A visual inspection method for airport noise reduction diversion walls

Abstract

The present invention relates to the technical field of visual inspection of guide walls, and specifically to a visual inspection method for airport noise reduction guide walls. The present invention first obtains all guide groove areas and two corresponding guide groove edges in the area to be inspected, and obtains the ground height of each guide groove area from the ground; further obtains the flexural deformation coefficient and the foundation settlement coefficient of the area to be inspected; according to the flexural deformation coefficient and the foundation settlement coefficient, obtains the deformation coefficient of the area to be inspected, and then obtains the deformation detection result of the windward surface of the airport noise reduction guide wall. Based on the groove texture characteristics of the noise reduction guide wall, combined with the groove light and shadow changes and texture deformation caused by the deformation, combined with the foundation settlement characteristics of the guide wall deformation, the present invention quantifies the deformation coefficient of the area to be inspected, and improves the deformation detection efficiency and accuracy of the airport noise reduction guide wall.

Description

Technical Field

[0001] The present invention relates to the technical field of visual inspection of guide walls, and in particular to a visual inspection method for noise reduction guide walls at airports. Background Art

[0002] When the aircraft engine is tested or taking off, it will generate huge thrust and high-temperature and high-speed exhaust gas, accompanied by high-decibel noise; the airport noise reduction diversion wall can block or disperse these noise and exhaust gas through its special structure and material design, reducing the impact on the surrounding environment and personnel. However, due to its long-term strong impact and high-temperature corrosion, as well as the influence of extreme weather conditions such as strong winds, heavy rains, ice and snow, the diversion wall structure may deform, destroy its structural stability, and then collapse and other serious accidents may occur, so the deformation detection of the airport noise reduction diversion wall is very important.

[0003] In the prior art, visual inspection technology can be used to collect images of airport noise reduction diversion walls, and then a large number of images of airport noise reduction diversion walls with different deformation degrees are obtained for manual annotation to construct a training sample set, and a deformation recognition model is trained based on a deep learning algorithm, so that the deformation degree of the current airport noise reduction diversion wall can be evaluated using the trained deformation recognition network; however, the process of collecting diversion wall samples with different deformation degrees to construct a training set and training the recognition model is complicated, and the recognition effect is easily affected by the training set and training hyperparameters, resulting in low efficiency in deformation detection of airport noise reduction diversion walls. Summary of the invention

[0004] In order to solve the technical problem that the deformation detection efficiency of the airport noise reduction diversion wall is low in the prior art, the purpose of the present invention is to provide a visual detection method for the airport noise reduction diversion wall. The technical solution adopted is as follows:

[0005] The present invention provides a visual detection method for an airport noise reduction diversion wall, the method comprising:

[0006] Obtain a grayscale image of the area to be detected on the windward surface of the airport noise reduction guide wall, obtain all guide groove areas and two guide groove edges in each of the guide groove areas in the grayscale image, and obtain the ground height of each of the guide groove areas from the ground;

[0007] According to the grayscale uniformity in all the guide groove areas and the difference in the number of adjacent pixels on both sides of each guide groove edge in each guide groove area, the flexural deformation coefficient of the area to be detected is obtained; according to the distance between the pixels on the two guide groove edges in each guide groove area and the ground height of each guide groove area, the basic settlement coefficient of the area to be detected is obtained;

[0008] According to the flexural deformation coefficient and the foundation settlement coefficient, the deformation coefficient of the area to be detected is obtained; according to the deformation coefficient of the area to be detected, the deformation detection result of the windward surface of the airport noise reduction guide wall is obtained.

[0009] Furthermore, the method for obtaining the guide groove area includes:

[0010] Obtain all closed edges in the grayscale image, and obtain the average area of ​​the regions enclosed by all the closed edges, and use the region enclosed by the closed edges whose area is smaller than the average area as a suspected guide groove region;

[0011] Taking the degree of the closed edge-like moment of each of the suspected guide groove regions as the first confidence level of each of the suspected guide groove regions;

[0012] According to the distribution regularity of each of the suspected guide groove regions and its adjacent guide groove regions, a second confidence level of each of the suspected guide groove regions is obtained;

[0013] The groove confidence of each of the suspected guide groove areas is obtained according to the first confidence and the second confidence, wherein the first confidence and the second confidence are both positively correlated with the groove confidence; and all the suspected guide groove areas whose groove confidence is greater than a preset threshold are taken as guide groove areas.

[0014] Furthermore, the method for obtaining the class moment degree includes:

[0015] Taking any edge pixel point on each of the closed edges as a starting point, obtaining the 8 chain codes corresponding to the closed edges; in the 8 chain codes corresponding to each of the closed edges, sorting the occurrence frequency of each chain code value in descending order, taking the top two chain code values ​​as the first group of chain code values, and taking the third and fourth chain code values ​​as the second group of chain code values;

[0016] On the closed edge, according to the difference between the chain code value of each group of chain code values ​​and the numerical difference of the preset positive integer, and the difference in the number of edge pixels corresponding to the chain code values, the confidence coefficient of each group of chain code values ​​corresponding to the chain code values ​​of the opposite sides of the rectangle is obtained; the numerical difference and the quantitative difference are both negatively correlated with the confidence coefficient;

[0017] The confidence coefficients of the two groups of chain code values ​​are multiplied to obtain the moment-like degree corresponding to the closed edge.

[0018] Furthermore, the method for obtaining the second confidence level includes:

[0019] Taking any of the suspected guide groove areas as the target area; within the neighborhood of the target area, taking the left and right suspected guide groove areas as horizontal reference areas, and taking the upper and lower suspected guide groove areas as vertical reference areas;

[0020] Respectively obtaining a first distance and a second distance between the two horizontal reference areas and the corresponding centroids of the target area, and mapping the difference between the first distance and the second distance to a negative correlation to obtain a horizontal uniform distribution parameter;

[0021] Respectively obtaining a third distance and a fourth distance between the two vertical reference areas and the corresponding centroids of the target area, and mapping the difference between the third distance and the fourth distance in a negative correlation manner to obtain a vertical uniform distribution parameter;

[0022] The horizontal uniform distribution parameter and the vertical uniform distribution parameter are combined to obtain a second confidence level of each target area.

[0023] Furthermore, the method for obtaining the flexural deformation coefficient includes:

[0024] According to the grayscale difference of the pixel points on both sides of the central axis of each guide groove area, a first deflection parameter of each guide groove area is obtained; the central axis is parallel to the edge of the guide groove, and the grayscale difference is positively correlated with the first deflection parameter;

[0025] According to the difference in the number of adjacent pixel points on both sides of each guide groove edge, the deformation parameter of each guide groove edge is obtained; according to the difference in the deformation parameters of the two guide groove edges in the guide groove area, the second deflection parameter of each guide groove area is obtained;

[0026] According to the first deflection parameter and the second deflection parameter, a deflection sub-parameter of each of the guide groove regions is obtained, wherein the first deflection parameter and the second deflection parameter are both positively correlated with the deflection sub-parameter;

[0027] The deflection sub-parameters of all the guide groove areas in the grayscale image are integrated to obtain the deflection deformation coefficient of the area to be detected.

[0028] Furthermore, the method for obtaining the deformation parameter includes:

[0029] In the grayscale image, all outermost pixel points and all innermost pixel points on the edge of each guide groove are respectively obtained, and all pixel points adjacent to the outermost pixel points are used as outer pixel points of the region, and all pixel points adjacent to the innermost pixel points are used as inner pixel points of the region, and neither the outer pixel points of the region nor the inner pixel points of the region are on the edge of the guide groove;

[0030] The deformation parameter of each guide groove edge is obtained according to the difference between the total number of pixel points outside the area of ​​each guide groove edge and the total number of pixel points inside the area.

[0031] Furthermore, the method for obtaining the basic sedimentation coefficient includes:

[0032] In each of the guide groove regions, according to the distance between all innermost pixel points on two edges of the guide groove, a flat distortion parameter of each of the guide groove regions is obtained;

[0033] The ground height of each diversion groove area is negatively correlated and mapped as a weight, and the corresponding flat distortion parameter is weighted to obtain a basic settlement sub-parameter of each diversion groove area;

[0034] The basic sedimentation sub-parameters of all the guide groove areas are combined to obtain the basic sedimentation coefficient of the area to be detected.

[0035] Furthermore, the method for obtaining the flat distortion parameter includes:

[0036] In each of the guide groove regions, all innermost pixel points on each guide groove edge are sequentially sorted, and according to the distance between all innermost pixel points with the same sorting sequence number on the two guide groove edges, the average groove width and width range of the corresponding guide groove are obtained;

[0037] According to the average groove width corresponding to each guide groove area and its deviation from the average groove width corresponding to all guide groove areas, the relative flatness of each guide groove area is obtained; the width range is used as the distortion coefficient of the guide groove area;

[0038] The flat distortion parameter of each guide groove area is obtained according to the relative flatness and the distortion coefficient, and both the relative flatness and the distortion coefficient are positively correlated with the flat distortion parameter.

[0039] Furthermore, the method for obtaining the deformation coefficient includes:

[0040] The product of the flexural deformation coefficient and the foundation settlement coefficient is used as the deformation coefficient of the area to be detected.

[0041] Furthermore, the method for obtaining the deformation detection result includes:

[0042] When the deformation coefficient is greater than a preset threshold, the deformation detection result is determined to be deformation, otherwise the deformation detection result is determined to be non-deformation.

[0043] The present invention has the following beneficial effects:

[0044] The present invention obtains a grayscale image of a to-be-detected area on the windward surface of an airport noise reduction guide wall, obtains all guide groove areas and two guide groove edges in each guide groove area in the grayscale image, and obtains the ground height of each guide groove area from the ground, so as to prepare for the subsequent analysis of groove texture deformation; according to the grayscale uniformity in all guide groove areas and the difference in the number of adjacent pixel points on both sides of each guide groove edge in each guide groove area, the flexural deformation coefficient of the to-be-detected area is obtained, and the flexural deformation coefficient is combined with the light and shadow change influence and texture distortion characteristics brought by the groove deformation to preliminarily evaluate the deformation degree of the to-be-detected area; according to the distance between the pixel points on the two guide groove edges in each guide groove area and the ground height of each guide groove area, the basic settlement coefficient of the to-be-detected area is obtained, and the basic settlement coefficient is combined with the settlement collapse characteristics and the influence of gravity on the shape extrusion degree of the groove to further evaluate the deformation degree of the to-be-detected area; and then according to the flexural deformation coefficient and the basic settlement coefficient, the deformation coefficient of the to-be-detected area is obtained; according to the deformation coefficient of the to-be-detected area, the deformation detection result of the windward surface of the airport noise reduction guide wall is obtained. The invention is based on the groove texture characteristics of the noise reduction and diversion wall, combined with the groove light and shadow changes and texture deformation caused by the deformation, combined with the foundation settlement characteristics of the deformation of the diversion wall, to quantify the deformation coefficient of the area to be detected, thereby improving the deformation detection efficiency and accuracy of the airport noise reduction and diversion wall. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A flow chart of a visual inspection method for an airport noise reduction diversion wall provided by one embodiment of the present invention;

[0047] Figure 2 An example diagram of a noise reduction diversion wall for an airport to be inspected provided by an embodiment of the present invention;

[0048] Figure 3 A flow chart of a method for obtaining a guide groove area provided by an embodiment of the present invention;

[0049] Figure 4 A flowchart of a method for obtaining a flexural deformation coefficient provided by an embodiment of the present invention;

[0050] Figure 5 A schematic diagram of pixel points outside and pixel points inside a guide groove region provided by an embodiment of the present invention;

[0051] Figure 6 A flow chart of a method for obtaining a basic sedimentation coefficient provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the visual detection method of the airport noise reduction diversion wall proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0053] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0054] The following is a detailed description of a specific scheme of a visual inspection method for an airport noise reduction diversion wall provided by the present invention in conjunction with the accompanying drawings.

[0055] See also Figure 1 , which shows a flow chart of a visual detection method for an airport noise reduction diversion wall provided by an embodiment of the present invention, specifically comprising:

[0056] Step S1, obtain a grayscale image of the area to be detected on the windward side of the airport noise reduction guide wall, obtain all guide groove areas and two guide groove edges in each guide groove area in the grayscale image, and obtain the ground height of each guide groove area from the ground.

[0057] It should be noted that the deformation detection object targeted by the embodiment of the present invention is a vertical noise reduction guide wall, that is, the windward surface of the noise reduction guide wall is perpendicular to the ground, and the windward surface panel of the noise reduction guide wall is grooved, which is mainly used for noise reduction and guide; please refer to Figure 2 , which shows an example diagram of an airport noise reduction guide wall to be detected provided by an embodiment of the present invention.

[0058] In one embodiment of the present invention, a high-resolution industrial camera is first installed directly in front of the noise reduction and diversion wall of the airport to collect high-definition color images of the area to be detected on the windward side of the noise reduction and diversion wall. Each area to be detected should include the entire area from the bottom to the top of the noise reduction and diversion wall. In order to facilitate the analysis of the deformation of the noise reduction and diversion wall according to the texture information of the diversion groove, the color image is further grayed to obtain a grayscale image.

[0059] It should be noted that, considering that the noise reduction guide wall is relatively long or its shape is U-shaped to accommodate various types of aircraft, a single acquisition of the entire noise reduction guide wall may not be able to effectively analyze its internal texture, so the implementer can divide the noise reduction guide wall into multiple areas to be detected according to needs; grayscale processing is a well-known technology and will not be repeated here.

[0060] Considering that the deformation of the airport noise reduction diversion wall is mainly due to long-term external impact and material aging, which leads to reduced structural stability, resulting in settlement deformation, which may lead to safety accidents such as collapse and failure; it is reflected in the texture deformation of the corresponding edge of the diversion groove. Therefore, all the diversion groove areas and the corresponding two diversion groove edges in the grayscale image are obtained to analyze the deformation.

[0061] Preferably, in one embodiment of the present invention, the method for obtaining the guide groove area includes:

[0062] See also Figure 3 , which shows a flow chart of a method for obtaining a guide groove area provided by an embodiment of the present invention, including:

[0063] Step S301, obtaining all closed edges in the grayscale image, and obtaining the average area of ​​the regions enclosed by all closed edges, and taking the region enclosed by closed edges whose area is smaller than the average area as the suspected guide groove region.

[0064] The concave-convex texture characteristics of the diversion wall are usually manifested as alternation of grooves and bulges, and all appear as a rectangular area. At the same time, the area of ​​the groove rectangular area is usually smaller than the area of ​​the bulge rectangular area; therefore, the shape and area characteristics can be combined to analyze and obtain the diversion groove area.

[0065] As an example, the canny edge detection algorithm is used to obtain all closed edges. When the area enclosed by a closed edge is smaller than the average area of ​​the areas enclosed by all closed edges, it means that it is more likely to correspond to the guide groove area; but there are also interfering edges such as stains or scratches that cause the area enclosed by the corresponding closed edge to be smaller, thus being misjudged, so further analysis and evaluation are needed.

[0066] In other examples, implementers may also use other algorithms to detect closed edges, or obtain the area of ​​the minimum circumscribed rectangle of the closed edge for evaluation, which are all well-known technologies and will not be described in detail here.

[0067] Step S302: taking the degree of the closed edge of each suspected guide groove region as the first confidence level of each suspected guide groove region.

[0068] Considering that even if settlement deformation occurs, the shape of the guide groove area should still be a quasi-rectangular shape, the greater the degree of rectangularity of the closed edge corresponding to the suspected guide groove area, the greater the confidence that it corresponds to the guide groove area.

[0069] In a preferred embodiment of the present invention, considering that chain codes are often used to describe curves or boundary contours, starting from any point on a rectangle, each point on each edge corresponds to a chain code value; the points on each edge have the same direction, while the directions of the opposite sides are opposite, which is reflected in the same or different corresponding chain code values; at the same time, the number of points corresponding to the chain code values ​​of each group of opposite sides should be similar or consistent; the acquisition method based on such moment degree includes:

[0070] Taking any edge pixel point on each closed edge as the starting point, obtain the 8-chain code corresponding to the closed edge; in the 8-chain code corresponding to each closed edge, sort the occurrence frequency of each chain code value in descending order, take the top two chain code values ​​as the first group of chain code values, and take the third and fourth chain code values ​​as the second group of chain code values;

[0071] On the closed edge, according to the difference between the chain code value of each group of chain code values ​​and the numerical difference of the preset positive integer, and the difference in the number of edge pixels corresponding to the chain code value, the confidence coefficient of each group of chain code values ​​corresponding to the opposite sides of the rectangle is obtained; the numerical difference and the quantitative difference are both negatively correlated with the confidence coefficient;

[0072] Multiply the confidence coefficients of the two sets of chain code values ​​to obtain the moment-like degree of the corresponding closed edge.

[0073] Each edge pixel on the closed edge corresponds to a chain code value of 0-7, among which the top four chain code values ​​are most likely to be the chain code values ​​corresponding to the pixels on the four sides of the suspected rectangle. After obtaining two groups of chain code values ​​suspected to be corresponding to the long sides of the rectangle, the first group of chain code values ​​suspected to be corresponding to the short sides of the rectangle, and the second group of chain codes suspected to be corresponding to the short sides of the rectangle, the confidence coefficient of each group of chain code values ​​corresponding to the opposite sides of the rectangle can be calculated in combination with the relevant feature information of the 8-chain code of the rectangle to determine its degree of rectangle-likeness; the 8-chain code is a prior art and will not be described in detail.

[0074] As an example, the confidence coefficient calculation formula includes:

[0075] Among them, A u,z B is the confidence coefficient of the chain code value of the zth group in the uth closed edge corresponding to the chain code value of the opposite side of the rectangle; z1 is the number of edge pixels corresponding to the chain code value with the highest frequency in the zth group of chain code values; B z2 is the number of edge pixels corresponding to the chain code value with the least frequency in the zth group of chain code values; C z is the number of all edge pixels corresponding to the zth group of chain code values; D z1is the chain code value with the highest frequency in the zth group of chain code values; D z2 is the chain code value with the least frequency in the zth group of chain code values; α is a preset positive integer, which is 4 in this example; exp( ) is an exponential function with the natural constant e as the base.

[0076] Step S303: obtaining a second confidence level of each suspected guide groove region according to the distribution regularity between each suspected guide groove region and its adjacent guide groove regions.

[0077] Considering that the regular arrangement of the diversion grooves remains unchanged even if settlement deformation occurs, the confidence can be evaluated based on the neighborhood distribution regularity of the suspected diversion groove area.

[0078] In a preferred embodiment of the present invention, considering that if the guide groove areas are arranged regularly, the guide groove areas above and below and left and right in the neighborhood should be symmetrical; therefore, the method for obtaining the second confidence includes:

[0079] Any suspected diversion groove area is taken as the target area; within the neighborhood of the target area, the left and right suspected diversion groove areas are taken as horizontal reference areas, and the upper and lower suspected diversion groove areas are taken as vertical reference areas;

[0080] The first distance and the second distance between the corresponding centroids of the two horizontal reference areas and the target area are respectively obtained, and the difference between the first distance and the second distance is negatively correlated and mapped to obtain a horizontal uniform distribution parameter;

[0081] Obtaining the third distance and the fourth distance between the corresponding centroids of the two vertical reference areas and the target area respectively, mapping the difference between the third distance and the fourth distance in a negative correlation manner, and obtaining a vertical uniform distribution parameter;

[0082] The horizontal uniform distribution parameters and the vertical uniform distribution parameters are integrated to obtain the second confidence level of each target area.

[0083] As an example, the horizontal uniform distribution parameter is evaluated by using the difference between the corresponding centroids of the two suspected guide groove areas that are closest to the centroid of the target area in the left and right directions of the target area. Specifically, the Euclidean distance between the centroid of the left adjacent suspected guide groove area of ​​the target area and the centroid of the target area is taken as the first distance, and the Euclidean distance between the centroid of the right adjacent suspected guide groove area of ​​the target area and the centroid of the target area is taken as the second distance. The absolute value of the difference between the first distance and the second distance is calculated. The smaller the absolute value of the difference, the symmetrical arrangement of the two suspected guide groove areas on the left and right is based on the target area. The absolute value of the difference is taken as the x in the exponential function exp(-x) with the natural constant e as the base, and negative correlation mapping normalization is performed; the vertical uniform distribution parameter is obtained in the same way, which will not be repeated; then the horizontal uniform distribution parameter is multiplied by the vertical uniform distribution parameter to obtain the second confidence.

[0084] In other examples, implementers may also use other normalization methods, such as adding the absolute value of the difference to the anti-zero parameter 1 and then performing a reciprocal operation; or they may use basic mathematical methods such as addition or weighted summation to combine two uniform distribution parameters, which will not be elaborated here.

[0085] In other embodiments of the present invention, the implementer may also construct a neighborhood with the target area as the center, obtain suspected guide groove areas on the horizontal line, vertical line and diagonal line respectively, and then obtain the centroid coordinate sequence of the corresponding suspected guide groove area on each line, and then perform differential to obtain the Euclidean distance sequence, and obtain the variance of all Euclidean distances to evaluate the discreteness. The smaller the variance, the more regular the distribution. Then, the corresponding variances on all lines are averaged to obtain the second confidence level.

[0086] Step S304, obtaining the groove confidence of each suspected guide groove area according to the first confidence and the second confidence, wherein the first confidence and the second confidence are both positively correlated with the groove confidence; and taking all suspected guide groove areas whose groove confidence is greater than a preset threshold as guide groove areas.

[0087] As an example, the first confidence level and the second confidence level are multiplied and combined to obtain the groove confidence level; the preset threshold is set to 0.8; in other examples, the two can be combined using basic mathematical methods such as addition or weighted summation, or the preset threshold can be set by oneself.

[0088] In other embodiments of the present invention, the implementer can manually annotate all groove areas in the grayscale image, construct a training sample set, and thus train the feature recognition network model, and then use the trained feature network recognition model to obtain all diversion groove areas in the grayscale image of the area to be detected; or use the light and dark differences between grooves and protrusions to obtain all diversion groove areas based on a threshold segmentation algorithm; these are all well-known technologies and will not be described in detail here.

[0089] After obtaining all the guide groove areas, the corresponding moment-like closed edge can be obtained, and the two long side corresponding edges on the moment-like closed edge are used as two guide groove edges.

[0090] Considering that settlement deformation is mainly due to material failure and under the influence of gravity, the diversion wall tends to collapse downward. The closer the diversion groove is to the foundation, the greater the possibility of settlement deformation. Therefore, the ground height of each diversion groove area is obtained.

[0091] In one embodiment of the present invention, the distance between the centroid of each guide groove area and the bottom end of the guide wall in the detection area is taken as its corresponding ground height; in other embodiments, the area center, center of gravity, etc. can also be used instead of the centroid, which will not be repeated here.

[0092] Step S2, according to the grayscale uniformity in all guide groove areas and the difference in the number of adjacent pixels on both sides of each guide groove edge in each guide groove area, obtain the flexural deformation coefficient of the area to be detected; according to the distance between the pixels on the two guide groove edges in each guide groove area and the ground height of each guide groove area, obtain the basic settlement coefficient of the area to be detected.

[0093] Considering that the deformation of the guide groove causes the change of its groove surface morphology, which in turn causes the change of light and shadow, the grayscale in the guide groove area will change unevenly; considering that the deformation of the groove texture will cause the deformation of the corresponding guide groove edge, which will lead to inconsistent number of pixels on both sides of the edge; therefore, the flexural deformation coefficient of the area to be detected can be obtained based on this.

[0094] Preferably, in one embodiment of the present invention, the method for obtaining the flexural deformation coefficient includes:

[0095] See also Figure 4 , which shows a flow chart of a method for obtaining a flexural deformation coefficient provided by an embodiment of the present invention, including:

[0096] Step S401, obtaining the first deflection parameter of each guide groove area according to the grayscale difference of the pixel points on both sides of the central axis of each guide groove area; the central axis is parallel to the edge of the guide groove, and the grayscale difference is positively correlated with the first deflection parameter.

[0097] As an example, first obtain the central axis of each guide groove area, and the central axis should be parallel to the edge of the guide groove; obtain the grayscale mean of the areas divided on both sides of the central axis respectively; the larger the absolute value of the difference in the grayscale mean of the two divided areas, the greater the light and shadow changes caused by the deflection change of the guide groove, and the larger the first deflection parameter.

[0098] It should be noted that the acquisition of the central axis is a well-known technology and will not be elaborated here.

[0099] Step S402, obtaining the deformation parameter of each guide groove edge according to the difference in the number of adjacent pixel points on both sides of each guide groove edge; obtaining the second deflection parameter of each guide groove area according to the difference in the deformation parameters of two guide groove edges in the guide groove area.

[0100] In a preferred embodiment of the present invention, the method for obtaining deformation parameters includes:

[0101] In the grayscale image, all outermost pixel points and all innermost pixel points on the edge of each guide groove are obtained respectively, and all pixel points adjacent to the outermost pixel points are taken as outer pixel points of the region, and all pixel points adjacent to the innermost pixel points are taken as inner pixel points of the region, wherein the outer pixel points of the region and the inner pixel points of the region are not on the edge of the guide groove;

[0102] The deformation parameter of each guide groove edge is obtained according to the difference between the total number of pixel points outside the region of each guide groove edge and the total number of pixel points inside the region.

[0103] For easier understanding, see Figure 5 , which shows a schematic diagram of the outer pixel points and the inner pixel points of a guide groove area provided by an embodiment of the present invention; wherein the gray area is a closed edge, the closed edge and the area enclosed by it correspond to a guide groove area, the upper and lower long sides are the guide groove edges, and taking one of the guide groove edges as an example, the pixel point labeled 1 is the inner pixel point of the area, and the pixel point labeled 2 is the outer pixel point of the area.

[0104] As an example, the second deflection parameter is calculated as: Among them, Q j is the second deflection parameter of the jth guide groove area; w j,1 is the total number of pixels outside the first groove edge of the jth groove area; n j,1 is the total number of inner pixels in the first guide groove edge of the jth guide groove area; is the deformation parameter of the first guide groove edge in the jth guide groove area; w j,2 is the total number of pixels outside the second groove edge of the jth groove area; n j,2 is the total number of inner pixels in the second guide groove edge of the jth guide groove area; is the deformation parameter of the second guide groove edge in the jth guide groove area.

[0105] The larger the second deflection parameter is, the greater the possibility that one of the two guide groove edges is concave and the other is convex, and the greater the possibility that the corresponding guide groove area will undergo deflection deformation due to settlement and collapse.

[0106] Step S403, obtaining a deflection sub-parameter of each guide groove area according to the first deflection parameter and the second deflection parameter, wherein the first deflection parameter and the second deflection parameter are both positively correlated with the deflection sub-parameter.

[0107] As an example, the first flexure parameter and the second flexure parameter are multiplied and combined to obtain the flexure sub-parameter of each guide groove area; in other examples, the implementer may also combine the two using basic mathematical means such as addition or weighted summation.

[0108] Step S404, the bending sub-parameters of all the guide groove regions in the grayscale image are integrated to obtain the bending deformation coefficient of the region to be detected.

[0109] As an example, the flexural sub-parameters of all guide groove areas in the grayscale image are averaged, and the average is normalized to obtain the flexural deformation coefficient of the area to be detected; the normalization function is a hyperbolic tangent function, and other normalization methods can also be used, or the maximum value of the flexural sub-parameters of all guide groove areas can be used as the flexural deformation coefficient.

[0110] Considering that the settlement deformation of the guide wall will cause the guide grooves to squeeze each other, thereby causing the distance between the noise reduction guide grooves to change, and due to the influence of gravity, the closer the guide grooves are to the bottom, the greater the impact of settlement squeezing; therefore, the embodiment of the present invention obtains the basic settlement coefficient of the area to be detected according to the distance between the pixel points on the edges of the two guide grooves in each guide groove area, and the ground height of each guide groove area.

[0111] Preferably, in one embodiment of the present invention, the method for obtaining the basic sedimentation coefficient includes:

[0112] See also Figure 6 , which shows a flow chart of a method for obtaining a basic sedimentation coefficient provided by an embodiment of the present invention, including:

[0113] Step S601: in each guide groove region, according to the distance between all innermost pixel points on the edges of two guide grooves, obtain the flat distortion parameter of each guide groove region.

[0114] In a preferred embodiment of the present invention, considering that when the guide groove undergoes extrusion deformation, its width becomes relatively smaller, and the extrusion effect may be different, there is a situation where the width of the guide groove is uneven; and the width of the guide groove can be evaluated based on the distance between the edges of the two guide grooves; therefore, the method for obtaining the flat distortion parameter includes:

[0115] In each guide groove area, all the innermost pixel points on the edge of each guide groove are sorted in sequence, and the average groove width and width range of the corresponding guide groove are obtained according to the distance between all the innermost pixel points with the same sorting number on the edges of the two guide grooves; the relative flattening of each guide groove area is obtained according to the average groove width corresponding to each guide groove area and its deviation from the average groove width corresponding to all guide groove areas; the width range is used as the distortion coefficient of the guide groove area; the flat distortion parameter of each guide groove area is obtained according to the relative flattening and the distortion coefficient, and the relative flattening and the distortion coefficient are positively correlated with the flat distortion parameter.

[0116] As an example, the distance between the innermost pixel points with the same serial number on the edge of different diversion grooves is first calculated. This distance reflects the groove width. The average distance is used as the average groove width, and the distance range is used as the width range to evaluate the settlement distortion of the groove.

[0117] The calculation formula of the flat distortion parameter is: Among them, B j is the flat distortion parameter of the jth guide groove area; norm( ) is the standard normalization function, specifically linear normalization is used; is the average groove width of the jth guide groove area; is the mean of the average groove width of all guide groove areas; is the relative flatness of the jth guide groove area; q j is the distortion coefficient of the jth guide groove area.

[0118] In the above formula, when the average groove width of each guide groove area is smaller and the groove width of all other guide groove areas deviates more, it means that it is flatter. Negative correlation mapping is performed to obtain relative flattening, which is further multiplied and combined with the distortion coefficient to obtain flat distortion parameters; in other examples, the implementer may further normalize the distortion coefficient and then add or weightedly combine it with the relative flattening, which will not be elaborated here.

[0119] Step S602, the ground height of each diversion groove area is negatively correlated and mapped as a weight, and the corresponding flat distortion parameter is weighted to obtain the basic settlement sub-parameter of each diversion groove area; the basic settlement sub-parameters of all diversion groove areas are combined to obtain the basic settlement coefficient of the area to be detected.

[0120] As an example, the ground height is reciprocally calculated and then multiplied with the corresponding flattening distortion coefficient as a weight to obtain the basic settlement sub-parameter of each diversion groove area; then the basic settlement sub-parameters of all diversion groove areas are averaged, and the mean is linearly normalized to obtain the basic settlement coefficient of the area to be detected.

[0121] In other examples, other negative correlation mapping means and normalization means may also be used, and the maximum value of the basic settlement sub-parameters of all guide groove areas may be used as the basic settlement parameter, which will not be described in detail here.

[0122] Step S3, obtaining the deformation coefficient of the area to be detected according to the flexural deformation coefficient and the foundation settlement coefficient; obtaining the deformation detection result of the windward surface of the airport noise reduction guide wall according to the deformation coefficient of the area to be detected.

[0123] Considering that both the flexural deformation coefficient and the foundation settlement coefficient are parameters reflecting the deformation degree of the corresponding noise reduction diversion wall in the area to be detected, the two can be comprehensively used to evaluate the deformation coefficient of the area to be detected, and then evaluate the deformation detection results.

[0124] Preferably, in one embodiment of the present invention, the product of the flexural deformation coefficient and the basic settlement coefficient is used as the deformation coefficient of the area to be detected. In other embodiments, the implementer may also combine the two by basic mathematical means such as addition or weighted summation to obtain the deformation coefficient of the area to be detected, which will not be described in detail here.

[0125] Preferably, in one embodiment of the present invention, when the deformation coefficient is greater than a preset threshold, the deformation detection result is determined to be deformation, otherwise it is determined to be non-deformation; wherein the preset threshold is set to 0.5, and the implementer can set other values ​​at will.

[0126] When the airport noise reduction and diversion wall is deformed, relevant technicians need to be arranged to repair it. The repair work may involve replacing damaged parts, reinforcing the structure, repairing surface damage or other necessary operations. After the repair is completed, further testing and verification work is required to ensure the repair effect so that the performance and function of the airport noise reduction and diversion wall can be restored to the design requirements.

[0127] In summary, the embodiment of the present invention obtains all guide groove areas and the corresponding two guide groove edges in the area to be detected, and obtains the ground height of each guide groove area from the ground; further obtains the flexural deformation coefficient and the basic settlement coefficient of the area to be detected; according to the flexural deformation coefficient and the basic settlement coefficient, obtains the deformation coefficient of the area to be detected, so as to obtain the deformation detection result of the windward surface of the airport noise reduction guide wall. Based on the groove texture characteristics of the noise reduction guide wall, combined with the groove light and shadow changes and texture deformation caused by the deformation, combined with the basic settlement characteristics of the guide wall deformation, the present invention quantifies the deformation coefficient of the area to be detected, and improves the deformation detection efficiency and accuracy of the airport noise reduction guide wall.

[0128] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0129] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A visual inspection method for airport noise reduction diversion walls, characterized in that: The method comprises: Obtain a grayscale image of the area to be detected on the windward surface of the airport noise reduction guide wall, obtain all guide groove areas and two guide groove edges in each of the guide groove areas in the grayscale image, and obtain the ground height of each of the guide groove areas from the ground; According to the grayscale uniformity in all the guide groove areas and the difference in the number of adjacent pixels on both sides of each guide groove edge in each guide groove area, the flexural deformation coefficient of the area to be detected is obtained; according to the distance between the pixels on the two guide groove edges in each guide groove area and the ground height of each guide groove area, the basic settlement coefficient of the area to be detected is obtained; According to the flexural deformation coefficient and the foundation settlement coefficient, the deformation coefficient of the area to be detected is obtained; according to the deformation coefficient of the area to be detected, the deformation detection result of the windward surface of the airport noise reduction guide wall is obtained; The method for obtaining the flexural deformation coefficient includes: According to the grayscale difference of the pixel points on both sides of the central axis of each guide groove area, a first deflection parameter of each guide groove area is obtained; the central axis is parallel to the edge of the guide groove, and the grayscale difference is positively correlated with the first deflection parameter; According to the difference in the number of adjacent pixel points on both sides of each guide groove edge, the deformation parameter of each guide groove edge is obtained; according to the difference in the deformation parameters of the two guide groove edges in the guide groove area, the second deflection parameter of each guide groove area is obtained; According to the first deflection parameter and the second deflection parameter, a deflection sub-parameter of each of the guide groove regions is obtained, wherein the first deflection parameter and the second deflection parameter are both positively correlated with the deflection sub-parameter; The deflection sub-parameters of all the guide groove regions in the grayscale image are integrated to obtain the deflection deformation coefficient of the area to be detected; The method for obtaining the basic sedimentation coefficient includes: In each of the guide groove regions, according to the distance between all innermost pixel points on two edges of the guide groove, a flat distortion parameter of each of the guide groove regions is obtained; The ground height of each diversion groove area is negatively correlated and mapped as a weight, and the corresponding flat distortion parameter is weighted to obtain a basic settlement sub-parameter of each diversion groove area; The basic sedimentation sub-parameters of all the guide groove areas are combined to obtain the basic sedimentation coefficient of the area to be detected.

2. The visual inspection method for airport noise reduction diversion wall according to claim 1 is characterized in that: The method for obtaining the guide groove area includes: Obtain all closed edges in the grayscale image, and obtain the average area of ​​the regions enclosed by all the closed edges, and use the region enclosed by the closed edges whose area is smaller than the average area as a suspected guide groove region; Taking the degree of the closed edge-like moment of each of the suspected guide groove regions as the first confidence level of each of the suspected guide groove regions; According to the distribution regularity of each of the suspected guide groove regions and its adjacent guide groove regions, a second confidence level of each of the suspected guide groove regions is obtained; The groove confidence of each of the suspected guide groove areas is obtained according to the first confidence and the second confidence, wherein the first confidence and the second confidence are both positively correlated with the groove confidence; and all the suspected guide groove areas whose groove confidence is greater than a preset threshold are taken as guide groove areas.

3. The visual inspection method for airport noise reduction diversion wall according to claim 2 is characterized in that: The method for obtaining the class moment degree includes: Taking any edge pixel point on each of the closed edges as a starting point, obtaining the 8 chain codes corresponding to the closed edges; in the 8 chain codes corresponding to each of the closed edges, sorting the occurrence frequency of each chain code value in descending order, taking the top two chain code values ​​as the first group of chain code values, and taking the third and fourth chain code values ​​as the second group of chain code values; On the closed edge, according to the difference between the chain code value of each group of chain code values ​​and the numerical difference of the preset positive integer, and the difference in the number of edge pixels corresponding to the chain code values, the confidence coefficient of each group of chain code values ​​corresponding to the chain code values ​​of the opposite sides of the rectangle is obtained; the numerical difference and the quantitative difference are both negatively correlated with the confidence coefficient; The confidence coefficients of the two groups of chain code values ​​are multiplied to obtain the moment-like degree corresponding to the closed edge.

4. The visual inspection method for airport noise reduction diversion wall according to claim 2 is characterized in that: The method for obtaining the second confidence level includes: Taking any of the suspected guide groove areas as the target area; within the neighborhood of the target area, taking the left and right suspected guide groove areas as horizontal reference areas, and taking the upper and lower suspected guide groove areas as vertical reference areas; Respectively obtaining a first distance and a second distance between the two horizontal reference areas and the corresponding centroids of the target area, and mapping the difference between the first distance and the second distance to a negative correlation to obtain a horizontal uniform distribution parameter; Respectively obtaining a third distance and a fourth distance between the two vertical reference areas and the corresponding centroids of the target area, and mapping the difference between the third distance and the fourth distance in a negative correlation manner to obtain a vertical uniform distribution parameter; The horizontal uniform distribution parameter and the vertical uniform distribution parameter are combined to obtain a second confidence level of each target area.

5. The visual inspection method for airport noise reduction diversion wall according to claim 1 is characterized in that: The method for obtaining the deformation parameters includes: In the grayscale image, all outermost pixel points and all innermost pixel points on the edge of each guide groove are respectively obtained, and all pixel points adjacent to the outermost pixel points are used as outer pixel points of the region, and all pixel points adjacent to the innermost pixel points are used as inner pixel points of the region, and neither the outer pixel points of the region nor the inner pixel points of the region are on the edge of the guide groove; The deformation parameter of each guide groove edge is obtained according to the difference between the total number of pixel points outside the area of ​​each guide groove edge and the total number of pixel points inside the area.

6. The visual inspection method for airport noise reduction diversion wall according to claim 1 is characterized in that: The method for obtaining the flat distortion parameter includes: In each of the guide groove regions, all innermost pixel points on each guide groove edge are sequentially sorted, and according to the distance between all innermost pixel points with the same sorting sequence number on the two guide groove edges, the average groove width and width range of the corresponding guide groove are obtained; According to the average groove width corresponding to each guide groove area and its deviation from the average groove width corresponding to all guide groove areas, the relative flatness of each guide groove area is obtained; the width range is used as the distortion coefficient of the guide groove area; The flat distortion parameter of each guide groove area is obtained according to the relative flatness and the distortion coefficient, and both the relative flatness and the distortion coefficient are positively correlated with the flat distortion parameter.

7. The visual inspection method for airport noise reduction diversion wall according to claim 1 is characterized in that: The method for obtaining the deformation coefficient includes: The product of the flexural deformation coefficient and the foundation settlement coefficient is used as the deformation coefficient of the area to be detected.

8. The visual inspection method for airport noise reduction diversion wall according to claim 1 is characterized in that: The method for obtaining the deformation detection result includes: When the deformation coefficient is greater than a preset threshold, the deformation detection result is determined to be deformation, otherwise the deformation detection result is determined to be non-deformation.

Images