CN111260616A - Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization - Google Patents
Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization Download PDFInfo
- Publication number
- CN111260616A CN111260616A CN202010032660.3A CN202010032660A CN111260616A CN 111260616 A CN111260616 A CN 111260616A CN 202010032660 A CN202010032660 A CN 202010032660A CN 111260616 A CN111260616 A CN 111260616A
- Authority
- CN
- China
- Prior art keywords
- image
- insulator
- dimensional
- noise
- edge
- 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
- 239000012212 insulator Substances 0.000 title claims abstract description 108
- 238000001514 detection method Methods 0.000 title claims abstract description 29
- 238000005457 optimization Methods 0.000 title claims abstract description 17
- 230000011218 segmentation Effects 0.000 title claims abstract description 15
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 30
- 238000001914 filtration Methods 0.000 claims abstract description 17
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 238000005516 engineering process Methods 0.000 claims abstract description 9
- 230000003313 weakening effect Effects 0.000 claims abstract description 5
- 238000013507 mapping Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 12
- 230000002708 enhancing effect Effects 0.000 claims description 9
- 230000000694 effects Effects 0.000 claims description 8
- 238000003708 edge detection Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000003321 amplification Effects 0.000 claims description 5
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 5
- 238000001228 spectrum Methods 0.000 claims description 5
- 230000001629 suppression Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000009499 grossing Methods 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 239000006185 dispersion Substances 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 230000009977 dual effect Effects 0.000 claims description 3
- 238000012886 linear function Methods 0.000 claims description 3
- 238000005728 strengthening Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 2
- 238000010276 construction Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 claims description 2
- 238000003706 image smoothing Methods 0.000 claims description 2
- 238000007689 inspection Methods 0.000 abstract description 15
- 238000012423 maintenance Methods 0.000 abstract description 2
- 230000005540 biological transmission Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000005764 inhibitory process Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 229910052573 porcelain Inorganic materials 0.000 description 2
- 239000003086 colorant Substances 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000000875 corresponding effect Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
Images
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
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
Abstract
An insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization comprises the following steps: and shooting the insulator on the overhead line, and inputting the shot insulator picture into a computer. The method comprises the steps of adopting a computer image processing technology to carry out preprocessing on a shot insulator image, wherein the preprocessing comprises the steps of carrying out background weakening, graying, contrast enhancement, filtering and the like on the shot insulator image, so that a target insulator is more prominent in the image, the influence of noise on the image is reduced, and the image quality is optimized. And finally, edge optimization is carried out on the insulator by Canny operator two-dimensional threshold segmentation, so that the influence of noise is reduced, and the extracted insulator edge detail features are clearer. And identifying whether the insulator has cracks or not. The invention can be widely applied to the inspection of the overhead line of the power grid, assists the field inspection and maintenance operation of the overhead line, reduces the labor intensity of inspection personnel, reduces the dangerous operation in the operation of the power grid and avoids the economic loss.
Description
Technical Field
The invention belongs to the technical field of insulator fault detection of a power transmission line, and particularly relates to an insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization.
Background
The insulator is widely applied to a power transmission line, is exposed in the atmosphere for a long time, works in severe environments such as a strong electric field, strong mechanical stress, wind, rain, snow, fog and the like, and cracks are generated on the insulator in an unavoidable ground. The insulator porcelain body cracks to reduce the insulating strength of the insulator, and can lead to porcelain bottle breakage even cause line faults to cause the action of a protection device in serious conditions, thereby causing the unplanned shutdown of power equipment. Therefore, fault detection, particularly crack detection, of the insulator is important. If the insulator cracks can be detected early, many power system faults will be reduced or avoided.
In recent years, computer monitoring systems have become a hotspot in the field of automation control, and have the advantages of strong functions, convenient management, good safety and real-time performance, and no need of excessive human intervention, so that the computer monitoring systems are more and more emphasized by various industries. With the continuous improvement of electric power informatization and automation level, the insulator crack detection has a new breakthrough. The computer image processing technology can extract the insulator from the complex background of the aerial photography insulator image and detect the crack of the insulator, and has the advantages of simple operation and low cost. However, the problem that the background of the aerial image is complex, and the edge detection result has false edges and a large amount of noise exists, so that the accurate detection of the edge of the insulator from the complex background becomes a difficult point of work.
Disclosure of Invention
In order to solve the technical problem, the invention provides an insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization, which can extract the edge of an insulator from an insulator image with a complex background and detect whether the insulator has cracks. The method is used for achieving the purposes of weakening the background, removing the false target and enabling the extracted detailed features of the target to be clearer in the insulator image with the complex background.
The technical scheme adopted by the invention is as follows:
an insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization comprises the following steps:
step 1: and shooting the insulator on the overhead line in a helicopter inspection mode, and inputting the shot insulator picture into a computer.
Step 2: the method comprises the steps of adopting a computer image processing technology to carry out preprocessing on a shot insulator image, wherein the preprocessing comprises the steps of carrying out background weakening, graying, contrast enhancement, filtering and the like on the shot insulator image, so that a target insulator is more prominent in the image, the influence of noise on the image is reduced, and the image quality is optimized.
And step 3: and finally, edge optimization is carried out on the insulator by Canny operator two-dimensional threshold segmentation, so that the influence of noise is reduced, and the extracted insulator edge detail features are clearer.
And 4, step 4: and identifying whether the insulator has cracks or not.
In the step 2, the insulator of the overhead transmission line is used for preventing a live part of the transmission line from forming a grounding channel, and is an important part of the overhead transmission line. The method for extracting the edge of the insulator from the aerial image by using the computer digital image processing technology is a good method for detecting the defect of the insulator. However, because the background in the aerial image is complex, a plurality of background objects similar to the characteristics of the insulator exist, and noise generated in the transmission process of the image influences the accurate detection of the edge of the insulator, the method firstly preprocesses the shot insulator image sample.
Characteristic analysis of aerial insulating sub-images:
currently, before processing an image, the image is generally mapped to an RGB color space or an HSI color space for analysis. Wherein, the RGB color space describes the pixel of any point in the image as R (red), G (green), B (blue) three characteristic quantities, the values of which are all [0,255], and the RGB color space represents the intensity of the three colors at the pixel (i, j); the RGB color space is described by three feature quantities, I (brightness), S (saturation), and H (chroma). The RGB color space and the RGB color space have a conversion relationship as shown in the following formula (1):
according to the method, a large number of insulator aerial images are mapped in RGB color space and HSI color space, and the result shows that in the aerial images, the insulator body and the background image have large difference in color saturation. Therefore, after the aerial image is mapped to the HSI color space, the S component, namely the saturation is enhanced, and then the graying processing is carried out, so that the edge of the insulator can be strengthened, and the subsequent image processing and noise weakening are facilitated.
After the image is mapped to the HSI color space, the insulator S component image extracted from the HSI color space weakens the background and highlights the insulator. But most of the color information contained in the image is useless information and occupies a large amount of calculation time for subsequent image processing, which is a problem to be solved firstly in the preprocessing stage; secondly, the contrast of the edge of the outline of the insulator may be poor due to the shooting of light rays and the like; meanwhile, noise is also a problem to be considered in an important way in the image preprocessing stage. Aiming at the problems, firstly, graying processing of an image is required, a color image is converted into a grayscale image, and the image is subjected to processing such as denoising and contrast enhancement, so that the calculation speed is increased and the contour edge of the insulator is more obvious. Through analysis, the preprocessing step of the aerial insulator image is divided into the following four steps: conversion to HSI color space, image graying, contrast enhancement and image denoising.
Step 2.1, mapping the HSI color space: firstly, representing an original image as digital description in an RGB space, and then completing the mapping from the RGB color space of the aerial image to an HSI color space through a formula (1);
step 2.2, extracting an S component diagram in the HSI color space and carrying out gray processing: after obtaining the HSI color space mapping result of the image, the weight of the S characteristic quantity is increased in the gray processing process, so that the purpose of strengthening the insulator body image is achieved, and the mapping relation of the gray processing is shown as the following formula (2):
gray(i,j)＝0×H+1×S+0×I (2)；
wherein: gray (i, j) represents a gray value at a point (i, j) in the gray image;
in the HSI color space: h represents a hue; s represents color saturation; i denotes brightness.
Step 2.3, enhancing the contrast of the gray level image: performing linear expansion on each pixel in the obtained insulator image by adopting a linear function, expanding the pixels, performing gray level stretching, and enhancing the image contrast; after graying and enhancing the contrast ratio of the original image, the edge outline of the insulator becomes clearer;
step 2.4, denoising treatment based on wiener filtering:
common filters are also gaussian, mean, etc. Both gaussian filtering and mean filtering are commonly used filters in image noise processing; gaussian noise is common noise in image processing, and Gaussian noise processing by adopting Gaussian filtering has a good effect, but noise generated in an aerial image is a result of comprehensive action of multiple factors, the noise in the aerial image is not subjected to Gaussian distribution, and the precision of the Gaussian filtering may be influenced by the influence of the aerial image and the multiple factors. The average filtering is to perform an average operation on surrounding pixels, the amplitudes are approximately equal and are randomly distributed on different positions, so that an image can be smoothed, the speed is high, the algorithm is simple, but noise cannot be removed, and the noise can be weakened only weakly. Such filters work poorly in aerial images with complex backgrounds. Through analysis and comparison of several filters, a wiener filter is finally adopted.
Wiener filtering is a least variance filter that is based on making f (x, y) and f (x, y) sum on the assumption that the image signal can be approximately seen as a stationary random process
The mean square error between the two is minimized, namely:
in the formula: f (x, y) represents an input signal, i.e., an original image;
an estimate representing the output signal, i.e. the non-degraded image; e denotes a parameter expectation value.
Into the frequency domain, there is a lagrange function:
in the formula: p is the Fourier transform of the Laplace operator; n is the Fourier transform of the noise;
represents an estimate of the non-degraded image, H (u, v) represents the transfer function; sn(u, v) represents a power spectrum of the noise; sf(u, v) represents the power spectrum of the undegraded image; g (u, v) represents the fourier transform of the degraded image; γ represents a parameter of the lagrangian function.
The wiener filter has an automatic suppression effect on noise amplification, if the transfer function H (u, v) is zero at a certain position, S exists at the position of a denominatorn(u,v)/Sf(u, v), so no singularity occurs, in a certain spectral region, if the signal-to-noise ratio is high, i.e. the signal-to-noise ratio is high
Sn(u,v)≤Sf(u,v) (6)；
The effect of the filter tends to be inverse filtering if:
Sn(u,v)>>Sf(u,v) (7)；
i.e. the signal-to-noise ratio is small, the performance of the filter is insensitive, which shows that the wiener filter avoids the amplification effect on the noise in the process of restoring the image. The noise is effectively filtered by the image after wiener filtering, and the contrast between the foreground and the background becomes more obvious and outstanding, so that the difference between the insulator and the background can be highlighted.
The step 3 comprises the following steps:
step 3.1, Canny operator algorithm construction:
the differential operator is a traditional edge detection method and is also the most common edge detection method. The edge is detected by using a first-order differential local extremum method, because at the edge, the first-order differential has an extremum point, and the local extremum of the first-order differential corresponds to the image edge, but the noise signal easily affects the detection of the weak edge of the image. The Canny operator can well solve the defect, is a dual-threshold detection method and has good effect on the detection of the strong edge and the weak edge of the color band.
Let the original input image be f (x, y), firstly, a gaussian function is used for smoothing operation, that is, the gradient of g (x, y) after smoothing is:
in formula (8): g (x, y) is a Gaussian function; f (x, y) is an input image;
is a gradient vector.
The image smoothing processing by adopting the Gaussian function can lead the edge of the original image to be blurred and the width to be increased, and a non-maximum value inhibition technology is introduced to sharpen the blurred edge.
Let a two-dimensional Gaussian filter function be
Decomposing the two-dimensional Gaussian filter function to obtain the gradient vector
The two filter convolution templates of (a) are decomposed into two one-dimensional row and column filters:
and (3) performing convolution calculation on the two convolution templates respectively in the image f (x, y) to obtain output:
in the formulas (10) and (11),
representing that the partial derivatives of x and y are respectively solved for a two-dimensional Gaussian function; ex(x, y) and Ex(x, y) represent convolution values in the x and y directions, respectively.
In the formula (12), a (x, y) is represented as the amplitude of the gradient, α (x, y) is represented as the direction of the gradient, the size of the edge intensity a (x, y) value at the (x, y) point on the image cannot determine whether the point is an edge point, the ridge zone in the amplitude image needs to be refined, and the point with the largest local amplitude change is reserved.
In the shot insulator image, the dual threshold method can remove these false edges due to the presence of noise and false targets. Selection of threshold value t by double threshold method1And t2As a dual threshold, and t2＝2t1，g1(x, y) and g2(x, y) the two dual-threshold edge images are obtained.
Step 3.2, selecting two-dimensional threshold values based on OTSU:
the original Canny algorithm does not give a threshold value selection mode, and the high threshold value and the low threshold value are not determined by characteristic information of image edges and are set respectively according to different situations. Aiming at the defects that a large amount of noise and false targets exist when the insulator edge is extracted from an aerial insulator image: when thresholding is carried out, the selection of high and low thresholds of a Canny operator is improved, the dimensionality is increased by a one-dimensional maximum inter-class variance method, the field average gray level of pixel points is increased to form a binary group, and the method specifically comprises the following steps:
an M × N gray image is obtained by dividing the average gray level of the pixel into L levels, and calculating the average gray level of the pixel area at each pixel point, thereby forming a binary group: the gray value of a pixel point and the average gray value of the field thereof are set as fijRespective joint probability densities p can be definedijComprises the following steps:
pij＝fij/N,i,j＝1,2,...,L (13)；
in formula (13): n is the number of pixel points of the image; (i, j) represents a certain pixel point.
Suppose there are two classes c in a two-dimensional histogram0And c1Which respectively represent an object and a background, have two different probability density function distributions, and set the threshold value as (s, t), then the probabilities of the two types of occurrences are respectively
In the formulae (14) and (15), w0Representing the probability of the occurrence of the target object; w is a1Indicating the probability of the background occurring.
Two classes of corresponding mean vectors;
in the formula (16), u0Representing a mean vector of the target object in the two-dimensional histogram; u. of1Representing the mean vector of the background in a two-dimensional histogram.
Total mean vector on two-dimensional histogram:
and has the following components:
w0+w1≈1,μT≈w0μ0+w1μ1(18)；
using tr σBAs a measure of the inter-class variance, then：
Optimum threshold value(s)*,t*) Comprises the following steps:
(s*,t*)＝argmax{trσB(s,t)} (20)；
according to a two-dimensional OTSU algorithm, the invention provides that an average gray scale-local variance two-dimensional histogram is used for carrying out edge optimization on an insulator, and the two-dimensional algorithm can effectively improve the influence degree of noise during edge detection;
selecting the area of the pixel point, using the area average gray level g (x, y) as the abscissa, and the local variance σ of the k × k area of (x, y)2(x, y) is a vertical coordinate, a new two-dimensional histogram is constructed, the local variance reflects the average difference degree of each pixel point of the gray level image, and the local variance is an index for describing the dispersion degree of each pixel point;
Cijrepresenting the frequency of occurrence of the mean gray level-local variance doublet, pijRepresenting a joint probability;
in the formula (21), M × N represents the image pixel size, the width is M, and the height is N;
in the formula (22), σ2(x, y) represents the domain local variance of the pixel; k is the selected local size of the domain.
The smaller the neighborhood local variance of a pixel point, the smaller the deviation difference from the selected center pixel point. In general, the gray level of the target area is highly correlated with the gray level inside the gray level of the background area, the local variance inside the target area and the background area is small, and the local variance of the edge area, the noise area and the texture area is large, so that the area division effectively reduces wrong division. The local variance more fully considers the discrete degree of the central pixel point and the k multiplied by k neighborhood pixel point relative to the gradient and the average value, and can better improve the phenomenon that the noise point is wrongly divided into a target region under the influence of high-intensity noise, thereby effectively removing the noise, more clearly distinguishing the target point and the background point and enabling the division result to be more accurate.
In the step 4, after the aerial insulator image is subjected to edge extraction, connection and refinement, the edge of the insulator is well extracted, and finally whether the insulator has cracks or not is identified. The insulator bottle body is usually provided with a narrow chain isolated from the image edge, the cracks comprise three types of transverse cracks, longitudinal cracks and oblique cracks, but in practice, completely horizontal or vertical cracks do not exist, and the cracks are generally longitudinally and obliquely inclined lines along the bottle body. Therefore, the insulator image is scanned in a certain direction by adopting the second-order spline biorthogonal wavelet transform of the Petrou slope model, and whether the insulator has cracks or not can be identified.
The insulator edge was characterized with a Petrou slope model:
and scanning different rows column by column respectively by adopting a Petrou slope model, recording results, and forming a crack if the scanned points which do not meet the slope model are communicated.
The invention discloses an insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization, which has the following technical effects:
1: the threshold selection of the traditional Canny operator is improved, and the threshold processing is carried out on the preprocessed image by adopting a two-dimensional histogram gray scale-local variance method, the method utilizes the domain average gray scale information of the pixels, considers the dispersion degree of each pixel point and the central pixel point data, can effectively reduce the noise at the edge, and enables the extracted target crack characteristics to be clearer.
2: the method provides an insulator image crack detection method finished by a computer image processing technology, can be widely applied to power grid overhead line inspection, assists overhead line field inspection and maintenance operation, reduces labor intensity of inspection personnel, reduces dangerous operation in power grid operation, and avoids economic loss.
3: when the invention is used for power grid line inspection, a quick and effective method can be provided for identifying whether the insulator has a fault, so that the power grid inspection efficiency and the fault detection accuracy are improved, the labor intensity of power grid inspection personnel is reduced, dangerous operation in power grid operation is reduced, and economic loss is avoided.
Drawings
FIG. 1 is a flow chart of insulator image crack detection according to the present invention.
FIG. 2 is a flow chart of image preprocessing according to the present invention.
Fig. 3 is a flow chart of the steps of the improved Canny algorithm of the present invention.
FIG. 4 is a flow chart of crack detection according to the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The embodiment of the invention discloses an insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization, which comprises the following implementation steps:
referring to fig. 1, a helicopter is used to patrol the field power grid overhead line, and the insulators on the overhead line are shot and sampled. Due to the complex field environment, the appearance of the whole insulator needs to be shot when the insulator is shot, and the picture quality is ensured; after the helicopter inspection task is completed, inputting the shot insulator picture into a computer image processing system, screening the picture, deleting the picture without shooting the complete insulator and with a fuzzy shooting effect, and avoiding influencing a subsequent detection result; and the screened pictures are compressed, so that the speed of processing the pictures by a computer is higher.
Referring to fig. 2, the insulator image is preprocessed, which specifically includes the steps of:
the method comprises the following steps: HSI color space mapping. Firstly, representing an original image as digital description in an RGB space, and then completing mapping from the RGB space of the aerial image to an HSI space through a formula (1);
step two: and extracting an S component map in the HSI space and carrying out gray processing. After the HSI color space mapping result of the image is obtained, the weight of the S characteristic quantity is increased in the gray processing process, so that the purpose of strengthening the insulator body image is achieved. The mapping relation of the graying processing is as shown in formula (2);
step three: enhancing the gray scale image contrast. Performing linear expansion on each pixel in the obtained insulator image by adopting a linear function, expanding the pixels for gray stretching, and enhancing the image contrast so as to enable the edge outline of the insulator to be clearer;
step four: and denoising based on wiener filtering. The wiener filter adopts the minimum variance function to process the mean square error among pixels, which is shown in the formula (3), the formula (4) and the formula (5), so that the wiener filter has an automatic suppression effect on noise amplification, the signal to noise ratio is reduced, the noise is effectively filtered by the image after the wiener filter, and the contrast between the foreground and the background becomes more obvious and outstanding, so that the difference between an insulator and the background can be highlighted.
Referring to fig. 3, the edge detection is performed on the preprocessed image by using a modified Canny algorithm, and the specific contents include:
step S1: the image is smoothed using a gaussian filter. The image is smoothed by adopting a Gaussian function, which is shown in a formula (8), so that Gaussian noise points of the image can be effectively filtered, and the edge of the original image is blurred and the width of the original image is increased.
Step S2: the magnitude and direction of the gradient is calculated. The magnitude and direction of the gradient are calculated by using the finite difference of the first order partial derivatives, an image containing derivatives in the horizontal and vertical directions is obtained, and then the gradient direction and magnitude are correspondingly solved according to the formula (11) and the formula (12) for each pixel on the two images.
Step S3: non-maximum suppression is performed on the gradient amplitudes. The edge cannot be determined by obtaining the global gradient, so a non-maximum value inhibition method is adopted to search the point of the local pixel maximum value, remove the non-edge point and inhibit the non-maximum value.
Step S4: and detecting, connecting and refining edges by a double-threshold method. Two thresholds are acted on the image subjected to non-maximum value inhibition, the selection of the thresholds is shown in formula (20), the two thresholds are defined as 1 and 2 after action, most of noise is removed from the image 2 detected by the larger threshold, but a lot of useful edge information is lost, more useful edge information is retained in the image 1 detected by the smaller threshold, and on the basis, the lost information in the image 2 is supplemented, and the image edge is connected.
Referring to fig. 4, the specific steps of the insulator crack detection are as follows:
and (3) scanning the extracted insulator edge image by adopting second-order spline biorthogonal wavelet transform of a Petrou slope model according to the formula (23), scanning different rows one by one and recording results, and forming a crack for the scanned points which do not meet the slope model if the scanned points are communicated, so that whether the insulator cracks or not can be detected.
In conclusion, the invention can effectively detect whether the insulator on the overhead line has faults or not by adopting the computer image processing technology to carry out edge extraction and crack detection on the shot insulator image based on the insulator image shot by the helicopter in the inspection. The method has the main functions that when the power grid line inspection is carried out, whether the insulator breaks down or not can be identified quickly and effectively, so that the power grid inspection efficiency and the fault detection accuracy are improved, the labor intensity of power grid inspection personnel is reduced, dangerous operation in power grid operation is reduced, and economic loss is avoided.
Claims (4)
1. An insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization is characterized by comprising the following steps:
step 1: shooting an insulator on an overhead line, and inputting a shot insulator picture into a computer;
step 2: preprocessing a shot insulator image by adopting a computer image processing technology, wherein the preprocessing comprises the steps of weakening the background, graying, enhancing the contrast, filtering and the like on the shot insulator image, so that a target insulator is more prominent in the image, and the influence of noise on the image is reduced;
and step 3: finally, performing edge optimization on the insulator by Canny operator two-dimensional threshold segmentation;
and 4, step 4: and identifying whether the insulator has cracks or not.
2. The insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization according to claim 1, characterized in that: the step 2 comprises the following steps:
step 2.1, mapping the HSI color space: firstly, representing an original image as digital description in an RGB space, and then completing the mapping from the RGB color space of the aerial image to an HSI color space through a formula (1);
the HSI color space and the RGB color space have a conversion relationship as shown in the following formula (1):
step 2.2, extracting an S component diagram in the HSI color space and carrying out gray processing: after obtaining the HSI color space mapping result of the image, the weight of the S characteristic quantity is increased in the gray processing process, so that the purpose of strengthening the insulator body image is achieved, and the mapping relation of the gray processing is shown as the following formula (2):
gray(i,j)＝0×H+1×S+0×I (2)；
wherein: gray (i, j) represents a gray value at a point (i, j) in the gray image;
in the HSI color space: h represents a hue; s represents color saturation; i represents brightness;
step 2.3, enhancing the contrast of the gray level image: performing linear expansion on each pixel in the obtained insulator image by adopting a linear function, expanding the pixels, performing gray level stretching, and enhancing the image contrast; after graying and enhancing the contrast ratio of the original image, the edge outline of the insulator becomes clearer;
step (ii) of2.4, denoising treatment based on wiener filtering: wiener filtering is a least variance filter that is based on making f (x, y) and f (x, y) sum on the assumption that the image signal can be approximately seen as a stationary random process
The mean square error between the two is minimized, namely:
in the formula: f (x, y) represents an input signal, i.e., an original image;
an estimate representing the output signal, i.e. the non-degraded image; e represents a parameter expected value;
into the frequency domain, there is a lagrange function:
in the formula: p is the Fourier transform of the Laplace operator; n is the Fourier transform of the noise;
represents an estimate of the non-degraded image, H (u, v) represents the transfer function; sn(u, v) represents a power spectrum of the noise; sf(u, v) represents the power spectrum of the undegraded image; g (u, v) represents the fourier transform of the degraded image; gamma represents a parameter of a lagrange function;
the wiener filter has an automatic suppression effect on noise amplification, if the transfer function H (u, v) is zero at a certain position, S exists at the position of a denominatorn(u,v)/Sf(u, v) so that it does not go outAt the singular point, in a certain frequency spectrum region, if the signal-to-noise ratio is high, the singular point is
Sn(u,v)≤Sf(u,v) (6)；
The effect of the filter tends to be inverse filtering if:
Sn(u,v)>>Sf(u,v) (7)；
the signal-to-noise ratio is small, and the performance of the filter is insensitive, so that the wiener filter avoids the amplification effect on noise in the process of recovering the image; the noise is effectively filtered by the image after the wiener filtering.
3. The insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization according to claim 1, characterized in that: the step 3 comprises the following steps:
step 3.1, Canny operator algorithm construction:
let the original input image be f (x, y), firstly, a gaussian function is used for smoothing operation, that is, the gradient of g (x, y) after smoothing is:
in formula (8): g (x, y) is a Gaussian function; f (x, y) is an input image;
is a gradient vector;
the image smoothing processing of the Gaussian function is adopted, the edge of the original image is blurred, the width of the original image is increased, and a non-maximum suppression technology is introduced to sharpen the blurred edge;
let the two-dimensional gaussian filter function be:
decomposing the two-dimensional Gaussian filter function to obtain the gradient vector
The two filter convolution templates of (a) are decomposed into two one-dimensional row and column filters:
and (3) performing convolution calculation on the two convolution templates respectively in the image f (x, y) to obtain output:
in the formulas (10) and (11),
representing that the partial derivatives of x and y are respectively solved for a two-dimensional Gaussian function; ex(x, y) and Ex(x, y) represent convolution values in x and y directions, respectively;
α (x, y) represents the direction of the gradient, the size of the edge strength A (x, y) value at the (x, y) point on the image cannot determine whether the point is an edge point, the ridge zone in the amplitude image needs to be refined, and the point with the maximum local amplitude change is reserved;
in the shot insulator image, due to the existence of noise and false targets, the double-threshold method can remove the false edges; selection of threshold value t by double threshold method1And t2As a dual threshold, and t2＝2t1，g1(x, y) and g2(x, y) the two dual-threshold edge images are obtained;
step 3.2, selecting two-dimensional threshold values based on OTSU:
when thresholding is carried out, the selection of high and low thresholds of a Canny operator is improved, the dimensionality is increased by a one-dimensional maximum inter-class variance method, the field average gray level of pixel points is increased to form a binary group, and the method specifically comprises the following steps:
an M × N gray image is obtained by dividing the average gray level of the pixel into L levels, and calculating the average gray level of the pixel area at each pixel point, thereby forming a binary group: the gray value of a pixel point and the average gray value of the field thereof are set as fijRespective joint probability densities p can be definedijComprises the following steps:
pij＝fij/N,i,j＝1,2,...,L (13)；
in formula (13): n is the number of pixel points of the image; (i, j) represents a certain pixel point;
setting in a two-dimensional histogram there are two classes c0And, c1They represent the object and the background, respectively, with two different probability density function distributions; assuming that the threshold is (s, t), the probabilities of two types of occurrences are:
in the formulae (14) and (15), w0Representing the probability of the occurrence of the target object; w is a1Representing the probability of the background occurring;
two classes of corresponding mean vectors;
in the formula (16), u0Representing a mean vector of the target object in the two-dimensional histogram; u. of1A mean vector representing the background in a two-dimensional histogram;
total mean vector on two-dimensional histogram:
and has the following components: w is a0+w1≈1,μT≈w0μ0+w1μ1(18)；
Using tr σBAs a measure of the inter-class variance, there are:
optimum threshold value(s)*,t*) Comprises the following steps:
(s*,t*)＝argmax{trσB(s,t)} (20)；
according to a two-dimensional OTSU algorithm, a mean gray scale-local variance two-dimensional histogram is proposed to carry out edge optimization on the insulator, and the two-dimensional algorithm can effectively improve the influence degree of noise during edge detection;
selecting the area of the pixel point, using the area average gray level g (x, y) as the abscissa, and the local variance σ of the k × k area of (x, y)2(x, y) is a vertical coordinate, a new two-dimensional histogram is constructed, the local variance reflects the average difference degree of each pixel point of the gray level image, and the local variance is an index for describing the dispersion degree of each pixel point;
Cijrepresenting the frequency of occurrence of the mean gray level-local variance doublet, pijRepresenting a joint probability;
in the formula (21), M × N represents the image pixel size, the width is M, and the height is N;
in the formula (22), σ2(x, y) represents the domain local variance of the pixel; k is the local size of the selected field;
the smaller the neighborhood local variance of a pixel point, the smaller the deviation difference from the selected center pixel point.
4. The insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization according to claim 1, characterized in that: in the step 4, scanning the insulator image in a certain direction by adopting second-order spline biorthogonal wavelet transform of a Petrou slope model, so as to identify whether the insulator has cracks or not;
the insulator edge was characterized with a Petrou slope model:
and scanning different rows column by column respectively by adopting a Petrou slope model, recording results, and forming a crack if the scanned points which do not meet the slope model are communicated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010032660.3A CN111260616A (en) | 2020-01-13 | 2020-01-13 | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010032660.3A CN111260616A (en) | 2020-01-13 | 2020-01-13 | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111260616A true CN111260616A (en) | 2020-06-09 |
Family
ID=70954097
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010032660.3A Pending CN111260616A (en) | 2020-01-13 | 2020-01-13 | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111260616A (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070717A (en) * | 2020-08-05 | 2020-12-11 | 三峡大学 | Power transmission line icing thickness detection method based on image processing |
| CN112150500A (en) * | 2020-09-17 | 2020-12-29 | 西安工程大学 | Insulator segmentation and extraction method based on joint component graying |
| CN112164060A (en) * | 2020-10-22 | 2021-01-01 | 江苏科技大学 | Diamond sawing wire fracture prediction method based on image processing |
| CN112347890A (en) * | 2020-10-30 | 2021-02-09 | 武汉理工大学 | Insulator robot operation identification method, storage medium and system |
| CN112669286A (en) * | 2020-12-29 | 2021-04-16 | 北京建筑材料检验研究院有限公司 | Infrared thermal image-based method for identifying defects and evaluating damage degree of external thermal insulation system of external wall |
| CN112686120A (en) * | 2020-12-25 | 2021-04-20 | 国家电网有限公司 | Power transmission line anomaly detection method based on unmanned aerial vehicle aerial image |
| CN112907460A (en) * | 2021-01-25 | 2021-06-04 | 宁波市鄞州区测绘院 | Remote sensing image enhancement method |
| CN113160255A (en) * | 2021-02-24 | 2021-07-23 | 国网福建省电力有限公司检修分公司 | Method for monitoring change of environment factor of operating line corridor |
| CN113313647A (en) * | 2021-05-31 | 2021-08-27 | 齐鲁工业大学 | Intelligent denoising method for leather fiber MCT tomography image and application |
| CN113643234A (en) * | 2021-07-05 | 2021-11-12 | 华北电力大学 | Composite insulator damage detection method, terminal equipment and readable storage medium |
| CN113658302A (en) * | 2021-08-23 | 2021-11-16 | 李帮音 | Three-dimensional animation data processing method and device |
| CN113680695A (en) * | 2021-08-24 | 2021-11-23 | 武昌工学院 | Robot-based machine vision garbage sorting system |
| CN113763397A (en) * | 2021-09-03 | 2021-12-07 | 国网山东省电力公司电力科学研究院 | Composite insulator fault detection method and system |
| CN114061476A (en) * | 2021-11-17 | 2022-02-18 | 国网宁夏电力有限公司建设分公司 | Deflection detection method for insulator of power transmission line |
| CN114264397A (en) * | 2021-12-21 | 2022-04-01 | 中南林业科技大学 | Wavelet and contour line-based instantaneous cable force identification method, electronic device and medium |
| CN114782354A (en) * | 2022-04-15 | 2022-07-22 | 华侨大学 | Explosive dynamic coke speckle characteristic identification and extraction and stress intensity factor self-adaptive algorithm |
| CN114862880A (en) * | 2022-07-06 | 2022-08-05 | 山东泰恒石材有限公司 | Cutting optimization method and system based on anisotropic stone |
| CN115661110A (en) * | 2022-11-08 | 2023-01-31 | 西南交通大学 | Method for identifying and positioning transparent workpiece |
| CN116310806A (en) * | 2023-02-28 | 2023-06-23 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
| CN116485350A (en) * | 2023-06-20 | 2023-07-25 | 山东鲁玻玻璃科技有限公司 | Intelligent production system of medium borosilicate glass based on image recognition |
| CN117173190A (en) * | 2023-11-03 | 2023-12-05 | 成都中轨轨道设备有限公司 | Insulator infrared damage inspection system based on image processing |
| CN117455911A (en) * | 2023-12-22 | 2024-01-26 | 广东胜高通信有限公司 | Production test method and system for network cable crystal head |
| CN117710597A (en) * | 2023-12-19 | 2024-03-15 | 中铁一局集团市政环保工程有限公司 | Three-dimensional modeling method and system based on oblique photographic data and electronic equipment |
| CN117765656A (en) * | 2024-02-21 | 2024-03-26 | 四川省肿瘤医院 | Control method and control system for gate of each ward of inpatient department |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400141A (en) * | 2013-07-24 | 2013-11-20 | 华南理工大学 | Method for calculating thickness of ice coated on transmission line on basis of improved image method |
| WO2016119410A1 (en) * | 2015-01-26 | 2016-08-04 | 广州广电运通金融电子股份有限公司 | Banknote crack detection method and apparatus |
| US20170301078A1 (en) * | 2016-04-15 | 2017-10-19 | ecoATM, Inc. | Methods and systems for detecting cracks in electronic devices |
| CN109360217A (en) * | 2018-09-29 | 2019-02-19 | 国电南瑞科技股份有限公司 | Power transmission and transforming equipment method for detecting image edge, apparatus and system |
| CN109785285A (en) * | 2018-12-11 | 2019-05-21 | 西安工程大学 | A kind of insulator damage testing method based on oval feature fitting |
-
2020
- 2020-01-13 CN CN202010032660.3A patent/CN111260616A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400141A (en) * | 2013-07-24 | 2013-11-20 | 华南理工大学 | Method for calculating thickness of ice coated on transmission line on basis of improved image method |
| WO2016119410A1 (en) * | 2015-01-26 | 2016-08-04 | 广州广电运通金融电子股份有限公司 | Banknote crack detection method and apparatus |
| US20170301078A1 (en) * | 2016-04-15 | 2017-10-19 | ecoATM, Inc. | Methods and systems for detecting cracks in electronic devices |
| CN109360217A (en) * | 2018-09-29 | 2019-02-19 | 国电南瑞科技股份有限公司 | Power transmission and transforming equipment method for detecting image edge, apparatus and system |
| CN109785285A (en) * | 2018-12-11 | 2019-05-21 | 西安工程大学 | A kind of insulator damage testing method based on oval feature fitting |
Non-Patent Citations (3)
| Title |
|---|
| 孙晋: "基于边缘检测的绝缘子裂纹诊断研究" * |
| 杨恢先: "改进的灰度-局部方差二维直方图图像分割" * |
| 黄宵宁: "直升机巡检航拍图像中绝缘子图像的提取算法" * |
Cited By (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070717B (en) * | 2020-08-05 | 2024-06-04 | 煜邦数字科技(广东)有限公司 | Power transmission line icing thickness detection method based on image processing |
| CN112070717A (en) * | 2020-08-05 | 2020-12-11 | 三峡大学 | Power transmission line icing thickness detection method based on image processing |
| CN112150500B (en) * | 2020-09-17 | 2023-12-26 | 西安工程大学 | Insulator segmentation extraction method based on joint component graying |
| CN112150500A (en) * | 2020-09-17 | 2020-12-29 | 西安工程大学 | Insulator segmentation and extraction method based on joint component graying |
| CN112164060A (en) * | 2020-10-22 | 2021-01-01 | 江苏科技大学 | Diamond sawing wire fracture prediction method based on image processing |
| CN112164060B (en) * | 2020-10-22 | 2024-02-06 | 江苏科技大学 | Diamond saw wire fracture prediction method based on image processing |
| CN112347890A (en) * | 2020-10-30 | 2021-02-09 | 武汉理工大学 | Insulator robot operation identification method, storage medium and system |
| CN112686120B (en) * | 2020-12-25 | 2023-03-03 | 国家电网有限公司 | Power transmission line abnormity detection method based on unmanned aerial vehicle aerial image |
| CN112686120A (en) * | 2020-12-25 | 2021-04-20 | 国家电网有限公司 | Power transmission line anomaly detection method based on unmanned aerial vehicle aerial image |
| CN112669286A (en) * | 2020-12-29 | 2021-04-16 | 北京建筑材料检验研究院有限公司 | Infrared thermal image-based method for identifying defects and evaluating damage degree of external thermal insulation system of external wall |
| CN112907460A (en) * | 2021-01-25 | 2021-06-04 | 宁波市鄞州区测绘院 | Remote sensing image enhancement method |
| CN112907460B (en) * | 2021-01-25 | 2022-07-29 | 宁波市鄞州区测绘院 | Remote sensing image enhancement method |
| CN113160255A (en) * | 2021-02-24 | 2021-07-23 | 国网福建省电力有限公司检修分公司 | Method for monitoring change of environment factor of operating line corridor |
| CN113313647A (en) * | 2021-05-31 | 2021-08-27 | 齐鲁工业大学 | Intelligent denoising method for leather fiber MCT tomography image and application |
| CN113643234B (en) * | 2021-07-05 | 2024-05-17 | 华北电力大学 | Composite insulator damage detection method, terminal equipment and readable storage medium |
| CN113643234A (en) * | 2021-07-05 | 2021-11-12 | 华北电力大学 | Composite insulator damage detection method, terminal equipment and readable storage medium |
| CN113658302B (en) * | 2021-08-23 | 2024-04-12 | 麦应俊 | Three-dimensional animation data processing method and device |
| CN113658302A (en) * | 2021-08-23 | 2021-11-16 | 李帮音 | Three-dimensional animation data processing method and device |
| CN113680695A (en) * | 2021-08-24 | 2021-11-23 | 武昌工学院 | Robot-based machine vision garbage sorting system |
| CN113763397A (en) * | 2021-09-03 | 2021-12-07 | 国网山东省电力公司电力科学研究院 | Composite insulator fault detection method and system |
| CN113763397B (en) * | 2021-09-03 | 2024-03-29 | 国网山东省电力公司电力科学研究院 | Composite insulator fault detection method and system |
| CN114061476A (en) * | 2021-11-17 | 2022-02-18 | 国网宁夏电力有限公司建设分公司 | Deflection detection method for insulator of power transmission line |
| CN114264397B (en) * | 2021-12-21 | 2023-12-19 | 中南林业科技大学 | Instantaneous cable force identification method based on wavelet and contour line, electronic equipment and medium |
| CN114264397A (en) * | 2021-12-21 | 2022-04-01 | 中南林业科技大学 | Wavelet and contour line-based instantaneous cable force identification method, electronic device and medium |
| CN114782354A (en) * | 2022-04-15 | 2022-07-22 | 华侨大学 | Explosive dynamic coke speckle characteristic identification and extraction and stress intensity factor self-adaptive algorithm |
| CN114862880B (en) * | 2022-07-06 | 2022-09-02 | 山东泰恒石材有限公司 | Cutting optimization method and system based on anisotropic stone |
| CN114862880A (en) * | 2022-07-06 | 2022-08-05 | 山东泰恒石材有限公司 | Cutting optimization method and system based on anisotropic stone |
| CN115661110A (en) * | 2022-11-08 | 2023-01-31 | 西南交通大学 | Method for identifying and positioning transparent workpiece |
| CN115661110B (en) * | 2022-11-08 | 2023-08-01 | 西南交通大学 | Transparent workpiece identification and positioning method |
| CN116310806A (en) * | 2023-02-28 | 2023-06-23 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
| CN116310806B (en) * | 2023-02-28 | 2023-08-29 | 北京理工大学珠海学院 | Intelligent agriculture integrated management system and method based on image recognition |
| CN116485350A (en) * | 2023-06-20 | 2023-07-25 | 山东鲁玻玻璃科技有限公司 | Intelligent production system of medium borosilicate glass based on image recognition |
| CN116485350B (en) * | 2023-06-20 | 2023-09-01 | 山东鲁玻玻璃科技有限公司 | Intelligent production system of medium borosilicate glass based on image recognition |
| CN117173190B (en) * | 2023-11-03 | 2024-02-02 | 成都中轨轨道设备有限公司 | Insulator infrared damage inspection system based on image processing |
| CN117173190A (en) * | 2023-11-03 | 2023-12-05 | 成都中轨轨道设备有限公司 | Insulator infrared damage inspection system based on image processing |
| CN117710597A (en) * | 2023-12-19 | 2024-03-15 | 中铁一局集团市政环保工程有限公司 | Three-dimensional modeling method and system based on oblique photographic data and electronic equipment |
| CN117455911A (en) * | 2023-12-22 | 2024-01-26 | 广东胜高通信有限公司 | Production test method and system for network cable crystal head |
| CN117455911B (en) * | 2023-12-22 | 2024-05-10 | 广东胜高通信有限公司 | Production test method and system for network cable crystal head |
| CN117765656A (en) * | 2024-02-21 | 2024-03-26 | 四川省肿瘤医院 | Control method and control system for gate of each ward of inpatient department |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111260616A (en) | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization | |
| CN114937055B (en) | Image self-adaptive segmentation method and system based on artificial intelligence | |
| CN106683075B (en) | Method for detecting bolt defects at cross arm of power transmission line tower | |
| CN106780486B (en) | Steel plate surface defect image extraction method | |
| CN109785285B (en) | Insulator damage detection method based on ellipse characteristic fitting | |
| WO2020224458A1 (en) | Method for detecting corona discharge employing image processing | |
| CN110009638B (en) | Bridge inhaul cable image appearance defect detection method based on local statistical characteristics | |
| CN107784669A (en) | A kind of method that hot spot extraction and its barycenter determine | |
| CN109523529B (en) | Power transmission line defect identification method based on SURF algorithm | |
| CN109472788B (en) | Method for detecting flaw on surface of airplane rivet | |
| CN115063430B (en) | Electric pipeline crack detection method based on image processing | |
| CN112070717B (en) | Power transmission line icing thickness detection method based on image processing | |
| CN108665464A (en) | A kind of foreign matter detecting method based on morphologic high tension electric tower and high-tension bus-bar | |
| CN110660065A (en) | Infrared fault detection and identification algorithm | |
| CN115631116B (en) | Aircraft power inspection system based on binocular vision | |
| CN112017109B (en) | Online ferrographic video image bubble elimination method | |
| CN114445331A (en) | Cable intermediate joint construction defect detection method, system and device based on image recognition | |
| CN115100077B (en) | Image enhancement method and device | |
| CN111008967B (en) | Insulator RTV coating defect identification method | |
| CN112529853A (en) | Method and device for detecting damage of netting of underwater aquaculture net cage | |
| CN110363749B (en) | Evaluation method for rusting degree of vibration damper based on image processing | |
| CN115272362A (en) | Method and device for segmenting effective area of digital pathology full-field image | |
| CN113744326B (en) | Fire detection method based on seed region growth rule in YCRCB color space | |
| CN113971681A (en) | Edge detection method for belt conveyor in complex environment | |
| US20060126136A1 (en) | Image inspecting apparatus, image inspecting method, control program and computer-readable storage medium |
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 |