CN108629347A - The wide tuning drive gear formula instrument automatic identifying method of substation - Google Patents
The wide tuning drive gear formula instrument automatic identifying method of substation Download PDFInfo
- Publication number
- CN108629347A CN108629347A CN201810207331.0A CN201810207331A CN108629347A CN 108629347 A CN108629347 A CN 108629347A CN 201810207331 A CN201810207331 A CN 201810207331A CN 108629347 A CN108629347 A CN 108629347A
- Authority
- CN
- China
- Prior art keywords
- rectangle
- scale
- instrument
- drive gear
- pointer
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/02—Recognising information on displays, dials, clocks
Abstract
The present invention relates to a kind of wide tuning drive gear formula instrument automatic identifying method of substation, digital calibration form is rectangle in the width tuning drive gear formula instrument, and 1 is divided between digital calibration；It is characterized in that：Including following key step：Instrument image is read in first, and image is pre-processed；Then it identifies meter dial, including each rectangle meter dial in image is automatically positioned, and carry out scale radial angle sequence, realize the foundation of meter dial coordinate system；Then, it is identified and is positioned into line pointer with based on LSD line detection algorithms and the oval circumferential confinement of scale distribution；With finally realizing instrument automatic interpretation, is read out registration.Subtract the dependence to people in identification process, identification process does not need the input of additional meter dial parameter, under the premise of ensureing accuracy of identification, with realizing width tuning drive gear formula instrument Rapid reading, it can be used for Intelligent Mobile Robot and execute patrol task, have preferable intelligent.
Description
Technical field
The present invention relates to the wide tuning drive gear formula instrument in image steganalysis field more particularly to a kind of substation to know automatically
Other method.
Technical background
Although digital instrument, using more and more extensive, for pointer instrument because of its strong antijamming capability, registration identification can
The features such as high by property, is still widely used in the substation of bad environments.Pointer instrument does not have data line interface usually
Function can not be read in automatic identification and return table.Therefore, the data of record pointer formula instrument are mainly by human eye short distance
Observation and manually record, repeat mechanical movement easily cause human visual fatigue, cause identification data, record data go out
Existing mistake.With propulsion of the automatic technology in substation, pointer instrument in substation is acquired using crusing robot
Picture, then have become a kind of trend from the registration in automatic identification table in picture.
For the automatic identification of pointer instrument, in the positioning of instrument dial plate and pointer identification usually using template matching method
With Hough transform method.For template matching method, need first manually to come out meter calibration on the picture containing instrument, then
The instrument of calibration is matched on picture to be identified, has not only limited to the type of instrument in this way, but also to manually demarcating dependence
Greatly, it is unfavorable for the development of the automation of transformation substations.For Hough transform, traditional Hough transform algorithm operation time is long, stores
Space is big so that identification process inefficiency.
Due to being influenced by light and crusing robot shooting angle, pointer in substation is acquired using crusing robot
When the picture of formula instrument, the instrument in picture is generally stretched prior, to be affected to the identification of instrument board and pin-point reading.
Invention content
This paper technical problems to be solved are：In view of the above technical problems, a kind of wide tuning drive gear formula instrument of substation is proposed
Table automatic identifying method, this method is using image processing techniques reconstruct dial plate and pointer, to reach the mesh for accurately identifying registration
's.The dependence to people in identification process is subtracted, identification process does not need additional meter dial parameter input, is ensureing identification essence
Under the premise of degree, Rapid reading, can be used for Intelligent Mobile Robot and executes patrol task with realizing wide tuning drive gear formula instrument,
With preferable intelligent.
The present invention is in order to solve the above technical problems, adopt the following technical scheme that：
A kind of wide tuning drive gear formula instrument automatic identifying method of substation, digital calibration in the width tuning drive gear formula instrument
Form is rectangle, and 1 is divided between digital calibration；It is characterized in that：Including following key step：Instrument image is read in first, to figure
As being pre-processed；Then it identifies meter dial, including each rectangle meter dial in image is automatically positioned, and carved
Radial angle sequence is spent, realizes the foundation of meter dial coordinate system；Then, with based on LSD line detection algorithms and scale distribution
Oval circumferential confinement is identified into line pointer and is positioned；With finally realizing instrument automatic interpretation, is read out registration.
Further, in the above method, described image pretreatment includes image gray processing and gaussian filtering.
Further, in the above method, identification meter dial specific method includes following three steps：
Step S11：Using Canny operator edge detections, obtain the edge contour information of meter dial, to edge contour into
8 neighborhood Contour extraction of row, obtains continuous edge contour；Qualified rectangle is filtered out by constraints again；It is described about
Beam condition includes：In the number of vertex of edge contour, profile elemental area, profile be flange shape or concave edge shape and profile most
Small angle；
Step S12：The centre distance d between the rectangle of all qualified meters dial is calculated, center is worked as in d≤30
When distance is less than a certain threshold value, it is believed that be the rectangle repeated, wiped；
Step S13：It takes the central point of the rectangle of each meter dial to carry out least square ellipse fitting, calculates each number and carve
The central point of rectangle is spent to the vectorial angle of elliptical center, and the sequence of tape index is carried out to each angle.
Further, in the above method, pointer identification positioning specifically includes following two steps：
Step S21：Since the ellipse of fitting is approximately round, using the mean value of oval major semiaxis and semi-minor axis as circle
Radius r, using the Line Segment Detection Algorithm detection based on LSD by pre-processing all line segments in obtained instrument gray-scale map, with
It crosses elliptical center and length l is in a certain threshold range, takeLine segment is screened for constraint, acquisition meets item
All line segment lines_std of part；
Step S22：The midpoint pt_m for calculating all line segment lines_std, asked the straight line of all line segment lines_std with
Elliptical intersection point takes the savepoint pt_t close from midpoint pt_m；The mean value of all savepoint pt_t is exactly pointer and elliptical friendship
Point pt_result.
Further, it in the above method, reads registration and specifically includes following three steps：
Step S31：Seek point pt_ on the basis of the midpoint on the instrument dial plate of ellipse fitting between 0 scale and n scales
base；N is maximum scale value in instrument；
Step S32：Obtain two angles and three points for describing each digital calibration；
Wherein three points include the central point pt1 of each digital calibration rectangle, the nearly datum mark pt_ of each digital calibration rectangle
The midpoint pt2 of the sides the base and midpoint pt3 of the sides each digital calibration rectangle remote datum mark pt_base；
Two angles include：1, datum mark pt_base is directed toward the vector at the center of fitted ellipse to each digital calibration square
It is directed toward the angle ang1 between the vector of elliptical center in the midpoint of the sides shape nearly datum mark pt_base；2, datum mark pt_base is directed toward
It is directed toward the vector of elliptical center in the midpoint of the vector at the center of fitted ellipse to the sides each scale rectangle remote datum mark pt_base
Between angle ang2；
Step S33：Calculating benchmark point pt_base is directed toward the vector at the center of fitted ellipse to pointer and elliptical intersection point
Pt_result is directed toward the angle ang_result between the vector of elliptical center；When pointer refers on scale rectangle, output should
The value of digital calibration rectangle is meter reading；When pointer refers between two digital calibration rectangles, export compared with decimal fractions scale
Value and pointer whereabouts account for the ratio of the angle between two digital calibrations and be meter reading.
In said program, the reading of image, gray processing processing, gaussian filtering, Canny algorithms, 8 neighborhood Contour extractions are calculated
Method, polygonal segments algorithm, the line detection algorithm based on LSD, the ellipse fitting based on least square are all based on OpenCV
The library function in library.
For the automatic identifying method of existing pointer instrument, the mould used in the positioning of instrument dial plate and pointer identification
Plate matching method and Hough transform method template matching method need first manually to come out meter calibration on the picture containing instrument, then
The instrument of calibration is matched on picture to be identified, has not only limited to the type of instrument in this way, but also to manually demarcating dependence
Property it is big, be unfavorable for the development of the automation of transformation substations, and traditional Hough transform algorithm operation time is long, storage space is big so that
Identification process inefficiency.A kind of automatic identifying method of the wide tuning drive gear formula instrument of substation proposed by the present invention, uses one
Kind improved ellipse fitting method based on least square is reduced in the identification of substation's pointer instrument to the dependence of people,
The automatic reading for realizing wide tuning drive gear formula instrument, proposes a kind of improved line detection method based on LSD algorithm, improves wide
The speed and precision of tuning drive gear formula Meter recognition.Since the measurement range of the wide tuning drive gear formula instrument of substation is 0~n,
Scale interval is 1, and the present invention only needs to identify the rectangle where scale, and gives rectangle sequence assignment, need not establish template
Number corresponding with identification scale, significantly simplifies program, improves the speed of service.Meanwhile identifying instrument rectangle quarter
Under the premise of degree, meter dial coordinate system is established based on meter dial and radial angle correspondence, the process of foundation is only
It is realized by primary simple sequence, independent of dial digital identification, process is simpler, the effect more Shandong of identification
Stick.
Description of the drawings
Fig. 1 is that the angle of the meter dial of the present invention describes schematic diagram；
Fig. 2 is the wide tuning drive gear formula instrument automatic identification overall procedure exemplary plot of the present invention；
Fig. 3 is the meter dial rectangle recognizer flow chart of the present invention；
Fig. 4 is gauge pointer identification and the pointer visualization procedure flow chart of the present invention；
Fig. 5 is the program flow diagram of the automatic identification total indicator reading of the present invention.
Specific implementation method
Technical solution in order to further illustrate the present invention carries out specifically this programme below in conjunction with attached drawing 2-5
It is bright.
According to the wide tuning drive gear formula instrument automatic identifying method of substation that the present invention is implemented, specific implementation such as Fig. 2 institutes
Show, including following flow：
Step S1：Meter dial rectangle is automatically positioned and scale visualization, as shown in Figure 3：
Step S11：Instrument edge contour is extracted using canny operator edge detections and 8 neighborhood contour following algorithms；
Step S111：Gray processing is carried out to original image：Gray processing is carried out to a width RGB figures, is exactly each according to image
The sampled value in channel is weighted average.
Method one：Gray=(R+G+B)/3；
Method two：Gray=0.299R+0.587G+0.114B；
Wherein, R, G, B respectively represent the color value in three channels of red, green, blue in coloured image.
Step S112：Gaussian filtering is carried out to image：The principle of gaussian filtering is rolled up to image using Gauss operator
Product operation, image gaussian filtering can be realized with convolution of a two-dimensional Gaussian kernel, as shown in formula (1)：
σ is variance in formula.Entire Gaussian kernel is normalized after acquiring Gaussian kernel；
Step S113：Calculate amplitude and the direction of gradient：The gradient of gray value of image can be used first difference point come into
Row is approximate, can obtain two matrixes of image partial derivative in the x and y direction in this way.Convolution operator used by Canny algorithms
As shown in formula (2)：
Its x to y to first-order partial derivative matrix, mathematic(al) representation such as formula (3) institute of gradient magnitude and gradient direction
Show：
In formula (3), p [i, j] is x direction gradient amplitudes, and Q [i, j] is y direction gradient amplitudes, and M [i, j] is the amplitude,
θ [i, j] is gradient direction；
Step S114：Non-maxima suppression is carried out to gradient magnitude：Pixel local maximum is found, it will be all non-very big
The corresponding gray value of value point is set to 0, can reject the point of most of non-edge in this way；
Step S115：Dual threshold detects and edge connection：Two threshold values of selection (choose the maximum in preceding 79% gray value
Gray value be high threshold, Low threshold is about the half of high threshold), an edge image is obtained according to high threshold, due to threshold
Value is higher, and image contains only a small amount of false edge, but also causes image border that may not be closed.Edge is connected in high threshold image
It is connected into profile, when reaching the endpoint of profile, which can find the point for meeting Low threshold, then root in 8 neighborhood points of breakpoint
Point collects new edge and can be obtained by instrument edge contour in this way until whole image edge closure accordingly.
Step S116：8 neighborhood Contour extractions：Since the edge contour obtained with canny operators is not necessarily continuously, institute
To need to obtain continuous edge contour with 8 neighborhood contour following algorithms.8 neighborhood contour following algorithms are 8 neighborhoods for utilizing point
Information selects next point as boundary point.Select a starting point, can select image most on most left point be target point,
Then the point for checking its 8 neighborhoods is begun look for from 45 ° of the position in lower right, if it is target point, will be made along 90 ° clockwise
Process above is repeated if finding, directly if it is not, then being continually looked for along 45 ° counterclockwise for the direction found next time
It traverses and completes to all the points, so that it may to obtain continuous edge contour.
Step S12：Meter dial positions：First, the continuous edge contour obtained in S11 is subjected to polygonal segments,
Obtain the minimum external contact zone of all edge contours；Then, with the number of edges of external contact zone (number of vertex) rect_v, polygon
Shape (convex-edge shape or concave edge shape) rect_t, the area (unit of polygon：Pixel) rect_s and polygon Minimum Internal Angle
Rect_c is constraint, works as rect_v=4, and rect_t is convex-edge shape, ＜=500 rect_s ＞=200＆rect_s (pixel) and
When rect_c=90 (degree), it is believed that polygon is rectangle, and the rectangle for the condition that meets is saved in rects.
Step S13：Reject the rectangle of overlapping：First, center point coordinate (four tops of each rectangle in step S12 are calculated
Point coordinates is averaged) pt_ms；Then, the distance d between calculating rectangular centre point two-by-two, if ＞=30 d, then it is assumed that be overlapping
Rectangle, then the rectangle of overlapping is rejected, obtains scale rectangle rect_result, thereby realize to instrument carve
The automatic positioning of degree ground；
Step S14：Meter dial visualizes：First, base is carried out with the center point coordinate pt_ms of each rectangle in S13
In the ellipse fitting of least square, instrument dial plate is visualized；Then, elliptical center (x is calculatedc, yc) arrive rectangular centre coordinate pt_
The azimuth (angular range (0,2 π]) of ms；Finally, azimuth is carried out to the sequence of tape index, you can obtain meter dial arrangement
Coordinate system；
Step S2：Gauge pointer positions, as shown in Figure 4；
Step S21：Detect line segment：The gray-scale map of the instrument picture generated in input step S1 is calculated using LSD straight-line detections
Method (a Line Segment Detector, line segment detector) detects all line segment lines_std in gray-scale map；
Step S211：Image down sampling：It is the 80% of original size by the image down sampling of input, sampling purpose is to subtract
It is weak even to be eliminated the sawtooth effect occurred in many images；
Step S212：Gradient calculates：Image gradient is calculated in 2 × 2 template, and masterplate such as formula (4) is given
It is shown：
In formula (4), i (x, y), i (x+1, y), i (x, y+1) and i (x+1, y+1) be respectively on gray level image point (x,
Y), the gradient calculation formula of (x+1, y), the pixel value on (x, y+1) and (x+1, y+1), image are as shown in formula (5)：
G in formula (5)x(x, y), gy(x, y) respectively indicate the position (x, y) pixel gradient x-component and y-component, i (x+1,
Y), i (x+1, y+1), i (x, y), i (x, y+1) indicate on gray level image respectively at point (x+1, y), (x+1, y+1), (x, y),
Pixel value on (x, y+1)；
The Grad of picture pixels is calculated by formula (6) and (7) respectively with gradient direction：
In formula (6) and (7), G (x, y) indicates that the Grad of the position (x, y) pixel, ang (x, y) indicate the position (x, y) picture
The gradient direction of element；
Step S213：Gradient puppet sorts：The gradient of usual picture point is bigger, which is that the probability of marginal point is also bigger, because
This, is equally spaced 1024 bin, by the pixel on instrument gray-scale map by the gray-scale map that pixel value range is (0,255)
It is categorized on this 1024 bin, this completes gradient puppet sorting operations；
Step S214：Line segment region increases：A not used picture is selected from the sorted lists obtained in step S213
Vegetarian refreshments is seed point, carries out region growth, other not used points are used as test point, when seed point angular deviation is more than a threshold
When value τ (taking τ=π/8 (radian) here), stop area increases；
Step S215：Rectangle is estimated：Line segmentation corresponds to each rectangle, the regional perspective point of rectangular centre point and rectangle
It is not calculated by formula (8) and (9)：
In formula, cxAnd cyIndicate that the x at rectangular area center, y-coordinate, G (j) indicate the gradient of the j point in rectangular area
Value, region_Ang and AngjThe gradient direction of j-th of pixel in rectangular area angle and rectangular area is indicated respectively；
Step S216：Rectangle adjusts：Calculate snap point in the minimum enclosed rectangle in each region described in step S215
The density of (aligned point) reduces region growing angle threshold or reduces radius, so if density is unsatisfactory for requiring
NFA (Number of False Alarms report number by mistake) value is recycled to be adjusted rectangle afterwards；
Step S22：Screen gauge pointer line segment：First, although the meter diagram piece of shooting has stretching, what is be fitted is ellipse
Circle instrument dial plate is still approximately round, radius r of the elliptical major semiaxis of digital simulation with the mean value of semi-minor axis as approximate circle；
Then, made with the line segment lines_std and the long line_l of approximate round heart distance line_d and line segment that are obtained in step S21
For constraints, line segment is screened, wherein constraints is distanceAnd line segment lengthThen think that the line segment lines_result for meeting constraint on gauge pointer, deletes ineligible line
Section；
Step S23：It determines the intersection point of pointer and instrument dial plate, visualizes pointer：First, the friendship of pointer and instrument dial plate is asked
Point needs to seek elliptical expression equation, shown in elliptical General Expression equation such as formula (10)：
Ax2+Bxy+Cy2+ Dx+Ey+f=0 (10)
It is converted into shown in oblique elliptic equation such as formula (11)：
In formula (10) and (11), (x, y) is the point on ellipse, and θ is oval deviation angle.
Then, it determines that straight line method calculates the linear equation by line segment lines_result using 2 points, calculates
The midpoint pt_m of line segment lines_result；Finally, simultaneous elliptic equation and linear equation, calculate straight line on pointer with it is ellipse
Round intersection point takes the point pts that offline section midpoint pt_m is close；Finally, the mean value for seeking point pts obtains the intersection point pt_ of pointer and dial plate
Result draws the straight line between intersection point pt_result and the center of circle, visualizes pointer；
Step S3：Automatic identification meter reading, as shown in Figure 5；
Step S31：Set datum mark：The point pt_base on the basis of the midpoint on ellipse between 0 scale and n scales, it is first
First, the vector that 0 scale rectangular centre point is directed toward the vector sum n scale rectangular centre points direction elliptical center of elliptical center is obtained,
The angle ang_2pt between two vectors is calculated, angle ang_pt=ang_2pt/2 where datum mark；Then, angle is utilized
Ang_pt and elliptical center point (xc, yc), seek linear equation using point slope form method；Finally ask straight according to the method in step S23
Line and elliptical intersection point, the point for taking 0 scale rectangle midpoint closer is exactly datum mark pt_base；
Step S32：Rectangle scale is described：Each scale rectangle is described with two angles and three points, two angles include：
Angle of the datum mark to the angle ang1 at the midpoint of the nearly datum mark side of each scale rectangle and the midpoint of remote datum mark side on ellipse
Ang2 can find out ang1 and ang2 using the method in step S31；Three points include the midpoint pt1 of each scale rectangle, are carved
Spend the midpoint pt3 of the midpoint pt2 and the remote datum mark side of scale rectangle of the nearly datum mark side of rectangle；As shown in Figure 1；
Step S33：Instrument automatic reading：First, the angle of the method calculating pointer in step S31 to datum mark is utilized
ang_result；Then, scale n-1 is arrived for scale 0, as ang_result ＞ ang1 and ang_result ＜ ang2, i.e.,
For pointer on scale rectangle, the value of scale where output is reading.As ang_result ＞ ang2 and ang_result ＜ ang1
When, i.e. pointer calculates the ratio of pointer position angle between two scales, i.e. (ang_result- between two scale rectangles
Ang2)/(ang1-ang2), the value of scale is with (ang_result-ang2)/(ang1-ang2's) and to read where output ang2
Number.For n scales, as ang_result ＞ ang1 and ang_result ＜ ang2, i.e. pointer exports institute on scale rectangle
It is reading in the value of scale.As ang_result ＞ ang2, output no to scale warning is so far completed meter reading and is known automatically
Not.
In said program, the automatic identifying method of above-mentioned substation's pointer instrument, 64 bit manipulations based on windows 7
The libraries OpenCV of increasing income that Visual Studio 2013 and version under system are 3.1.0.
In said program, the reading of picture, Canny operator edge detections, polygonal segments, is based on most gray processing processing
Small two multiply ellipse fitting, LSD line detection algorithms, be all based on the library function in the libraries OpenCV.
Claims (6)
1. a kind of wide tuning drive gear formula instrument automatic identifying method of substation, digital calibration shape in the width tuning drive gear formula instrument
State is rectangle, and 1 is divided between digital calibration；It is characterized in that：Including following key step：Instrument image is read in first, to image
It is pre-processed；Then it identifies meter dial, including each rectangle meter dial in image is automatically positioned, and carry out scale
Radial angle sorts, and realizes the foundation of meter dial coordinate system；Then, with ellipse based on LSD line detection algorithms and scale distribution
Circle circumferential confinement is identified into line pointer and is positioned；With finally realizing instrument automatic interpretation, is read out registration.
2. the wide tuning drive gear formula instrument automatic identifying method of substation according to claim 1, it is characterised in that：The figure
As pretreatment includes image gray processing and gaussian filtering.
3. the wide tuning drive gear formula instrument automatic identifying method of substation according to claim 1, it is characterised in that：Identifier
Table scale specifically includes following three steps：
Step S11：Using Canny operator edge detections, the edge contour information of meter dial is obtained, 8 are carried out to edge contour
Neighborhood Contour extraction obtains continuous edge contour；Qualified rectangle is filtered out by constraints again；The constraint item
Part includes：Elemental area, profile are that flange shape or concave edge shape and profile are minimum in the number of vertex of edge contour, profile
Angle；
Step S12：The centre distance d between the rectangle of all qualified meters dial is calculated, centre distance is worked as in d≤30
When less than a certain threshold value, it is believed that be the rectangle repeated, wiped；
Step S13：It takes the central point of the rectangle of each meter dial to carry out least square ellipse fitting, calculates each digital calibration square
The central point of shape carries out each angle the sequence of tape index to the angle of the vector of elliptical center.
4. the wide tuning drive gear formula instrument automatic identifying method of substation according to claim 3, it is characterised in that：Pointer is known
Following two steps Ding Wei not specifically included：
Step S21：Since the ellipse of fitting is approximately round, using the mean value of oval major semiaxis and semi-minor axis as the radius of circle
R, using the Line Segment Detection Algorithm detection based on LSD by pre-processing all line segments in obtained instrument gray-scale map, with excessively ellipse
Circle center and length l is in a certain threshold range, takesLine segment is screened for constraint, acquisition meets condition
All line segment lines_std；
Step S22：The midpoint pt_m for calculating all line segment lines_std asked the straight line and ellipse of all line segment lines_std
Intersection point, take the savepoint pt_t close from midpoint pt_m；The mean value of all savepoint pt_t is exactly pointer and elliptical intersection point pt_
result。
5. the wide tuning drive gear formula instrument automatic identifying method of substation according to claim 4, it is characterised in that：Reading is shown
Number specifically includes following three steps：
Step S31：Seek point pt_base on the basis of the midpoint on the instrument dial plate of ellipse fitting between 0 scale and n scales；N is
Maximum scale value in instrument；
Step S32：Obtain two angles and three points for describing each digital calibration；
Wherein three points include the central point pt1 of each digital calibration rectangle, the nearly datum mark pt_base of each digital calibration rectangle
The midpoint pt2 of the side and midpoint pt3 of each sides digital calibration rectangle remote datum mark pt_base；
Two angles include：1, the vector at the center of datum mark pt_base directions fitted ellipse is close to each digital calibration rectangle
It is directed toward the angle ang1 between the vector of elliptical center in the midpoint of the sides datum mark pt_base；2, datum mark pt_base is directed toward fitting
It is directed toward between the vector of elliptical center at the midpoint of the vector at elliptical center to the sides each scale rectangle remote datum mark pt_base
Angle ang2；
Step S33：Calculating benchmark point pt_base is directed toward the vector at the center of fitted ellipse to pointer and elliptical intersection point pt_
Result is directed toward the angle ang_result between the vector of elliptical center；When pointer refers on scale rectangle, the number is exported
The value of word scale rectangle is meter reading；When pointer refers between two digital calibration rectangles, the value compared with decimal fractions scale is exported
With pointer whereabouts account for the ratio of the angle between two digital calibrations and be meter reading.
6. the wide tuning drive gear formula instrument automatic identifying method of substation according to claim 5, it is characterised in that：Image
Reading, gray processing processing, gaussian filtering, Canny algorithms, 8 neighborhood contour following algorithms, polygonal segments algorithm, based on LSD's
Line detection algorithm, the ellipse fitting based on least square are all based on the library function in the libraries OpenCV.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810207331.0A CN108629347B (en) | 2018-03-14 | 2018-03-14 | Automatic identification method for wide-scale pointer instrument of transformer substation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810207331.0A CN108629347B (en) | 2018-03-14 | 2018-03-14 | Automatic identification method for wide-scale pointer instrument of transformer substation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108629347A true CN108629347A (en) | 2018-10-09 |
| CN108629347B CN108629347B (en) | 2021-12-21 |
Family
ID=63706239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810207331.0A Active CN108629347B (en) | 2018-03-14 | 2018-03-14 | Automatic identification method for wide-scale pointer instrument of transformer substation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108629347B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110490145A (en) * | 2019-08-22 | 2019-11-22 | 国网四川省电力公司信息通信公司 | A kind of readings of pointer type meters recognition methods |
| CN110503144A (en) * | 2019-08-15 | 2019-11-26 | 南京天创电子技术有限公司 | A kind of pointer instrument recognition methods for crusing robot |
| CN112232209A (en) * | 2020-10-16 | 2021-01-15 | 蓝普金睛（北京）科技有限公司 | Pointer type instrument panel reading identification method for transformer substation inspection robot |
| CN113628131A (en) * | 2021-07-22 | 2021-11-09 | 济南驰昊电力科技有限公司 | Intelligent identification method for pointer type oil level meter of transformer substation in foggy weather scene |
| CN113723179A (en) * | 2021-07-19 | 2021-11-30 | 湘潭大学 | Digital display instrument data recording method using target tracking algorithm |
| US20220284690A1 (en) * | 2018-03-23 | 2022-09-08 | Kabushiki Kaisha Toshiba | Reading system, reading method, and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1693852A (en) * | 2005-01-28 | 2005-11-09 | 华南理工大学 | Instrument pointer automatic detection identification method and automatic reading method |
| CN101620682A (en) * | 2008-06-30 | 2010-01-06 | 汉王科技股份有限公司 | Method and system for automatically identifying readings of pointer type meters |
| WO2013035890A2 (en) * | 2011-09-07 | 2013-03-14 | Yazaki Corporation | Speed displaying device and speed displaying method |
| CN107145890A (en) * | 2017-05-02 | 2017-09-08 | 南通大学 | A kind of pointer dashboard automatic reading method under remote various visual angles environment |
| CN107590498A (en) * | 2017-09-27 | 2018-01-16 | 哈尔滨工业大学 | A kind of self-adapted car instrument detecting method based on Character segmentation level di- grader |
| CN107729853A (en) * | 2017-10-24 | 2018-02-23 | 武汉科技大学 | A kind of automatic identifying method suitable for the narrow tuning drive gear formula instrument of transformer station |
| CN107729896A (en) * | 2017-10-24 | 2018-02-23 | 武汉科技大学 | A kind of automatic identifying method suitable for the wide tuning drive gear formula instrument of transformer station |
-
2018
- 2018-03-14 CN CN201810207331.0A patent/CN108629347B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1693852A (en) * | 2005-01-28 | 2005-11-09 | 华南理工大学 | Instrument pointer automatic detection identification method and automatic reading method |
| CN101620682A (en) * | 2008-06-30 | 2010-01-06 | 汉王科技股份有限公司 | Method and system for automatically identifying readings of pointer type meters |
| WO2013035890A2 (en) * | 2011-09-07 | 2013-03-14 | Yazaki Corporation | Speed displaying device and speed displaying method |
| CN107145890A (en) * | 2017-05-02 | 2017-09-08 | 南通大学 | A kind of pointer dashboard automatic reading method under remote various visual angles environment |
| CN107590498A (en) * | 2017-09-27 | 2018-01-16 | 哈尔滨工业大学 | A kind of self-adapted car instrument detecting method based on Character segmentation level di- grader |
| CN107729853A (en) * | 2017-10-24 | 2018-02-23 | 武汉科技大学 | A kind of automatic identifying method suitable for the narrow tuning drive gear formula instrument of transformer station |
| CN107729896A (en) * | 2017-10-24 | 2018-02-23 | 武汉科技大学 | A kind of automatic identifying method suitable for the wide tuning drive gear formula instrument of transformer station |
Non-Patent Citations (2)
| Title |
|---|
| RAFAEL GROMPONE VON GIOI等: ""LSD: A Fast Line Segment Detector with a False Detection Control"", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 》 * |
| 张文杰等: ""基于视觉显著性的指针式仪表读数识别算法"", 《计算机辅助设计与图形学学报》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220284690A1 (en) * | 2018-03-23 | 2022-09-08 | Kabushiki Kaisha Toshiba | Reading system, reading method, and storage medium |
| CN110503144A (en) * | 2019-08-15 | 2019-11-26 | 南京天创电子技术有限公司 | A kind of pointer instrument recognition methods for crusing robot |
| CN110490145A (en) * | 2019-08-22 | 2019-11-22 | 国网四川省电力公司信息通信公司 | A kind of readings of pointer type meters recognition methods |
| CN110490145B (en) * | 2019-08-22 | 2021-09-17 | 国网四川省电力公司信息通信公司 | Pointer instrument reading identification method |
| CN112232209A (en) * | 2020-10-16 | 2021-01-15 | 蓝普金睛（北京）科技有限公司 | Pointer type instrument panel reading identification method for transformer substation inspection robot |
| CN113723179A (en) * | 2021-07-19 | 2021-11-30 | 湘潭大学 | Digital display instrument data recording method using target tracking algorithm |
| CN113723179B (en) * | 2021-07-19 | 2023-07-28 | 湘潭大学 | Digital display instrument data recording method using target tracking algorithm |
| CN113628131A (en) * | 2021-07-22 | 2021-11-09 | 济南驰昊电力科技有限公司 | Intelligent identification method for pointer type oil level meter of transformer substation in foggy weather scene |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108629347B (en) | 2021-12-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108629347A (en) | The wide tuning drive gear formula instrument automatic identifying method of substation | |
| CN106951900B (en) | A kind of automatic identifying method of arrester meter reading | |
| CN107590498B (en) | Self-adaptive automobile instrument detection method based on character segmentation cascade two classifiers | |
| CN103034838B (en) | A kind of special vehicle instrument type identification based on characteristics of image and scaling method | |
| CN110807355B (en) | Pointer instrument detection and reading identification method based on mobile robot | |
| CN102704215B (en) | Automatic cutting method of embroidery cloth based on combination of DST file parsing and machine vision | |
| CN108121985A (en) | A kind of double pointer meters number reading methods based on machine vision | |
| CN108921176A (en) | A kind of pointer instrument positioning and recognition methods based on machine vision | |
| CN104008399B (en) | The recognition methodss of the gauge pointer shake based on support vector machine in a kind of instrument detection | |
| CN109214420A (en) | The high texture image classification method and system of view-based access control model conspicuousness detection | |
| CN110148162A (en) | A kind of heterologous image matching method based on composition operators | |
| CN108319973A (en) | Citrusfruit detection method on a kind of tree | |
| CN110286124A (en) | Refractory brick measuring system based on machine vision | |
| WO2003021533A9 (en) | Color image segmentation in an object recognition system | |
| CN109635806A (en) | Ammeter technique for partitioning based on residual error network | |
| CN112699876A (en) | Automatic reading method for various meters of gas collecting station | |
| Veras et al. | Discriminability tests for visualization effectiveness and scalability | |
| CN110057828A (en) | A kind of surface inspecting method and system based on machine vision | |
| CN107609564A (en) | Submarine target image-recognizing method based on joint segmentation and Fourier descriptor storehouse | |
| CN108154496B (en) | Electric equipment appearance change identification method suitable for electric power robot | |
| CN106709523B (en) | Optical remote sensing image ship identification method based on S-HOG characteristics | |
| CN106023201A (en) | Rectangular transparent plastic bag-oriented dimension measurement method | |
| Khamael et al. | Using adapted JSEG algorithm with fuzzy C mean for segmentation and counting of white blood cell and nucleus images | |
| CN109784257B (en) | Transformer thermometer detection and identification method | |
| Hori et al. | Quantitative measurement of the performance of raster-to-vector conversion algorithms |
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 |