CN113763379A - Method and device for detecting broken string of dropper, computer equipment and storage medium - Google Patents
Method and device for detecting broken string of dropper, computer equipment and storage medium Download PDFInfo
- Publication number
- CN113763379A CN113763379A CN202111110795.8A CN202111110795A CN113763379A CN 113763379 A CN113763379 A CN 113763379A CN 202111110795 A CN202111110795 A CN 202111110795A CN 113763379 A CN113763379 A CN 113763379A
- Authority
- CN
- China
- Prior art keywords
- dropper
- image
- contour
- region
- area
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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 OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06T2207/30236—Traffic on road, railway or crossing
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 Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method and a device for detecting broken strands of a dropper, computer equipment and a storage medium, and relates to the technical field of image detection. The method comprises the steps of sequentially carrying out image filtering, edge detection, binarization and connected domain processing on a dropper image to be detected to obtain an ROI (region of interest), setting a search template, sliding the search template on the ROI to locate two ends of the dropper from the ROI to obtain a dropper primary location area, sequentially carrying out transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the dropper primary location area, screening out dropper contour lines according to the included angles of contour fitting straight lines of the connected domain and standard straight lines to obtain a dropper area image, extracting all area center lines in the dropper area image, respectively carrying out straight line fitting on all the center lines, calculating the included angles of the fitting straight lines of the center lines and the standard straight lines, and carrying out strand breakage judgment. The intelligent recognition of the broken string of the dropper is realized, and the recognition efficiency and the recognition accuracy are both considered.
Description
Technical Field
The invention relates to the technical field of image detection, in particular to a method and a device for detecting broken strands of a dropper, computer equipment and a storage medium.
Background
The railway contact net dropper is equivalent to a bracket on the whole contact net and is a key part for safe operation of the train. The sag of the contact line can be adjusted by adjusting the dropper, the elasticity of the contact suspension is improved, good sliding friction between the contact line and the pantograph is ensured, and the current taking quality of the pantograph of the electric locomotive is improved. The dropper therefore plays a very important role in the whole railway catenary. With the continuous acceleration of the train in China, the fatigue and strand breaking accidents of the catenary dropper occur, the dropper strand breaking is detected accurately in time, and the method has very important significance for guiding the implementation and maintenance.
At present, the defect of broken string strand of the dropper is mainly detected by a manual inspection method, but the problems of long inspection period, low efficiency, easy omission and the like exist, the omission of broken string strand of the dropper can generate great negative effects on the normalized operation of the railway, and the accuracy and the efficiency of the detection of broken string strand of the dropper still need to be improved in consideration of the data volume of images of the dropper and the characteristic that the broken string strand of the dropper is difficult to distinguish.
Disclosure of Invention
In order to overcome the defects and shortcomings in the prior art, the invention provides a method and a device for detecting broken strands of a dropper, computer equipment and a storage medium. The invention aims to solve the problems of long detection time, low efficiency and easy omission of detection of broken string strands of a dropper in the prior art. The method comprises the steps of sequentially carrying out image filtering, edge detection, binarization and connected domain processing on a dropper image to be detected to obtain an ROI (region of interest), setting a search template, sliding the search template on the ROI to obtain a union set of the ROI and the search template so as to position the upper end part and the lower end part of the dropper from the ROI; extracting a region between two end parts of the dropper as a dropper primary positioning region, sequentially performing transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the dropper primary positioning region, screening dropper contour lines according to included angles between a connected domain contour fitting straight line and a standard straight line, and obtaining a dropper region image by using a minimum circumscribed rectangle of the dropper contour lines; extracting all area center lines in the dropper area image, respectively performing linear fitting on all the center lines, calculating the included angle between the center line fitting straight line and the standard straight line, and if the included angle is larger than a set threshold value, judging that the dropper has a broken strand. The method has the advantages of less calculation amount, no need of long-time model training and lower cost.
In a first aspect of the present invention, a method for detecting a broken string of a dropper is provided, the method comprising:
s1, acquiring an image to be detected, wherein the image to be detected comprises a dropper;
s2, sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI (region of interest);
s3, setting a search template, sliding the search template on the ROI area, obtaining a union of the ROI area and the search template, and if the geometric dimension characteristics of the union area are larger than the search template, determining that the union area is the upper end part of the hanger or the lower end part of the hanger; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is a circle center or an ellipse;
s4, extracting a region between the upper end part and the lower end part of the hanger to be a hanger primary positioning region, and then sequentially carrying out transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanger primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
s5, extracting all area center lines in the dropper area image, respectively performing straight line fitting on all the center lines, calculating the included angle between the center line fitting straight line and the standard straight line, and if the included angle is larger than a set threshold value, judging that the dropper has a broken strand.
Further, the above-mentioned upper end positioning frame and lower end positioning frame of the hanger are the minimum outside rectangular frames for positioning the upper end and lower end of the hanger from the ROI region in step S3, respectively.
And S4, screening out dropper contour lines according to the included angle between the connected domain contour fitting straight line and the standard straight line, specifically, obtaining the connected domain contour lines according to the connected domain contour fitting straight line, traversing each contour line to be selected, selecting contour points of the middle part of the contour lines, performing straight line fitting on the contour points of the middle part, calculating the included angle between the fitted straight line and the standard straight line, wherein the straight line with the included angle smaller than a set threshold range is the dropper contour line, extracting coordinate points of all dropper contour lines, and calculating the minimum circumscribed rectangle of the coordinate points to obtain a dropper region image.
And step S4, further comprising skeleton extraction, namely extracting regional center lines of all regions of the connected domain, traversing all regional center lines, and extracting each contour point on the center line, wherein if points which do not belong to the dropper positioning frame exist on the center line, the judgment of dropper broken strand of the line is not carried out.
In the step S2, the image filtering of the image to be detected means that a bilateral filtering mode is adopted to perform filtering and noise reduction processing on the image to be detected; the edge detection means that a second-order difference method is adopted to calculate the edge of the image; the binarization and connected domain processing in the step S2 specifically includes performing fixed threshold binarization on the image after the edge extraction, extracting connected domains by using an eight-connected domain method, calculating geometric features of each region, and filtering the interference region by setting conditions, thereby obtaining an ROI region.
In the step S3, the search template is set according to the regional characteristics of the connecting region of the dropper with the conductor and the catenary, and the diameter of the search template is the same as the width of the catenary wire clamp region or the catenary wire clamp region.
And S4, sequentially performing transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the dropper primary positioning area, specifically, performing image filtering on the dropper primary positioning area by adopting a transverse filtering mode, performing transverse gradient calculation on the filtered image, performing binarization processing on the transverse gradient image by using a fixed threshold value, and extracting the connected domain from the binarized image by adopting an 8-neighborhood mode.
In a second aspect of the present invention, there is provided a dropper strand breakage detection apparatus, comprising:
the device comprises an image acquisition module to be detected, a detection module and a control module, wherein the image acquisition module to be detected is used for acquiring an image to be detected, and the image to be detected comprises a dropper;
the ROI area positioning module is used for sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI area;
the hanging string end positioning module is provided with a search template and is used for sliding the search template on the ROI area to obtain a union set of the ROI area and the search template, and if the geometric size characteristic of the union set area is larger than that of the search template, the union set area is the upper end part or the lower end part of the hanging string; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is a circle center or an ellipse;
the hanging string region image positioning module extracts a region between the upper end part and the lower end part of the hanging string in the hanging string end part positioning module as a hanging string primary positioning region, and then sequentially performs transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanging string primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
and the strand breakage detection module is used for extracting all regional center lines in the dropper regional image positioning module, respectively performing linear fitting on all the center lines, calculating an included angle between a central line fitting linear line and the standard linear line, and judging that the dropper is a strand breakage if the included angle is larger than a set threshold value.
Furthermore, the upper end positioning frame and the lower end positioning frame of the hanger are respectively the smallest external rectangular frames for positioning the upper end and the lower end of the hanger from the ROI area in the hanger end positioning module.
In the dropper area image positioning module, dropper contour lines are screened out according to the included angle between the connected domain contour fitting straight line and the standard straight line, specifically, the connected domain contour lines are obtained according to the connected domain contour fitting straight line, each contour line to be selected is traversed, contour points of the middle part of the contour lines are selected, straight line fitting is carried out on the contour points of the middle part, the included angle is calculated between the fitted straight line and the standard straight line, the straight line with the included angle smaller than a set threshold range is the dropper contour line, coordinate points of all dropper contour lines are extracted, the minimum circumscribed rectangle of the dropper contour lines is calculated, and a dropper area image is obtained.
The image positioning module for the dropper area also comprises a framework extraction step, wherein the framework extraction step is used for extracting area center lines of all areas of a connected area, traversing all the area center lines, and extracting each contour point on the center line, and if points which do not belong to the dropper positioning frame exist on the center line, the judgment of dropper broken strands is not carried out on the line.
In the ROI area positioning module, image filtering on an image to be detected means that filtering and noise reduction processing is carried out on the image to be detected in a bilateral filtering mode; the edge detection means that a second-order difference method is adopted to calculate the edge of the image; the binarization and connected domain processing in the step S2 specifically includes performing fixed threshold binarization on the image after the edge extraction, extracting connected domains by using an eight-connected domain method, calculating geometric features of each region, and filtering the interference region by setting conditions, thereby obtaining an ROI region.
In the ROI area positioning module, the search template is set according to the area characteristics of a connecting area of a dropper, a lead and a catenary, and the diameter of the search template is the same as the width of a catenary wire clamp area or a dropper wire clamp area.
In the dropper area image positioning module, transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction are sequentially carried out on the dropper initial positioning area, specifically, image filtering is carried out on the dropper initial positioning area in a transverse filtering mode, transverse gradient calculation is carried out on the filtered image, binarization processing is carried out on the transverse gradient image by using a fixed threshold value, and the connected domain is extracted from the binarized image in an 8-neighborhood mode.
A third aspect of the present invention provides a computer apparatus comprising a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is adapted to store a computer program comprising program instructions, and the processor is configured to invoke the program instructions to perform some or all of the steps as described in the first aspect of the present invention.
A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program causes a computer to perform some or all of the steps as described in the first aspect of the present invention.
Compared with the prior art, the beneficial technical effects brought by the invention are as follows:
1. according to the method, the upper end point and the lower end point of the dropper are extracted in a template searching matching mode, the standard straight line is extracted according to the upper end point positioning frame and the lower end point positioning frame, the dropper area is extracted according to the included angle between the area contour line and the standard straight line, and the dropper broken strand is detected according to the included angle between the area center line and the standard straight line. The detection process of the broken string of the dropper is refined, a large number of samples are not needed for network model training, and the calculation efficiency is high.
2. The hanger broken strand defect is fine, and the defects such as hanger looseness and breakage defects are more difficult to identify. According to the method, the broken string of the dropper is identified through the angle comparison between all central lines in the dropper area and the standard straight line, the intelligent identification of the broken string of the dropper is realized, and the identification efficiency and the identification accuracy are considered.
3. In the step of S4, extracting a region between the upper end part and the lower end part of the hanger as a hanger primary positioning region, and then sequentially carrying out transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanger primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line; compared with the neural network learning and other modes in the prior art, the method has the advantages that the computation amount of the step of extracting the dropper contour lines is small, and the extraction efficiency is high.
4. In order to prevent false recognition, the invention also carries out skeleton extraction in the step S4, extracts the region central line, and eliminates the lines which do not belong to the dropper line according to the spatial relationship between the region central line and the dropper region, thereby further improving the precision of the broken strand detection judgment.
Drawings
FIG. 1 is a flow chart of a method for detecting broken strands of a dropper in accordance with the present invention;
FIG. 2 is a flow chart of the output of the end region of the dropper of the present invention;
FIG. 3 is a flow chart of the dropper area output of the present invention;
FIG. 4 shows a dropper end positioning result image processing flow;
FIG. 5 shows a dropper area positioning image;
FIG. 6 shows a standard line locate image;
fig. 7 shows a dropper area positioning image processing flow.
Detailed Description
The technical solution of the present invention is further elaborated below with reference to specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example 1
As a preferred embodiment of the present invention, the present embodiment discloses: a dropper strand breakage detection method, the method comprising:
s1, acquiring an image to be detected, wherein the image to be detected comprises a dropper;
s2, sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI (region of interest);
s3, setting a search template, sliding the search template on the ROI area, obtaining a union of the ROI area and the search template, and if the geometric dimension characteristics of the union area are larger than the search template, determining that the union area is the upper end part of the hanger or the lower end part of the hanger; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is a circle center or an ellipse;
s4, extracting a region between the upper end part and the lower end part of the hanger to be a hanger primary positioning region, and then sequentially carrying out transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanger primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
s5, extracting all area center lines in the dropper area image, respectively performing straight line fitting on all the center lines, calculating the included angle between the center line fitting straight line and the standard straight line, and if the included angle is larger than a set threshold value, judging that the dropper has a broken strand.
As an embodiment of this embodiment, the dropper upper end positioning frame and the dropper lower end positioning frame are the minimum outside rectangular frames for positioning the dropper upper end and the dropper lower end from the ROI in step S3, respectively.
As another implementation manner of this embodiment, in the step S4, string contour lines are screened out according to an included angle between a connected domain contour fitting straight line and a standard straight line, specifically, a connected domain contour line is obtained according to a connected domain contour fitting straight line, each contour line to be selected is traversed, contour points of a middle portion of the contour lines are selected, straight line fitting is performed on the contour points of the middle portion, an included angle is calculated between the fitted straight line and the standard straight line, a straight line with an included angle smaller than a set threshold range is a string contour line, coordinate points of all string contour lines are extracted, and a minimum circumscribed rectangle thereof is calculated, so that a string region image is obtained.
As another embodiment of this embodiment, the step S4 further includes skeleton extraction, which includes extracting area centerlines of all areas of the connected domain, traversing all the area centerlines, and extracting each contour point on the centerline, and if there is a point on the centerline that does not belong to the dropper positioning frame, then the determination of dropper broken strand for the line is not performed.
As another implementation manner of this embodiment, in step S2, the image filtering on the image to be detected means that a bilateral filtering manner is adopted to perform filtering and noise reduction processing on the image to be detected; the edge detection means that a second-order difference method is adopted to calculate the edge of the image; the binarization and connected domain processing in the step S2 specifically includes performing fixed threshold binarization on the image after the edge extraction, extracting connected domains by using an eight-connected domain method, calculating geometric features of each region, and filtering the interference region by setting conditions, thereby obtaining an ROI region.
As a further implementation manner of this embodiment, in step S3, the search template is set according to the regional characteristics of the dropper-conductor-catenary connection region, and the diameter of the search template is the same as the width of the dropper-catenary wire clamp region or dropper-conductor wire clamp region.
As another implementation manner of this embodiment, in step S4, the transverse filtering, the transverse gradient calculation, the binarization processing, and the connected component extraction are sequentially performed on the string initial positioning area, specifically, the image filtering is performed on the string initial positioning area by adopting a transverse filtering method, the transverse gradient calculation is performed on the filtered image, the binarization processing is performed on the transverse gradient image by using a fixed threshold, and the connected component is extracted from the binarized image by adopting an 8-neighborhood method.
Example 2
As another preferred embodiment of the present invention, referring to fig. 1 to 7 in the specification, the present embodiment discloses a method for detecting a broken string of a dropper, and the specific flow is as shown in fig. 1, and the method includes steps of inputting an image, positioning a dropper, extracting a dropper area, and determining a broken string of a dropper.
The basis of the extraction of the dropper line is the accurate positioning of the dropper, as shown in fig. 2, the dropper positioning steps are as follows: inputting an image, filtering the image, detecting edges, binarizing, processing a connected domain, positioning a dropper end and outputting the area of the dropper end.
As shown in fig. 4, the original image is filtered and denoised in a bilateral filtering manner. The image edges of the filtered image are calculated using a second order difference method. In the case of a continuous function, at a maximum or minimum in the first order derivative plot, it is considered an edge; the zero crossing between a maximum and a minimum in the second order derivative plot is considered an edge. First order difference:
(ii) a Second order difference:
(ii) a From this can be obtained
Extracting the coefficients of the previous: [1, -2,1]. In the case of two dimensions, the same reasoning applies
Extracting each coefficient of the function, and writing the coefficient into a template form:
. Since the image is expressed in the form of two-dimensional data, the two-dimensional template can be convolved with the image to obtain the image edge.
And extracting the edge image, carrying out fixed threshold binarization processing, and extracting a connected domain by adopting an 8-neighborhood mode. And sequentially calculating the geometric characteristics of each region: zone perimeter (number of zone boundary contour points); the length and the width of the region (obtaining a region external rectangle and obtaining the length and the width of the region according to the external rectangle); region quality (zeroth order moment of the image:
wherein V (i, j) represents the gray scale value of the image at the (i, j) point; when the image is a two-value image,
is the sum of some white area pixels of the image, which can be used to findSolving the gray quality of the white area); region centroid (first moment of image:
，
) When the image is a binary image, V (i, j) has only two values of 0 (black) and 255 (white),
is the accumulation of the x-coordinates of the white area on the image,
is the accumulation of the y-coordinate of the white area on the image, so the image centroid:
，
). And filtering other interference areas based on the calculated geometric characteristics by setting a certain condition, so as to obtain the area of interest of the contact network.
Two ends of the dropper are respectively connected with the lead and the carrier cable. Because the difficulty of directly finding the dropper line is high, the text turns to find the connection point of the dropper line and the lead and the carrier cable. From the binarization result, the connection point area is wider than other areas, and a circular template m with the diameter almost equal to the width of the area is manufactured by combining the characteristics. And (3) sliding a template m on the graph 7, respectively solving a union set of m and the area, wherein if the area, the length and the width of the union set area are all larger than the corresponding characteristics of the template, the area is the end part of the positioned dropper, and the dropper is positioned.
As shown in fig. 5, according to the positioning positions of the two ends of the dropper, a dropper area can be preliminarily obtained, and fine positioning of the dropper area is performed on the basis. The dropper region extraction algorithm flow is shown in fig. 3.
Because the dropper lines are shown in the direction of y in fig. 7, according to the characteristic, the invention adopts a transverse filtering mode to reduce the interference of background objects on subsequent processing results as much as possible.
And (4) obtaining the gradient of the image in the x direction according to the dropper characteristics (the trend in the y direction), and removing the interference objects in the y direction to the greatest extent. Pixel points in an image
The gradient in the x-direction is:
。
carrying out binarization processing on the gradient image by using a fixed threshold value; extracting a connected domain: and extracting a connected domain from the binarized image by adopting an 8-neighborhood mode. And performing skeleton extraction on the image after binarization and connected domain processing, wherein the purpose of skeleton extraction is to extract the region center line. Filtering the regional lines, wherein the line filtering is mainly based on the following two characteristics: length, angle. Line length: the number of pixel points on the line. Line angle: and calculating the angle between the line and the standard line.
Definition of standard line: as shown in fig. 6, the upper left corner of the positioning frame for the upper end of the dropper is defined as p1, the upper left corner of the positioning frame for the lower end of the dropper is defined as p2, and the line formed by connecting the two points p1 and p2 is the standard line. Traversing each contour line to be selected, selecting contour points of the middle part (sorting the y coordinates of the contour points from low to high, removing a small number of points at two ends to avoid error calculation), fitting the contour points with straight lines, and calculating the included angle between the fitted straight lines and the standard line. And setting a smaller threshold, wherein the line with the included angle smaller than a certain threshold range is the output of the dropper line. And extracting all coordinate points of the dropper line, and calculating the minimum external rectangle of the coordinate points, namely the dropper area.
And obtaining the central lines of all the regions according to the skeleton extraction result. And traversing all the lines, extracting each contour point on the line, and if the line has a point which does not belong to the dropper frame, not judging the broken string of the line by the dropper, so as to avoid false detection of the broken string. On the basis, linear fitting is carried out on the line, an included angle between the fitted linear line and the standard line is calculated, and when the included angle is larger than a certain threshold value, the phenomenon that the strand of the dropper is broken is judged.
Example 3
As another preferred embodiment of the present invention, this embodiment provides a dropper strand breakage detection apparatus, including:
the device comprises an image acquisition module to be detected, a detection module and a control module, wherein the image acquisition module to be detected is used for acquiring an image to be detected, and the image to be detected comprises a dropper;
the ROI area positioning module is used for sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI area;
the hanging string end positioning module is provided with a search template and is used for sliding the search template on the ROI area to obtain a union set of the ROI area and the search template, and if the geometric size characteristic of the union set area is larger than that of the search template, the union set area is the upper end part or the lower end part of the hanging string; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is a circle center or an ellipse;
the hanging string region image positioning module extracts a region between the upper end part and the lower end part of the hanging string in the hanging string end part positioning module as a hanging string primary positioning region, and then sequentially performs transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanging string primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
and the strand breakage detection module is used for extracting all regional center lines in the dropper regional image positioning module, respectively performing linear fitting on all the center lines, calculating an included angle between a central line fitting linear line and the standard linear line, and judging that the dropper is a strand breakage if the included angle is larger than a set threshold value.
Further, as an implementation manner of this embodiment, the dropper upper end positioning frame and the dropper lower end positioning frame are minimum outside rectangular frames for positioning the dropper upper end and the dropper lower end from the ROI region in the dropper end positioning module, respectively.
As another implementation manner of this embodiment, in the dropper area image positioning module, dropper contour lines are screened out according to an included angle between a connected domain contour fitting straight line and a standard straight line, specifically, a connected domain contour line is obtained according to a connected domain contour fitting straight line, each contour line to be selected is traversed, a contour point of a middle part of a contour line is selected, straight line fitting is performed on the contour point of the middle part, an included angle is calculated between the fitted straight line and the standard straight line, a straight line with an included angle smaller than a set threshold range is a dropper contour line, coordinate points of all dropper contour lines are extracted, and a minimum circumscribed rectangle is calculated, so that a dropper area image is obtained.
As another implementation manner of this embodiment, the dropper area image positioning module further includes skeleton extraction, which is to extract area center lines of all areas in the connected domain, traverse all area center lines, and extract each contour point on the center line, and if there is a point on the center line that does not belong to the dropper positioning frame, the determination of dropper broken strand for the line is not performed.
As another implementation manner of this embodiment, in the ROI positioning module, the performing image filtering on the image to be detected means performing filtering and noise reduction processing on the image to be detected by using a bilateral filtering manner; the edge detection means that a second-order difference method is adopted to calculate the edge of the image; the binarization and connected domain processing in the step S2 specifically includes performing fixed threshold binarization on the image after the edge extraction, extracting connected domains by using an eight-connected domain method, calculating geometric features of each region, and filtering the interference region by setting conditions, thereby obtaining an ROI region.
In yet another embodiment of this embodiment, the ROI positioning module sets the search template according to the regional characteristics of the dropper-conductor-catenary connection region, and the search template has the same diameter as the width of the dropper-catenary wire clamp region or the dropper-conductor wire clamp region.
As another implementation manner of this embodiment, in the dropper area image positioning module, the horizontal filtering, the horizontal gradient calculation, the binarization processing, and the connected component extraction are sequentially performed on the dropper initial positioning area, specifically, the horizontal filtering is performed on the dropper initial positioning area, the horizontal gradient calculation is performed on the filtered image, the binarization processing is performed on the horizontal gradient image by using a fixed threshold, and the connected component is extracted from the binarized image by using an 8-neighborhood method.
Example 4
In order to achieve the above object, according to another aspect of the present application, there is also provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the above-mentioned dropper strand detection method when executing the computer program.
The processor may be a Central Processing Unit (CPU) in this embodiment. The Processor may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, or a combination thereof.
The memory, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and units, such as the corresponding program units in the above-described method embodiments of the present invention. The processor executes various functional applications of the processor and the processing of the work data by executing the non-transitory software programs, instructions and modules stored in the memory, that is, the method in the above method embodiment is realized.
The memory may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created by the processor, and the like. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory located remotely from the processor, and such remote memory may be coupled to the processor via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The one or more units are stored in the memory and, when executed by the processor, perform the method of embodiment 1 above.
Example 5
As another preferred embodiment of the present invention, this embodiment discloses a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps of embodiment 1 above.
Claims (10)
1. A method for detecting broken strands of a dropper is characterized by comprising the following steps:
s1, acquiring an image to be detected, wherein the image to be detected comprises a dropper;
s2, sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI (region of interest);
s3, setting a search template, sliding the search template on the ROI area, obtaining a union of the ROI area and the search template, and if the geometric dimension characteristics of the union area are larger than the search template, determining that the union area is the upper end part of the hanger or the lower end part of the hanger; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is circular or elliptical;
s4, extracting a region between the upper end part and the lower end part of the hanger to be a hanger primary positioning region, and then sequentially carrying out transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanger primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
s5, extracting all area center lines in the dropper area image, respectively performing straight line fitting on all the center lines, calculating the included angle between the center line fitting straight line and the standard straight line, and if the included angle is larger than a set threshold value, judging that the dropper has a broken strand.
2. The method for detecting broken strands of dropper according to claim 1, wherein: and the step S3 is that the minimum external rectangular frames of the upper end part and the lower end part of the hanger are positioned from the ROI area.
3. The method for detecting broken strands of dropper according to claim 1, wherein: and S4, screening out dropper contour lines according to the included angle between the connected domain contour fitting straight line and the standard straight line, specifically, obtaining the connected domain contour lines according to the connected domain contour fitting straight line, traversing each contour line to be selected, selecting contour points of the middle part of the contour lines, performing straight line fitting on the contour points of the middle part, calculating the included angle between the fitted straight line and the standard straight line, wherein the straight line with the included angle smaller than a set threshold range is the dropper contour line, extracting coordinate points of all dropper contour lines, and calculating the minimum circumscribed rectangle of the coordinate points to obtain a dropper region image.
4. A dropper strand breakage detection method as claimed in claim 1, 2 or 3, wherein: and step S4, further comprising skeleton extraction, namely extracting regional center lines of all regions of the connected domain, traversing all regional center lines, and extracting each contour point on the center line, wherein if points which do not belong to the dropper positioning frame exist on the center line, the judgment of dropper broken strand of the line is not carried out.
5. The method for detecting broken strands of dropper according to claim 1, wherein: in the step S2, the image filtering of the image to be detected means that a bilateral filtering mode is adopted to perform filtering and noise reduction processing on the image to be detected; the edge detection means that a second-order difference method is adopted to calculate the edge of the image; the binarization and connected domain processing in the step S2 specifically includes performing fixed threshold binarization on the image after the edge extraction, extracting connected domains by using an eight-connected domain method, calculating geometric features of each region, and filtering the interference region by setting conditions, thereby obtaining an ROI region.
6. The method for detecting broken strands of dropper according to claim 1, wherein: in the step S3, the search template is set according to the regional characteristics of the connecting region of the dropper with the conductor and the catenary, and the diameter of the search template is the same as the width of the catenary wire clamp region or the catenary wire clamp region.
7. The method for detecting broken strands of dropper according to claim 1, wherein: and S4, sequentially performing transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the dropper primary positioning area, specifically, performing image filtering on the dropper primary positioning area by adopting a transverse filtering mode, performing transverse gradient calculation on the filtered image, performing binarization processing on the transverse gradient image by using a threshold method, and extracting the connected domain from the binarized image by adopting an 8-neighborhood mode.
8. A dropper strand breakage detection device, comprising:
the device comprises an image acquisition module to be detected, a detection module and a control module, wherein the image acquisition module to be detected is used for acquiring an image to be detected, and the image to be detected comprises a dropper;
the ROI area positioning module is used for sequentially carrying out image filtering, edge detection, binarization and connected domain processing on an image to be detected to obtain an ROI area;
the hanging string end positioning module is provided with a search template and is used for sliding the search template on the ROI area to obtain a union set of the ROI area and the search template, and if the geometric size characteristic of the union set area is larger than that of the search template, the union set area is the upper end part or the lower end part of the hanging string; positioning an upper end part and a lower end part of the hanger from the ROI; the search template is a circle center or an ellipse;
the hanging string region image positioning module extracts a region between the upper end part and the lower end part of the hanging string in the hanging string end part positioning module as a hanging string primary positioning region, and then sequentially performs transverse filtering, transverse gradient calculation, binarization processing and connected domain extraction on the hanging string primary positioning region; screening out a dropper contour line according to an included angle between the connected domain contour fitting straight line and the standard straight line, and acquiring a dropper region image according to the minimum external rectangle of the dropper contour line; the standard straight line is a connecting line between the same angles on the upper end positioning frame and the lower end positioning frame of the dropper;
and the strand breakage detection module is used for extracting all regional center lines in the dropper regional image positioning module, respectively performing linear fitting on all the center lines, calculating an included angle between a central line fitting linear line and the standard linear line, and judging that the dropper is a strand breakage if the included angle is larger than a set threshold value.
9. A computer device comprising a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is configured to store a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of any of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program comprising program instructions that, when executed by a processor, cause the processor to carry out the method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111110795.8A CN113763379B (en) | 2021-09-23 | 2021-09-23 | Suspension string broken strand detection method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111110795.8A CN113763379B (en) | 2021-09-23 | 2021-09-23 | Suspension string broken strand detection method, device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113763379A true CN113763379A (en) | 2021-12-07 |
| CN113763379B CN113763379B (en) | 2023-07-18 |
Family
ID=78796838
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111110795.8A Active CN113763379B (en) | 2021-09-23 | 2021-09-23 | Suspension string broken strand detection method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113763379B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612480A (en) * | 2022-02-21 | 2022-06-10 | 北京百度网讯科技有限公司 | Method, device, equipment, medium and product for detecting the bending state of anti-vibration hammer |
| CN114627203A (en) * | 2022-03-16 | 2022-06-14 | 北京京东方技术开发有限公司 | Image sample production method and system, computer readable storage medium and device |
| CN115909330A (en) * | 2022-09-07 | 2023-04-04 | 江苏音律未来乐器科技有限公司 | A method and system for identifying strings |
| CN118379317A (en) * | 2024-06-24 | 2024-07-23 | 常州市宏发纵横新材料科技股份有限公司 | A glass fiber cloth edge recognition method, device and storage medium |
| CN118799810A (en) * | 2024-08-05 | 2024-10-18 | 矿冶科技集团有限公司 | A method and device for thickening machine mud layer height image recognition based on regional attention mechanism |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140200426A1 (en) * | 2011-02-28 | 2014-07-17 | Abbott Diabetes Care Inc. | Devices, Systems, and Methods Associated with Analyte Monitoring Devices and Devices Incorporating the Same |
| CN104392212A (en) * | 2014-11-14 | 2015-03-04 | 北京工业大学 | Method for detecting road information and identifying forward vehicles based on vision |
| CN104400265A (en) * | 2014-10-08 | 2015-03-11 | 吴长兴 | Feature extraction method applicable to corner weld of laser vision guided welding robot |
| CN108106562A (en) * | 2017-12-05 | 2018-06-01 | 浙江维思无线网络技术有限公司 | A kind of contact net measuring method and device |
| CN108898147A (en) * | 2018-06-27 | 2018-11-27 | 清华大学 | A kind of two dimensional image edge straightened method, apparatus based on Corner Detection |
| WO2019041712A1 (en) * | 2017-08-29 | 2019-03-07 | 广州微盾科技股份有限公司 | Ultrathin finger vein recognition apparatus and recognition method |
| US20190347497A1 (en) * | 2017-01-25 | 2019-11-14 | Wuhan Jimu Intelligent Technology Co., Ltd. | Road sign recognition method and system |
| CN110910443A (en) * | 2019-12-04 | 2020-03-24 | 成都唐源电气股份有限公司 | Contact net geometric parameter real-time measuring method and device based on single monitoring camera |
| CN111476246A (en) * | 2020-04-07 | 2020-07-31 | 杭州国彪超声设备有限公司 | Robust and efficient intelligent reading method for pointer instrument applied to complex environment |
| CN111667477A (en) * | 2020-06-10 | 2020-09-15 | 创新奇智(广州)科技有限公司 | Magnetic material size defect detection method and device, detection equipment and readable storage medium |
| US20210150729A1 (en) * | 2018-08-30 | 2021-05-20 | Boe Technology Group Co., Ltd. | Line segment detection method, apparatus, device, and computer-readable storage medium |
| CN113310450A (en) * | 2020-03-12 | 2021-08-27 | 广东中科如铁技术有限公司 | Contact net dropper detection method based on point cloud training model |
-
2021
- 2021-09-23 CN CN202111110795.8A patent/CN113763379B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140200426A1 (en) * | 2011-02-28 | 2014-07-17 | Abbott Diabetes Care Inc. | Devices, Systems, and Methods Associated with Analyte Monitoring Devices and Devices Incorporating the Same |
| CN104400265A (en) * | 2014-10-08 | 2015-03-11 | 吴长兴 | Feature extraction method applicable to corner weld of laser vision guided welding robot |
| CN104392212A (en) * | 2014-11-14 | 2015-03-04 | 北京工业大学 | Method for detecting road information and identifying forward vehicles based on vision |
| US20190347497A1 (en) * | 2017-01-25 | 2019-11-14 | Wuhan Jimu Intelligent Technology Co., Ltd. | Road sign recognition method and system |
| WO2019041712A1 (en) * | 2017-08-29 | 2019-03-07 | 广州微盾科技股份有限公司 | Ultrathin finger vein recognition apparatus and recognition method |
| CN108106562A (en) * | 2017-12-05 | 2018-06-01 | 浙江维思无线网络技术有限公司 | A kind of contact net measuring method and device |
| CN108898147A (en) * | 2018-06-27 | 2018-11-27 | 清华大学 | A kind of two dimensional image edge straightened method, apparatus based on Corner Detection |
| US20210150729A1 (en) * | 2018-08-30 | 2021-05-20 | Boe Technology Group Co., Ltd. | Line segment detection method, apparatus, device, and computer-readable storage medium |
| CN110910443A (en) * | 2019-12-04 | 2020-03-24 | 成都唐源电气股份有限公司 | Contact net geometric parameter real-time measuring method and device based on single monitoring camera |
| CN113310450A (en) * | 2020-03-12 | 2021-08-27 | 广东中科如铁技术有限公司 | Contact net dropper detection method based on point cloud training model |
| CN111476246A (en) * | 2020-04-07 | 2020-07-31 | 杭州国彪超声设备有限公司 | Robust and efficient intelligent reading method for pointer instrument applied to complex environment |
| CN111667477A (en) * | 2020-06-10 | 2020-09-15 | 创新奇智(广州)科技有限公司 | Magnetic material size defect detection method and device, detection equipment and readable storage medium |
Non-Patent Citations (6)
| Title |
|---|
| 刘射德: "基于GNSS与图像特征识别的列车组合定位算法研究", 中国优秀硕士学位论文全文数据库工程科技Ⅱ辑 * |
| 张润彤: "高速动车组弓网运行状态的数字图像处理和识别研究", 中国优秀硕士学位论文全文数据库工程科技Ⅱ辑 * |
| 彭檐波: "基于机器视觉的激光切向整形V形凹面砂轮测控系统设计", 中国优秀硕士学位论文全文数据库信息科技辑 * |
| 朱挺;: "新型国产轨道交通接触网检测系统研究", 都市快轨交通 * |
| 朱欢欢;郝永平;刘扬;郭鹏跃;: "平板类微小型零件的移动位姿调整", 仪表技术与传感器 * |
| 武福;豆辉;武云;李忠学;杨喜娟;: "基于机器视觉的尖轨轮廓检测方法", 光学技术 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114612480A (en) * | 2022-02-21 | 2022-06-10 | 北京百度网讯科技有限公司 | Method, device, equipment, medium and product for detecting the bending state of anti-vibration hammer |
| CN114627203A (en) * | 2022-03-16 | 2022-06-14 | 北京京东方技术开发有限公司 | Image sample production method and system, computer readable storage medium and device |
| CN115909330A (en) * | 2022-09-07 | 2023-04-04 | 江苏音律未来乐器科技有限公司 | A method and system for identifying strings |
| CN115909330B (en) * | 2022-09-07 | 2025-09-23 | 江苏音律未来乐器科技有限公司 | A method and system for identifying musical strings |
| CN118379317A (en) * | 2024-06-24 | 2024-07-23 | 常州市宏发纵横新材料科技股份有限公司 | A glass fiber cloth edge recognition method, device and storage medium |
| CN118799810A (en) * | 2024-08-05 | 2024-10-18 | 矿冶科技集团有限公司 | A method and device for thickening machine mud layer height image recognition based on regional attention mechanism |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113763379B (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113763379B (en) | Suspension string broken strand detection method, device, computer equipment and storage medium | |
| CN107808378B (en) | Method for detecting potential defects of complex-structure casting based on vertical longitudinal and transverse line profile features | |
| CN103927526B (en) | Vehicle detection method based on Gaussian difference multi-scale edge fusion | |
| CN104700072B (en) | Recognition methods based on lane line historical frames | |
| CN110910443B (en) | Contact net geometric parameter real-time measuring method and device based on single monitoring camera | |
| CN109685766B (en) | A cloth defect detection method based on region fusion features | |
| CN109671058B (en) | Defect detection method and system for large-resolution image | |
| CN113498528A (en) | Image defect determining method and device, electronic equipment and storage medium | |
| CN107895492A (en) | A kind of express highway intelligent analysis method based on conventional video | |
| CN110321933B (en) | Fault identification method and device based on deep learning | |
| CN116660286A (en) | Wire harness head peeling measurement and defect detection method and system based on image segmentation | |
| CN113744200B (en) | Camera dirt detection method, device and equipment | |
| Farag | Real-time detection of road lane-lines for autonomous driving | |
| CN110473184A (en) | A kind of pcb board defect inspection method | |
| CN107644417B (en) | Method for detecting appearance defect of strain gauge | |
| Hoseini et al. | Fabric defect detection using auto-correlation function | |
| CN102324099A (en) | A Step Edge Detection Method for Humanoid Robot | |
| CN117830287A (en) | A circuit board online detection re-judgment algorithm | |
| CN118230265A (en) | A method for detecting foreign matter on rail transit track bed | |
| CN109766750A (en) | A kind of table line position finding and detection method of financial statement | |
| CN115661157A (en) | Panel circle defect detection method, device, medium, equipment and program product | |
| CN113378847B (en) | Character segmentation method, system, computer device and storage medium | |
| CN108242060A (en) | A Method of Image Edge Detection Based on Sobel Operator | |
| CN110516527B (en) | Visual SLAM loop detection improvement method based on instance segmentation | |
| CN111597939A (en) | High-speed rail line nest defect detection method based on deep 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 |