CN110310276B - Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics - Google Patents
Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics Download PDFInfo
- Publication number
- CN110310276B CN110310276B CN201910601591.0A CN201910601591A CN110310276B CN 110310276 B CN110310276 B CN 110310276B CN 201910601591 A CN201910601591 A CN 201910601591A CN 110310276 B CN110310276 B CN 110310276B
- Authority
- CN
- China
- Prior art keywords
- image
- communication areas
- supporting device
- straight lines
- straight line
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
Abstract
The invention relates to a method for positioning a pipe cap of an electrified railway supporting device based on geometric characteristics, which comprises the steps of preprocessing a global image of the supporting device obtained by shooting, carrying out binarization processing, calculating the central coordinates of communicated areas of the binarized image, determining the corresponding straight lines of each communicated area, combining the corresponding straight lines, and finally positioning the pipe cap. The invention adopts a method for detecting the pipe cap on the arm pipe based on the geometric characteristics of the supporting device, and the pipe cap is accurately positioned by pixel-by-pixel search. The method can accurately complete the positioning of the pipe cap area, the detection result is not influenced by illumination and angle during shooting and the shooting integrity of the supporting device, the robustness is strong, the processing speed is high, the requirements of practical application are met, and the method has higher practical prospect and research significance.
Description
Technical Field
The invention relates to a detection method of an electrified railway supporting device, in particular to a method for positioning a pipe cap of the electrified railway supporting device based on geometric characteristics.
Background
The power supply system of the contact network is used as an important component of the electrified railway and directly influences the normal operation of the railway system. A contact net suspension state detection and monitoring device (4C) in the system is difficult to realize automatic identification due to the fact that the fault condition of shot field equipment is complex and the image data volume is huge. The problems of cantilever linear detection of a supporting suspension device, accurate positioning of a pipe cap and automatic fault identification in automatic 4C image identification by using an image processing technology have become one of hot research directions of contact network information intellectualization.
For the fault identification of the 4C image, many relevant studies have been made by scholars at home and abroad. The state recognition of the positioning wire clamp of the contact net is completed by machine learning and geometric characteristics in the spring of Zhang Chun; the Yangming completes the positioning of the insulator by using HOG + SVM, and then completes the fault identification by using the gray statistical characteristics; korean utilizes HOG characteristics and two-dimensional Gabor wavelet transform to complete the fracture fault identification of the ear of the supporting device. However, no study has been made by scholars on the problem of identifying the positioning of the pipe cap of the supporting device. Due to the diversity of field conditions, various conditions can exist in the acquired images of the supporting device, for example, the supporting device in the images presents various angles due to different shooting angles; the image acquisition is carried out by a high-speed camera arranged on a train running at a high speed according to a certain frequency, so that the situation that the structure of the supporting device in the image is incomplete can occur, and even a small part of the supporting device can be shot. These different situations all add to the difficulty of positioning the cap.
Disclosure of Invention
The invention aims to provide a method for positioning a pipe cap of a supporting device of an electrified railway based on geometric characteristics, and aims to solve the problems of poor detection result and low positioning accuracy when the pipe cap of the supporting device of the electrified railway is positioned by adopting the prior art.
The invention is realized by the following steps: a method for positioning a pipe cap of an electrified railway supporting device based on geometric characteristics comprises the following steps:
a. inputting a global image of the supporting device, and inputting the global image of the supporting device obtained by shooting by the camera device into the image processing equipment;
b. preprocessing a global image of the supporting device, and performing filtering, denoising, histogram equalization and reduction operations on the image to enable the image to meet the requirements of subsequent image processing;
c. binarization of the image, namely performing binarization processing on the preprocessed image, selecting a gray value slightly smaller than a galvanized part in the image as a threshold value, comparing the gray value of each pixel point in the image with the size relation of a set threshold value, setting the pixel point to be white if the gray value of the pixel point is larger than the threshold value, otherwise setting the pixel point to be black, obtaining a plurality of white connected regions, and obtaining the white regions as the extracted galvanized part (R region);
d. calculating the center of the communicated area, and calculating the center coordinate of each communicated area according to the obtained binary image;
e. calculating a straight line corresponding to each communication area, setting a smaller threshold value for the original image to carry out binarization processing, marking the central coordinates of each communication area obtained in the step d on the binarized original image, and calculating the straight line corresponding to each communication area by using a nearest pixel method;
f. combining corresponding straight lines, grouping all straight lines corresponding to the communication areas according to slopes, dividing the straight lines corresponding to the communication areas with the slopes close to each other into a group, leaving out the group containing only one straight line corresponding to the communication areas, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of each group to belong to the arm pipe with the pipe cap, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of the same group to belong to the same arm pipe, taking the average value of the slopes of the corresponding straight lines of each group of the remaining communication areas as the slope of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, taking the average value of the central coordinates of the communication areas corresponding to the corresponding straight lines of each group of the remaining communication areas as the central point of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, and obtaining the unique straight line corresponding to each arm pipe with the pipe cap.
g. And positioning the tube cap end, starting from the central point of the obtained corresponding unique straight line, searching the two ends of each arm tube with the tube cap pixel by pixel along the obtained unique straight line, stopping searching when the first black pixel point is encountered or the image range is exceeded, recording the coordinate of the searched last pixel point as an end point coordinate, discarding the end point coordinate of the non-tube cap end, and intercepting the tube cap area on the original drawing by taking the reserved end point coordinate as the center.
In the step b, denoising the global image of the supporting device by adopting a bilateral filtering method, enhancing the contrast of the image by adopting a contrast-limiting self-adaptive histogram equalization algorithm, and finally reducing the image by using a near interpolation method to obtain the required image.
In step c, the threshold value is gradually increased from 100 to 255 in a certain step length, each threshold value is used for carrying out binarization processing on the preprocessed image, the area of the largest communication area in each binarized image is counted, the communication area of each image, the area of which is less than 1/25 of the area of the largest communication area in each image, is omitted, the number of reserved communication areas in each image is counted, and the binarized image with the largest number of communication areas is selected as the finally required binarized image.
In step d, the calculated center coordinates of each communication area are the coordinates of the reduced image, the corresponding center coordinates in the original image are calculated according to the zooming relation of the image, and the corresponding center coordinates in the original image are marked in step e.
In step e, 10 is selected as the threshold.
In step g, a distance value D is obtained by counting the distance from the tail end of the pipe cap in the image to the center coordinate of the nearest communication area on the arm pipe where the pipe cap is located, the distance D is slightly larger than the maximum distance from the tail end of the pipe cap in the image to the center coordinate of the nearest communication area on the arm pipe where the pipe cap is located, the distance from each endpoint coordinate to the center coordinate of the nearest communication area on the arm pipe where the endpoint coordinate is located is calculated, if the distance is far larger than the distance D, or the endpoint coordinate is judged to be located at the edge of the image, the endpoint coordinate is omitted, and the reserved endpoint coordinate is the vertex coordinate of the end of the pipe cap.
In the step f, the centers of the communication areas are connected with each other pairwise to form straight lines, whether the straight lines with similar slopes correspond to the same arm pipe or not is distinguished according to the straight lines formed by connecting the centers of the communication areas, and if the slopes of the straight lines determined by connecting the centers of the two communication areas are very close to the slopes of the straight lines corresponding to the communication areas which are respectively corresponding to the two communication areas, the straight lines corresponding to the two communication areas belong to the same arm pipe; if the slope of the straight line determined by connecting the centers of the two communication areas is greatly different from the slope of the straight line corresponding to the communication areas corresponding to the two communication areas separately, the straight lines corresponding to the communication areas corresponding to the two communication areas separately belong to different arm tubes.
The method comprises the steps of preprocessing a global image of a support device obtained by shooting, carrying out binarization processing, calculating the central coordinates of communicated areas of the binarized image, determining corresponding straight lines of each communicated area, combining the corresponding straight lines, and finally positioning a pipe cap. The method for detecting the pipe cap on the arm pipe based on the geometrical characteristics of the supporting device is adopted, and the pipe cap is accurately positioned through pixel-by-pixel search.
The method can accurately complete the positioning of the pipe cap area, the detection result is not influenced by illumination and angle during shooting and the shooting integrity of the supporting device, the robustness is strong, the processing speed is high, the requirements of practical application are met, and the method has higher practical prospect and research significance.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a global image of the support device of the present invention.
Fig. 3 is an image subjected to binarization processing in accordance with the present invention.
Fig. 4 is a small threshold binarized image labeled with the center of the connected region according to the present invention.
Fig. 5 is a schematic diagram of calculating a straight line corresponding to a communication area according to the present invention.
Fig. 6 is a schematic view of all communication zones of the present invention in line with one another.
FIG. 7 is a schematic diagram of the present invention after linear combination.
Fig. 8 is a schematic diagram of endpoints on a full image of the present invention.
FIG. 9 is a schematic diagram of the endpoints on an incomplete image of the present invention.
FIGS. 10-12 are images of caps detected by the present invention.
FIG. 13 shows experimental comparison results between the Hough transform method of the present invention and the conventional Hough transform method.
In the figure: 1. a center point of the region of connectivity; 2. a connected region boundary; 3. the communication area corresponds to a straight line.
Detailed Description
As shown in fig. 1, the present invention comprises the steps of:
a. inputting a global image of the supporting device, and inputting the global image of the supporting device, which is obtained by shooting by the camera, into the image processing equipment;
b. preprocessing a global image of the supporting device, and performing filtering, denoising, histogram equalization and reduction operations on the image to enable the image to meet the requirements of subsequent image processing;
c. binarization of the image, namely performing binarization processing on the preprocessed image, selecting a gray value slightly smaller than a galvanized part in the image as a threshold value, comparing the gray value of each pixel point in the image with the size relation of a set threshold value, setting the pixel point to be white if the gray value of the pixel point is larger than the threshold value, otherwise setting the pixel point to be black, obtaining a plurality of white connected regions, and obtaining the white regions as the extracted galvanized part (R region);
d. calculating the center of the communication area, and calculating the center coordinate of each communication area according to the obtained binary image;
e. calculating a straight line corresponding to each communication area, setting a smaller threshold value for the original image to carry out binarization processing, marking the central coordinates of each communication area obtained in the step d on the binarized original image, and calculating the straight line corresponding to each communication area by using a nearest pixel method;
f. combining corresponding straight lines, grouping all straight lines corresponding to the communication areas according to the slope, dividing the straight lines corresponding to the communication areas with the slope close to each other into a group, leaving out the group only containing one straight line corresponding to the communication area, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of each group to belong to the arm pipes with the pipe caps, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of the same group to belong to the same arm pipe, taking the average value of the slopes of the corresponding straight lines of each group of the remaining communication areas as the slope of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, taking the average value of the central coordinates of the communication areas corresponding to the corresponding straight lines of each group of the remaining communication areas as the central point of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, and obtaining the unique straight line corresponding to each arm pipe with the pipe caps.
g. And positioning the tube cap end, starting from the central point of the obtained corresponding unique straight line, searching the two ends of each arm tube with the tube cap pixel by pixel along the obtained unique straight line, stopping searching when the first black pixel point is encountered or the image range is exceeded, recording the coordinate of the searched last pixel point as an end point coordinate, discarding the end point coordinate of the non-tube cap end, and intercepting the tube cap area on the original drawing by taking the reserved end point coordinate as the center.
The surface of the arm tube of the supporting device main body is provided with the bright silvery galvanized coating, so that the gray value of the part in a picture is generally larger than that of other parts, and the characteristic belongs to the structural characteristic of the supporting device, and the arm tube cannot change due to incomplete shooting or change of shooting angles of the supporting device, so that the invention realizes the linear measurement of the arm tube of the 4C image supporting device by taking the structural characteristic as a starting point.
As shown in fig. 2, since maintenance and repair work of the railway system can be performed only at a specially set night time, 4C images can be acquired only at night. Therefore, before processing the image, preprocessing such as filtering and denoising needs to be performed on the image, and when the information expression required in the image is weak or the image contrast is low, the image needs to be enhanced to meet the requirement of subsequent image processing. The invention adopts a bilateral filtering method aiming at the characteristics of the image, and the method can better retain edge detail information while filtering and denoising. Meanwhile, the contrast of the image is enhanced by adopting a contrast-limiting self-adaptive histogram equalization algorithm. Then, the image is reduced by selecting a proximity interpolation method, so that the time required by calculation can be shortened to the greatest extent on the basis of ensuring the accuracy of the algorithm identification result.
As shown in fig. 3, in the whole image, the R region is the region with the largest gray scale value, but there are some non-R regions (such as the localization sleeve) at the same time, although the gray scale value is smaller than that of the R region, the gray scale values of the two regions are relatively close. Compared with the main structure of the supporting device and known from actual operation, if the binary threshold value is gradually changed from small to large, the number of the communication areas formed by the bright silver part is the largest, and the difference between the areas of the minimum communication area and the maximum communication area in the formed communication areas does not exceed 25 times. From a large number of picture analyses, it can be seen that the grayscale values of the R region are all above 120 degrees. Therefore, the threshold value is gradually increased to 255 from 100 by a certain step length, the preprocessed image is binarized by each threshold value, the area of the maximum communication area in each image is counted, and in order to eliminate the interference of the maximum value of some areas, the communication area in the image is reserved if the area of the communication area is larger than 1/25 of the area of the maximum communication area, otherwise, the communication area is discarded. And finally, counting the number of all the remaining communication areas in the image, selecting the image with the largest number of the communication areas as an optimal binary image, wherein the corresponding threshold value is the optimal binary threshold value, so that the R area can be extracted very accurately.
The center of the communication area is obtained by calculation according to all points in the communication area, and the calculation formula is as follows:
wherein, x and y are pixel point coordinates, and A is the total number of the pixel points in the area.
And calculating the center coordinate of each connected region in the binarized image obtained in the last step according to a formula. Because the binarized image is preprocessed and the image is obtained by reducing the original image, after the central coordinate of the link region in the binarized image is obtained, the corresponding central coordinate in the original image can be deduced according to the corresponding relation of pixel point coordinates in the image before and after reduction.
As shown in fig. 4, the original image is binarized by selecting a threshold value of 10, the center of the connected region obtained in the previous step is marked in the obtained binarized image, and a straight line 3 corresponding to each connected region is obtained by using the nearest pixel method (as shown in fig. 5). According to the set principle, a connecting line of a black pixel point closest to the central point 1 of the communication area and the central point 1 of the communication area is necessarily perpendicular to the boundary 2 of the communication area, and a straight line which passes through the central point 1 of the communication area and is perpendicular to the connecting line is a corresponding straight line 3 of the communication area. Since the method is a pixel-level operation, the higher the image resolution is, the higher the accuracy is, when the nearest pixel method is used for solving the straight line 3 corresponding to the connected region, the method is carried out on the binary image of the original image, and 10 with a smaller value is selected as the binary threshold.
As shown in fig. 6, the straight lines corresponding to all the connected regions are obtained according to the previous step. According to observation, the straight line corresponding to the communication area corresponding to the arm pipe with the pipe cap is larger than one straight line, and the arm pipe without the pipe cap only corresponds to one straight line corresponding to the communication area, so that all straight lines corresponding to the communication areas can be classified and screened. Grouping all the straight lines corresponding to the communication areas, grouping the straight lines corresponding to the communication areas with the similar slope into the same group, if no straight line corresponding to the communication areas with the similar slope exists, independently grouping the straight lines corresponding to the communication areas into one group, after grouping the straight lines corresponding to the communication areas, taking the right of the number of the straight lines corresponding to the communication areas in each group, leaving the group with the number of the straight lines corresponding to the communication areas being one, and reserving the group with the number of the straight lines corresponding to the communication areas being more than one. The arm pipes corresponding to the groups of which the number of the straight lines is more than one in the communication area are the arm pipes with pipe caps, and the arm pipes corresponding to the left straight lines in the communication area are the arm pipes without the pipe caps.
Because the boundary of the communication area is composed of pixels, the boundary is not a smooth straight line, the slope of the straight line corresponding to each communication area obtained through calculation has certain slight error, and due to the geometrical characteristics of the supporting device, two arm pipes are approximately parallel, the slope of the straight line corresponding to the obtained communication area can be very close, so that the communication areas can be mistakenly classified into one group. In order to determine whether the straight line corresponding to the communication area is the straight line corresponding to the same arm pipe, the straight line formed by connecting the centers of all the communication areas in pairs is used for judgment. The center of any two communication areas can determine a straight line, if the two communication areas are positioned on the same arm pipe, the slope of the straight line determined by the center of the communication area is very close to the slope of the straight line corresponding to the communication areas which are respectively corresponding to the two communication areas, and the straight lines corresponding to the two communication areas can be judged to belong to the same arm pipe; if the slope of the straight line determined by the center of the communication area is greatly different from the slopes of the straight lines corresponding to the communication areas which are respectively corresponding to the two communication areas, the straight lines corresponding to the communication areas which are respectively corresponding to the two communication areas belong to different arm tubes. By the method, whether the straight lines corresponding to the communication areas in the same group correspond to the same arm pipe or not can be further judged.
As shown in fig. 7, the straight lines corresponding to the communication areas corresponding to each remaining arm tube with the tube cap are merged, the average value of the slopes of the straight lines corresponding to each remaining set of communication areas is calculated, and the average value of the center coordinates of the center of the communication area corresponding to each set of straight lines is obtained. And taking the obtained mean value of the central coordinates as the central point of the combined straight line, and taking the obtained mean value of the slopes as the slopes of the combined straight line, thereby obtaining the combined unique straight line of each arm tube with the tube cap.
When the pipe cap is positioned, starting from the central point of the combined unique straight line, searching pixel by pixel from the straight line to two ends, stopping searching when the first black pixel point is met or the image range is exceeded, and recording the coordinates of the point. As shown in fig. 8 and 9, for the positioning of several typical types of the end points of the boom pipe, due to the incomplete shooting of the supporting device or the absence of the pipe cap part in the drawing, the pipe cap area needs to be further screened out from the positioned end points.
According to the analysis of a large number of pictures, the size of the supporting device is not changed greatly in all the pictures, the distance between the tail end of the pipe cap and the center of the nearest communication area on the arm pipe where the supporting device is located is within a certain range, and if the determined end point is not the end of the pipe cap, the distance between the end point and the center of the nearest communication area on the arm pipe where the supporting device is located is far larger than the range value. In this embodiment, since the size of the picture is 6600 × 4400, the distance between the end of the tube cap and the center of the nearest communication area on the arm tube where the tube cap is located is 30 to 280 (pixels), and if the determined end point is not the end of the tube cap, the distance between the end point and the center of the nearest communication area on the arm tube where the tube cap is located is far greater than 280 (pixels). Setting a distance value D =300, counting the distance between the end point obtained by positioning and the center of the nearest communication area on the arm tube where the end point is located, and if the distance is smaller than D, the end point is the end of the tube cap; if the distance is much greater than D or the endpoint coordinates are located at the edge of the image, the endpoint is dropped. By this method, the pipe cap can be accurately positioned, as shown in fig. 10, 11 and 12, and the pipe cap image finally intercepted by the obtained pipe cap end coordinate is obtained.
In order to analyze the performance of the method, the method is compared with the traditional Hough transformation method in the stage of calculating the corresponding straight line of the arm tube:
randomly selecting 50 different types of on-site supporting device pictures, wherein the pictures comprise supporting device images with different integrity degrees, images with different front and back positioning degrees and images with different fill light intensities, and performing tests by using two methods respectively, wherein the test result is shown in fig. 13.
From the experimental results it can be seen that: although the Hough transform detection method can detect the linear features in the image, the detected result is only the linear features of the edge part of the wrist arm, the detection effect is extremely poor, many false detection and missing detection situations exist, and the detection time is longer compared with that of the method. The method can well detect the straight line of the wrist arm, the straight line is positioned on the wrist arm, the straight line of the wrist arm can be well represented, the detection accuracy is far higher than that of the Hough method, and the detection time is also better than that of the Hough method. Therefore, the line detection method proposed herein is more suitable for cantilever straight line detection of the support device.
According to experiments, the algorithm can well complete the positioning of the pipe cap area, the detection result is not influenced by illumination, angle and the integrity of shooting by a supporting device, and the robustness is strong; the average processing time of each image in the aspect of detection time is less than 1s, so that the method meets the requirements of practical application and has higher practical prospect and research significance.
Claims (4)
1. A method for positioning a pipe cap of an electrified railway supporting device based on geometric characteristics is characterized by comprising the following steps:
a. inputting a global image of the supporting device, and inputting the global image of the supporting device obtained by shooting by the camera device into the image processing equipment;
b. preprocessing a global image of the supporting device, and performing filtering, denoising, histogram equalization and reduction operation on the image to enable the image to meet the requirement of subsequent image processing;
c. the method comprises the following steps of carrying out binarization on a preprocessed image, selecting a gray value smaller than a galvanized part in the image as a threshold value, comparing the gray value of each pixel point in the image with the magnitude relation of a set threshold value, if the gray value of the pixel point is larger than the threshold value, setting the pixel point to be white, otherwise, setting the pixel point to be black, obtaining a plurality of white connected regions, wherein the obtained white regions are extracted galvanized parts;
d. calculating the center of the communicated area, and calculating the center coordinate of each communicated area according to the obtained binary image;
e. calculating a straight line corresponding to each communication area, setting a threshold value of the original image to be 10, carrying out binarization processing, marking the central coordinates of the communication areas obtained in the step d on the binarized original image, and calculating the straight line corresponding to each communication area by using a nearest pixel method;
f. combining corresponding straight lines, grouping all straight lines corresponding to the communication areas according to slopes, dividing the straight lines corresponding to the communication areas with the slopes close to each other into a group, leaving out the group containing only one straight line corresponding to the communication areas, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of each group to belong to the arm pipe with the pipe cap, leaving the communication areas corresponding to the straight lines corresponding to the communication areas of the same group to belong to the same arm pipe, taking the average value of the slopes of the corresponding straight lines of each group of the remaining communication areas as the slope of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, taking the average value of the central coordinates of the communication areas corresponding to the corresponding straight lines of each group of the remaining communication areas as the central point of the unique straight line after the combination of the corresponding straight lines of each group of the communication areas, and obtaining the unique straight line corresponding to each arm pipe with the pipe cap;
g. and positioning the tube cap end, starting from the central point of the obtained corresponding unique straight line, searching the two ends of each arm tube with the tube cap pixel by pixel along the obtained unique straight line, stopping searching when the first black pixel point is encountered or the image range is exceeded, recording the coordinate of the searched last pixel point as an end point coordinate, discarding the end point coordinate of the non-tube cap end, and intercepting the tube cap area on the original drawing by taking the reserved end point coordinate as the center.
2. The method for positioning the pipe cap of the supporting device of the electrified railway based on the geometric characteristics as claimed in claim 1, wherein in the step b, the global image of the supporting device is denoised by adopting a bilateral filtering method, the image is subjected to contrast enhancement by adopting a contrast-limiting adaptive histogram equalization algorithm, and finally the image is reduced by using a near interpolation method to obtain the required image.
3. The method as claimed in claim 1, wherein in step c, the threshold is gradually increased from 100 to 255 in a certain step, each threshold is used to perform binarization processing on the preprocessed image, the area of the largest connected region in each binarized image is counted, the connected region with the area smaller than 1/25 of the area of the largest connected region in each image is discarded, the number of the reserved connected regions in each image is counted, and the binarized image with the largest number of connected regions is selected as the finally-obtained binarized image.
4. The method as claimed in claim 1, wherein the calculated central coordinates of each communication area are the coordinates of the reduced image in step d, the corresponding central coordinates in the original image are calculated according to the scaling relationship of the image, and the corresponding central coordinates in the original image are marked in step e.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910601591.0A CN110310276B (en) | 2019-07-05 | 2019-07-05 | Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910601591.0A CN110310276B (en) | 2019-07-05 | 2019-07-05 | Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110310276A CN110310276A (en) | 2019-10-08 |
| CN110310276B true CN110310276B (en) | 2022-11-18 |
Family
ID=68078017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910601591.0A Active CN110310276B (en) | 2019-07-05 | 2019-07-05 | Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110310276B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004004974A (en) * | 2003-08-25 | 2004-01-08 | Seiko Epson Corp | Image forming apparatus |
| CN104236498A (en) * | 2014-09-30 | 2014-12-24 | 中铁山桥集团有限公司 | Method for measuring end size of switch rail |
| CN106290393A (en) * | 2016-08-31 | 2017-01-04 | 西南交通大学 | A kind of high ferro contact net inclined cantilever pipe cap loss detection method |
| CN107169951A (en) * | 2016-03-03 | 2017-09-15 | 成都交大光芒科技股份有限公司 | A kind of the missing detection method and system of the inclined cantilever end pipe cap based on image |
| CN206703995U (en) * | 2017-05-08 | 2017-12-05 | 山东齐林电力设备股份有限公司 | Electrification railway contact net wire support N-type positioner |
| CN108297887A (en) * | 2017-09-26 | 2018-07-20 | 山西聚鑫顺科技有限公司 | A kind of lock device of fixed bow-backed transport closed-type semitrailer |
| CN109146852A (en) * | 2018-07-30 | 2019-01-04 | 国网江苏省电力有限公司电力科学院研究院 | A kind of porcelain insulating substring chapeau de fer disk edge detection method based on infrared image |
-
2019
- 2019-07-05 CN CN201910601591.0A patent/CN110310276B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004004974A (en) * | 2003-08-25 | 2004-01-08 | Seiko Epson Corp | Image forming apparatus |
| CN104236498A (en) * | 2014-09-30 | 2014-12-24 | 中铁山桥集团有限公司 | Method for measuring end size of switch rail |
| CN107169951A (en) * | 2016-03-03 | 2017-09-15 | 成都交大光芒科技股份有限公司 | A kind of the missing detection method and system of the inclined cantilever end pipe cap based on image |
| CN106290393A (en) * | 2016-08-31 | 2017-01-04 | 西南交通大学 | A kind of high ferro contact net inclined cantilever pipe cap loss detection method |
| CN206703995U (en) * | 2017-05-08 | 2017-12-05 | 山东齐林电力设备股份有限公司 | Electrification railway contact net wire support N-type positioner |
| CN108297887A (en) * | 2017-09-26 | 2018-07-20 | 山西聚鑫顺科技有限公司 | A kind of lock device of fixed bow-backed transport closed-type semitrailer |
| CN109146852A (en) * | 2018-07-30 | 2019-01-04 | 国网江苏省电力有限公司电力科学院研究院 | A kind of porcelain insulating substring chapeau de fer disk edge detection method based on infrared image |
Non-Patent Citations (3)
| Title |
|---|
| Identification Algorithm of Electrical Equipment Type and Fault State Based on FastPCA and SVM;Shuohe Wang等;《2019 IEEE 3rd Information Technology,Networking,Electronic and Automation Control Conference》;20190606;第33-37页 * |
| 基于ANSYS的高速铁路接触网系统静态找形研究;李安;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20131125;第C033-87页 * |
| 基于机器视觉的接触网巡检关键技术研究;吴光龙;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20170515;第C033-21页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110310276A (en) | 2019-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110163853B (en) | Edge defect detection method | |
| CN108596165B (en) | Road traffic marking detection method and system based on unmanned plane low latitude Aerial Images | |
| CN112232391B (en) | Dam crack detection method based on U-net network and SC-SAM attention mechanism | |
| CN109325935B (en) | Power transmission line detection method based on unmanned aerial vehicle image | |
| CN110930357B (en) | In-service steel wire rope surface defect detection method and system based on deep learning | |
| EP3296953A1 (en) | Method and device for processing depth images | |
| CN110321933B (en) | Fault identification method and device based on deep learning | |
| CN111814686A (en) | Vision-based power transmission line identification and foreign matter invasion online detection method | |
| CN106446894A (en) | Method for recognizing position of spherical object based on contour | |
| CN107358628B (en) | Linear array image processing method based on target | |
| CN111814919A (en) | Instrument positioning and identifying system based on deep learning | |
| CN110288571B (en) | High-speed rail contact net insulator abnormity detection method based on image processing | |
| CN110232682B (en) | Image-based track foreign matter detection method | |
| CN107133623A (en) | A kind of pointer position accurate detecting method positioned based on background subtraction and the center of circle | |
| CN109829911A (en) | A kind of pcb board surface inspecting method based on the overproof algorithm of profile | |
| CN110866915A (en) | Circular inkstone quality detection method based on metric learning | |
| CN105678737A (en) | Digital image corner point detection method based on Radon transform | |
| CN110956078B (en) | Power line detection method and device | |
| CN104715250A (en) | Cross laser detection method and device | |
| CN113657339A (en) | Instrument pointer counting and reading method and medium based on machine vision | |
| CN110310276B (en) | Method for positioning pipe cap of supporting device of electrified railway based on geometric characteristics | |
| CN112132820A (en) | Catenary positioning pipe stay cable fault detection method based on YOLOv2 algorithm | |
| Zhu et al. | Fast and robust 2D-shape extraction using discrete-point sampling and centerline grouping in complex images | |
| CN112991327B (en) | Steel grid welding system, method and terminal equipment based on machine vision | |
| CN111161264B (en) | Method for segmenting TFT circuit image with defects |
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 |