell having a state s _i I represents the index of the neighboring cells;

s2.2.2, judging the next state:

judging the result of weighted summation, taking the majority value of the states in the Moore neighborhood N as the state of the next step:

1 represents an active state, 0 represents an inactive state;

s2.3, predicting the information vector DV of the next time step ₊₁ ：

DV ₊₁ ＝[f(n),f(n),f(n)......f(n)]Totally k

The number of f (n) is k.

S3, executing D-S evidence theory verification

Taking the information vector DV obtained by visual detection as evidence A in the current time step, and taking the information vector DV predicted in the previous time step ₊₁ As evidence B. Combining the evidence A and the evidence B by using a Dempster's combination principle, mapping the evidence A and the evidence B into a correction factor alpha, and correcting a transfer function f in a cellular automaton model.

In the step S3, the Dempster' S combination principle includes:

s3.1, calculating the credibility by using a membership function:

bel (A) and Bel (B) represent the credibility of the evidence A and the evidence B, respectively, A _i And B _i Representing the i-th element in said information vector DV respectively corresponding thereto:

s3.2, calculating the complementarity of the credibility by using a complementarity function:

Pl(A)＝1-Bel(A)

Pl(B)＝1-Bel(B)

pl (A) and Pl (B) represent complementarity of the credibility of the evidence A and the evidence B, respectively;

S3.3, merging into a joint membership function:

s3.3.1 calculating the intersection m (A.u.B) of two evidences, for measuring the degree of conflict of the two evidences:

bel (X) represents the degree of confidence for subset X;

s3.3.2, calculating a joint membership function Bel (a n B):

m (a n B) represents the intersection of the evidence a and the evidence B.

S4, formulating maintenance operation strategy

Based on the corrected information vector DV ₊₁ And a substantial maintenance direction and basis are provided for a worker to formulate a maintenance operation strategy of the road sign.

Second aspect

Road sign self-adaptation formula fortune dimension monitoring devices

The road sign self-adaptive operation and maintenance monitoring device is used for executing the road sign self-adaptive operation and maintenance judging method, and comprises the following steps:

the base is fixedly connected to the road sign support column, and an adjusting mechanism is mounted on the base; the adjusting mechanism comprises a rotational degree of freedom and a first linear degree of freedom, and the rotational degree of freedom is used for pitching an output angle of the first linear degree of freedom; when the first linear degree of freedom is output, the first linear degree of freedom drives all the blooming components to approach or separate the mutual spacing and included angle; when the moving away is performed, under the front view of the road sign, the projection of all the moving-on assemblies covers the sign body of the road sign;

The device comprises a blooming assembly, wherein at least two detection units are arranged on the blooming assembly in an array mode, the detection units comprise at least three second linear degrees of freedom which are arranged in an annular array mode along the coaxial direction, and the second linear degrees of freedom are connected with a CCD industrial vision camera for detecting the surface defects of the road signs in an action mode.

In the above embodiment, the following embodiments are described. The road sign self-adaptive operation and maintenance monitoring device adopts a base which is firmly connected to the supporting column of the road sign. An adjustment mechanism is mounted on the base, the adjustment mechanism having a rotational degree of freedom and a first linear degree of freedom. The rotational degrees of freedom are responsible for adjusting the output angle of the first linear degrees of freedom. When the first linear degree of freedom is output, the plurality of blooming components are driven to be close to or far away from each other through linkage, and the spacing and the included angle between the components are adjusted. The projection of the blooming components is covered on the front view of the road sign, so that the road sign surface is covered.

Wherein in one embodiment: the adjusting mechanism comprises a pitching motor for outputting the first linear degree of freedom, the pitching motor is connected with a shaft head at one end of a screw rod for adjusting pitching angles, and the screw rod is used for outputting the first linear degree of freedom; the other end shaft head of the screw rod is in running fit with a base through a bearing, and an execution motor for driving the screw rod to rotate is mounted on the base; a star wheel is connected to the thread surface of the screw rod in a threaded manner, and the screw rod is driven to slide up and down when the screw rod rotates to output the first linear degree of freedom; the star wheel and the base are provided with the blooming assembly.

In the above embodiment, the following embodiments are described. The adjustment mechanism includes a pitch motor for outputting an angular adjustment of the first linear degree of freedom. The pitching motor is connected to one end shaft head of the screw rod to adjust pitching angle. The screw is responsible for outputting the first linear degree of freedom. The other end of the screw rod is connected to a base through a bearing, and an execution motor is mounted on the base and used for driving the screw rod to rotate.

Wherein in one embodiment: the blooming assembly comprises a beam framework hinged on the base, a supporting plate is carried on the outer part of the beam framework, and one end and the other end of a hinge arm are respectively hinged on the middle part of the beam framework and the star wheel; the supporting plate is provided with the detection unit through the commodity shelf.

In the above embodiment, the following embodiments are described. The spindle assembly consists of a beam framework which is hinged on the base. The outside of beam skeleton carries on the backup pad, and the detecting element is installed through the supporter to the backup pad. The hinge arms are respectively hinged to the middle part of the beam framework and the star wheel.

Wherein in one embodiment: when the star wheel moves up and down, the star wheel transmits the first linear degree of freedom to the beam framework through the hinge arm, and only one end of the beam framework is hinged to the base, so that the star wheel uses the base as a supporting reference for angle adjustment; and because the star wheel is all the blooming components driven synchronously, when the star wheel moves up and down, the distance and the included angle between all the blooming components can be adjusted to be close to or far away from each other. At this time, all the fraying assemblies form a fraying gesture, the tile body of the road sign is covered, and the detection unit mounted on the fraying assemblies can execute the detection procedure of the road sign self-adaptive operation and maintenance judging method provided in the foregoing.

In the above embodiment, the following embodiments are described. The up-and-down movement of the star wheel transfers the first linear degree of freedom to the beam skeleton via the hinge arms. Only one end of the beam skeleton is hinged to the base, so that the base is used as a supporting reference for angle adjustment. Because the star wheel synchronously drives all the blooming components, the movement of the star wheel can adjust the spacing and the included angle of all the blooming components to enable the blooming components to be close to or far away from each other. Thus, all the fraying components form a fraying gesture which covers the body of the road sign.

Wherein in one embodiment: the detection unit comprises two opposite frame bodies, six servo electric cylinders for outputting the second linear degree of freedom are arranged between the frame bodies in a ring array mode, and a cylinder body and a piston rod of each servo electric cylinder are respectively and universally hinged to one surface of each of the two opposite frame bodies through a universal joint coupling; and one frame body is provided with the CCD industrial vision camera. The other frame body is fixedly arranged on the storage frame.

Compared with the prior art, the invention has the beneficial effects that:

(1) High-efficient automated detection: the invention realizes the rapid and efficient detection of the surface defects of the road indication board through the machine vision detection and the automatic device, and greatly improves the detection efficiency compared with the traditional manual detection. By adopting the self-adaptive device structure, the detection unit can be ensured to cover each position of the road indication board, and the information such as cracks, fading and the like can be accurately detected, so that omission is avoided.

(2) Eliminating human subjective errors: the automatic operation of the mechanical device eliminates the subjectivity of people, so that the detection result is more objective and accurate and is independent of subjective judgment of individuals.

(3) Predictive guidance information: according to the invention, through the processing of the S2 and the S3, the information vector on the surface of the road sign is extracted and analyzed, predictive guidance information can be provided for formulating maintenance operation strategies, and the planning of more effective maintenance strategies is facilitated.

(4) Human resources are saved: the automatic detection and operation and maintenance method reduces the labor intensity, thereby saving human resources, reducing the operation and maintenance cost and enabling the implementation of large-scale guideboard operation and maintenance to become more feasible.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic flow chart of the method of the present invention;

FIG. 2 is a schematic perspective view of the present invention;

FIG. 3 is a perspective view of an adjusting mechanism according to the present invention;

FIG. 4 is a perspective view of an alternative view of the adjustment mechanism of the present invention;

fig. 5 is a perspective view of the blooming assembly of the present invention;

FIG. 6 is a schematic perspective view of a detection unit according to the present invention;

FIG. 7 is a control program diagram of a sixth embodiment of the present invention;

FIG. 8 is a control program diagram of a sixth embodiment of the present invention;

reference numerals: 1. road signs; 2. a base; 3. an adjusting mechanism; 301. a pitch motor; 302. a base; 303. a screw rod; 304. star wheels; 305. a blooming assembly; 3051. a beam skeleton; 3052. a support plate; 3053. an arm hinge; 3054. a commodity shelf; 306. executing a motor; 4. a detection unit; 401. a frame body; 402. a servo electric cylinder; 403. a universal joint coupling; 404. CCD industrial vision camera;

Detailed Description

In order that the above objects, features and advantages of the invention will be readily understood, a more particular description of the invention will be rendered by reference to the appended drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may be embodied in many other forms than described herein and similarly modified by those skilled in the art without departing from the spirit of the invention, whereby the invention is not limited to the specific embodiments disclosed below;

It is noted that terms like "degree of freedom" refer to a relationship of connection and application of a force of at least one component, e.g. "linear degree of freedom" refers to a relationship in which a component is connected to and applies a force to another component or components through the linear degree of freedom such that it is capable of sliding fit or application of a force in a straight direction; "rotational freedom" means that a component is free to rotate about at least one axis of rotation and can apply or receive torque.

It will be further appreciated by those of skill in the art that the various example elements and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the various example elements and steps have been described generally in terms of function in the foregoing description to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

It is noted that the steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

Embodiment one:

referring to fig. 1, the present embodiment provides: a self-adaptive operation and maintenance judging method for a road sign comprises the following steps:

s1, visual detection: performing machine vision detection on the surface of the road sign, and outputting an information vector DV at each time step t, wherein the information vector DV comprises crack information and fading information;

s2, executing cellular automata: the surface of the road sign is abstracted into a two-dimensional grid N, the two-dimensional grid N is divided into a plurality of small areas with the same size, each small area is regarded as a cell N, the cell N is defined to execute Moore neighborhood, the Moore neighborhood is subjected to probability transfer by a transfer function f, and the transfer function f predicts and outputs an information vector DV of the next time step ₊₁ ；

S3, executing D-S evidence theory verification: in the current time step, the information vector DV of S1 is taken as evidence A, and the predicted current information vector DV of S2 in the previous time step is taken as evidence A ₊₁ As evidence B, the evidence A and the evidence B are combined by using the Dempster's combination principle and mapped into a [0,1 ]]Correction factor α in the section, which corrects the transfer function f in S2 at each time step.

S4, in each time step, the information vector DV based on the next time step ₊₁ Providing basis for working staff to make maintenance operation strategy.

In this embodiment, the visual detection is the first step of the road sign adaptive operation and maintenance determination method, and the machine vision technology is adopted to detect the surface of the road sign. And inputting the surface image of the road sign into an algorithm through visual equipment such as a camera and the like, and performing image processing, feature extraction and analysis. And through algorithm analysis, the crack and fading condition of the road sign surface can be detected. These pieces of information are integrated into an information vector DV, which includes crack information and fading information. Visual inspection by means of machine vision techniques, an image of the surface of the road sign is first acquired. Then, by image processing techniques, cracks and discoloration in the image are identified and analyzed. The purpose of this step is to quantify the visual characteristics of the road sign into data that can be processed by the machine, providing basis for subsequent steps. The obtained information vector DV will become the input of the subsequent cellular automaton for the prediction and verification of the model. For a specific hardware principle and implementation, please refer to example seven.

In this embodiment, the surface of the road sign is abstracted into a two-dimensional grid N, dividing the grid into several small areas, each of which is regarded as a cell N. And defining the state and neighborhood relation of the cell n by adopting a cellular automaton model, wherein Moore neighborhood is a neighborhood definition in the cellular automaton field. Through the transfer function f, the cell n carries out probability transfer according to the state of the neighborhood thereof, predicts and outputs the information vector DV of the next time step ₊₁ . The cellular automaton model provides a basic framework for the self-adaptive operation and maintenance of the road indication board. Dividing the surface of the road sign into a plurality of small areas, and carrying out information exchange and state update on each small area and the adjacent areas through cellular automata. The transfer function f predicts the state of the next step by taking into account the information of the neighborhood, i.e. predicts the information vector DV of the next time step ₊₁ . The modeling based on the cellular automaton can better capture the change and trend of the road sign surface.

In this embodiment, the D-S evidence theory verification is to enhance the robustness and reliability of the model. In this step, the information vector DV obtained by visual inspection is regarded as the data A, and the information vector DV predicted in the previous time step is regarded as ₊₁ And is regarded as data B. Evidence a and evidence B are combined using the Dempster's combination principle to produce a correction factor α. The correction factor alpha is used for adjusting the transfer function f of the cellular automaton model, and the prediction accuracy of the model on the surface state of the road sign is enhanced.

In this embodiment, the D-S evidence theory verification considers the correlation between information of different sources (visual inspection and automaton model prediction), and improves the accuracy and reliability of the system by combining these information. And combining information obtained by visual detection and information predicted by a model by using a Dempster's combination principle to generate a correction factor alpha. The correction factor alpha is used for adjusting the transfer function f of the automaton model so as to improve the prediction accuracy of the model on the surface state of the road sign.

In the present embodiment, the verified information vector DV is utilized ₊₁ On the basis, a maintenance operation strategy is formulated for the staff. And (3) according to the predicted information of the surface of the road sign, making a proper maintenance scheme, including repairing cracks, repainting and the like. Formulating maintenance work policies depends on verified information vector DV ₊₁ The information vector accurately reflects the state of the road sign surface. According to the prediction information, the staff can carry out maintenance operation in a targeted manner, and blind maintenance and resource waste are avoided. This step will provide a useful guide for maintenance personnel to ensure the well-being and safety of the road sign.

Specifically, the information vector DV mentioned in S4 ₊₁ The prediction information vector of the next time step is fused with the visual detection information (DV) of the current time step and the prediction result of the cellular automaton in the calculation process of the D-S evidence theory. This information vector is used as a basis for formulating maintenance work policies, which can be implemented by:

p1, analyzing information vectors: parsing DV ₊₁ Specific information contained therein, such as predicted crack information and discoloration information of the road sign surface, and the like, is obtained. And analyzing the analyzed information to know the state of the surface of the road sign, including crack condition, fading condition and the like. Such information may be indicative of the health of the roadway sign.

P2, formulating a maintenance strategy: and according to the information obtained by analysis, a corresponding maintenance operation strategy is formulated. For example, if a crack in the roadway sign is detected, periodic maintenance or replacement may be required; recoating or replacement may be required if there is discoloration.

P3, planning and task: the maintenance strategy is embodied as a maintenance plan and task. Specific times, places, personnel and required resources are determined, as well as steps and procedures for performing maintenance jobs.

P4, implementing maintenance operation: according to the formulated plan and task, the organization personnel performs maintenance work, repairs, updates or replaces the road signs to ensure good status and effective functions thereof.

Embodiment two:

the embodiment further discloses a specific implementation manner of S1 of the road sign adaptive operation and maintenance determination method, which includes:

s1.1, machine vision detection: shooting the surface of the road indication board to generate an image matrix; optionally, the image may be further preprocessed to reduce noise and enhance features. Including conventional graying, denoising, etc.

s1.2.1, converting a color image into a grayscale image:

GrayImage(x,y)＝0.299×R(x,y)+0.587×G(x,y)+0.114×B(x,y)

(x, y) represents the pixel position in the image, R (x, y), G (x, y) and B (x, y) are the intensities of the red, green and blue channels, respectively, at that position; the color channel values (R, G, B) for each pixel are weighted and summed according to a weighted average coefficient to obtain a gray value representing the brightness of the pixel. Thus, each pixel point in the color image corresponds to one pixel point in the gray image, and conversion from the color image to the gray image is completed. This step simplifies subsequent image processing, making image analysis more efficient.

The purpose of converting a color image into a grayscale image is to simplify the image processing process, reduce the amount of computation, while retaining the main information of the image. A gray image means that each pixel is represented by only one gray value, which may be a weighted average of the color channel intensities. Common weighted average coefficients are 0.299, 0.587 and 0.114, corresponding to red, green and blue of the RGB color channels. These coefficients are derived from the sensitivity of the human eye to different colors.

S1.2.2, calculating the gradient magnitude GM (x, y) and gradient direction GD (x, y) of the image:

G _x (x, y) and G _y (x, y) is the gradient of the image in the horizontal and vertical directions at (x, y), and can be calculated by a convolution operator such as Sobel; arctan is an arctangent trigonometric function;

gradients are the direction and rate of change of image intensity. The gradient amplitude and gradient direction of the image can reveal the edge information of the image, and are the basis of edge detection. The gradient magnitude is used to represent the intensity of the edge and the gradient direction is used to represent the direction of the edge. By calculating the gradient of the image in the horizontal and vertical directions, the horizontal component Gx (x, y) and the vertical component Gy (x, y) of the gradient can be obtained by using convolution operators such as Sobel. Then, the gradient magnitude and gradient direction of each pixel point are calculated from these two components. The gradient magnitude indicates the degree of intensity change of the pixel point, and the gradient direction indicates the direction of this change. This information provides a basis for subsequent edge detection.

S1.2.3, non-maximum suppression NMSG (Non-Maximum Suppression):

non-maxima suppression is a common edge refinement technique used to eliminate redundant edge responses in gradient images, preserving the most significant edge responses. This process helps to accurately detect thin lines and edges in the image. When non-maximum suppression is performed on the gradient image, for each pixel, two adjacent pixels in the gradient direction are inspected along the direction. The gradient magnitude of the current pixel is kept if it is a local maximum in this direction, otherwise it is suppressed to 0. Thus, the image obtained after non-maximum suppression only maintains the position with the largest gradient change, and suppresses other positions to 0, thereby forming a thinned edge image.

S1.2.4, dual threshold edge detection: converting the non-maxima suppressed image NMSG (x, y) into a binary image EM (x, y), and marking strong, weak and non-edges:

the dual-threshold edge detection is a threshold segmentation method based on a gradient image, and pixels in the image are divided into three types of strong edges, weak edges and non-edges. Such segmentation helps to determine which edges in the image are true edges and which are noise only. Dual-threshold edge detection gradient images after non-maximum suppression are classified into three classes according to two preset thresholds (HighThreshold and LowThreshold): strong edges, weak edges, and non-edges. If the gradient magnitude is greater than HighThreshold, it is classified as a strong edge; if the gradient magnitude is between LowThreshold and HighThreshold, it is classified as a weak edge; if the gradient magnitude is less than LowThreshold, it is classified as non-edge. The binary image thus obtained represents the distribution of the edges.

S1.2.5, edge tracking: connecting weak edge pixels to strong edge pixels, forming connected edges:

edge tracking is to connect weak edges into strong edges, forming connected edge lines to better express real edge information in the image. The edge tracking algorithm starts with weak edge pixels and gradually connects strong edge pixels along the direction of the strong edge to form a connected edge path. In this way, weak edges in the image can be connected to long edges, enhancing edge continuity. The edge tracking results are connected edge lines that accurately represent the true edge position in the image.

The image preprocessing and edge detection steps above provide a basis for subsequent extraction of crack and fading information.

s1.3.1, normalized RGB values: the R, G and B values are divided by 255, normalized to the interval [0,1]: for subsequent computation. After normalization, the maximum maxRGB and minimum minRGB of the normalized RGB components are calculated. The RGB components are normalized, i.e. the R, G and B values are divided by 255, respectively, to obtain normalized RGB values (R ', G ', B '). Then, the maximum value maxRGB and the minimum value minRGB of the normalized RGB components are calculated:

Calculating the maximum value maxRGB and the minimum value minRGB of the normalized RGB components:

maxRGB＝max(R'，G'，B')

minRGB＝min(R'，G'，B')

by the method, normalized RGB components, maximum values and minimum values are obtained, and a basis is provided for subsequent calculation of hue, saturation and brightness.

S1.3.2, calculate Hue: the position on the color ring is represented, and the value range is [0,360] degrees or [0,1]:

specifically, the calculated hue needs to be determined according to the normalized RGB components and the maximum maxRGB and minimum minRGB.

mod6 represents the modulo operation, to beDivided by 6 to give a hue value of [0,360]]Degree (or [0,1]]) Within the range;

if it is0 or more and less than 1, the modulo operation does not affect the result, whereas the modulo operation can limit the result to [0,5 ]]And in the range, the condition of exceeding the range is avoided. Finally multiply by 60, ensuring that the hue is at [0,360]In the range of degrees.

S1.3.3, calculating saturation S: saturation (Saturation) represents the vividness of the color, and the value range is [0,1]. The saturation is calculated from the normalized RGB components and the maximum maxRGB and minimum minRGB.

And calculating the saturation S according to the maximum value maxRGB and the minimum value minRGB of the normalized RGB components. The specific calculation method is as follows:

S1.3.4, calculating the lightness V: the brightness (Value) represents the brightness of the color, and the range of values is [0,1]. The brightness V is the maximum value maxRGB of the normalized RGB component, so that the calculated V is the brightness value, and necessary data is provided for calculating the fading information:

V＝maxRGB

s1.3.5, construction of fade information: the fade information FA is constructed based on the normalized RGB components and the calculated Hue, saturation, value (HSV color space), and by a set threshold T. FA may reflect the degree of color fade.

FA＝T*HSV

Generating an information vector DV: the fading information FA is constructed by multiplying Hue, saturation, value by the coefficients T, respectively:

DV＝[EM(x,y)，HSV]

where symbol "is a vector concatenation operation. In this way, the fading information FA contains information calculated through the HSV color space, which is used in the subsequent fading recognition process, and provides beneficial information for the operation and maintenance of the guideboard.

In S1.1 of the present embodiment, machine vision detection is an important step in the road sign adaptive operation and maintenance determination method, and visual information of the road sign is converted into processable data by photographing the surface of the road sign and generating an image matrix. This step takes a picture of the road sign using an imaging device (e.g., a camera or CCD camera) to produce an image of the road sign surface. The image is converted to digital form, generating an image matrix in which each element corresponds to a luminance value or color value of a pixel. This matrix is the basis for subsequent image processing. The logic of machine vision inspection is to convert visual information of the road sign surface into a computer-processable data format. The image pickup apparatus collects an image of a road sign, divides the image into pixels, and converts information of each pixel into a digital value. In this way the whole image is converted into a matrix of numbers, where each number represents the property of a pixel. The matrix is used as the input of the subsequent step, and the analysis and detection of the characteristics of cracks, fading and the like on the surface of the road sign can be realized by further processing the image matrix.

In S1.2.2 of this embodiment, the Sobel operator is a convolution operator for calculating the gradient of an image. Gradients reflect the rate of change of the image and can help the present embodiments locate edges in the image. The Sobel operator carries out edge detection on the image through convolution operation, and the gradients of the image in the horizontal direction and the vertical direction are calculated respectively, so that the gradient amplitude and the gradient direction are obtained.

Illustratively, the present embodiment provides the following Sobel operator scheme: a 3x3 convolution kernel matrix has two convolution kernels for computing the horizontal gradient (Gx) and the vertical gradient (Gy) of the image, respectively. These two convolution kernels are each shown below:

convolution kernel Gx:

-1 0 1

-2 0 2

-1 0 1

convolution kernel Gy:

-1 -2 -1

0 0 0

1 2 1

for a given image, the present embodiments first apply a Sobel operator convolution kernel to each pixel point of the image. And respectively carrying out convolution operation on the image and the Gx convolution kernel and the Gy convolution kernel. The two convolution results obtained by calculation represent the gradient components of the image in the horizontal and vertical directions, respectively. Then, the Gradient Magnitude (GM) and Gradient Direction (GD) may be calculated using the following formulas:

gradient amplitude:

GM(x,y)＝sqrt(Gx(x,y)^2+Gy(x,y)^2)

gradient direction:

GD(x,y)＝arctan(Gy(x,y)/Gx(x,y))

wherein Gx (x, y) and Gy (x, y) represent gradient values of the image in the horizontal and vertical directions at the position (x, y), respectively. In this way, the embodiments can obtain the gradient amplitude and gradient direction of each pixel point of the image, and provide a basis for subsequent edge detection.

In the present embodiment, the information vector DV is obtained by photographing and analyzing the surface of the road sign by machine vision detection. The information contained in DV mainly covers crack information and fading information of the guideboard. Cracks refer to cracks, breakage or damage that may occur to the surface of the guideboard. By means of machine vision technology, DV is able to identify and record the crack condition of the guideboard surface. These crack information assess the health status of the guideboard, providing an important clue to the health status of the guideboard structure.

Fading information refers to the change in color of a guideboard, and generally relates to the lightness, saturation, etc. of the color. DV analyzes the change condition of the color of the guideboard through the technologies of color space conversion and the like, including the brightness of the color, the vividness of the color and the like. The fading information reflects whether the color of the guideboard deviates from the original design, possibly due to sun and rain, long time, etc.

In connection with this embodiment, the combination of crack information and fade information, the information vector DV provides an assessment of the road surface structure and color status. The information has important significance for formulating maintenance strategies, judging operation and maintenance requirements and guaranteeing the visibility and the accuracy of the guideboard. By analyzing the crack and fading information in DV, the self-adaptive operation and maintenance judgment of the road sign can be realized, and effective support is provided for guaranteeing road safety and transportation.

Embodiment III:

the embodiment further discloses a specific implementation manner of S2 of the road sign adaptive operation and maintenance determination method, where in S2, the method includes:

s2.1, moore neighborhood: the Moore neighborhood around the cell n is formed by 8 adjacent cells;

moore neighborhood refers to a layout centered on cell n, including 8 adjacent cells around it. Specifically, eight adjacent cells are arranged around the cell, namely, upper, lower, left, right, upper left, lower left, upper right and lower right. The neighborhood structure can more comprehensively consider the relationship between the cells and surrounding cells, and is convenient for carrying out state transition calculation in a transfer function. This neighborhood form enables cell n to interact with, influence and be influenced by cells in the eight directions around.

S2.2, transfer function f:

weight w _i By taking into account the distance of the adjacent cells from the central cell, the closer the distance is, the greater the weight.

d _i Representing adjacent cells n _i Distance to central cell n;

p is an adjustable parameter for adjusting the influence degree of the distance and takes a value larger than or equal to 1;

for each adjacent cell, calculate its distance d to the center cell _i The weights are then calculated using a formula. The purpose of this design is to make the closer cells have a greater impact on the center cell. The status of the road sign is not only dependent on the status of all surrounding cells, but it is emphasized that cells closer to the central cell have an influence on the status thereof. This enhances the information correlation in the local area, so that the state determination of the road sign is more accurate. The influence of remote cells is reduced by means of weight decay, which avoids remote interference. In the operation and maintenance of the road sign, the state of a far distance may not be greatly related to the actual condition of the road sign, and by reducing the influence of the state, the attention of the system to a local area can be improved. The closer cells have a greater influence on the central cell, meaning that the change in state of the road sign in the local area is more likely to be affected by the local situation. Based on the characteristics, the conditions of the road signs at different positions can be predicted more accurately, so that the establishment of operation and maintenance strategies is optimized, and the efficiency and pertinence of operation and maintenance are improved. Most importantly, the state of the road sign may change due to local conditions, and the influence of the closer cells on the state is greater so that the system can adapt to local changes more. This is particularly important for guideboard operation and maintenance, as the environment in which the road signs are located may change due to traffic, light, etc., and adapting to these changes is critical to maintaining accuracy of operation and maintenance decisions.

S2.2.2, calculating a weighted sum of states in Moore neighborhood N:

s _i ＝[0 1]

1 represents an active state; 0 represents an inactive state;

for each neighboring cell in Moore neighborhood N, its state is multiplied by the corresponding weight, and then these weights are summed.

S2.2.2, judging the next state:

judging the weighted summation result, and taking the majority value of the states in Moore neighborhood N as the next state:

and determining the state of the next cell by judging the weighted summation result. If the weighted sum is more than or equal to half of the total number of adjacent cells, taking 1 to represent an active state; otherwise taking 0 indicates an inactive state. The weighted sum result is compared to half of the total number of neighboring cells. And if the number is more than or equal to half, setting the device to be in an active state (1), otherwise setting the device to be in an inactive state (0).

S2.3, predicting the information vector DV of the next time step ₊₁ ：

DV ₊₁ ＝[f(n),f(n),f(n)......f(n)]Totally k

Predicting and generating information vector DV of next time step based on next state obtained by transfer function ₊₁ This information vector will be used as a basis for prediction and decision at the next time step. The number of f (n) is k.

In the present embodiment, DV ₊₁ The flow of the solution of (2) can be regarded as:

p1, moore neighborhood weight calculation: for each cell, the weights w of its surrounding neighboring cells are calculated _i Wherein the weight represents the effect of distance from the center cell on the center cell. This effect may be quantified by the inverse of the distance or other function.

P2、State weighted summation: for each cell, the states s of its surrounding neighboring cells are determined _i Multiplied by the corresponding weight w _i Then, summing the weighted states of all adjacent cells to obtain a weighted sum result f;

p3, judging the next state: judging a weighted summation result f, and if f is more than or equal to half of the total number k of adjacent cells, setting the state f (n) of the next time step to be 1 to represent an active state; otherwise, it will be set to 0, indicating an inactive state.

Through this process, the present embodiments predict and determine the state of each cell of the road sign surface at the next time step based on the states of the adjacent cells and their distances from the center cell to form the information vector DV of the next time step ₊₁ . Such predictions may be implemented based on current time-of-day status and weights to facilitate formulating maintenance work policies.

It should be noted that, in the present embodiment, the information vector DV of the next time step ₊₁ The solution flow of (1) reserves the information of the previous time step DV, and specifically comprises the following steps:

(1) The weighted summation takes into account the surrounding cell states: in calculating DV ₊₁ In this case, a weighted summation approach is used, where the state of each neighboring cell is considered and its influence is determined by the distance from the center cell. This ensures that the state of the surrounding cells is opposite to DV ₊₁ An influence is generated.

(2) Relationship between weight and distance: weight w _i The calculation of (2) takes into account the distance from the central cell, and weights are typically defined using the inverse of the distance or other function. This means that cells closer in distance have a greater impact on the central cell, preserving the impact of the state of the surrounding cells on the next time step.

(3) Judging the next state to consider neighborhood multi-numerical values: and when judging the state of the next step, adopting a multi-numerical mode. If the weighted sum f is equal to or greater than half the total number k of neighboring cells, it indicates that the surrounding cells tend to be active, thereby determining the state of the next time step to be active. This approach takes into account the majority of the opinion of the neighborhood, preserving the consistency of information.

Therefore, the set of calculation flow maintains the information of the previous time step DV, and takes the states of surrounding cells and the distance between the surrounding cells and the central cell into consideration, so as to reasonably predict and form the information vector DV of the next time step ₊₁ 。

Embodiment four:

the embodiment further discloses a specific implementation mode of S3 of the road sign adaptive operation and maintenance determination method, and in S3, the Dempster' S combination principle includes:

s3.1, calculating the credibility by using a membership function:

bel (A) and Bel (B) represent the confidence levels of evidence A and evidence B, respectively, A _i And B _i Representing the state of the i-th element in the information vector DV respectively corresponding thereto, i.e. the i-th neighbor cell in the Moore neighborhood, k of the formula is the total number of neighbor cells. Bel (A) and Bel (B) are defined as the average of all elements in the information vector DV:

specifically, in the Dempster's combination principle, the credibility of each evidence is first calculated. The confidence level is calculated by the average value of each element in the information vector DV.

S3.2, calculating the complementarity of the credibility by using a complementarity function: complementarity is a complement to confidence, representing a measure of confidence opposite:

Pl(A)＝1-Bel(A)

Pl(B)＝1-Bel(B)

pl (A) and Pl (B) represent complementarity of credibility of evidence A and evidence B, respectively;

s3.3, merging into a joint membership function: this step combines the confidence and complementarity of the two pieces of evidence to form a joint membership function. The method mainly comprises two substeps:

S3.3.1 calculating the intersection m (A.u.B) of two evidences, also called mutual confidence, for measuring the degree of conflict of the two evidences:

bel (X) represents the degree of trust for subset X, which can be understood as the importance or degree of trust of the event represented by subset X in the overall event;

s3.3.2, calculating a joint membership function Bel (a n B):

m (a n B) represents the intersection of evidence a and evidence B (also known as the portion of mutual trust or conflict), i.e. the portion where two evidence in some cases produce contradictions or inconsistencies; m (A.cndot.B) reflects the degree of this contradiction. Specifically, for any subset X, bel (X) is the confidence of that subset, and m (A n B) is the uncertainty measure of the intersection of A and B, used to measure the degree of conflict of the two. This value reflects the trustworthiness of the two evidence after merging. The above steps provide a method of combining different evidence into a joint membership function so that the contribution and impact of different evidence on the guideboard operation and maintenance can be more comprehensively considered.

In this embodiment, complementarity is a concept in D-S evidence theory, used as a complementary or inverse measure of confidence. It is related to the degree of credibility (Belief) to represent the uncertainty of a proposition and can also be regarded as the degree of credibility of the proposition. In the D-S evidence theory of this embodiment, the confidence Bel (a) of a proposition represents the degree of confidence in the proposition, while the complementarity Pl (a) represents a measure of the degree of distrust or uncertainty in the proposition. Complementarity is defined in terms of the complement of confidence, namely:

Pl(A)＝1-Bel(A)

Pl(B)＝1-Bel(B)

If the confidence level of a proposition is high, its complementarity will be low and vice versa. Complementarity is used to provide a comprehensive view of the credibility of the proposition from an untrusted or uncertain perspective. In this way, the present embodiments are better able to understand and handle uncertainties, especially in the case of multi-source information fusion or multiple evidence.

Fifth embodiment:

the embodiment further discloses a specific implementation mode of the S3 of the road sign adaptive operation and maintenance determination method, in S3, the correction factor α is obtained by calculating mutual trust of intersection of evidence a and evidence B:

the correction factor alpha is the interval value between [0,1 ]. Bel (A.u.B) represents the confidence level of the intersection of evidence A and evidence B, and m (A.u.B) represents the mutual confidence level of the two evidences at the intersection part, i.e. the degree of conflict of the two evidences. The correction factor alpha can provide an index for measuring the importance of the credibility of the intersection of two evidences, and if the credibility of the intersection is high, the influence of the intersection part on the whole is large.

Further, the ensured value of the correction factor alpha is ensured between [0,1] by the denominator (1-m (A.cndot.B)) in the calculation formula. In the calculation formula of the correction factor alpha, the value range of the denominator is [0,1], which is based on m (A and B), namely the mutual trust degree of the intersection of the evidence A and the evidence B, namely the conflict degree of the evidence A and the evidence B. In particular, m (A.cndot.B) reflects the degree to which two pieces of evidence contradict or disagree in some cases. In this case, if m (A.cndot.B) is close to 0, indicating that the intersection of the two is small and the degree of conflict is low, 1-m (A.cndot.B) is close. And m (A.cndot.B) is close to 1, which means that the intersection of the two is large and the degree of conflict is high, 1-m (A.cndot.B) is close to 0.

Thus, the denominator (1-m (A.cndot.B)) ensures that the correction factor α has a value between [0,1], since the value range of (1-m (A.cndot.B)) is [0,1], and α is the result of dividing the numerator Bel (A.cndot.B) by the denominator (1-m (A.cndot.B)), α is also necessarily between [0,1 ]. Thus α can also be considered as a percentage correction factor, for example, α is 0.65 and 65%.

In S3, the correction includes:

the correction factor alpha is used for adjusting the weight w _i To modify the output of the new transfer function f'. The correction process affects the output of 'f' so that the output is more practical, and the credibility and mutual credibility of the intersection part are considered. The effect of the correction factor alpha is to adjust the weight, so that the evidence with higher reliability has greater influence on the transfer function, thereby improving the overall accuracy and reliability.

In this embodiment, in the D-S evidence theory, the merge rule is to determine the trustworthiness of the composite evidence by combining the trustworthiness of two or more evidences. D-S evidence theory allows the present embodiments to combine evidence from different sources to produce a comprehensive, modified, more reliable message.

In the calculation of the correction factor alpha, the merging rule of the D-S evidence theory, in particular the Dempster' S combination principle, is utilized. This principle uses the confidence Bel (a n B) and the collision degree m (a n B) of the intersection a n B of two evidences to calculate the correction factor α. In the calculation formula of the correction factor alpha, the denominator (1-m (A.u.B)) is used for adjusting the size of alpha, so that the range of alpha is ensured to be within [0,1 ].

The correction factor α retains the confidence Bel (a n B) of the intersection, which is the composite information of evidence of the intersection. In the calculation of α, the molecule Bel (A.cndot.B) represents a part of the integrated information of the intersection. Thus, α contains the information of intersection A.cndot.B, which is obtained by the merging rules of the D-S evidence theory. The introduction and calculation of the correction factor α is essentially a correction of the transfer function. Alpha is obtained by adjusting the weight w _i The output of the new transfer function f' (n) is affected to reflect the effect of the combined evidence on the cell state. In this way the first and second light sources,the correction factor alpha can correct the original transfer function to a certain extent so that the change of the cell state can be reflected more accurately.

In summary, α corrects the transfer function by preserving the credibility information of the intersection and combining with the merging rule of the D-S evidence theory, so as to obtain a new transfer function more in line with the actual situation.

Example six:

the present embodiment provides a storage medium, please refer to fig. 7 to 8, in which program instructions for implementing a road sign adaptive operation and maintenance determination method as disclosed in embodiments one to six are stored, which only show logic in the form of C language pseudo code, and the principle thereof includes:

S1, machine vision detection

Principle of: the machine vision detection is to preprocess and extract the characteristics of the road sign image so as to determine the following self-adaption operation and maintenance. The process comprises the steps of image graying, gradient calculation, non-maximum value suppression, edge detection and the like.

Key functions:

convertToGrayScale (image) the color image is converted to a gray scale image and the values of the RGB channels are converted to gray scale values using a formula.

calculateGradient (grayImage) the gradient magnitude and gradient direction of the image are calculated, typically using the Sobel operator.

nonmaxsupport (gradientDirection) performs non-maximum suppression, preserving local maxima in the gradient direction.

doubleThreshold (nonMaxSuppressedImage), double threshold edge detection, classifying pixels according to gradient amplitude values to obtain strong edges, weak edges and non-edges.

edgeTracking (edgeMap) edge tracking, connecting weak edges to strong edges, forming connected edges.

S2, cellular automaton

Principle of: the cellular automaton simulates the dynamic evolution process of the road sign image, and predicts the state of the next time step based on the state of the cells and the neighbor state. Moore neighborhood and transfer function are used to model intercellular state propagation.

Key functions:

initializeCellularGrid (image) initializing a cell state grid to divide the image into cells.

update cellul state (). Update cell state, update cell state according to transfer function and neighbor state.

predictionvector (). The information vector for the next time step is predicted, calculated based on the cell state.

S3.D-S evidence theory

Principle of: the D-S evidence theory is used to combine information from different sources, and calculate the confidence level and mutual confidence level of the evidence. And correcting the transfer function by calculating a correction factor alpha so as to improve the accuracy of the model.

Key functions:

calculateBelief (informationVector) the confidence level of the information vector is calculated.

calcualiterstaction (informationVectorA, informationVectorB) calculates the intersection of two information vectors, i.e. mutual confidence.

calcualteBelief Interaction (belifefa, belifefb, interaction) calculates joint membership functions of intersections for correction factor calculation.

correctTransitionFunction (alpha) the transfer function is modified according to the modification factor alpha.

Embodiment seven:

referring to fig. 2 to 6, the present embodiment discloses a road sign adaptive operation and maintenance monitoring device: the road sign self-adaptive operation and maintenance monitoring device is used for executing the road sign self-adaptive operation and maintenance judging method, and comprises the following steps:

A base 2 fixedly connected to the support column of the road sign 1, wherein an adjusting mechanism 3 is carried on the base 2; the adjusting mechanism 3 comprises a rotational degree of freedom and a first linear degree of freedom, and the rotational degree of freedom is used for pitching the output angle of the first linear degree of freedom; when the first linear degree of freedom is output, the first linear degree of freedom drives all the blooming components 305 to approach or separate from each other at intervals and included angles; when the distance is carried out, under the front view of the road sign 1, the projection of all the blooming assemblies 305 are covered on the body of the road sign 1;

the blooming assembly 305 is provided with at least two detection units 4 in an array, and the detection units 4 comprise at least three second linear degrees of freedom which are arranged in a coaxial annular array, and the second linear degrees of freedom are in operative connection with the CCD industrial vision camera 404 for detecting the defects of the surface of the road sign.

In the scheme, the method comprises the following steps: the self-adaptive operation and maintenance monitoring device for the road sign adopts a base 2, and the base 2 is firmly connected to a supporting column of the road sign 1. An adjustment mechanism 3 is mounted on the base 2, and the adjustment mechanism 3 has a rotational degree of freedom and a first linear degree of freedom. The rotational degrees of freedom are responsible for adjusting the output angle of the first linear degrees of freedom. The first linear degree of freedom is output by driving the plurality of displacement assemblies 305 toward or away from each other in a coordinated fashion to adjust the spacing and angle between the assemblies. The projection of these blooming assemblies 305 is overlaid on the front view of the road sign 1, enabling coverage on the road sign surface.

In the scheme, all electric elements of the whole device are powered by mains supply; specifically, the electric elements of the whole device are in conventional electrical connection with the commercial power output port through the relay, the transformer, the button panel and other devices, so that the energy supply requirements of all the electric elements of the device are met.

Specifically, a controller is further arranged outside the device and is used for connecting and controlling all electrical elements of the whole device to drive according to a preset program as a preset value and a drive mode; it should be noted that the driving mode corresponds to output parameters such as start-stop time interval, rotation speed, power and the like between related electrical components, and meets the requirement that related electrical components drive related mechanical devices to operate according to the functions described in the related electrical components.

Preferably, the controller is also provided with a wireless transmitting module and a wireless receiving module, and the wireless transmitting module sends out an instruction signal of working or suspending to the wireless receiving module through a medium; when necessary, a worker can input an instruction to the wireless transceiver module through a background wireless remote control device so as to remotely control a controller, and further, all electric elements of the device are remotely controlled to drive according to a related driving mode; meanwhile, the wireless transceiver module can also transmit the relevant coefficients or other information detected by the relevant sensing elements or the servo driving element system in the device to the background staff.

Specific: the adjusting mechanism 3 achieves accurate positioning of the monitoring device through the rotational degree of freedom and the first linear degree of freedom. The linkage of the first linear degree of freedom drives the blooming assembly 305 to adjust its spacing and angle relative to each other to accommodate the size and shape of the different pavement indicators 1. Each blooming assembly 305 carries at least two detection units 4, and each detection unit 4 comprises at least three second linear degrees of freedom which are annularly arranged along the coaxial direction. These second degrees of linear freedom are connected to a CCD industrial vision camera 404 for detecting road sign surface defects.

It will be appreciated that in this embodiment: the device realizes the self-adaptive monitoring of the road sign through the rotational freedom degree and the first linear freedom degree of the adjusting mechanism 3. The flexible adjustment of the blooming assembly 305 can cover the surface of different pavement indicators to ensure comprehensive detection. The detection unit 4 on each blooming assembly 305 realizes multi-azimuth and multi-angle defect detection through a plurality of second linear degrees of freedom, and the accuracy and the comprehensiveness of detection are ensured. The whole device can adapt to the sizes and the shapes of different road signs, and high-efficiency and accurate operation and maintenance monitoring is realized.

In some embodiments of the present application, please refer to fig. 2-6 in combination: the adjusting mechanism 3 comprises a pitching motor 301 for outputting a first linear degree of freedom, the pitching motor 301 is connected with a shaft head at one end of a screw rod 303 for pitching angle adjustment, and the screw rod 303 is used for outputting the first linear degree of freedom; the other end shaft head of the screw rod 303 is rotatably matched with a base 302 through a bearing, and an execution motor 306 for driving the screw rod 303 to rotate is mounted on the base 302; a star wheel 304 is connected on the thread surface of the screw rod 303 in a threaded manner, and the star wheel 304 is driven to slide up and down when the screw rod 303 rotates to output a first linear degree of freedom; the star wheel 304 and the base 302 are mounted with a spindle assembly 305.

In the scheme, the method comprises the following steps: the adjustment mechanism 3 comprises a pitch motor 301 for outputting an angular adjustment of the first linear degree of freedom. The pitch motor 301 achieves adjustment of the pitch angle by being connected to one end stub shaft of the lead screw 303. The screw 303 is responsible for outputting the first linear degree of freedom. The other end of the screw 303 is connected to a base 302 through a bearing, and an actuator motor 306 is mounted on the base 302 to drive the screw 303 to rotate.

Specific: the pitch motor 301 is responsible for adjusting the first linear degree of freedom, which is connected to one end head of the screw 303, the angle of which is changed by rotating the screw 303. The screw 303 is connected to the base 302 through a bearing, and the other end is connected to the execution motor 306. The actuator motor 306 drives the screw 303 to rotate, and the screw 303 is screwed with the star wheel 304. When the screw 303 is rotated, the star wheel 304 slides up and down, thereby adjusting the first linear degree of freedom.

It will be appreciated that in this embodiment: this embodiment enables angular adjustment of the first linear degree of freedom by the pitch motor 301 and the lead screw 303, enabling the device to accommodate different road sign sizes and angles. The design of the star 304 makes the adjustment more precise and stable, ensuring precise adjustment of the first linear degree of freedom. By the mode, the device can accurately cover different areas of the road sign, and comprehensive monitoring and operation and maintenance judgment are realized.

In some embodiments of the present application, please refer to fig. 2-6 in combination: the spindle assembly 305 comprises a beam framework 3051 hinged on the base 302, a supporting plate 3052 is carried on the outer part of the beam framework 3051, and one end and the other end of a hinge arm 3053 are respectively hinged on the middle part of the beam framework 3051 and the star wheel 304; the support plate 3052 is mounted with a detection unit 4 through a shelf 3054.

In the scheme, the method comprises the following steps: the spindle assembly 305 is comprised of a beam frame 3051 that is hinged to the base 302. The beam frame 3051 is externally mounted with a support plate 3052, and the support plate is mounted with a detection unit 4 via a shelf 3054. Hinge arms 3053 are hinged to the middle portion of the beam frame 3051 and the star wheel 304, respectively.

Specific: the beam frame 3051, which is the primary structure of the blooming assembly, ensures the stability of the device by being hinged to the base 302. The hinge arm 3053 is connected to the support plate 3052 by being hinged to the beam frame 3051 and the star wheel 304. The support plate 3052 is mounted with a detection unit 4 through a shelf 3054. The design enables the detection unit 4 to adapt to road signs at different positions, and multi-angle and multi-azimuth detection is realized.

It will be appreciated that in this embodiment: the blooming assembly 305 in this embodiment provides good adjustability and adaptability. The beam frame 3051 is designed such that the support plate 3052 can be rotated and adjusted in position, thereby installing the sensing units 4 at different positions. This design ensures that the detection unit 4 can cover different areas of the road sign and enables multi-angle monitoring. Meanwhile, the rotation of the supporting plate 3052 can also adapt to the surface characteristics of different road signs, so that the accuracy and the comprehensiveness of detection are ensured.

In some embodiments of the present application, please refer to fig. 2-6 in combination: when the star wheel 304 moves up and down, the star wheel transmits a first linear degree of freedom to the beam frame 3051 through the hinge arm 3053, and only one end of the beam frame 3051 is hinged to the base 302, so that the star wheel uses the base 302 as a supporting reference for angle adjustment; because the star wheel 304 is all the blooming components 305 driven synchronously, when the star wheel 304 moves up and down, the distance and the included angle between all the blooming components 305 can be adjusted to approach or separate. At this time, all the fraying modules 305 form a fraying posture covering the body of the road sign 1, and the detection unit 4 mounted thereon can perform the detection process of the road sign adaptive operation and maintenance determination method provided above.

Specific: the up and down movement of the star wheel 304 transfers the first linear degree of freedom to the beam frame 3051 via the hinge arm 3053, effecting angular adjustment. Only one end of the hinge of the beam frame 3051 is connected to the base 302, and the angle adjustment is performed with reference to the base 302. Synchronous driving of the star wheels 304 results in simultaneous adjustment of all the fraying assemblies 305, achieving adjustment of the mutual spacing and angle, forming a fraying posture.

It will be appreciated that in this embodiment: by the synchronous movement of the star wheel 304, all the fraying assemblies 305 can form a fraying posture covering the body of the road sign 1. The layout ensures that all the carried detection units 4 can fully cover the surface of the road sign board 1, realizes multi-angle and multi-azimuth self-adaptive operation and maintenance monitoring, and accords with the detection procedure of the road sign board self-adaptive operation and maintenance judging method provided in the prior art.

Further, by covering the surface of the road sign 1, all the mounted detection units 4 can detect the road sign in all directions and at multiple angles, so as to ensure that no possible defects or problems are missed. The "blooming" arrangement of the blooming assembly 305 allows the detection units 4 to operate simultaneously, improving detection efficiency and speed. The high efficiency has important significance for the operation and maintenance of the large-scale road sign, can rapidly finish detection, and reduces operation and maintenance time and cost. The layout of the bursting module 305 can adapt to different shapes and sizes of the road signs 1, so that the monitoring device is ensured to be applicable to various types of road signs, and has strong universality and adaptability. The accuracy of the monitoring result can be improved through multi-angle and multi-azimuth monitoring, the surface defect of the road indication board can be identified more accurately through comparing the detection results from different angles, the misjudgment rate is reduced, and the judgment reliability is enhanced. Through multi-angle, diversified self-adaptation fortune dimension monitoring, can evaluate the health condition of road sign more comprehensively, accurately, help making more scientific, accurate maintenance strategy, improved fortune dimension quality and effect.

Furthermore, in this embodiment, only one degree of freedom (the first linear degree of freedom) is needed to control the detection unit 4, so as to realize multi-angle and multi-azimuth self-adaptive operation and maintenance monitoring, and the complexity of the system is reduced and the design, installation and operation processes are simplified by using single degree of freedom input. Compared with multi-degree-of-freedom input, the system has the advantages that the technical threshold of the system is reduced, and the operation is more visual and easier to master. The use of a single degree of freedom input can reduce the number of actuators and control systems required, reducing manufacturing and maintenance costs. This cost saving is very important for the feasibility of the solution and the sustainability of the practical application. The single degree of freedom input reduces the complexity and possible mechanical noise of the system and helps to improve the stability and accuracy of the movement device. The system is easier to control and adjust, ensuring accurate movement to cover the surface of the road sign. The resource utilization can be better optimized by controlling the plurality of detection units with a single degree of freedom input. One degree of freedom can fully cover the movement of a plurality of detection units, effectively utilizes the potential functions of the mechanical structure and improves the efficiency of the equipment. The mode of single degree of freedom input can adapt to different shapes and sizes of the road indication board more quickly, and the self-adaptability of the system is enhanced. Meanwhile, the response speed is faster, the position of the detection unit can be quickly adjusted, and the detection requirement of real-time change is met.

In some embodiments of the present application, please refer to fig. 2-6 in combination: the detection unit 4 comprises two opposite frame bodies 401, six servo electric cylinders 402 for outputting a second linear degree of freedom are arranged between the frame bodies 401 in a ring array form, and a cylinder body and a piston rod of each servo electric cylinder 402 are respectively and universally hinged on one surface of each opposite frame body 401 through a universal joint coupling 403; a CCD industrial vision camera 404 is mounted on one frame 401. The other frame 401 is fixedly arranged on the storage frame 3054.

Specific: the frames 401 of the detection unit 4 form an annular array, and each frame is provided with a servo cylinder 402 for outputting a second linear degree of freedom. The servo cylinders 402 are designed to move the carriage in a two-dimensional plane. The CCD industrial vision camera 404 is installed on one of the frame bodies 401, and can realize multi-directional defect detection through the motion of the servo electric cylinder 402.

It will be appreciated that in this embodiment: the detection unit 4 in the embodiment adopts a ring array architecture, and six servo electric cylinders 402 enable the frame body to flexibly move on a two-dimensional plane, so that a second multi-azimuth linear degree of freedom is realized. By installing the CCD industrial vision camera 404, the high-efficiency and comprehensive detection of the surface defects of the road sign is realized. The design ensures multi-angle and multi-azimuth coverage of the detection unit 4, has strong adaptability, and can accurately monitor the surface condition of the road sign.

It should be noted that, in this embodiment: the combination of "macro-movement" of the spindle assembly 305 with "micro-movement" of the detection unit 4 may bring about the following beneficial effects:

(1) Efficient coverage and detection: the macro movement of the blooming assembly is responsible for large-scale movement and adjustment, so that the surface of the whole road sign is ensured to be covered. The micro-movement of the detection unit is responsible for fine surface detection, and through the combination, each part of the road sign, including a large area and a fine area, can be efficiently and comprehensively detected.

(2) Self-adaptive coverage and accurate detection: the combination of macro movement and micro movement can realize self-adaptive coverage, and the position and the angle of the blooming assembly can be dynamically adjusted so as to adapt to road signs with different shapes and sizes. Meanwhile, the accurate detection at each specific position is ensured by inching, and the subtle changes and defects on the surface of the road sign are effectively captured.

(3) All-round data acquisition: through the combination of macro movement and micro movement, the omnibearing data of the road sign surface can be collected, including information of different angles, different heights and different positions. Such multi-angle, multi-location data acquisition facilitates a more comprehensive understanding of the condition of the roadway sign, providing more information for subsequent maintenance and operation.

(4) Energy conservation and efficiency improvement: when the movement and adjustment are needed, the macro-movement combined with the micro-movement can complete the detection task with minimum energy consumption. The micro-motion can accurately move aiming at a specific small area, so that unnecessary energy consumption is reduced, and the efficiency is improved.

(5) Easy and convenient to handle and maintenance: the macro-movement and micro-movement are comprehensively utilized, so that the operation flow and maintenance process of the equipment can be simplified. The staff can realize comprehensive coverage and detection through simple operation, and is more convenient during maintenance simultaneously, has reduced the time and the cost of maintaining.

All of the above examples merely represent embodiments of the invention which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.

Claims

1. A self-adaptive operation and maintenance judging method for a road sign is characterized in that: the method comprises the following steps:

S3, executing D-S evidence theory verification: in the current time step, taking the information vector DV of the S1 as evidence A, taking the predicted current information vector DV of the S2 in the last time step ₊₁ As evidence B, the evidence A and the evidence B are combined by using the Dempster's combination principle and mapped into a [0,1 ]]A correction factor α within a section, which corrects the transfer function f in the S2 at each time step.

2. The road sign adaptive operation and maintenance determination method according to claim 1, wherein: in the S1, it includes:

generating an information vector DV:

DV＝[EM(x,y)，HSV]

where symbol "is a vector concatenation operation.

3. The road sign adaptive operation and maintenance determination method according to claim 1, wherein: in the S2, it includes:

s2.2, the transfer function f:

d _i representing adjacent cells n _i A distance to the central cell n;

p is an adjustable parameter, and takes a value greater than or equal to 1;

s2.2.2, calculating a weighted sum of states in Moore neighborhood N:

s2.2.2, judging the next state:

1 represents an active state, 0 represents an inactive state;

S2.3, predicting the information vector DV of the next time step ₊₁ ：

DV ₊₁ ＝[f(n),f(n),f(n)......f(n)]Totally k

The number of f (n) is k.

4. The road sign adaptive operation and maintenance determination method according to claim 3, wherein: in the step S3, the Dempster' S combination principle includes:

s3.1, calculating the credibility by using a membership function:

Pl(A)＝1-Bel(A)

Pl(B)＝1-Bel(B)

s3.3, merging into a joint membership function:

bel (X) represents the degree of confidence for subset X;

s3.3.2, calculating a joint membership function Bel (a n B):

m (a n B) represents the intersection of the evidence a and the evidence B.

5. The method for adaptively determining the operation and maintenance of the road sign according to claim 4, wherein: in the step S3, the correction factor α is obtained by calculating the mutual trust of the intersection of the evidence a and the evidence B:

The correction factor alpha is a value of a section between [0,1 ].

6. The method for adaptively determining the operation and maintenance of the road sign according to claim 5, wherein: in the S3, the correcting includes:

the correction factor alpha is used for adjusting the weight w _i To modify the output of the new transfer function f'.

7. The utility model provides a road sign self-adaptation fortune dimension monitoring devices which characterized in that: the road sign self-adaptive operation and maintenance monitoring device is used for executing the road sign self-adaptive operation and maintenance judging method according to any one of claims 1 to 6, and comprises the following steps: a base (2) fixedly connected to a support column of the road sign (1), wherein an adjusting mechanism (3) is carried on the base (2);

the adjusting mechanism (3) comprises a rotational degree of freedom and a first linear degree of freedom, the rotational degree of freedom being used for pitching an output angle of the first linear degree of freedom; when the first linear degree of freedom is output, the first linear degree of freedom drives all the blooming components (305) to approach or separate from each other in terms of distance and included angle; when the distance is carried out, all projections of the blooming assembly (305) cover the tile body of the road sign (1) under the front view of the road sign (1);

The detection unit (4) comprises at least three second linear degrees of freedom which are arranged in a coaxial annular array manner, and the second linear degrees of freedom are connected with a CCD industrial vision camera (404) for detecting the surface defects of the road indication board.

8. The roadway sign adaptive operation and maintenance monitoring device of claim 7, wherein: the adjusting mechanism (3) comprises a pitching motor (301) for outputting the first linear degree of freedom, the pitching motor (301) is connected with a shaft head at one end of a screw rod (303) to adjust the pitching angle, and the screw rod (303) is used for outputting the first linear degree of freedom;

the other end shaft head of the screw rod (303) is in rotary fit with a base (302), and an execution motor (306) for driving the screw rod (303) to rotate is mounted on the base (302);

a star wheel (304) is connected to the threaded surface of the screw rod (303) in a threaded manner, and the screw rod (303) drives the star wheel (304) to slide up and down when rotating to output the first linear degree of freedom;

the star wheel (304) and the base (302) are provided with the spindle assembly (305).

9. The roadway sign adaptive operation and maintenance monitoring device of claim 8, wherein: the blooming assembly (305) comprises a beam framework (3051) hinged on the base (302), a supporting plate (3052) is carried on the outer part of the beam framework (3051), and one end and the other end of a hinge arm (3053) are respectively hinged on the middle part of the beam framework (3051) and the star wheel (304);

The supporting plate (3052) is provided with the detection unit (4) through a commodity shelf (3054).

10. The roadway sign adaptive operation and maintenance monitoring device of claim 7, wherein: the detection unit (4) comprises two mutually opposite frame bodies (401), six servo electric cylinders (402) for outputting the second linear degree of freedom are arranged between the frame bodies (401) in a ring-shaped array, and a cylinder body and a piston rod of each servo electric cylinder (402) are respectively and universally hinged on one surface of each of the two mutually opposite frame bodies (401) through a universal joint coupling (403);

one of the frame bodies (401) is provided with the CCD industrial vision camera (404).