CN117437191A - Rail surface detection method and system integrating inspection image and magnetic flux leakage detection - Google Patents
Rail surface detection method and system integrating inspection image and magnetic flux leakage detection Download PDFInfo
- Publication number
- CN117437191A CN117437191A CN202311383836.XA CN202311383836A CN117437191A CN 117437191 A CN117437191 A CN 117437191A CN 202311383836 A CN202311383836 A CN 202311383836A CN 117437191 A CN117437191 A CN 117437191A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- steel rail
- pixel point
- gloss
- 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.)
- Pending
Links
- 230000004907 flux Effects 0.000 title claims abstract description 71
- 238000001514 detection method Methods 0.000 title claims abstract description 62
- 238000007689 inspection Methods 0.000 title claims abstract description 19
- 229910000831 Steel Inorganic materials 0.000 claims abstract description 112
- 239000010959 steel Substances 0.000 claims abstract description 112
- 230000007547 defect Effects 0.000 claims abstract description 100
- 239000002245 particle Substances 0.000 claims abstract description 61
- 238000000034 method Methods 0.000 claims abstract description 52
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000012937 correction Methods 0.000 claims abstract description 8
- 230000008859 change Effects 0.000 claims description 30
- 238000005516 engineering process Methods 0.000 claims description 8
- 238000003709 image segmentation Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000012986 modification Methods 0.000 claims description 2
- 230000004048 modification Effects 0.000 claims description 2
- 238000003672 processing method Methods 0.000 claims description 2
- 230000005856 abnormality Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 9
- 230000002950 deficient Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000000149 argon plasma sintering Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
- G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
- G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- Image Processing (AREA)
Abstract
The application belongs to the technical field of computer vision and missing detection, and provides a steel rail surface detection method and system for fusing inspection images and missing magnetic detection, wherein the method comprises the following steps: acquiring and processing to obtain a gray level image and a magnetic flux leakage image of the steel rail; dividing pixel points in the gray level image to obtain light-sensitive particles; according to the distribution condition of light-sensitive particles around the pixel points, the glossiness of the pixel points is obtained; further analyzing to obtain the gloss gradient of the gray level image and the gloss abnormality index of the pixel point in the distinguishing direction; analyzing the amplitude value of the magnetic leakage image to construct the magnetic flux coefficient of the steel rail; fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence coefficient of the steel rail surface; correcting an SSR image enhancement algorithm based on the rail defect confidence; and dividing the steel rail defect area based on a correction SSR image enhancement algorithm. The method solves the problems of single and inaccurate detection mode of the traditional steel rail defects, and improves the detection efficiency of the steel rail surface defects.
Description
Technical Field
The application relates to the technical field of computer vision and leakage detection, in particular to a steel rail surface detection method and system for fusing inspection images and leakage detection.
Background
Along with the development of railway technology, high-speed rails and subways become one of important modes of going out of people in China, so that the high-speed rails and subways are very important for the overhaul and maintenance work of railways. The rail is repeatedly impacted, rubbed, impacted and the like by the wheels in the use process, and the rail can be cracked, worn, bent and other defects after long-term use. At present, a plurality of steel rail detection methods exist, wherein the method comprises the steps of periodically using a hand-push type detection device to patrol the steel rail by a overhauling worker, detecting whether the steel rail has defects or not by magnetic leakage, ultrasonic waves and the like, and then repairing the steel rail.
With the development of computer vision technology in recent years, a method for detecting rail defects through images is also one of important methods for detecting rail defects. However, there are certain problems in detecting the defects of the steel rail by using magnetic flux leakage or detecting the defects of the steel rail by using images, such as the need of professional staff to read information, different accuracy for various defects, and difficulty in accurately detecting and positioning the defects by a single method.
Therefore, a method for detecting the surface of a steel rail is needed to solve the problem of inaccurate detection of the defects of the traditional steel rail.
Disclosure of Invention
The application provides a steel rail surface detection method and system for fusing a patrol image and magnetic flux leakage detection, which improves the detection accuracy of steel rail surface defects.
According to a first aspect of an embodiment of the present application, there is provided a method for detecting a surface of a rail by fusing a patrol image with a magnetic flux leakage detection, including:
acquiring RGB images and magnetic flux leakage images of the steel rail, and preprocessing the RGB images to obtain gray images;
dividing pixel points in the gray level image according to the pixel gray level value of the gray level image to obtain light-sensitive particles;
according to the distribution condition of the light-sensitive particles around the pixel points, the glossiness of the pixel points is obtained;
analyzing the glossiness of the pixel points to obtain the glossiness gradient of the gray level image;
analyzing the gloss gradient to obtain a gloss anomaly index of the pixel point in the distinguishing direction;
analyzing the amplitude value of the magnetic leakage image to construct a steel rail magnetic flux coefficient;
fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence coefficient of the steel rail surface;
modifying a sigma value of the single-scale SSR image enhancement algorithm based on the rail defect confidence coefficient to obtain a modified SSR image enhancement algorithm;
and dividing the steel rail defect area by using an image segmentation technology based on the correction SSR image enhancement algorithm.
In some embodiments of the present invention, dividing pixel points in the gray scale image according to the pixel gray scale value of the gray scale image to obtain light sensitive particles includes:
counting the pixel gray values of the gray image to obtain a maximum pixel gray value T max And a minimum pixel gray value T min ；
Constructing a division threshold T of the light-sensitive particles according to the maximum pixel gray value and the minimum pixel gray value:
wherein T represents the division threshold of the photosensitive particles, T min T is the minimum pixel gray value in the gray image max A is a maximum pixel gray value in a gray image, and a is a threshold adjustment coefficient;
and judging whether the pixel gray value of the pixel point is larger than or equal to the dividing threshold value, and if so, dividing the pixel point into light-sensitive particles.
In some embodiments of the present invention, obtaining the glossiness of the pixel according to the distribution of the light-sensitive particles around the pixel includes:
counting the pixel gray values of the gray images, and determining the pixel point corresponding to the maximum pixel gray value as a central pixel point;
constructing a first pixel window by taking each pixel point as a center, acquiring light-sensitive particles contained in the first pixel window, and calculating a distance value d between the light-sensitive particles and the pixel point of the center point i ；
According to the distance value d i Obtaining the glossiness LD of the pixel point:
where LD is the glossiness of the pixel, d i And n is the total number of the light-sensitive particles in the pixel window, wherein the distance between the ith light-sensitive particle and the central pixel point is the distance between the ith light-sensitive particle and the central pixel point.
In some embodiments of the present invention, analyzing the glossiness of the pixel to obtain a glossiness gradient of the gray image includes:
constructing a second pixel window of 3*3 by taking a pixel point corresponding to the maximum pixel gray value in the gray image as a central pixel point;
in the second pixel window, a center pixel point (x 0 ,y 0 ) Is the direction of the four adjacent domains;
the method for respectively acquiring the gloss gradient in the direction of the four adjacent domains comprises the following steps:
with pixels (x) i ,y i ) Constructing a third pixel window for the center;
calculating the pixel glossiness difference value of three pixel points on the neighborhood direction in the third pixel window and three pixel points on the opposite neighborhood direction, and obtaining the glossiness gradient LDB of the pixel points, namely:
LDB(x i ,y i )＝[LD(x i -1,y i +1)+LD(x i -1,y i )+LD(x i -1,y i -1)]
-[LD(x i +1,y i +1)+LD(x i +1,y i )+LD(x i +1,y i -1)]
in the formula, LDB (x) i ,y i ) Representing pixel points (x) i ,y i ) Is a gradient of gloss gradation, LD (x i -1,y i +1) represents the pixel point (x) i ,y i ) Point gloss at a point with a distance of 1 in the negative x-axis direction and a point with a distance of 1 in the positive y-axis direction, the remaining LD values and so on, (x) i ,y i ) The ith point coordinate in the negative x-axis direction is represented, wherein the negative x-axis direction represents one of the four neighborhood directions.
In some embodiments of the present invention, analyzing the gradient of gradual gloss to obtain a gloss anomaly index of the pixel in the distinguishing direction includes:
a discrimination window is constructed by taking the pixel point as the center, and the line between the pixel point and the center pixel point of the gray image is taken as the discrimination direction;
analyzing the gloss gradient in the judging window and in the judging direction to obtain a gloss anomaly index LND of the pixel point in the judging direction, wherein the LND calculating method comprises the following steps:
in the formula, LND represents the gloss anomaly index of the pixel point in the distinguishing direction and LDB 0 Indicating the gradient of the gradual change of the brightness of the pixel point, LDB i＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative y-axis component, LDB j＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative direction component of the x axis and LDB i Representing the gradient of gradual gloss change of the ith pixel point on the negative y-axis component, LDB i+1 Representing the gradient of gradual gloss change of the (i+1) th pixel point in the negative y-axis direction component, LDB j Representing the gradient of gradual gloss change of the jth pixel point on the negative x-axis component, LDB j+1 The gloss gradient of the j+1th pixel point on the x-axis negative direction component is represented, h represents the number of the pixel points of the pixel point on the y-axis negative direction component, and l represents the number of the pixel points of the pixel point on the x-axis negative direction component.
In some embodiments of the present invention, the rail magnetic flux coefficient calculation method includes:
wherein PNB represents the magnetic flux coefficient of the steel rail,is pixel (x) i ,y i ) Amplitude value of>Is pixel (x) i ,y i ) The magnetic flux leakage image amplitude value at the position of the abscissa is the average value.
In some embodiments of the present invention, the method for calculating the confidence coefficient of the rail defect is as follows:
QBL＝LND*PNB
where QBL is the rail defect confidence, LND is the gloss anomaly index, PNB is the rail magnetic flux coefficient.
In some embodiments of the present invention, modifying a sigma value of a single-scale SSR image enhancement algorithm based on the rail defect confidence, to obtain a modified SSR image enhancement algorithm, including:
normalizing the value of the confidence coefficient of the steel rail defect, wherein the processing method comprises the following steps:
δ′＝FQBL×δ
wherein FQBL is normalized rail defect confidence, c and d are normalized adjustment parameters, QBL represents rail defect confidence, min (QBL) represents minimum value of rail defect confidence, max (QBL) represents maximum value of rail defect confidence, initial input delta value is set, and delta' is sigma value finally input into SSR algorithm.
According to a second aspect of embodiments of the present application, there is provided a rail surface detection system that fuses a patrol image with magnetic leakage detection, the system comprising a memory module and a processor module, wherein:
the memory module is used for storing program codes;
the processor module is configured to read the program code stored in the memory module and perform the method according to the first aspect of the embodiments of the present application.
In some embodiments of the invention, the processor module comprises an image acquisition module, a gray level image processing module, a magnetic flux leakage image processing module and a steel rail defect detection module.
As can be seen from the above embodiments, the method and system for detecting the surface of the steel rail by fusing the inspection image and the magnetic flux leakage detection provided by the embodiments of the present application have the following beneficial effects:
according to the invention, by fusing the inspection image and the magnetic flux leakage detection image and utilizing the characteristics of the rail defects in the two images, the rail defect confidence is constructed, the rail surface defects are accurately identified and positioned, finally, the SSR algorithm is modified to carry out targeted image enhancement on the rail defect surfaces, the problem that the traditional rail defect detection mode is single and inaccurate is solved, and the rail surface defect detection efficiency is improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings that are needed in the embodiments will be briefly described below, and it will be obvious to those skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a basic flow diagram of a method for detecting a surface of a steel rail by fusing a patrol image and magnetic flux leakage detection according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a basic flow chart of a method for obtaining light-sensitive particles according to an embodiment of the present application;
fig. 3 is a basic flow chart of a method for obtaining glossiness of a pixel according to an embodiment of the present application;
fig. 4 is a basic flow diagram of a method for acquiring a gradient of gradual gloss change of a gray scale image according to an embodiment of the present application;
fig. 5 is a schematic diagram of four-neighboring domain direction setting of a center pixel according to an embodiment of the present application;
fig. 6 is a basic flow chart of a method for obtaining a gloss anomaly index of a pixel in a distinguishing direction according to an embodiment of the present application;
fig. 7 is a schematic diagram of a discrimination window according to an embodiment of the present application;
fig. 8 is a schematic diagram of a steel rail surface detection system with integrated inspection images and magnetic flux leakage detection according to an embodiment of the present application.
Detailed Description
In order to make the technical solution of the present invention better understood by those skilled in the art, the technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The following describes in detail a method and a system for detecting a surface of a steel rail by fusing a patrol image and magnetic flux leakage detection according to this embodiment with reference to the accompanying drawings.
Fig. 1 is a basic flow chart of a steel rail surface detection method for fusing a patrol image and magnetic flux leakage detection, provided in an embodiment of the present application, and as shown in fig. 1, the steel rail surface detection method specifically includes the following steps:
s100: and acquiring RGB images and magnetic flux leakage images of the steel rail, and preprocessing the RGB images to obtain gray images.
A CCD camera and a shooting light source are arranged on a hand-push type inspection device of the steel rail, the steel rail is shot in a overlooking mode, and the light source is turned on during shooting, so that RGB images of the surface of the steel rail are obtained. Converting the RGB image into an initial gray image by using a gray value average method, denoising the initial gray image by using median filtering, sharpening the initial gray image by using a Laplacian operator, finally obtaining a gray image, and recording the gray image as a gray image A.
Then the magnetic leakage detection equipment is used for carrying out magnetic leakage detection on the steel rail, the whole region to be detected is subjected to magnetic leakage scanning, a plurality of magnetic leakage signal images are obtained, each magnetic leakage image represents the magnetic leakage detection condition of a row of pixels, and the magnetic leakage detection condition is recorded as P i (i＝1,2,…g)。
S200: and dividing pixel points in the gray level image according to the pixel gray level value of the gray level image to obtain the light-sensitive particles.
At present, high-speed rails, subways and the like are used as important travel tools for people, the degree of loss suffered by the steel rail is greatly improved, and the probability of defect occurrence of the steel rail is increased due to daily and repeated use. The surface of the steel rail is usually cracked and worn due to rolling friction of wheels, and rusted due to untimely maintenance or environmental factors, so that potential safety hazards are generated for running of the train. When the defects appear on the surface of the steel rail, the defects are mainly expressed as concave areas with stripe shapes and spot shapes on the surface of the steel rail, the surface of the steel rail in a normal state is smooth and flat and sensitive to the response of a light source, the whole gloss is bright, but if the defects appear in a certain area, the whole gloss of the area is reduced, the textures of the areas at the positions of the defects are complex, the light scattering is serious, and a large number of scattering points are formed in the local area. Under the irradiation of the shooting auxiliary light source, the surface gloss of the steel rail is gradually decreased in a diffusing manner in the middle to all directions, and when defects occur, jump occurs in the area with original gradually changed gloss. Based on the above features, the following processing analysis is performed on the gradation image.
Fig. 2 is a basic flow chart of a light-sensitive particle acquiring method provided in the embodiment of the present application, as shown in fig. 2, in some embodiments of the present invention, a light-sensitive particle may be obtained by dividing a pixel point in a gray image according to a pixel gray value of the gray image. The method specifically comprises the following steps:
s201: counting the pixel gray values of the gray image to obtain a maximum pixel gray value T max And a minimum pixel gray value T min 。
S202: and constructing a division threshold T of the photosensitive particles according to the maximum pixel gray value and the minimum pixel gray value.
And constructing a division threshold T of the photosensitive particles according to the maximum pixel gray value and the minimum pixel gray value. Considering that the steel rail belongs to metal, the whole light source reaction is sensitive, the real light-sensitive particles and the light-weak sensitive particles are difficult to distinguish only by the average value of the maximum and minimum pixel gray values of the gray image A, and the threshold value needs to be corrected to a certain extent in order to better distinguish the light-sensitive particles and the light-weak sensitive particles, so that the adjustment parameter a is added. Therefore, the method for calculating the division threshold T of the photosensitive particles is as follows:
wherein T represents the division threshold of the photosensitive particles, T min T is the minimum pixel gray value in the gray image max And a is a threshold adjustment coefficient, which is the maximum pixel gray value in the gray image. Wherein, the empirical value of a can be 0.8.
S203: judging whether the pixel gray value of the pixel point is larger than or equal to a dividing threshold value, and if so, dividing the pixel point into light-sensitive particles.
According to the dividing threshold T obtained in step S202, when the pixel gray value of a certain pixel point is greater than or equal to the threshold T, dividing the pixel point into light-sensitive particles; when the pixel gray value of a certain pixel point is smaller than the threshold value T, dividing the pixel point into light weakly sensitive particles. To this end, the pixel point in the grayscale image a is divided into two parts of light-sensitive particles and light-weakly-sensitive particles.
S300: and obtaining the glossiness of the pixel according to the distribution condition of the light-sensitive particles around the pixel.
Fig. 3 is a basic flow chart of a method for obtaining glossiness of a pixel according to an embodiment of the present application, as shown in fig. 3, in some embodiments of the present invention, glossiness of the pixel is obtained according to distribution of light-sensitive particles around the pixel, including the following steps:
s301: and counting the pixel gray values of the gray image, and determining the pixel point corresponding to the maximum pixel gray value as a central pixel point.
Due to the existence of the auxiliary light source, the brightest point in the steel rail image is the center point of the image. Counting the pixel gray value in the gray image A, and finding out the point record gray value with the maximum pixel gray value as T max The point gray value with the smallest gray value is recorded as T min The pixel with the largest gray value is used as the center point pixel of the gray image A, and the coordinates of the pixel are recorded as (x 0 ,y 0 )。
S302: constructing a first pixel window by taking each pixel point as the center, acquiring light-sensitive particles contained in the first pixel window, and calculating a distance value d between the light-sensitive particles and the pixel point of the center point i 。
Constructing a first pixel window of 5*5 by taking each pixel point as the center, acquiring light-sensitive particles contained in the first pixel window, and calculating a distance value d between the light-sensitive particles and the pixel point of the center point i (i＝1,2,…n)。
S303: according to the distance value d i The glossiness LD of the pixel is obtained.
According to the distance value d i The method for calculating the glossiness LD of the pixel point comprises the following steps:
where LD is the glossiness of the pixel, d i Is the distance between the ith light-sensitive particle and the central pixel point, and n is the pixelTotal number of light-sensitive particles within the window.
When the glossiness of the pixel point is larger, the light-sensitive particles around the pixel point are densely distributed, the density is larger, and the overall glossiness is larger; when the glossiness of the pixel point is smaller, the light-sensitive particles around the pixel point are scattered, the density is smaller, and the glossiness is smaller.
S400: and analyzing the glossiness of the pixel points to obtain the glossiness gradient of the gray level image.
Fig. 4 is a basic flow chart of a method for obtaining a gradient of brightness gradient of a gray image according to an embodiment of the present application, as shown in fig. 4, in some embodiments of the present invention, the method includes the following steps:
s401: and constructing a second pixel window 3*3 by taking the pixel point corresponding to the maximum pixel gray value in the gray image as a central pixel point.
S402: in the second pixel window, a center pixel point (x 0 ,y 0 ) Is defined in the four-neighbor direction of the mobile terminal.
Fig. 5 is a schematic diagram of four-neighborhood direction setting of a central pixel according to an embodiment of the present application, as shown in fig. 5, in a second pixel window, a central pixel (x 0 ,y 0 ) Is defined in the four-neighbor direction of the mobile terminal. I.e. with a central pixel point (x 0 ,y 0 ) The positive and negative directions of the x axis and the positive and negative directions of the y axis which are the origins are the directions of the four adjacent domains.
S403: and respectively obtaining the gloss gradient in the directions of the four adjacent domains.
At the center pixel point (x 0 ,y 0 ) The brightness gradient condition of the gray level image A is obtained in the direction of the four adjacent domains, namely the brightness gradient in the direction of the four adjacent domains is respectively obtained. The method for acquiring the gloss gradient in the direction of the four adjacent domains comprises the following steps: with pixels (x) i ,y i ) Constructing a third pixel window for the center; and calculating pixel glossiness differences of three pixel points on a neighborhood direction in the third pixel window and three pixel points on an opposite neighborhood direction to obtain a glossiness gradient LDB of the pixel points. Gradual change of gloss in the negative x-axis directionTaking the case of a pixel point (x i ,y i ) A third pixel window is built for the center, each point in the x-axis negative direction calculates the pixel glossiness difference value of three points in the x-axis negative direction and three points in the x-axis positive direction in the third pixel window, and the glossiness gradient LDB of the pixel point is obtained, namely:
LDB(x i ,y i )＝[LD(x i -1,y i +1)+LD(x i -1,y i )+LD(x i -1,y i -1)]
-[LD(x i +1,y i +1)+LD(x i +1,y i )+LD(x i +1,y i -1)]
in the formula, LDB (x) i ,y i ) Representing pixel points (x) i ,y i ) Is a gradient of gloss gradation, LD (x i -1,y i +1) represents the pixel point (x) i ,y i ) Point gloss at a point with a distance of 1 in the negative x-axis direction and a point with a distance of 1 in the positive y-axis direction, the remaining LD values and so on, (x) i ,y i ) Represents the ith point coordinate in the negative x-axis direction, where the negative x-axis direction represents one of the four neighborhood directions.
The larger the gradient of the gradual change in gloss is, the larger the difference in gloss of the pixel point in the direction is, and the smaller the gradient of the gradual change in gloss is, the smaller the difference in gloss of the pixel point in the direction is. And so on, the gloss gradient in the center point 4 directions was obtained.
S500: and analyzing the gradient of the gradual change of the gloss to obtain the gloss abnormality index of the pixel point in the distinguishing direction.
Under the influence of an auxiliary light source, the surface gloss gradient of the steel rail in a normal state meets the gradient state that the center is diffused to the periphery, and if the surface of the steel rail is defective, the surface of a defective area is uneven, so that the glossiness of the area is changed, and the glossiness gradient rule of the steel rail area is deviated. Based on this, the gloss anomaly index of the pixel in the discrimination direction can be obtained by analyzing the gloss gradient.
Fig. 6 is a basic flow chart of a method for obtaining a gloss anomaly index of a pixel in a distinguishing direction according to an embodiment of the present application, as shown in fig. 6, in some embodiments of the present invention, a gloss gradient is analyzed to obtain a gloss anomaly index of the pixel in the distinguishing direction, including the following steps:
s501: and constructing a judging window by taking the pixel point as a center, and taking the connecting line of the pixel point and the center pixel point of the gray level image as a judging direction.
Aiming at the defect characteristics of the surface of the steel rail, a 3*3 distinguishing window is built by taking each pixel point in the gray image A as the center. And the connection line between the pixel point and the central point of the gray image A is used as the distinguishing direction.
S502: and analyzing the gradient of the gradual change of the gloss in the judging window and the judging direction to obtain the gloss anomaly index LND of the pixel point in the judging direction.
Fig. 7 is a schematic diagram of a discrimination window according to an embodiment of the present application, as shown in fig. 7, so as to show the difference between the center pixel point (x 0 ,y 0 ) For example, a discrimination window constructed by pixels in the x-axis negative direction and the y-axis negative direction. The distinguishing direction can be regarded as combining one gloss gradient direction component in the x-axis negative direction and the gloss gradient component in the y-axis negative direction, then based on the gloss gradient components in the two directions, obtaining the gloss gradient value of the pixel point which is on the y-axis negative direction component and has the same abscissa as the pixel point as the direction component, recording as LDB i (i=1, 2, … h), and the pixel point which is classified in the negative x-axis direction and has the same ordinate as the pixel point is the direction component gloss gradient value, recorded as LDB j (j＝1,2,…l)。
The gloss anomaly index LND of the pixel point in the distinguishing direction is obtained, and the LND calculating method comprises the following steps:
in the formula, LND represents the gloss anomaly index of the pixel point in the distinguishing direction and LDB 0 Indicating the gloss gradation of the pixel pointGradient, LDB i＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative y-axis component, LDB j＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative direction component of the x axis and LDB i Representing the gradient of gradual gloss change of the ith pixel point on the negative y-axis component, LDB i+1 Representing the gradient of gradual gloss change of the (i+1) th pixel point in the negative y-axis direction component, LDB j Representing the gradient of gradual gloss change of the jth pixel point on the negative x-axis component, LDB j+1 The gloss gradient of the j+1th pixel point on the x-axis negative direction component is represented, h represents the number of the pixel points of the pixel point on the y-axis negative direction component, and l represents the number of the pixel points of the pixel point on the x-axis negative direction component.
When the gloss anomaly index is larger, the change degree of the gloss gradient at the position of the pixel point is larger compared with that of the pixel point in the component direction, and the position of the pixel point is more likely to generate rail defects; when the gloss anomaly index is smaller, the gloss gradient of the position of the pixel point is similar to the change degree of the pixel point in the component direction, and the probability of rail defects at the position of the pixel point is smaller.
Since there may be one or more defects in the rail local area, in the case of multiple defects, deviation may occur in the gradual change discrimination of the rail surface, resulting in inaccurate detection. In order to accurately detect the defects on the surface of the steel rail, the invention combines a series of magnetic leakage images obtained by the magnetic leakage detection equipment to further judge the defects on the steel rail. The specific distinguishing method comprises the following steps.
S600: and analyzing the amplitude value of the magnetic leakage image to construct the magnetic flux coefficient of the steel rail.
The main principle of magnetic leakage detection is that a local area is magnetized, when the surface of a steel rail is free from defects, magnetic flux can completely pass through the surface of the steel rail to form a relatively flat magnetic flux image, and when the surface of the steel rail is defective, partial magnetic flux is distorted at the defective part due to the defects, and the magnetic flux is mainly expressed as abrupt change of wave bands on the magnetic leakage image. Based on the acquired magnetic flux leakage image P i And analyzing the position of the mutation in the magnetic flux leakage image.
For each pixel point, the coordinates (x i ,y i ) Are all corresponding to the magnetic leakage image P i And corresponding abscissa x i Is the leakage magnetic image amplitude value of (2)For the magnetic flux leakage image of the surface of the steel rail, when the amplitude peak value of a certain pixel point is obviously higher than the amplitude peak value of other pixel points, the pixel point is more likely to be in a defect area of the surface of the steel rail. Based on the acquired magnetic flux leakage image P i (i=1, 2, … g), each corresponding abscissa x is found i Mean value of leakage image amplitude values +.>Ordinate y of each pixel point i Is +.>According to the mean value and the amplitude value of the leakage magnetic image corresponding to the abscissa of each pixel point, constructing a steel rail magnetic flux coefficient PNB, wherein the steel rail magnetic flux coefficient calculating method comprises the following steps:
wherein PNB represents the magnetic flux coefficient of the steel rail,is pixel (x) i ,y i ) Amplitude value of>Is pixel (x) i ,y i ) The magnetic flux leakage image amplitude value at the position of the abscissa is the average value.
When the magnetic flux coefficient of the steel rail is larger, the more serious the magnetic flux leakage phenomenon of the position of the pixel point is, the more likely the steel rail defect is; when the magnetic flux coefficient of the steel rail is smaller, the magnetic flux condition of the position of the pixel point is normal, and the surface state of the steel rail is good.
S700: and fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence of the steel rail surface.
In order to accurately judge the defect condition of the steel rail surface, a gloss anomaly index LND acquired in a patrol shot image and a steel rail magnetic flux coefficient PNB acquired in a steel rail magnetic flux leakage image are fused, and a defect confidence level QBL of the steel rail surface is constructed, wherein the steel rail defect confidence level calculating method comprises the following steps:
QBL＝LND*PNB
where QBL is the rail defect confidence, LND is the gloss anomaly index, PNB is the rail magnetic flux coefficient.
When the confidence of the rail defect is larger, the probability that the defect exists in the inspection image and the magnetic flux leakage image at the position of the pixel point is larger, and the rail defect is more likely to occur at the position; and otherwise, when the confidence of the steel rail is smaller, the probability that the position of the pixel point is in defects in the inspection image and the magnetic flux leakage image is smaller, and the position is more likely to be the surface of the normal steel rail.
S800: based on the rail defect confidence, modifying sigma values of the single-scale SSR image enhancement algorithm to obtain a corrected SSR image enhancement algorithm.
Based on the rail defect confidence, modifying sigma values of the single-scale SSR image enhancement algorithm to obtain a corrected SSR image enhancement algorithm.
It should be noted that the value of sigma in the single-scale SSR algorithm is in the value range of [80,100], so the value of sigma after adjustment should also be in the value range of [80,100], so that normalization operation on the value of QBL is required.
δ′＝FQBL×δ
Wherein FQBL is normalized rail defect confidence, the normalization range is [0.8,1.0], c and d are normalization adjustment parameters, wherein the values of c and d can be c=0.2, d=0.8, min (QBL) represents the minimum value of the rail defect confidence, max (QBL) represents the maximum value of the rail defect confidence, the value of initial input delta is set to be 100, and delta' is the value of sigma finally input into an SSR algorithm.
When the normalized rail defect confidence FQBL is smaller, the defect probability of the pixel point position is smaller, so that the final value of delta' is smaller, and additional enhancement is not needed; when the value of the normalized rail defect confidence FQBL is larger, the defect probability of the pixel point position is larger, so that the final value of delta' is larger, and the reinforcing effect on the region is stronger.
S900: based on a correction SSR image enhancement algorithm, an image segmentation technology is used for dividing a steel rail defect region.
Modifying the sigma value in the SSR algorithm based on the constructed rail defect confidence, carrying out stronger image enhancement on the region more likely to have the rail defect, carrying out common enhancement on the region with lower defect probability, and finally, making the region with the defect in the image more obvious in characteristic. And then based on a correction SSR image enhancement algorithm, an image segmentation technology such as OTSU is used for rapidly dividing the rail defect area, so that the rail defect detection speed and accuracy are improved.
Based on the same inventive concept as the above method, this embodiment also provides a rail surface detection system that fuses the inspection image and the magnetic flux leakage detection, and fig. 8 is a basic composition schematic diagram of the rail surface detection system that fuses the inspection image and the magnetic flux leakage detection, as shown in fig. 8, provided in the embodiment of the present application:
the system includes a memory module 10 and a processor module 20, wherein:
a memory module 10 for storing program codes;
the processor module 20 is used for reading the program codes stored in the memory module 10, collecting and obtaining RGB images and magnetic flux leakage images of the steel rail, and preprocessing the RGB images to obtain gray images; dividing pixel points in the gray level image according to the pixel gray level value of the gray level image to obtain light-sensitive particles; according to the distribution condition of light-sensitive particles around the pixel points, the glossiness of the pixel points is obtained; analyzing the glossiness of the pixel points to obtain the glossiness gradient of the gray level image; analyzing the gradient of gradual change of gloss to obtain a gloss anomaly index of the pixel point in the distinguishing direction; analyzing the amplitude value of the magnetic leakage image to construct the magnetic flux coefficient of the steel rail; fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence coefficient of the steel rail surface; modifying sigma values of the single-scale SSR image enhancement algorithm based on the rail defect confidence level to obtain a modified SSR image enhancement algorithm; based on a correction SSR image enhancement algorithm, an image segmentation technology is used for dividing a steel rail defect region.
In some embodiments of the invention, the processor module 20 includes: an image acquisition sub-module 21, a gray level image processing sub-module 22, a magnetic flux leakage image processing sub-module 23 and a rail defect detection sub-module 24.
Wherein the image acquisition sub-module 21 is configured to acquire RGB images and magnetic flux leakage images of the steel rail, and preprocess the RGB images to acquire gray scale images.
The gradation image processing sub-module 22 includes a light-sensitive particle acquiring unit 221, a glossiness acquiring unit 222, a gloss gradient acquiring unit 223, and a gloss anomaly index acquiring unit 224.
Further, the light-sensitive particle acquisition unit 221 is configured to: counting the pixel gray values of the gray image to obtain a maximum pixel gray value T max And a minimum pixel gray value T min The method comprises the steps of carrying out a first treatment on the surface of the Constructing a dividing threshold T of the photosensitive particles according to the maximum pixel gray value and the minimum pixel gray value; judging whether the pixel gray value of the pixel point is larger than or equal to a dividing threshold value, and if so, dividing the pixel point into light-sensitive particles.
The glossiness acquisition unit 222 is configured to: counting the pixel gray values of the gray image, and determining the pixel point corresponding to the maximum pixel gray value as a central pixel point; constructing a first pixel window by taking each pixel point as the center, acquiring light-sensitive particles contained in the first pixel window, and calculating a distance value d between the light-sensitive particles and the pixel point of the center point i The method comprises the steps of carrying out a first treatment on the surface of the According to the distance value d i The glossiness LD of the pixel is obtained.
The gloss-gradation gradient acquisition unit 223 is configured to: with the maximum in the gray imageConstructing a second pixel window 3*3 by taking the pixel point corresponding to the pixel gray value as the central pixel point; in the second pixel window, a center pixel point (x 0 ,y 0 ) Is the direction of the four adjacent domains; the method for respectively acquiring the gloss gradient in the direction of the four adjacent domains comprises the following steps: with pixels (x) i ,y i ) Constructing a third pixel window for the center; and calculating pixel glossiness differences of three pixel points on a neighborhood direction in the third pixel window and three pixel points on an opposite neighborhood direction to obtain a glossiness gradient LDB of the pixel points.
The gloss anomaly index acquisition unit 224 is configured to: a discrimination window is constructed by taking the pixel point as the center, and the line between the pixel point and the center pixel point of the gray image is taken as the discrimination direction; and analyzing the gradient of the gradual change of the gloss in the judging window and the judging direction to obtain the gloss anomaly index LND of the pixel point in the judging direction.
The magnetic flux leakage image processing sub-module 23 is configured to: and analyzing the amplitude value of the magnetic leakage image to construct the magnetic flux coefficient of the steel rail.
The rail defect detection sub-module 24 comprises a rail defect confidence acquiring unit 241, an SSR image enhancement algorithm correcting unit 242 and a rail defect region detecting unit 243.
Further, the rail defect confidence acquiring unit 241 is configured to: and fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence of the steel rail surface. And the method is also configured to normalize the value of the rail defect confidence.
The SSR image enhancement algorithm correction unit 242 is configured to: based on the rail defect confidence, modifying sigma values of the single-scale SSR image enhancement algorithm to obtain a corrected SSR image enhancement algorithm.
The rail defect region detection unit 243 is configured to: based on a correction SSR image enhancement algorithm, an image segmentation technology is used for dividing a steel rail defect region.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
It is noted that unless specified and limited otherwise, relational terms such as "first" and "second", and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such article or apparatus. Without further limitation, the statement "comprises one … …" does not exclude that an additional identical element is present in an article or device that comprises the element. In addition, the term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It is to be understood that the invention is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
Claims (10)
1. A steel rail surface detection method integrating inspection images and magnetic flux leakage detection is characterized by comprising the following steps:
acquiring RGB images and magnetic flux leakage images of the steel rail, and preprocessing the RGB images to obtain gray images;
dividing pixel points in the gray level image according to the pixel gray level value of the gray level image to obtain light-sensitive particles;
according to the distribution condition of the light-sensitive particles around the pixel points, the glossiness of the pixel points is obtained;
analyzing the glossiness of the pixel points to obtain the glossiness gradient of the gray level image;
analyzing the gloss gradient to obtain a gloss anomaly index of the pixel point in the distinguishing direction;
analyzing the amplitude value of the magnetic leakage image to construct a steel rail magnetic flux coefficient;
fusing the gloss anomaly index and the steel rail magnetic flux coefficient to construct the steel rail defect confidence coefficient of the steel rail surface;
modifying a sigma value of the single-scale SSR image enhancement algorithm based on the rail defect confidence coefficient to obtain a modified SSR image enhancement algorithm;
and dividing the steel rail defect area by using an image segmentation technology based on the correction SSR image enhancement algorithm.
2. The method for detecting the surface of the steel rail by fusing a patrol image and magnetic flux leakage detection according to claim 1, wherein dividing pixel points in the gray image to obtain light-sensitive particles according to pixel gray values of the gray image comprises the following steps:
counting the pixel gray values of the gray image to obtain a maximum pixel gray value T max And a minimum pixel gray value T min ；
Constructing a division threshold T of the light-sensitive particles according to the maximum pixel gray value and the minimum pixel gray value:
wherein T represents the division threshold of the photosensitive particles, T min T is the minimum pixel gray value in the gray image max A is a maximum pixel gray value in a gray image, and a is a threshold adjustment coefficient;
and judging whether the pixel gray value of the pixel point is larger than or equal to the dividing threshold value, and if so, dividing the pixel point into light-sensitive particles.
3. The method for detecting the surface of the steel rail by fusing a patrol image and magnetic leakage detection according to claim 1, wherein the step of obtaining the glossiness of the pixel according to the distribution condition of the light-sensitive particles around the pixel comprises the following steps:
counting the pixel gray values of the gray images, and determining the pixel point corresponding to the maximum pixel gray value as a central pixel point;
constructing a first pixel window by taking each pixel point as a center, acquiring light-sensitive particles contained in the first pixel window, and calculating a distance value d between the light-sensitive particles and the pixel point of the center point i ；
According to the distance value d i Obtaining the glossiness LD of the pixel point:
where LD is the glossiness of the pixel, d i And n is the total number of the light-sensitive particles in the pixel window, wherein the distance between the ith light-sensitive particle and the central pixel point is the distance between the ith light-sensitive particle and the central pixel point.
4. The method for detecting the surface of the steel rail by fusing a patrol image and magnetic flux leakage detection according to claim 1, wherein analyzing the glossiness of the pixel points to obtain the glossiness gradient of the gray image comprises the following steps:
constructing a second pixel window of 3*3 by taking a pixel point corresponding to the maximum pixel gray value in the gray image as a central pixel point;
in the second pixel window, a center pixel point (x 0 ,y 0 ) Is the direction of the four adjacent domains;
the method for respectively acquiring the gloss gradient in the direction of the four adjacent domains comprises the following steps:
with pixels (x) i ,y i ) Constructing a third pixel window for the center;
calculating the pixel glossiness difference value of three pixel points on the neighborhood direction in the third pixel window and three pixel points on the opposite neighborhood direction, and obtaining the glossiness gradient LDB of the pixel points, namely:
LDB(x i ,y i )＝[LD(x i -1，y i +1)+LD(x i -1，y i )+LD(x i -1，y i -1)]-[LD(x i +1，y i +1)+LD(x i +1，y i )+LD(x i +1，y i -1)]
in the formula, LDB (x) i ,y i ) Representing pixel points (x) i ,y i ) Is a gradient of gloss gradation, LD (x i -1,y i +1) represents the pixel point (x) i ,y i ) Point gloss at a point with a distance of 1 in the negative x-axis direction and a point with a distance of 1 in the positive y-axis direction, the remaining LD values and so on, (x) i ,y i ) The ith point coordinate in the negative x-axis direction is represented, wherein the negative x-axis direction represents one of the four neighborhood directions.
5. The method for detecting the surface of the steel rail by fusing the inspection image and the magnetic flux leakage detection according to claim 1, wherein analyzing the gradient of gradual change of gloss to obtain the gloss anomaly index of the pixel point in the distinguishing direction comprises the following steps:
a discrimination window is constructed by taking the pixel point as the center, and the line between the pixel point and the center pixel point of the gray image is taken as the discrimination direction;
analyzing the gloss gradient in the judging window and in the judging direction to obtain a gloss anomaly index LND of the pixel point in the judging direction, wherein the LND calculating method comprises the following steps:
in the formula, LND represents the gloss anomaly index of the pixel point in the distinguishing direction and LDB 0 Indicating the gradient of the gradual change of the brightness of the pixel point, LDB i＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative y-axis component, LDB j＝1 Representing the gradient of gradual gloss change of the 1 st pixel point on the negative direction component of the x axis and LDB i Representing the gradient of gradual gloss change of the ith pixel point on the negative y-axis component, LDB i+1 Representing the gradient of gradual gloss change of the (i+1) th pixel point in the negative y-axis direction component, LDB j Representing the gradient of gradual gloss change of the jth pixel point on the negative x-axis component, LDB j+1 The gloss gradient of the j+1th pixel point on the x-axis negative direction component is represented, h represents the number of the pixel points of the pixel point on the y-axis negative direction component, and l represents the number of the pixel points of the pixel point on the x-axis negative direction component.
6. The method for detecting the surface of the steel rail by fusing inspection images and magnetic flux leakage detection according to claim 1, wherein the method for calculating the magnetic flux coefficient of the steel rail is as follows:
wherein PNB represents the magnetic flux coefficient of the steel rail,is pixel (x) i ,y i ) Amplitude value of>Is pixel (x) i ,y i ) The magnetic flux leakage image amplitude value at the position of the abscissa is the average value.
7. The method for detecting the surface of the steel rail by fusing the inspection image and the magnetic flux leakage detection according to claim 1, wherein the method for calculating the confidence coefficient of the steel rail defect is as follows:
QBL＝LND*PNB
where QBL is the rail defect confidence, LND is the gloss anomaly index, PNB is the rail magnetic flux coefficient.
8. The method for detecting the surface of the steel rail by fusing the inspection image and the magnetic flux leakage detection according to claim 1, wherein the modification of sigma values is performed on a single-scale SSR image enhancement algorithm based on the steel rail defect confidence coefficient to obtain a modified SSR image enhancement algorithm, and the method comprises the following steps:
normalizing the value of the confidence coefficient of the steel rail defect, wherein the processing method comprises the following steps:
δ＝FQBL×δ
wherein FQBL is normalized rail defect confidence, c and d are normalized adjustment parameters, QBL represents rail defect confidence, min (QBL) represents minimum value of rail defect confidence, max (QBL) represents maximum value of rail defect confidence, initial input delta value is set, and delta' is sigma value finally input into SSR algorithm.
9. A rail surface detection system integrating inspection images with magnetic flux leakage detection, characterized in that the system comprises a memory module (10) and a processor module (20), wherein:
-said memory module (10) for storing program code;
the processor module (20) for reading the program code stored in the memory module (10) and performing the method of any of claims 1-8.
10. The steel rail surface detection system for fusing inspection images and magnetic leakage detection according to claim 9, wherein the processor module (20) comprises an image acquisition module (21), a gray level image processing module (22), a magnetic leakage image processing module (23) and a steel rail defect detection module (24).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311383836.XA CN117437191A (en) | 2023-10-24 | 2023-10-24 | Rail surface detection method and system integrating inspection image and magnetic flux leakage detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311383836.XA CN117437191A (en) | 2023-10-24 | 2023-10-24 | Rail surface detection method and system integrating inspection image and magnetic flux leakage detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117437191A true CN117437191A (en) | 2024-01-23 |
Family
ID=89557799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311383836.XA Pending CN117437191A (en) | 2023-10-24 | 2023-10-24 | Rail surface detection method and system integrating inspection image and magnetic flux leakage detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117437191A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117808799A (en) * | 2024-02-29 | 2024-04-02 | 天津戎军航空科技发展有限公司 | Chamfering equipment processing quality detection method based on artificial intelligence |
| CN117934455A (en) * | 2024-03-19 | 2024-04-26 | 江苏盖亚环境科技股份有限公司 | River water flow purification effect-based detection method and system |
-
2023
- 2023-10-24 CN CN202311383836.XA patent/CN117437191A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117808799A (en) * | 2024-02-29 | 2024-04-02 | 天津戎军航空科技发展有限公司 | Chamfering equipment processing quality detection method based on artificial intelligence |
| CN117808799B (en) * | 2024-02-29 | 2024-05-07 | 天津戎军航空科技发展有限公司 | Chamfering equipment processing quality detection method based on artificial intelligence |
| CN117934455A (en) * | 2024-03-19 | 2024-04-26 | 江苏盖亚环境科技股份有限公司 | River water flow purification effect-based detection method and system |
| CN117934455B (en) * | 2024-03-19 | 2024-06-11 | 江苏盖亚环境科技股份有限公司 | River water flow purification effect-based detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115409833B (en) | Hot spot defect detection method of photovoltaic panel based on unsharp mask algorithm | |
| CN117437191A (en) | Rail surface detection method and system integrating inspection image and magnetic flux leakage detection | |
| Wei et al. | Defect detection of pantograph slide based on deep learning and image processing technology | |
| Min et al. | Real time detection system for rail surface defects based on machine vision | |
| CN105069790B (en) | A kind of gear open defect fast image detection method | |
| CN110210477B (en) | Digital instrument reading identification method | |
| CN102589435B (en) | Efficient and accurate detection method of laser beam center under noise environment | |
| CN117237368B (en) | Bridge crack detection method and system | |
| KR100382577B1 (en) | Wheel measuring apparatus | |
| CN109682839B (en) | Online detection method for surface defects of metal arc-shaped workpiece | |
| CN109490316A (en) | A kind of surface defects detection algorithm based on machine vision | |
| CN107341802A (en) | It is a kind of based on curvature and the compound angular-point sub-pixel localization method of gray scale | |
| CN104718428A (en) | Pattern inspecting and measuring device and program | |
| CN116563282B (en) | Drilling tool detection method and system based on machine vision | |
| CN117095004B (en) | Excavator walking frame main body welding deformation detection method based on computer vision | |
| CN108596872A (en) | The detection method of rail disease based on Gabor wavelet and SVM | |
| CN113465541B (en) | Contact line abrasion measuring method and system | |
| CN116523923B (en) | Battery case defect identification method | |
| CN112001917A (en) | Machine vision-based geometric tolerance detection method for circular perforated part | |
| CN113155839A (en) | Steel plate outer surface defect online detection method based on machine vision | |
| CN115909256B (en) | Road disease detection method based on road visual image | |
| CN115240146B (en) | Intelligent machine tool assembly acceptance method based on computer vision | |
| CN115880280B (en) | Method for detecting quality of welding seam of steel structure | |
| CN115526864A (en) | Steel rail surface defect detection method based on improved characteristic pyramid network and metric learning | |
| CN112150375A (en) | Tape inspection system, tape inspection method, and storage medium with tape inspection program |
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 |