CN117788462B - Nondestructive testing method for breakage defect of self-unloading semitrailer axle - Google Patents
Nondestructive testing method for breakage defect of self-unloading semitrailer axle Download PDFInfo
- Publication number
- CN117788462B CN117788462B CN202410204913.9A CN202410204913A CN117788462B CN 117788462 B CN117788462 B CN 117788462B CN 202410204913 A CN202410204913 A CN 202410204913A CN 117788462 B CN117788462 B CN 117788462B
- Authority
- CN
- China
- Prior art keywords
- axle
- pixel
- difference
- reflection
- wavelength
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000007547 defect Effects 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000009659 non-destructive testing Methods 0.000 title claims abstract description 21
- 239000003973 paint Substances 0.000 claims abstract description 42
- 230000000877 morphologic effect Effects 0.000 claims abstract description 19
- 238000011156 evaluation Methods 0.000 claims abstract description 16
- 238000012795 verification Methods 0.000 claims abstract description 15
- 230000000694 effects Effects 0.000 claims description 22
- 238000001228 spectrum Methods 0.000 claims description 21
- 238000005070 sampling Methods 0.000 claims description 16
- 230000009467 reduction Effects 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 5
- 230000003595 spectral effect Effects 0.000 claims description 5
- 238000003786 synthesis reaction Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 3
- 230000002829 reductive effect Effects 0.000 claims description 3
- 230000001066 destructive effect Effects 0.000 claims 1
- 238000007689 inspection Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 19
- 238000007599 discharging Methods 0.000 abstract description 3
- 239000011324 bead Substances 0.000 description 12
- 230000035772 mutation Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 5
- 238000002310 reflectometry Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000005484 gravity Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 239000007769 metal material Substances 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 239000002519 antifouling agent Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003638 chemical reducing agent Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000000149 penetrating effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to the technical field of axle performance detection, in particular to a nondestructive testing method for a fracture defect of a self-unloading semitrailer axle, which comprises the following steps: collecting hyperspectral data of a vehicle chassis; acquiring an axle pixel set according to the hyperspectral data; obtaining a paint surface drop verification coefficient according to the reflection intensity of pixels in the axle pixel set on all wavelengths; acquiring a potential variation region according to the paint surface drop verification coefficient; combining hyperspectral data and potential variation regions to obtain morphological variation evaluation coefficients of the potential variation regions; acquiring a suspected defect area according to the morphological variation evaluation coefficient, and acquiring an axle fracture index by combining the reflection intensity of pixels in the suspected defect area to each wavelength; and acquiring the axle condition of the self-discharging semi-trailer according to the axle fracture index. The invention can realize the detection of the axle breakage defect of the self-unloading semitrailer and improve the detection precision.
Description
Technical Field
The application relates to the technical field of axle performance detection, in particular to a nondestructive detection method for a fracture defect of a self-unloading semitrailer axle.
Background
A self-unloading semitrailer is a truck for transporting bulk goods, which is generally composed of a traction head and a trailer type semitrailer, and a hydraulic system is arranged on the semitrailer body to form a hydraulic bracket for unloading goods by itself. The axle usually bears the load of the vehicle, realizes power transmission and maintains the stable control and braking of the vehicle on the road, and because the self-unloading semitrailer has larger load capacity, the axle has higher requirements on the structural strength of the axle, thereby the defect detection of the self-unloading semitrailer axle is required.
The hyperspectral detection has the advantages of no damage, rapidness and environmental protection, and is an important method for nondestructive detection of the axle fracture defect. The semi-trailing axle mainly comprises a main speed reducer, a differential mechanism, a wheel transmission device, a driving axle housing and the like. These parts are all wrapped by the driving axle housing, in order to prevent corrosion and durability, paint is usually required to be coated on the surface of the axle housing, the paint is deformed under the action of gravity, and the phenomenon of dropping a barrel and dropping a bead is mainly generated, so that the thickness of the paint on the surface of the axle is uneven, and the phenomenon is uncontrollable, therefore, when the hyperspectral is adopted for detecting the axle fracture defect, the edge position of the paint dropping barrel and the bead hanging area can be mistakenly detected as the fracture defect, and the accuracy and the reliability of the axle fracture detection are reduced. Aiming at the problems, the application provides a nondestructive testing method for the breakage defect of the self-unloading semitrailer axle, which aims to eliminate the influence of the drop bucket effect and the drop bead hanging phenomenon on the detection result of the breakage defect of the self-unloading semitrailer axle through the analysis of the hyperspectral data of the axle.
Disclosure of Invention
In order to solve the technical problems, the application provides a nondestructive testing method for the breakage defect of a self-unloading semitrailer axle, which aims to solve the existing problems.
The nondestructive testing method for the breakage defect of the self-unloading semitrailer axle adopts the following technical scheme:
The embodiment of the application provides a nondestructive testing method for the breakage defect of a self-unloading semitrailer axle, which comprises the following steps:
Collecting hyperspectral data of a vehicle chassis of the self-unloading semitrailer at each sampling moment, and obtaining the reflection intensity of each pixel in the hyperspectral data on each wavelength;
Acquiring an axle pixel set according to hyperspectral data of a vehicle chassis at each sampling moment; obtaining local fitting offset of pixels to each wavelength according to reflection intensities of pixels in the axle pixel set to all wavelengths; obtaining a reflection local difference intensity coefficient of the pixel to each wavelength according to the local fitting offset and the reflection intensity of the pixel to each wavelength; obtaining neighborhood information confusion of each pixel according to the reflection local difference intensity coefficients of each pixel to all wavelengths in the axle pixel set; obtaining a paint surface drop verification coefficient according to the maximum value of the reflection local difference intensity coefficients of all the wavelengths of each pixel in the axle pixel set and the neighborhood information confusion degree; acquiring each potential variation region according to the paint surface drop verification coefficient; combining the axle pixel set and the potential variation region to obtain a morphological variation evaluation coefficient of the potential variation region; acquiring standard axle spectrum data in a database; marking a set of potential variation regions with morphological variation evaluation coefficients greater than or equal to a preset variation threshold as suspected defect regions; the suspected defect area eliminates the variation area caused by the drip barrel effect and the drip hanging phenomenon; denoising the reflection intensity sequence of each pixel in the suspected defect area by adopting a Kalman filtering algorithm to obtain a reflection intensity sequence of each pixel after noise reduction; combining the reflection intensity sequence of the pixel after noise reduction in the suspected defect area with standard axle spectrum data to obtain an axle fracture index;
And acquiring the axle condition of the self-discharging semi-trailer according to the axle fracture index.
Further, the obtaining the axle pixel set according to the hyperspectral data of the vehicle chassis at each sampling moment includes:
The hyperspectral data of the vehicle chassis at each sampling moment are synthesized into a true color synthesized image by adopting a true color synthesis algorithm; the true color synthesized images at all sampling moments are spliced into a whole chassis image of the vehicle by using UnsupDIS unsupervised image splicing algorithm;
Acquiring a vehicle chassis background image in a database, and dividing the whole vehicle chassis image by using a background difference method according to the vehicle chassis background image to obtain an axle region;
The set of pixels in the axle region in the hyperspectral data of the vehicle chassis is denoted as the axle pixel set.
Further, the obtaining the local fitting offset of the pixels to each wavelength according to the reflection intensities of the pixels in the axle pixel set to all wavelengths includes:
for each pixel in the axle pixel set, fitting the reflection intensity of the pixel to all wavelengths by adopting a cubic spline interpolation method to obtain the fitting reflection intensity of the pixel to each wavelength; calculate the pixel pair Reflection intensity of individual wavelength and pixel pair/>The absolute value of the difference between the fitting reflection intensities of the wavelengths is recorded as a first absolute value of the difference; calculate the pixel pairReflection intensity of individual wavelength and pixel pair/>The absolute value of the difference between the fitting reflection intensities of the wavelengths is recorded as a second absolute value of the difference;
taking the sum of the absolute value of the first difference and the absolute value of the second difference as a pixel pair Local fitting offset for each wavelength.
Further, the obtaining the reflection local difference intensity coefficient of the pixel for each wavelength according to the local fitting offset and the reflection intensity of the pixel for each wavelength includes:
Calculate the pixel pair Reflection intensity of individual wavelength and pixel pair/>The absolute value of the difference value of the reflection intensity of each wavelength is recorded as a third absolute value of the difference value;
Calculate 1 and pixel pair Obtaining a sum value of local fitting offset of each wavelength, and obtaining a logarithmic function taking a natural constant as a base number and taking the sum value as a true number; taking the ratio of the third difference absolute value to the calculation result of the logarithmic function as a pixel pair/>The reflected local differential intensity coefficients for the individual wavelengths.
Further, the obtaining the neighborhood information confusion of each pixel according to the local difference intensity coefficient of reflection of each pixel to all wavelengths in the axle pixel set includes:
For the ith pixel in the axle pixel set, the maximum value of the reflection local difference intensity coefficients of the pixel for all wavelengths is marked as a maximum reflection difference value, and the wavelength corresponding to the maximum reflection difference value is marked as a maximum reflection difference wavelength; the array formed by the maximum reflection difference value and the maximum reflection difference wavelength is marked as a covariant conjugate pair of the pixels;
And constructing a neighborhood window with a preset side length by taking the ith pixel as a center, and recording the information entropy of the covariant conjugate pairs of all pixels in the neighborhood window as the neighborhood information chaos of the ith pixel.
Further, the paint drop verification coefficient includes:
For each pixel in the neighborhood window, calculating the difference absolute value of the maximum reflection difference value of the pixel and the maximum reflection difference value of the central pixel of the neighborhood window, and recording the difference absolute value as a fourth difference absolute value; acquiring an exponential function taking a natural constant as a base and taking a fourth difference absolute value as an exponent; calculating the absolute value of the difference between the maximum reflection difference wavelength of the pixel and the maximum reflection difference wavelength of the central pixel of the neighborhood window, and recording the absolute value as a fifth absolute value of the difference; calculating the product of the fifth difference absolute value and the calculation result of the exponential function;
and acquiring the average value of the products of all the pixels in the neighborhood window, and taking the product of the average value and the neighborhood information confusion degree of the ith pixel as a paint surface hanging check coefficient of the ith pixel.
Further, the obtaining each potential variation region according to the paint dripping check coefficient includes:
Recording pixels with paint surface drop verification coefficients larger than or equal to a preset drop threshold value as variant pixels; acquiring the areas of all the variation pixels corresponding to the interior of the axle area, and marking the areas as an integral potential variation area; and carrying out connected domain analysis on the whole potential variation region, wherein each obtained connected domain is used as each potential variation region.
Further, the combining the axle pixel set and the potential variation region to obtain the morphological variation evaluation coefficient of the potential variation region includes:
For each potential variation area, marking a set formed by pixels positioned at the edge of the potential variation area in the axle pixel set as a boundary pixel sequence of the potential variation area;
For each pixel in the boundary pixel sequence, the reflection intensity of the pixel to all wavelengths is arranged from short to long according to the length of the wavelength to obtain a reflection intensity sequence of the pixel; calculating a dynamic time warping distance between the pixel and the reflection intensity sequences of other pixels;
and taking the average value of all the distances in the boundary pixel sequence of the potential variation region as a morphological variation evaluation coefficient of the potential variation region.
Further, the obtaining the axle fracture index by combining the reflection intensity sequence of the pixel after noise reduction in the suspected defect area and the standard axle spectrum data comprises the following steps:
Calculating cosine similarity between the reflection intensity sequence of each pixel after noise reduction and standard axle spectrum data; acquiring the average value of all cosine similarities in the suspected defect area, and marking the average value as a first average value; taking the reciprocal of the first mean value as an axle fracture index.
Further, the obtaining the axle condition of the dump semi-trailer according to the axle fracture index includes:
When the axle fracture index is smaller than a preset fracture threshold value, the axle of the self-unloading semitrailer has no fracture defect; otherwise, the axle of the self-unloading semitrailer has a fracture defect.
The application has at least the following beneficial effects:
According to the application, through analyzing the paint characteristics of the surface of the vehicle axle, in order to eliminate the influence of the dropping barrel effect and the drop bead hanging phenomenon generated by the gravity action on axle fracture detection, the paint surface drop hanging verification coefficient is obtained through calculation through the difference between the reflection intensity changes of each wavelength of the pixels and is used for judging whether the abnormal condition occurs to the vehicle axle; in order to distinguish a variation region of a drop bucket effect and a drop bead hanging phenomenon and a suspected noise region, calculating a morphological variation evaluation coefficient to screen and exclude the potential variation region by utilizing the characteristic that the drop bucket effect paint presents banded distribution and the drop bead hanging phenomenon paint presents punctiform distribution, so as to obtain a suspected defect region, wherein the suspected defect region comprises noise and an axle fracture region; denoising pixels in the suspected defect area, and performing similarity calculation on the denoised result and standard axle hyperspectral data to obtain an axle fracture index, wherein the axle fracture index reflects the possibility of axle fracture; according to the method, whether the axle breaks or not is judged according to the axle breaking index, and compared with the traditional method that the axle breaking detection is carried out by directly carrying out the axle breaking detection on the actual axle and the hyperspectral data of the database, the method can eliminate detection defects caused by uneven distribution of the paint on the surface of the axle according to the detailed characteristics of the paint on the surface of the axle, and improve the detection precision of the axle breaking.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of a nondestructive testing method for the breakage defect of a self-unloading semitrailer axle;
FIG. 2 is a flow chart of acquisition of an axle set of pixels;
FIG. 3 is a specific flow chart of axle breakage index acquisition.
Detailed Description
In order to further explain the technical means and effects adopted by the application to achieve the preset aim, the following is a nondestructive testing method for the breakage defect of the axle of the self-unloading semitrailer according to the application, which is provided by combining the accompanying drawings and the preferred embodiment, and the specific implementation, the structure, the characteristics and the effects thereof are described in detail below. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
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 application belongs.
The application provides a specific scheme of a nondestructive testing method for the breakage defect of a self-discharging semitrailer axle, which is specifically described below with reference to the accompanying drawings.
The application provides a nondestructive testing method for a fracture defect of a self-unloading semitrailer axle, in particular to a nondestructive testing method for the fracture defect of the self-unloading semitrailer axle, referring to fig. 1, which comprises the following steps:
and S001, acquiring hyperspectral data of the vehicle chassis through a spectrum analyzer, and preprocessing the data.
The self-unloading semitrailer is often loaded more, and the load and power transmission of the whole vehicle bear more pressure, so the integrity and the nondestructive performance of the vehicle axle are necessary for ensuring the safe operation of the vehicle. The dump semitrailer usually needs to pass through a wagon balance when transporting goods and is used for calculating the weight of the load of the vehicle, so that the Ocean Optics USB 2000+series hyperspectral analyzer is deployed on the wagon balance to collect hyperspectral data of the vehicle. When the vehicle passes over the wagon balance, the vehicle chassis is photographed at intervals of 10s from the detection of the vehicle until the vehicle is not detected, whereby the vehicle chassis hyperspectral data is acquired.
The semitrailer chassis data obtained at the kth sampling moment is recorded asIn particular/>Wherein/>Represents the/>, of hyperspectral dataReflection intensity sequence of individual pixels,/>Indicating the total number of pixels in the hyperspectral data. Reflection intensity sequence for individual picture elements/>Wherein/>Indicating that the ith pixel sequence is at the/>Reflection intensity information for each wavelength, L, represents the number of wavelengths in the spectrum analyzer. In addition, the hyperspectral data is inevitably interfered by environmental noise in the process of acquisition, and the embodiment adopts wavelet threshold denoising to carry out noise reduction treatment on the acquired hyperspectral signals.
So far, the reflection intensity of each pixel in the hyperspectral data of the vehicle chassis on each wavelength is obtained.
Step S002, curve fitting is carried out on the reflection intensity sequences of the pixels of the axle hyperspectral data, a reflection local difference intensity coefficient is constructed according to the change characteristics of the spectrum curve of the paint surface, a comparison neighborhood is divided for each pixel to obtain a paint surface drip check coefficient, a potential variation region is obtained, a morphological variation evaluation coefficient is obtained by utilizing the morphological characteristics of the potential variation region, and a suspected defect region is obtained; and comparing the spectral information of the pixels in the suspected defect area with axle hyperspectral data of a database to obtain an axle fracture index.
Various components in the axle are precisely connected through a transmission shaft and a gear, and if a certain component breaks due to overload, the whole operation effect of the vehicle is affected, so that the axle of the vehicle needs to be subjected to nondestructive testing. Since the structural composition of the axle is substantially the same, the reflected intensity at each wavelength is substantially constant throughout the axle. When a vehicle axle is broken or cracked, the reflectivity of the damaged part for each wavelength is different, and abnormal spectral reflection occurs.
The axles are typically positioned under the vehicle chassis for carrying the weight of the vehicle and for power transmission, and in order to improve the wear and corrosion resistance of the axles, it is often necessary to paint the axles with a protective paint. The paint forms a drip barrel effect and a drip hanging phenomenon under the action of gravity. The drop bucket effect refers to a strip paint block formed when paint flows along the surface of the axle and finally drops, the drop hanging beads refer to drops of the paint hanging at the local lowest point of the surface of the axle, the drop hanging beads are generally distributed in a punctiform manner, and the drop bucket effect and the drop hanging beads phenomenon can cause interference to the detection of the breakage defect of the axle, so that the hyperspectral data of the axle part, namely an axle pixel set, is obtained by processing the hyperspectral data, and the specific flow is shown in fig. 2.
The method comprises the steps of performing true color synthesis on the hyperspectral data of the vehicle chassis at each sampling moment through a GRBS true color synthesis algorithm, inputting the hyperspectral data of the vehicle chassis at a single sampling moment, and outputting a true color synthesized image at the corresponding sampling moment. The specific flow of GRBS true color synthesis algorithm is a known technology and will not be described in detail.
And all true color synthesized images in the sampling period are spliced into an integral chassis image of the vehicle through UnsupDIS unsupervised image splicing algorithm. And acquiring a vehicle chassis background image in the database, and acquiring an axle region by using a background difference method on the whole vehicle chassis image according to the vehicle chassis background image. The UnsupDIS unsupervised image stitching algorithm and the background difference method are known techniques and will not be described in detail.
The set of pixels in the axle region in the hyperspectral data of the vehicle chassis is recorded as an axle pixel set。
In order to further eliminate the interference of the dropping barrel effect, the drop bead hanging phenomenon and noise on axle crack detection, in this embodiment, the axle fracture index is obtained by performing the reverse processing on the spectral data of each pixel in the axle pixel set to perform the nondestructive detection on the axle, and the specific flow is shown in fig. 3. Paint coated on the surface of the axle can influence the reflection intensity of the wavelength, and components in the paint can absorb the energy of the wavelength, so that the energy of the reflection intensity of the wavelength is reduced. In addition, the longer the wavelength is, the stronger the penetrating energy of the object is, so after the specific wavelength is reached, the reflection information of the wavelength represents the real structural data of the axle, and because the axle is composed of metal materials, the metal materials have higher reflectivity to the specific wavelength, and therefore, the phenomenon that the reflectivity of adjacent wavelengths is greatly different can be formed. If the pixels are located in the variation area caused by the dropping barrel effect and the drop hanging phenomenon, the thickness of the surface paint is different from that of the normal area, so that the reflectivity of the pixels in the variation area to longer wavelength is greatly different, and the degree of the difference is weakened due to the thickness.
For measuring the change characteristics of the pixel reflection intensity sequence, aiming at the axle pixel setIn the embodiment, a cubic spline interpolation method is adopted to fit the reflection intensity of the pixel to each wavelength to obtain a pixel spectrum curve/>The pixel spectral curves correspond to the fitted reflection intensity of the pixel for each wavelength. Since cubic spline interpolation is a well-known technique in the field of curve fitting, no further description is given. From this, according to the reflection intensity of the pixel for each wavelength and the fitting reflection intensity, a reflection local difference intensity coefficient is obtained:
In the method, in the process of the invention, Representing that the ith pixel in the axle pixel set is at the/>Reflection local difference intensity coefficient of individual wavelength,/>And/>Respectively represent the i-th pixel pair/>, in the axle pixel setAnd/>Reflection intensity of individual wavelength,/>Representing a logarithmic function based on a natural constant e,/>Represents the i-th pair of pixels in the axle set of pixels/>Partial fitting offset for individual wavelengths,/>And/>Respectively represent the i-th pixel pair/>, in the axle pixel setAnd/>Fitting the reflection intensity for each wavelength.
Local difference intensity coefficient of pixel reflectionReflects the current pixel pair/>The local difference of the reflection intensity of each wavelength, when the difference between the reflection intensity of the pixel for each wavelength and the fitting reflection intensity is smaller, indicates that the lower the noise level is, the closer the reflection intensity is to the true value, and at this time, if for the/>Reflection intensity of individual wavelengths and pair of the first/>When the reflection intensity difference of the wavelengths is large, the fitted spectrum curve is at the first/>The vicinity of each wavelength is close to a straight line, and the reflection local difference intensity coefficient is larger.
Traversing the ith pixel to obtain corresponding reflection local difference intensity coefficients for each wavelength, and selecting the maximum value in all reflection local difference intensity coefficientsWavelength/> corresponding to maximum valueTogether form a covariant conjugate pair. And further measuring whether the current pixel has a barrel dropping effect and a drop hanging phenomenon by using a covariant conjugate pair. Because the drip barrel effect and the drip hanging beads respectively show the banded and punctiform distribution, the current pixel is taken as the center to divide/>Neighborhood window/>In this embodiment/>The value of (2) is 5. Calculating the information entropy of covariant conjugate pairs of all pixels in a neighborhood window, and recording the information entropy as the neighborhood information confusion/>, of the ith pixel. Comparing differences of all pixels in the neighborhood and the center pixel, and combining the reflection intensity sequences of the pixels to obtain a paint surface drop and hang check coefficient:
In the method, in the process of the invention, Paint drop verification coefficient representing ith pixel in axle spectrum data,/>Neighborhood information chaos representing the i-th pixel,/>Representing the number of pixels contained in the neighborhood window of the ith pixel,/>And/>Respectively representing the maximum pixel reflection local difference intensity coefficient and the corresponding wavelength of the ith pixel in the axle spectrum data,/>AndRespectively representing the maximum pixel reflection local difference intensity coefficient and the corresponding wavelength of the j pixel in the neighborhood window of the i pixel,/>An exponential function based on a natural constant e is represented.
Paint drop coefficient of verificationReflecting the possibility that the corresponding pixels have the phenomenon of dropping the paint on the axle position of the corresponding truck and the phenomenon of dropping the paint on the axle. If the current pixel is positioned in a banded region of a drip barrel effect or a punctiform region of a drip hanging phenomenon, the difference of covariant conjugate pairs between pixels in a neighborhood window of the pixel is larger, namely the neighborhood information confusion is larger; in addition, if the current pixel is positioned in the variation area of the dropping barrel effect and the drop hanging bead phenomenon, the larger the variation of the degree of difference of the covariant conjugate pair of the current pixel and each pixel in the neighborhood window is, the calculation is carried out to obtain/>The value of (2) is larger, and finally, the paint surface is enabled to drop and hang the check coefficient/>The value of (2) is larger.
In order to separate and screen the variation region and the noise and crack region caused by the drip barrel effect and the drip hanging phenomenon, the embodiment sets the drip hanging threshold valueDropping the paint surface with a verification coefficient/>The pixels of the image sensor are marked as variant pixels, and the areas of all variant pixels corresponding to the interior of the axle area are marked as an integral potential variant area; and carrying out connected domain analysis on the whole potential variation region to obtain each connected domain, and taking each connected domain as each potential variation region, wherein the specific process of the connected domain analysis is the prior art, and the embodiment is not repeated. In the potential mutation area, there is a real mutation area caused by the drip barrel effect and the drip hanging phenomenon, and there is also a pseudo mutation area caused by local sampling environment noise, so that further analysis on the potential mutation area is needed. The set of pixels located at the edge of the potential variation region in the axle pixel set is recorded as the boundary pixel sequence/>, of the potential variation region。
The variation area formed by the drip barrel effect is in strip distribution, the variation area caused by the drip hanging phenomenon is in dot distribution, the thickness of the edge position is uniform, and then the pixels positioned at the edge of the variation area have consistency on the reflection intensity of electromagnetic waves with the same wavelength. And screening and eliminating the potential variation region by utilizing the dot band distribution characteristic of the variation region. Calculating morphological variation evaluation coefficients:
In the method, in the process of the invention, Morphological mutation assessment coefficient representing the r-th potential mutation region,/>Representing the sequence of boundary pixels of the (r) th potential variation region/>Number of picture elements,/>Representing the number of permutations of combinations of all pixels in a sequence of boundary pixels,/>Representing the calculation of the sequence of boundary pixels of the (r) th potential variation region/>Reflection intensity sequence/>, of the i-th pixel in (2)And the reflection intensity sequence/>, of the j-th pixelDTW dynamic time warping distance between.
Morphological variation evaluation coefficientThe point-like and band-like distribution of the (r) th potential variation region in the axle hyperspectral data is measured. If the r potential variation region is a suspected defect region, the suspected defect region comprises noise and the condition of axle fracture. Due to the randomness of local noise, the local noise randomly affects the reflection intensity of the boundary pixels on partial wavelengths, so that the similarity degree between the reflection intensity sequences of the pixels is lower, namely the DTW distance of the reflection intensity sequences of the pixels is larger, and finally the morphological variation evaluation coefficient/>The value of (2) is larger. In contrast, if the (r) th potential variation region is a true variation region caused by a barrel drop effect and a drop hanging phenomenon, the shape of the region is uniformly distributed in a dot shape and a strip shape, and the thickness of paint at the boundary is relatively uniform, so that/>The value of (2) is small.
In the axle hyperspectral data, the corresponding morphological variation evaluation coefficient is calculated by traversing all potential variation areas, and the variation threshold is set in the embodimentAssessment coefficient of morphological variation/>Marking the potential variant region of (a) as the true variant region, will/>The set of potential variant regions of (a) is noted as suspected defect regions.
The suspected defect area comprises noise and an axle fracture area, a Kalman filtering algorithm is adopted for denoising the reflection intensity sequence of the pixel in the suspected defect area to eliminate the influence of the noise, the algorithm is input into the reflection intensity sequence of the pixel in the suspected defect area, and the algorithm is output into the reflection intensity sequence of the pixel after noise reduction.
The standard axle spectrum data in the database is acquired, and it should be noted that, the database implementer of the standard axle spectrum data can construct by himself, and can acquire spectrum information by manually selecting the existing standard axle to obtain the spectrum data of the standard axle, so as to form the database of the standard axle spectrum data, which is not limited in this embodiment. Calculating the average value of cosine similarity between the reflection intensity sequences of all the pixels after noise reduction and standard axle spectrum data, and taking the reciprocal of the average value of the cosine similarity as an axle fracture index. If the reflection intensity sequence of the pixel after noise reduction is more similar to that of the standard axle spectrum data, the axle fracture index is smaller, which indicates that the possibility of fracture of an axle of an actual semitrailer is lower. Conversely, if the difference between the two is greater, the more serious the deviation of the highlight data of the actual axle from the standard axle is, the greater the axle fracture index, and the greater the possibility of axle fracture.
Thus, the axle fracture index of the dump semi-trailer is obtained.
And step S003, acquiring the axle condition of the self-unloading semitrailer according to the axle fracture index.
Finally, the embodiment analyzes the axle fracture condition of the dump semitrailer according to the axle fracture index of the dump semitrailerThe smaller indicates that the axle of the vehicle is more complete and the larger indicates that the axle of the vehicle may be at risk of breaking. Thus, the present embodiment sets the fracture threshold/>; If the axle fracture index/>The self-unloading semitrailer axle has the defect of fracture, and the axle can be normally on the road only after being overhauled normally; if the axle fracture index/>The self-unloading semitrailer axle is free from fracture defects.
It should be noted that: the sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and the same or similar parts of each embodiment are referred to each other, and each embodiment mainly describes differences from other embodiments.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; the technical solutions described in the foregoing embodiments are modified or some of the technical features are replaced equivalently, so that the essence of the corresponding technical solutions does not deviate from the scope of the technical solutions of the embodiments of the present application, and all the technical solutions are included in the protection scope of the present application.
Claims (9)
1. The nondestructive testing method for the breakage defect of the self-unloading semitrailer axle is characterized by comprising the following steps of:
Collecting hyperspectral data of a vehicle chassis of the self-unloading semitrailer at each sampling moment, and obtaining the reflection intensity of each pixel in the hyperspectral data on each wavelength;
Acquiring an axle pixel set according to hyperspectral data of a vehicle chassis at each sampling moment; obtaining local fitting offset of pixels to each wavelength according to reflection intensities of pixels in the axle pixel set to all wavelengths; obtaining a reflection local difference intensity coefficient of the pixel to each wavelength according to the local fitting offset and the reflection intensity of the pixel to each wavelength; obtaining neighborhood information confusion of each pixel according to the reflection local difference intensity coefficients of each pixel to all wavelengths in the axle pixel set; obtaining a paint surface drop verification coefficient according to the maximum value of the reflection local difference intensity coefficients of all the wavelengths of each pixel in the axle pixel set and the neighborhood information confusion degree; acquiring each potential variation region according to the paint surface drop verification coefficient; combining the axle pixel set and the potential variation region to obtain a morphological variation evaluation coefficient of the potential variation region; acquiring standard axle spectrum data in a database; marking a set of potential variation regions with morphological variation evaluation coefficients greater than or equal to a preset variation threshold as suspected defect regions; the suspected defect area eliminates the variation area caused by the drip barrel effect and the drip hanging phenomenon; denoising the reflection intensity sequence of each pixel in the suspected defect area by adopting a Kalman filtering algorithm to obtain a reflection intensity sequence of each pixel after noise reduction; combining the reflection intensity sequence of the pixel after noise reduction in the suspected defect area with standard axle spectrum data to obtain an axle fracture index;
acquiring the axle condition of the self-unloading semitrailer according to the axle fracture index;
the paint surface drop verification coefficient comprises:
For each pixel in the neighborhood window, calculating the difference absolute value of the maximum reflection difference value of the pixel and the maximum reflection difference value of the central pixel of the neighborhood window, and recording the difference absolute value as a fourth difference absolute value; acquiring an exponential function taking a natural constant as a base and taking a fourth difference absolute value as an exponent; calculating the absolute value of the difference between the maximum reflection difference wavelength of the pixel and the maximum reflection difference wavelength of the central pixel of the neighborhood window, and recording the absolute value as a fifth absolute value of the difference; calculating the product of the fifth difference absolute value and the calculation result of the exponential function;
and acquiring the average value of the products of all the pixels in the neighborhood window, and taking the product of the average value and the neighborhood information confusion degree of the ith pixel as a paint surface hanging check coefficient of the ith pixel.
2. The nondestructive testing method for the axle breakage defect of the self-unloading semitrailer according to claim 1, wherein the obtaining the axle pixel set according to the hyperspectral data of the vehicle chassis at each sampling moment comprises the following steps:
The hyperspectral data of the vehicle chassis at each sampling moment are synthesized into a true color synthesized image by adopting a true color synthesis algorithm; the true color synthesized images at all sampling moments are spliced into a whole chassis image of the vehicle by using UnsupDIS unsupervised image splicing algorithm;
Acquiring a vehicle chassis background image in a database, and dividing the whole vehicle chassis image by using a background difference method according to the vehicle chassis background image to obtain an axle region;
The set of pixels in the axle region in the hyperspectral data of the vehicle chassis is denoted as the axle pixel set.
3. The method for non-destructive testing of axle breakage defects of a self-unloading semitrailer according to claim 1, wherein the obtaining the local fitting offset of the pixels to each wavelength according to the reflection intensity of the pixels in the axle pixel set to all wavelengths comprises:
for each pixel in the axle pixel set, fitting the reflection intensity of the pixel to all wavelengths by adopting a cubic spline interpolation method to obtain the fitting reflection intensity of the pixel to each wavelength; calculate the pixel pair Reflection intensity of individual wavelength and pixel pair/>The absolute value of the difference between the fitting reflection intensities of the wavelengths is recorded as a first absolute value of the difference; calculate pixel pair/>Reflection intensity of individual wavelength and pixel pair/>The absolute value of the difference between the fitting reflection intensities of the wavelengths is recorded as a second absolute value of the difference;
taking the sum of the absolute value of the first difference and the absolute value of the second difference as a pixel pair Local fitting offset for each wavelength.
4. The method for non-destructive testing of a broken axle defect of a self-unloading semitrailer according to claim 3, wherein the obtaining the reflection local difference intensity coefficient of the pixel for each wavelength according to the local fitting offset and the reflection intensity of the pixel for each wavelength comprises:
Calculate the pixel pair Reflection intensity of individual wavelength and pixel pair/>The absolute value of the difference value of the reflection intensity of each wavelength is recorded as a third absolute value of the difference value;
Calculate 1 and pixel pair Obtaining a sum value of local fitting offset of each wavelength, and obtaining a logarithmic function taking a natural constant as a base number and taking the sum value as a true number; taking the ratio of the third difference absolute value to the calculation result of the logarithmic function as a pixel pairThe reflected local differential intensity coefficients for the individual wavelengths.
5. The method for non-destructive testing of axle breakage defects of a self-unloading semitrailer according to claim 4, wherein the obtaining the neighborhood information confusion of each pixel according to the reflection local difference intensity coefficient of each pixel to all wavelengths in the axle pixel set comprises:
For the ith pixel in the axle pixel set, the maximum value of the reflection local difference intensity coefficients of the pixel for all wavelengths is marked as a maximum reflection difference value, and the wavelength corresponding to the maximum reflection difference value is marked as a maximum reflection difference wavelength; the array formed by the maximum reflection difference value and the maximum reflection difference wavelength is marked as a covariant conjugate pair of the pixels;
And constructing a neighborhood window with a preset side length by taking the ith pixel as a center, and recording the information entropy of the covariant conjugate pairs of all pixels in the neighborhood window as the neighborhood information chaos of the ith pixel.
6. The method for non-destructive testing of a breakage defect of a self-unloading semitrailer axle according to claim 5, wherein said obtaining each potential variation region according to paint drop verification coefficients comprises:
Recording pixels with paint surface drop verification coefficients larger than or equal to a preset drop threshold value as variant pixels; acquiring the areas of all the variation pixels corresponding to the interior of the axle area, and marking the areas as an integral potential variation area; and carrying out connected domain analysis on the whole potential variation region, wherein each obtained connected domain is used as each potential variation region.
7. The method for non-destructive testing of a broken axle defect of a self-unloading semitrailer according to claim 6, wherein the step of combining the set of axle pixels and the potential variation region to obtain the morphological variation evaluation coefficient of the potential variation region comprises the steps of:
For each potential variation area, marking a set formed by pixels positioned at the edge of the potential variation area in the axle pixel set as a boundary pixel sequence of the potential variation area;
For each pixel in the boundary pixel sequence, the reflection intensity of the pixel to all wavelengths is arranged from short to long according to the length of the wavelength to obtain a reflection intensity sequence of the pixel; calculating a dynamic time warping distance between the pixel and the reflection intensity sequences of other pixels;
and taking the average value of all the distances in the boundary pixel sequence of the potential variation region as a morphological variation evaluation coefficient of the potential variation region.
8. The method for non-destructive testing of axle breakage defects of a dump trailer according to claim 7, wherein said combining the reflection intensity sequence of the noise-reduced pixels in the suspected defect area with the standard axle spectral data to obtain the axle breakage index comprises:
Calculating cosine similarity between the reflection intensity sequence of each pixel after noise reduction and standard axle spectrum data; acquiring the average value of all cosine similarities in the suspected defect area, and marking the average value as a first average value; taking the reciprocal of the first mean value as an axle fracture index.
9. The method for non-destructive inspection of a dump truck axle fracture defect according to claim 1, wherein the step of obtaining the axle condition of the dump truck according to the axle fracture index comprises the steps of:
When the axle fracture index is smaller than a preset fracture threshold value, the axle of the self-unloading semitrailer has no fracture defect; otherwise, the axle of the self-unloading semitrailer has a fracture defect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410204913.9A CN117788462B (en) | 2024-02-26 | 2024-02-26 | Nondestructive testing method for breakage defect of self-unloading semitrailer axle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410204913.9A CN117788462B (en) | 2024-02-26 | 2024-02-26 | Nondestructive testing method for breakage defect of self-unloading semitrailer axle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117788462A CN117788462A (en) | 2024-03-29 |
| CN117788462B true CN117788462B (en) | 2024-05-10 |
Family
ID=90389441
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410204913.9A Active CN117788462B (en) | 2024-02-26 | 2024-02-26 | Nondestructive testing method for breakage defect of self-unloading semitrailer axle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117788462B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117969535B (en) * | 2024-04-01 | 2024-06-07 | 黑龙江省泰斯特森工程检测有限公司 | Road bridge concrete defect nondestructive testing method and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002063939A2 (en) * | 2001-02-09 | 2002-08-22 | The United States Of America, As Represented By The Secretary Of Agriculture | Method and system for fecal and ingesta detection during meat and poultry processing |
| EP3591838A2 (en) * | 2018-07-06 | 2020-01-08 | Hampe & Schellhorn Service GbR | Vehicle system and method for autonomous and automated maintenance or care of a photovoltaic free surface system |
| CN111239073A (en) * | 2020-01-17 | 2020-06-05 | 西南交通大学 | Composite insulator aging degree detection method based on hyperspectral technology |
| CN114092838A (en) * | 2021-11-12 | 2022-02-25 | 前郭县富邦能源科技服务有限公司 | Photovoltaic defect detection method based on deep learning target detection |
| CN114719966A (en) * | 2020-12-22 | 2022-07-08 | 富泰华工业（深圳）有限公司 | Light source determination method and device, electronic equipment and storage medium |
| CN114998356A (en) * | 2022-08-08 | 2022-09-02 | 山东恩信特种车辆制造有限公司 | Axle defect detection method based on image processing |
| CN115984272A (en) * | 2023-03-20 | 2023-04-18 | 山东杨嘉汽车制造有限公司 | Semitrailer axle defect identification method based on computer vision |
| CN116452598A (en) * | 2023-06-20 | 2023-07-18 | 曼德惟尔(山东)智能制造有限公司 | Axle production quality rapid detection method and system based on computer vision |
-
2024
- 2024-02-26 CN CN202410204913.9A patent/CN117788462B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002063939A2 (en) * | 2001-02-09 | 2002-08-22 | The United States Of America, As Represented By The Secretary Of Agriculture | Method and system for fecal and ingesta detection during meat and poultry processing |
| EP3591838A2 (en) * | 2018-07-06 | 2020-01-08 | Hampe & Schellhorn Service GbR | Vehicle system and method for autonomous and automated maintenance or care of a photovoltaic free surface system |
| CN111239073A (en) * | 2020-01-17 | 2020-06-05 | 西南交通大学 | Composite insulator aging degree detection method based on hyperspectral technology |
| CN114719966A (en) * | 2020-12-22 | 2022-07-08 | 富泰华工业（深圳）有限公司 | Light source determination method and device, electronic equipment and storage medium |
| CN114092838A (en) * | 2021-11-12 | 2022-02-25 | 前郭县富邦能源科技服务有限公司 | Photovoltaic defect detection method based on deep learning target detection |
| CN114998356A (en) * | 2022-08-08 | 2022-09-02 | 山东恩信特种车辆制造有限公司 | Axle defect detection method based on image processing |
| CN115984272A (en) * | 2023-03-20 | 2023-04-18 | 山东杨嘉汽车制造有限公司 | Semitrailer axle defect identification method based on computer vision |
| CN116452598A (en) * | 2023-06-20 | 2023-07-18 | 曼德惟尔(山东)智能制造有限公司 | Axle production quality rapid detection method and system based on computer vision |
Non-Patent Citations (1)
| Title |
|---|
| 基于HALCON的贴片IC焊接缺陷检测算法研究;李文静;张虎山;焦键;;兵器装备工程学报;20200825(第08期);250-254 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117788462A (en) | 2024-03-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117788462B (en) | Nondestructive testing method for breakage defect of self-unloading semitrailer axle | |
| EP2573732B1 (en) | Reflection removal system | |
| CN102788642B (en) | Spectrum infers parameter generation device and method and spectrum apparatus for predicting | |
| US7027149B2 (en) | Photometric analysis of natural waters | |
| CN109671082B (en) | Sealing ring detection method, device, equipment and computer readable storage medium | |
| WO2021248962A1 (en) | Non-destructive testing method and device for testing and distinguishing internal and external defects of steel wire rope | |
| CN111879709B (en) | Lake water body spectral reflectivity inspection method and device | |
| Corrêa et al. | Fractal structure of digital speckle patterns produced by rough surfaces | |
| US7955858B2 (en) | Quantum dot-based environmental indicators | |
| CN112362544A (en) | Particle organic carbon monitoring method and system based on hyperspectral remote sensing | |
| CN116958036A (en) | Method for inspecting pollution state of insulator by unmanned aerial vehicle based on hyperspectral technology | |
| US20100166333A1 (en) | Noise filter | |
| CN114324329A (en) | Nondestructive testing and evaluation method for strong laser damage characteristics of optical element | |
| DE102017101309A1 (en) | A method for quantitatively determining a fluorescent component in blood and a meter | |
| CN113516608A (en) | Tire defect detection method and device, and tire detection device | |
| JP7370355B2 (en) | Evaluation method and device for roughening metal surfaces | |
| EP4070082B1 (en) | Method for assessing the quality of varnished wood surfaces | |
| CN111831872A (en) | Magnetic flux leakage data self-adaptive gray scale visualization method with enhanced small defect characteristics | |
| US20120179392A1 (en) | Method and system for measuring strain in twisted cord | |
| KR20220154345A (en) | Contour detection device using gaussian-weighted least squares and contour detection method thereof | |
| CN117783051B (en) | Methane gas leakage detection method based on multi-sensor data fusion | |
| CN110274620A (en) | A kind of brillouin scattering signal denoising method based on spectral centroid alignment | |
| WO2018207265A1 (en) | Tire degradation evaluation system, and method and program thereof | |
| CN107782797A (en) | Ultrasonic imaging processing method, device, storage medium and imaging system | |
| CN117367845B (en) | Health diagnosis method for army equipment maintenance equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |