CN118037735B - Cable insulation skin damage identification method - Google Patents
Cable insulation skin damage identification method Download PDFInfo
- Publication number
- CN118037735B CN118037735B CN202410438979.4A CN202410438979A CN118037735B CN 118037735 B CN118037735 B CN 118037735B CN 202410438979 A CN202410438979 A CN 202410438979A CN 118037735 B CN118037735 B CN 118037735B
- Authority
- CN
- China
- Prior art keywords
- target
- side length
- area
- gray level
- segmentation
- 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
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000009413 insulation Methods 0.000 title claims abstract description 41
- 230000037380 skin damage Effects 0.000 title abstract description 5
- 230000011218 segmentation Effects 0.000 claims abstract description 155
- 230000000694 effects Effects 0.000 claims abstract description 45
- 230000006378 damage Effects 0.000 claims abstract description 37
- 238000011156 evaluation Methods 0.000 claims abstract description 32
- 230000008569 process Effects 0.000 claims description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 230000008859 change Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000010606 normalization Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 5
- 238000003064 k means clustering Methods 0.000 claims description 4
- 238000005192 partition Methods 0.000 claims description 3
- 238000003909 pattern recognition Methods 0.000 abstract description 3
- 210000003491 skin Anatomy 0.000 description 36
- 238000012545 processing Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000001124 body fluid Anatomy 0.000 description 1
- 239000010839 body fluid Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 210000002615 epidermis Anatomy 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction 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
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of pattern recognition, in particular to a cable insulation skin damage recognition method, which comprises the following steps: capturing a gray image of an insulating surface skin of a cable to be detected; extracting local variation parameters of each divided region under the side length of each reference window; obtaining a plurality of target segmentation areas of the gray level image according to the segmentation effect evaluation coefficients; extracting the possibility of each target segmentation area belonging to the damaged area; extracting the effectiveness of the Ojin threshold of each target segmentation area; extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster; acquiring all damage identification thresholds according to the degree of preference of the Ojin threshold; dividing the gray level image for a plurality of times according to the damage identification threshold value to obtain an integrally divided gray level image; and carrying out damage identification on the cable insulation skin to be detected according to the wholly segmented gray level image. The invention improves the accuracy of identifying the damaged area of the cable insulation surface.
Description
Technical Field
The invention relates to the technical field of pattern recognition, in particular to a cable insulation skin damage recognition method.
Background
The cable is used as important energy transmission equipment and bears the transmission tasks of power, communication and control signals; the insulation surface of the cable may be damaged or broken due to long-term use or external environmental factors, which may cause the safety and reliability of the cable to be affected; the pattern recognition algorithm is utilized to analyze and process the pattern, so that the damaged area on the surface of the cable can be effectively recognized, the safe and reliable operation of the cable can be ensured, the occurrence of cable faults can be reduced, and the service life of equipment can be prolonged.
In the prior art, the Ojin threshold segmentation algorithm is a commonly used cable insulation skin damage recognition algorithm, and has the main effects that the insulation skin of the cable is segmented by using an inter-class variance, and then damage recognition is performed according to a segmentation result; however, when the insulation surface of the cable is damaged, a plurality of damage states coexist, so that the whole cable image space structure and gray value structure are complex, and when the cable is segmented by using the Ojin threshold method, a good segmentation effect cannot be obtained, and further, the recognition effect in the subsequent cable insulation surface damage recognition process is poor.
Disclosure of Invention
In order to solve the above problems, the present invention provides a method for identifying damage to an insulating skin of a cable, the method comprising:
capturing a gray image of an insulating surface skin of a cable to be detected;
acquiring a plurality of reference window side lengths, and acquiring a plurality of dividing areas of the gray level image under each reference window side length; extracting local variation parameters of each divided region under the side length of each reference window according to the gray value distribution condition of pixel points in each divided region of the gray image under the side length of each reference window; extracting a segmentation effect evaluation coefficient under the side length of each reference window according to the side length of each reference window and the local variation parameters of each segmentation area under the side length of the reference window; obtaining a plurality of target segmentation areas of the gray level image according to the segmentation effect evaluation coefficients;
Extracting the possibility of each target division area belonging to the damaged area according to the gray level distribution contrast condition between each target division area and the gray level image; acquiring an oxford threshold value of each target segmentation area; extracting the effectiveness of the Ojin threshold of each target segmentation area according to the possibility of belonging to the damaged area and the local change parameters; clustering all the target segmentation areas to obtain a plurality of clusters, and screening all the clusters according to the effectiveness of the Ojin threshold to obtain target clusters; extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster according to the difference of the oxford threshold value between each target segmentation area and other target segmentation areas in the target cluster; acquiring all damage identification thresholds according to the degree of preference of the Ojin threshold;
Dividing the gray level image for a plurality of times according to the damage identification threshold value to obtain an integrally divided gray level image; and carrying out damage identification on the cable insulation skin to be detected according to the wholly segmented gray level image.
Preferably, the acquiring a plurality of reference window side lengths, and acquiring a plurality of division areas of the gray image under each reference window side length, includes the specific method that:
presetting two window side lengths To/>For the initial window side length, the step length is 1, the window side length is sequentially increased, the initial window side length and the window side length after each increase are recorded as the reference window side length until the reference window side length isWhen the increment is stopped, a plurality of reference window side lengths are obtained; for any one reference window side length/>Size/>, useThe step length of the sliding window of the cable insulation surface to be detected is/>And taking each sliding window as a division area under the side length of the reference window, and obtaining a plurality of division areas of the gray level image under the side length of the reference window.
Preferably, the extracting the local variation parameter of each divided region under the side length of each reference window according to the gray value distribution condition of the pixel point in each divided region under the side length of each reference window, includes the specific method that:
Will be the first Maximum value of inter-class variance of gray values of all pixel points in all divided regions under the side length of each reference window and/>/>, Under a side length of a reference windowThe ratio of the variances of the gray values of all the pixel points in the divided areas is taken as the/>/>, Under a side length of a reference windowLocal variation factors of the individual segmented regions;
Will be the first And carrying out linear normalization on local change factors of all the divided areas under the side length of each reference window, and recording the normalized local change factors as local change parameters.
Preferably, the extracting the evaluation coefficient of the segmentation effect under the side length of each reference window according to the side length of each reference window and the local variation parameter of each segmentation area under the side length of the reference window comprises the following specific steps:
Will be the first Side length of each reference window and/>The ratio of the total number of all the divided areas under the side length of each reference window is recorded as a first ratio; will/>The accumulated sum of the local variation parameters of all the divided areas under the side length of each reference window is recorded as a first accumulated sum; taking the product of the first ratio and the first accumulated sum as the first/>Dividing effect evaluation factors under the side length of each reference window;
And carrying out linear normalization on the segmentation effect evaluation factors under the side length of all the reference windows, and marking the normalized segmentation effect evaluation factors as segmentation effect evaluation coefficients.
Preferably, the method for obtaining a plurality of target division areas of the gray image according to the division effect evaluation coefficient includes the following specific steps:
Taking the side length of the reference window with the maximum segmentation effect evaluation coefficient as the side length of the target window; and taking each divided area under the side length of the target window as a target divided area.
Preferably, the extracting the possibility of each target division area belonging to the damaged area according to the gray level distribution contrast condition between each target division area and the gray level image includes the following specific steps:
the calculation method for extracting the gray level integral discrete degree of the gray level image comprises the following steps:
In the method, in the process of the invention, The gray level overall discrete degree of the gray level image is represented; /(I)Representing the average value of gray values of all pixel points in the gray image; /(I)Representing the total number of all pixel points in the gray scale image; /(I)Representing the/>, in a gray scale imageGray values of the individual pixels;
for any one target segmentation area of the gray level image; the calculation method for extracting the gray level local discrete degree of the target segmentation area comprises the following steps:
In the method, in the process of the invention, Representing the gray level local discrete degree of the target segmentation area; /(I)Representing the average value of gray values of all pixel points in the target segmentation area; /(I)Representing the total number of all pixel points in the target segmentation area; /(I)Representing the/>, in the target partitionGray values of the individual pixels;
and taking the ratio of the gray level local discrete degree of the target divided region to the gray level integral discrete degree of the gray level image as the possibility of the target divided region belonging to the damaged region.
Preferably, the specific formula for extracting the effectiveness of the oxford threshold value of each target segmentation area according to the probability of belonging to the damaged area and the local variation parameters is as follows:
In the method, in the process of the invention, Represents the/>The oxford threshold effectiveness of the individual target segmentation areas; /(I)Represents the/>The possibility of the target divided regions belonging to the damaged region; /(I)Represents the/>Local variation parameters of the individual target segmentation areas; Representing the total number of all target segmentation areas; /(I) Represents the/>An oxford threshold for each target segment; /(I)Representation except for the (th) >First/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value; An exponential function based on a natural constant is represented.
Preferably, the clustering is performed on all the target segmentation areas to obtain a plurality of clusters, and the screening is performed on all the clusters according to the effectiveness of the Ojin threshold to obtain the target clusters, which comprises the following specific methods:
Clustering the effectiveness of the Ojin threshold values of all the target segmentation areas by using a self-adaptive k-means clustering algorithm to obtain a plurality of clustering clusters; taking the average value of the effectiveness of all the Ojin thresholds in each cluster as the first effectiveness average value of each cluster; and marking the cluster with the largest first validity mean as a target cluster.
Preferably, the specific formula for extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster according to the difference of the oxford threshold value between each target segmentation area and other target segmentation areas in the target cluster is as follows:
In the method, in the process of the invention, Representing the/>, in the target clusterThe degree of preference of the oxford threshold of each target segmentation area; /(I)Representing the total number of all target segmentation areas in the target cluster; /(I)Representing the/>, in the target clusterAn oxford threshold for each target segment; /(I)Representing the division number/>, in the target clusterFirst/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value.
Preferably, the method for obtaining all damage recognition thresholds according to the degree of preference of the oxford threshold includes the following specific steps:
presetting a preferred parameter For any one target segmentation area in the target cluster, if the degree of preference of the oxford threshold of the target segmentation area is smaller than or equal to a preference parameter/>And taking the Ojin threshold value of the target segmentation area as a breakage recognition threshold value.
The technical scheme of the invention has the beneficial effects that: according to the side length of each reference window and the local variation parameters of each divided area under the side length of the reference window, the dividing effect evaluation coefficient under the side length of each reference window is extracted; obtaining a plurality of target segmentation areas of the gray level image according to the segmentation effect evaluation coefficients; the target segmentation area is simpler in spatial structure and gray value distribution relative to the gray image, and can be segmented by utilizing the Ojin threshold value better; acquiring all damage identification thresholds according to the degree of preference of the Ojin threshold; dividing the gray level image for a plurality of times according to the damage identification threshold value to obtain an integrally divided gray level image; carrying out damage identification on the cable insulation skin to be detected according to the integrally segmented gray level image; the method has the advantages that the optimal segmentation threshold of the gray level image of the cable insulation skin to be detected is obtained by correcting the Ojin threshold of the target segmentation areas with the possibility of damage of different target segmentation areas, so that the accuracy of identifying the damaged area of the cable insulation skin is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart showing the steps of a method for identifying damage to an insulating surface of a cable according to the present invention;
Fig. 2 is a flow chart of the characteristic relation of a method for identifying damage to an insulating surface of a cable according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description refers to specific implementation, structure, characteristics and effects of a cable insulation skin breakage identification method according to the invention by combining the accompanying drawings and preferred embodiments. 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 invention belongs.
The specific scheme of the cable insulation skin breakage identification method provided by the invention is specifically described below with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a method for identifying damage to an insulating surface of a cable according to an embodiment of the invention is shown, the method includes the following steps:
step S001: and capturing a gray level image of the insulation surface of the cable to be detected.
It should be noted that, the main purpose of the present embodiment is to perform the overall segmentation processing on the image of the cable insulation skin, so that the image of the cable insulation skin needs to be captured first before the segmentation processing is performed on the image; in order to enable accurate segmentation of the cable insulation skin by means of machine vision, it is then necessary to arrange a device with a high definition camera for capturing an image of the cable insulation skin of the cable.
Specifically, firstly, a gray image of an insulating surface skin of a cable to be detected needs to be captured, and the specific process is as follows:
Constructing a defect recognition device comprising a high definition industrial camera, a light source and a processing unit; and placing the cable insulation skin to be detected on defect identification equipment, capturing an image of the cable insulation skin to be detected, and performing median filtering denoising and graying operation on the image of the cable insulation skin to be detected to obtain a gray image of the cable insulation skin to be detected.
The embodiment describes that the size of a captured gray image of an insulating skin of a cable to be detected is 800×800; the median filtering and graying operation are in the prior art, and the description of this embodiment is omitted here.
Thus, the gray level image of the insulation surface of the cable to be detected is captured through the method.
Step S002: acquiring a plurality of dividing areas of the gray image under the side length of each reference window; extracting local variation parameters of each divided region under the side length of each reference window; extracting a segmentation effect evaluation coefficient under the side length of each reference window; and acquiring a plurality of target segmentation areas of the gray level image according to the segmentation effect evaluation coefficients.
When the gray level image of the cable insulation skin to be detected is segmented by using the oxford segmentation algorithm, the difference between the different gray level value classifications in the image is calculated in a traversing manner by using the oxford threshold method, and then the segmentation processing is performed by taking the maximum difference between the classes as an optimal threshold, but when the damage degree of the cable insulation skin to be detected is different, the gray level value is different on the image, namely, the gray level value distribution of the whole gray level image and the space structure of the gray level image are complex, so that the accurate segmentation of the damaged area in the cable insulation skin cannot be effectively completed by using the conventional oxford threshold segmentation method.
It should be further noted that, in order to enable a better segmentation effect to be achieved on the gray level image of the cable insulation skin to be detected, the embodiment uses the local variation parameters to segment the whole region of the gray level image of the cable insulation skin to be detected by obtaining the local variation parameters based on the gray level image of the cable insulation skin to be detected, so that the image features of each segmented region are more consistent with the division of the oxford threshold value, and further the optimal oxford threshold value of each segmented region is obtained, then obtains the possibility of breakage based on different segmented regions, and uses the possibility of breakage of each segmented region and adjusts the optimal segmentation threshold value of the whole gray level image of the cable insulation skin to be detected by combining the current division of the oxford threshold value of the current segmented region so as to achieve the subsequent breakage identification of the cable insulation skin to be detected.
Presetting two window side lengthsWherein the present embodiment is described as/>To describe the example, the present embodiment is not particularly limited, wherein/>Depending on the particular implementation.
In particular, toFor the initial window side length, the step length is 1, the window side length is sequentially increased, the initial window side length and the window side length after each increase are recorded as the reference window side length until the reference window side length is/>When the increment is stopped, a plurality of reference window side lengths are obtained; for any one reference window side length/>Size/>, useThe step length of the sliding window of the cable insulation surface to be detected is/>And taking each sliding window as a division area under the side length of the reference window, and obtaining a plurality of division areas of the gray level image under the side length of the reference window.
And when the number of the edge pixel points is less than one sliding window in the process of sliding window of the gray level image, performing image edge zero padding operation.
So far, a plurality of dividing areas of the gray image under the side length of each reference window are obtained.
1. And extracting local variation parameters of each divided region of the gray image under the side length of each reference window.
Specifically, according to the gray value distribution condition of the pixel points in each divided region of the gray image under the side length of each reference window, extracting the local variation parameters of each divided region under the side length of each reference window.
As an example, extract the first/>, Under a side length of a reference windowThe calculation method of the local change factors of the individual divided regions comprises the following steps:
In the method, in the process of the invention, Represents the/>/>, Under a side length of a reference windowLocal variation factors of the individual segmented regions; Represents the/> Maximum value of inter-class variance of gray values of all pixel points in all divided areas under the side length of each reference window; /(I)Represents the/>/>, Under a side length of a reference windowVariance of gray values of all pixel points in each divided region.
Further, the first step isAnd carrying out linear normalization on local change factors of all the divided areas under the side length of each reference window, and recording the normalized local change factors as local change parameters.
It should be noted that, whenWhen the gray value of the whole pixel point in each divided area is relatively simple to constructIf the segmentation is smaller, the obtained segmented image is more approximate to the ideal segmentation effect after the segmentation is carried out by the theoretical Ojin threshold; and utilize the maximum inter-class variance/>Is aimed at judging the/>The effect of the division of the individual division regions on the body fluid is described as the greater the value, the greater the valueThe more obvious the foreground and background separation of the individual segmented regions is; the more obvious the separation of foreground from background, and the/>The image gray scale structure in each division area is simpler, and the better the overall division effect is; in summary, the/>, can be extracted/>, Under a side length of a reference windowLocal variation parameters of the individual segmented regions, the larger the value, the more/>/>, Under a side length of a reference windowThe better the segmentation effect of each segmented region, the opposite is true.
So far, the local variation parameters of each divided region of the gray image under the side length of each reference window are extracted.
2. And extracting a segmentation effect evaluation coefficient of the gray level image under the side length of each reference window.
Specifically, according to the side length of each reference window and the local variation parameters of each divided area under the side length of the reference window, the dividing effect evaluation coefficient under the side length of each reference window is extracted.
As an example, extract the firstThe calculation method of the segmentation effect evaluation factors under the side length of each reference window comprises the following steps:
In the method, in the process of the invention, Represents the/>Dividing effect evaluation factors under the side length of each reference window; /(I)Represents the/>The edge length of each reference window; /(I)Represents the/>The total number of all segmented regions at the side length of the reference window; /(I)Represents the/>/>, Under a side length of a reference windowLocal variation parameters of the individual segmented regions.
Further, the segmentation effect evaluation factors under the side length of all the reference windows are subjected to linear normalization, and the normalized segmentation effect evaluation factors are recorded as segmentation effect evaluation coefficients.
In the first aspect, the following is theEach divided area after being divided under the side length of each reference window is subjected to weighted average calculation, wherein the weight value is the first/>The reference window side sizes are correspondingly aimed at limiting the calculation amount of overall processing, when the larger reference window side length is used for dividing, the number of dividing areas existing in the gray level image is smaller, the subsequent overall calculation amount is smaller, so that the corresponding dividing effect under the reference window side length is better, and the opposite is true.
So far, the segmentation effect evaluation coefficients under the side lengths of all the reference windows are extracted.
Specifically, the side length of the reference window with the maximum evaluation coefficient of the segmentation effect is used as the side length of the target window; and taking each divided area under the side length of the target window as a target divided area, and further obtaining a plurality of target divided areas of the gray level image.
So far, a plurality of target segmentation areas of the gray level image are obtained through the method.
Step S003: extracting the possibility of each target segmentation area belonging to the damaged area; extracting the effectiveness of the Ojin threshold of each target segmentation area; clustering all the target segmentation areas to obtain a plurality of clusters, and screening all the clusters according to the effectiveness of the Ojin threshold to obtain target clusters; extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster; and obtaining all damage identification thresholds according to the degree of preference of the Ojin threshold.
1. And extracting the possibility of each target segmentation area belonging to the damaged area.
It should be noted that, compared with the gray image, the overall image structure and gray value distribution of each target division area are simpler, the target division areas can be better divided by using the oxford threshold value, and further, each target division area can effectively separate the foreground from the background, but for the insulating skin of the cable, each target division area does not contain broken flaws, so that not all the division thresholds are valid thresholds for the whole gray image, so that the validity of the threshold of each target division area needs to be judged, and then the damage threshold section corresponding to the cable skin is acquired by using the validity judgment result of the division thresholds of different target division areas, and further, the subsequent damage identification of the cable skin is performed.
Specifically, according to the gray distribution contrast condition between each target division area and the gray image, the possibility that each target division area belongs to a damaged area is extracted.
As an example, the calculation method for extracting the gray scale integral discrete degree of the gray scale image is as follows:
In the method, in the process of the invention, The gray level overall discrete degree of the gray level image is represented; /(I)Representing the average value of gray values of all pixel points in the gray image; /(I)Representing the total number of all pixel points in the gray scale image; /(I)Representing the/>, in a gray scale imageGray values of individual pixels.
For any one target segmentation area of the gray level image; as an example, the method for calculating the gray local discrete degree of the extraction target divided region is as follows:
In the method, in the process of the invention, Representing the gray level local discrete degree of the target segmentation area; /(I)Representing the average value of gray values of all pixel points in the target segmentation area; /(I)Representing the total number of all pixel points in the target segmentation area; /(I)Representing the/>, in the target partitionGray values of individual pixels.
As an example, the calculation method of the possibility of the target divided region belonging to the damaged region is:
In the method, in the process of the invention, Representing the possibility of the target divided region belonging to the damaged region; /(I)Representing the gray level local discrete degree of the target segmentation area; /(I)The gray scale of the gray scale image is expressed as a whole as discrete degree.
It should be noted that, although there is a case of skin breakage on the insulating skin of the cable, in the whole gray-scale image, the number of generally broken areas is smaller than that of normal areas, that is, in the gray-scale image, the normal areas are far larger than the broken areas, and the gray-scale value overall continuity of the pixel points corresponding to the normal areas is relatively high because the gray-scale values of the pixel points corresponding to the normal areas are relatively close; the continuity of the whole gray value of the damaged area is lower, because the gray value distribution of the pixel points corresponding to the damaged area is more discrete; therefore, in the embodiment, the gray level local discrete degree of the target division area is calculated and compared with the gray level integral discrete degree of the gray level image, and the larger the value is, the higher the possibility that the target division area is a cable breakage area is indicated, and the opposite is indicated; the local gray level discrete degree is obtained by quantifying the difference value between the gray level values of the pixel points, and dividing the gray level local discrete degree by the average value is constrained by the average value because the number of the pixel points of the target division area and the number of the pixel points of the gray level image have large differences.
To this end, the possibility of belonging to the damaged area is extracted for each target divided area.
2. And extracting the oxford threshold effectiveness of each target segmentation area.
Specifically, an oxford threshold value is obtained for all target segmentation areas of the gray image by using an oxford threshold value segmentation algorithm, and the oxford threshold value of each target segmentation area of the gray image is obtained; and extracting the oxford threshold value effectiveness of each target segmentation area according to the possibility of belonging to the damaged area and the local change parameters.
The oxford threshold segmentation algorithm is in the prior art, and the description of this embodiment is not repeated here.
For the firstThe target segmentation region, as an example, extract the/>The method for calculating the effectiveness of the Ojin threshold of each target segmentation area comprises the following steps:
In the method, in the process of the invention, Represents the/>The oxford threshold effectiveness of the individual target segmentation areas; /(I)Represents the/>The possibility of the target divided regions belonging to the damaged region; /(I)Represents the/>Local variation parameters of the individual target segmentation areas; Representing the total number of all target segmentation areas; /(I) Represents the/>An oxford threshold for each target segment; /(I)Representation except for the (th) >First/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value; representing an exponential function based on natural constants, the examples employ/> The model is used to present the inverse proportional relationship,For model input, the practitioner may choose the inverse proportion function according to the actual situation.
In the case of the first embodimentThe target divided region is a cable skin damaged region, and then the/>The Ojin threshold of each target divided region is a relatively important research region without considering the dividing effect, and is the first/>If the target divided region is a non-cable skin-broken region, then even the second/>The dividing effect of the individual target divided regions is good, and since the final object of the present embodiment is to perform breakage identification of the cable skin, the first/>The effectiveness of the oxford threshold corresponding to each target segmentation area is correspondingly poor; namely/>Possibility of belonging to damaged area and the first of the target divided areasThe larger the local variation parameter of each target segmentation area is, the/>The greater the effectiveness of the oxford threshold of the individual target segmentation areas; if/>The greater difference between the Ojin threshold of each target segmentation area and the Ojin threshold of other target segmentation areas indicates the/>The greater the likelihood that the oxford threshold of the individual target segment is the cable breakage segment threshold, i.e./>The greater the oxford threshold effectiveness of each target segmentation region.
Up to this point, the oxford threshold validity to each target division is extracted.
Specifically, clustering the oxford threshold effectiveness of all the target segmentation areas by using a self-adaptive k-means clustering algorithm to obtain a plurality of clustering clusters; taking the average value of the effectiveness of all the Ojin thresholds in each cluster as the first effectiveness average value of each cluster; and marking the cluster with the largest first validity mean as a target cluster.
The adaptive k-means clustering algorithm is an existing algorithm, and the embodiment is not described herein in detail.
Thus, the target cluster is obtained.
3. And extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster.
And extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster according to the difference of the oxford threshold value between each target segmentation area and other target segmentation areas in the target cluster.
For the first of the target clusterThe target segmentation areas, as an example, extract the/>, of the target clusterThe method for calculating the optimal degree of the Ojin threshold value of each target segmentation area comprises the following steps:
In the method, in the process of the invention, Representing the/>, in the target clusterThe degree of preference of the oxford threshold of each target segmentation area; /(I)Representing the total number of all target segmentation areas in the target cluster; /(I)Representing the/>, in the target clusterAn oxford threshold for each target segment; /(I)Representing the division number/>, in the target clusterFirst/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value.
The gray values corresponding to the broken surfaces of the cables show different levels after the broken surfaces are broken, but the gray values shown by the cables made of the same material have the similarity after the broken surfaces are broken, and the Ojin threshold is divided by the fixed gray value, so thatThe more similar the size of the oxford threshold of each target segmentation area is to the oxford threshold of other target segmentation areas in the target cluster, the/>The probability that the oxford threshold of each target divided region is a breakage threshold is high, and vice versa.
So far, extracting the optimal degree of the Ojin threshold value of each target segmentation area in the target cluster.
Presetting a preferred parameterWherein the present embodiment is described as/>To describe the example, the present embodiment is not particularly limited, wherein/>Depending on the particular implementation.
Specifically, for any one target segmentation region in the target cluster, if the degree of preference of the oxford threshold of the target segmentation region is less than or equal to a preference parameterAnd taking the Ojin threshold value of the target segmentation area as a breakage recognition threshold value.
So far, all damage recognition thresholds are obtained through the method.
Step S004: dividing the gray level image for a plurality of times according to the damage identification threshold value to obtain an integrally divided gray level image; and carrying out damage identification on the cable insulation skin to be detected according to the wholly segmented gray level image.
Specifically, all damage recognition thresholds are used as Ojin thresholds, and the grey-scale image is divided for a plurality of times by using an Ojin threshold dividing algorithm to obtain a plurality of divided grey-scale images; and carrying out image overlapping reconstruction on the plurality of segmented gray images by using an image overlapping algorithm to obtain an integrally segmented gray image, and taking a segmented region in the integrally segmented gray image as a damaged region of the insulating epidermis of the cable to be detected.
The image stacking algorithm is in the prior art, and the description of this embodiment is not repeated here.
Thus, the present embodiment is completed; referring to fig. 2, a characteristic relation flowchart of a method for identifying damage to an insulating skin of a cable is shown.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (6)
1. The method for identifying the damage of the insulating surface of the cable is characterized by comprising the following steps of:
capturing a gray image of an insulating surface skin of a cable to be detected;
acquiring a plurality of reference window side lengths, and acquiring a plurality of dividing areas of the gray level image under each reference window side length; extracting local variation parameters of each divided region under the side length of each reference window according to the gray value distribution condition of pixel points in each divided region of the gray image under the side length of each reference window; extracting a segmentation effect evaluation coefficient under the side length of each reference window according to the side length of each reference window and the local variation parameters of each segmentation area under the side length of the reference window; obtaining a plurality of target segmentation areas of the gray level image according to the segmentation effect evaluation coefficients;
Extracting the possibility of each target division area belonging to the damaged area according to the gray level distribution contrast condition between each target division area and the gray level image; acquiring an oxford threshold value of each target segmentation area; extracting the effectiveness of the Ojin threshold of each target segmentation area according to the possibility of belonging to the damaged area and the local change parameters; clustering all the target segmentation areas to obtain a plurality of clusters, and screening all the clusters according to the effectiveness of the Ojin threshold to obtain target clusters; extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster according to the difference of the oxford threshold value between each target segmentation area and other target segmentation areas in the target cluster; acquiring all damage identification thresholds according to the degree of preference of the Ojin threshold;
Dividing the gray level image for a plurality of times according to the damage identification threshold value to obtain an integrally divided gray level image; carrying out damage identification on the cable insulation skin to be detected according to the integrally segmented gray level image;
According to the gray value distribution condition of pixel points in each divided area of the gray image under the side length of each reference window, extracting the local variation parameters of each divided area under the side length of each reference window, comprising the following specific methods:
Will be the first Maximum value of inter-class variance of gray values of all pixel points in all divided regions under the side length of each reference window and/>/>, Under a side length of a reference windowThe ratio of the variances of the gray values of all the pixel points in the divided areas is taken as the first/>, Under a side length of a reference windowLocal variation factors of the individual segmented regions;
Will be the first Carrying out linear normalization on local change factors of all the divided areas under the side length of each reference window, and marking the normalized local change factors as local change parameters;
Extracting a segmentation effect evaluation coefficient under the side length of each reference window according to the side length of each reference window and the local variation parameters of each segmentation area under the side length of the reference window, wherein the specific method comprises the following steps:
Will be the first Side length of each reference window and/>The ratio of the total number of all the divided areas under the side length of each reference window is recorded as a first ratio; will/>The accumulated sum of the local variation parameters of all the divided areas under the side length of each reference window is recorded as a first accumulated sum; taking the product of the first ratio and the first accumulated sum as the first/>Dividing effect evaluation factors under the side length of each reference window;
carrying out linear normalization on the segmentation effect evaluation factors under the side length of all the reference windows, and marking the normalized segmentation effect evaluation factors as segmentation effect evaluation coefficients;
the method for extracting the possibility of each target division region belonging to the damaged region according to the gray level distribution contrast condition between each target division region and the gray level image comprises the following specific steps:
the calculation method for extracting the gray level integral discrete degree of the gray level image comprises the following steps:
In the method, in the process of the invention, The gray level overall discrete degree of the gray level image is represented; /(I)Representing the average value of gray values of all pixel points in the gray image; /(I)Representing the total number of all pixel points in the gray scale image; /(I)Representing the/>, in a gray scale imageGray values of the individual pixels;
for any one target segmentation area of the gray level image; the calculation method for extracting the gray level local discrete degree of the target segmentation area comprises the following steps:
In the method, in the process of the invention, Representing the gray level local discrete degree of the target segmentation area; /(I)Representing the average value of gray values of all pixel points in the target segmentation area; /(I)Representing the total number of all pixel points in the target segmentation area; /(I)Representing the/>, in the target partitionGray values of the individual pixels;
Taking the ratio of the gray level local discrete degree of the target divided area to the gray level integral discrete degree of the gray level image as the possibility of the target divided area belonging to the damaged area;
the specific formula for extracting the optimal degree of the oxford threshold value of each target segmentation area in the target cluster according to the difference of the oxford threshold value between each target segmentation area and other target segmentation areas in the target cluster is as follows:
In the method, in the process of the invention, Representing the/>, in the target clusterThe degree of preference of the oxford threshold of each target segmentation area; /(I)Representing the total number of all target segmentation areas in the target cluster; /(I)Representing the/>, in the target clusterAn oxford threshold for each target segment; /(I)Representing the division number/>, in the target clusterFirst/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value.
2. The method for identifying the damage of the insulating surface of the cable according to claim 1, wherein the steps of obtaining the side lengths of the plurality of reference windows and obtaining the plurality of divided areas of the gray image under the side length of each reference window comprise the following specific steps:
presetting two window side lengths To/>For the initial window side length, the step length is 1, the window side length is sequentially increased, the initial window side length and the window side length after each increase are recorded as the reference window side length until the reference window side length is/>When the increment is stopped, a plurality of reference window side lengths are obtained; for any one reference window side length/>Size/>, useThe step length of the sliding window of the cable insulation surface to be detected is/>And taking each sliding window as a division area under the side length of the reference window, and obtaining a plurality of division areas of the gray level image under the side length of the reference window.
3. The method for identifying the damage of the insulating surface of the cable according to claim 1, wherein the obtaining a plurality of target division areas of the gray level image according to the division effect evaluation coefficient comprises the following specific steps:
Taking the side length of the reference window with the maximum segmentation effect evaluation coefficient as the side length of the target window; and taking each divided area under the side length of the target window as a target divided area.
4. The method for identifying damage to an insulating surface of a cable according to claim 1, wherein the specific formula for extracting the validity of the oxford threshold value of each target division area according to the probability of belonging to the damaged area and the local variation parameter is as follows:
In the method, in the process of the invention, Represents the/>The oxford threshold effectiveness of the individual target segmentation areas; /(I)Represents the/>The possibility of the target divided regions belonging to the damaged region; /(I)Represents the/>Local variation parameters of the individual target segmentation areas; /(I)Representing the total number of all target segmentation areas; /(I)Represents the/>An oxford threshold for each target segment; /(I)Representation except the firstFirst/>, outside the individual target segmentation regionsAn oxford threshold for each target segment; /(I)The representation takes absolute value; /(I)An exponential function based on a natural constant is represented.
5. The method for identifying the damage of the cable insulation surface according to claim 1, wherein the clustering is performed on all the target segmentation areas to obtain a plurality of clusters, and the target clusters are obtained by screening all the clusters according to the effectiveness of the Ojin threshold, comprising the following specific steps:
Clustering the effectiveness of the Ojin threshold values of all the target segmentation areas by using a self-adaptive k-means clustering algorithm to obtain a plurality of clustering clusters; taking the average value of the effectiveness of all the Ojin thresholds in each cluster as the first effectiveness average value of each cluster; and marking the cluster with the largest first validity mean as a target cluster.
6. The method for identifying damage to the insulating surface of the cable according to claim 1, wherein the step of obtaining all damage identification thresholds according to the degree of preference of the oxford threshold comprises the following specific steps:
presetting a preferred parameter For any one target segmentation area in the target cluster, if the degree of preference of the oxford threshold of the target segmentation area is smaller than or equal to a preference parameter/>And taking the Ojin threshold value of the target segmentation area as a breakage recognition threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410438979.4A CN118037735B (en) | 2024-04-12 | 2024-04-12 | Cable insulation skin damage identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410438979.4A CN118037735B (en) | 2024-04-12 | 2024-04-12 | Cable insulation skin damage identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118037735A CN118037735A (en) | 2024-05-14 |
| CN118037735B true CN118037735B (en) | 2024-06-25 |
Family
ID=90984440
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410438979.4A Active CN118037735B (en) | 2024-04-12 | 2024-04-12 | Cable insulation skin damage identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118037735B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20210109134A (en) * | 2020-02-27 | 2021-09-06 | 사단법인 빛가람행복나눔복지회 | protection cover for underground cable and manufacturing method thereof |
| CN115825075A (en) * | 2022-11-03 | 2023-03-21 | 中核检修有限公司 | Cable breakage detection method based on infrared signals |
-
2024
- 2024-04-12 CN CN202410438979.4A patent/CN118037735B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20210109134A (en) * | 2020-02-27 | 2021-09-06 | 사단법인 빛가람행복나눔복지회 | protection cover for underground cable and manufacturing method thereof |
| CN115825075A (en) * | 2022-11-03 | 2023-03-21 | 中核检修有限公司 | Cable breakage detection method based on infrared signals |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118037735A (en) | 2024-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114140462A (en) | Bearing wear degree evaluation method based on image processing | |
| CN107657209B (en) | Template image registration mechanism based on finger vein image quality | |
| CN108734704B (en) | Transmission conductor strand breakage detection method based on gray variance normalization | |
| CN107481374B (en) | Intelligent terminal fingerprint unblock door opener | |
| CN115619793B (en) | Power adapter appearance quality detection method based on computer vision | |
| CN116385450B (en) | PS sheet wear resistance detection method based on image processing | |
| CN116309600B (en) | Environment-friendly textile quality detection method based on image processing | |
| CN117422712B (en) | Plastic master batch visual detection method and system based on image filtering processing | |
| CN116092013A (en) | Dangerous road condition identification method for intelligent monitoring | |
| CN116385472B (en) | Hardware stamping part deburring effect evaluation method | |
| CN115511907B (en) | Scratch detection method for LED screen | |
| CN117522719B (en) | Bronchoscope image auxiliary optimization system based on machine learning | |
| CN112819784A (en) | Method and system for detecting broken strands and scattered strands of wires of distribution line | |
| CN117541582B (en) | IGBT insulation quality detection method for high-frequency converter | |
| CN114881960A (en) | Feature enhancement-based cloth linear defect detection method and system | |
| CN111008967B (en) | Insulator RTV coating defect identification method | |
| CN116958182B (en) | Quick concrete crack detection method based on image data | |
| CN117274820B (en) | Map data acquisition method and system for mapping geographic information | |
| CN117974637A (en) | Intelligent detection method and system for guide wheel crack defects based on image features | |
| CN118037735B (en) | Cable insulation skin damage identification method | |
| CN116777845B (en) | Building site safety risk intelligent assessment method and system based on artificial intelligence | |
| CN111898464A (en) | Face recognition method based on 5G network | |
| CN114119577B (en) | High-speed rail tunnel leakage cable buckle detection method | |
| CN118150585B (en) | Crimping terminal detection system and method thereof | |
| CN113538369B (en) | Pressing plate state detection algorithm based on ellipse measurement learning |
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 |