CN107239742A - A kind of gauge pointer scale value calculating method - Google Patents
A kind of gauge pointer scale value calculating method Download PDFInfo
- Publication number
- CN107239742A CN107239742A CN201710331133.0A CN201710331133A CN107239742A CN 107239742 A CN107239742 A CN 107239742A CN 201710331133 A CN201710331133 A CN 201710331133A CN 107239742 A CN107239742 A CN 107239742A
- Authority
- CN
- China
- Prior art keywords
- pointer
- scale value
- pixel
- edge
- equation
- 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
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D7/00—Indicating measured values
-
- 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
- G06V10/443—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 by matching or filtering
-
- 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
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention belongs to gauge pointer reading technical field, and in particular to a kind of gauge pointer scale value calculating method, specifically include following methods：1）, make Background；2）, pass through the current figure of Prototype drawing positioning；3）, to pointer pixel carry out color training；4）, present image pointer pixel extraction；5）, pointer connected component analysis；6）, pointer edge extracting；7）, pointer edge lines fitting；8）, pointer center line equation calculate；9）, pointer scale value calculate.The present invention is by calculating the pixel extraction and connected component analysis of the gauge pointer in present image, edge extracting, fitting a straight line, center line equation, realize the quick calculating of the pointer scale value on the dial for known scale value, the characteristics of with efficiently and accurately, overcome the deficiency of artificial reading；The present invention, which will be produced and examined and determine to instrument, has huge impetus.
Description
Technical field
The invention belongs to gauge pointer reading technical field, and in particular to a kind of gauge pointer scale value calculating method.
Background technology
For the higher instrument of the instrument of some industries, the particularly degree of accuracy examination requirements testing person in each need
By comparing the actual read number of pointer gauge and the output of reference power supply on the scale to be examined and determine, between difference it is accredited to analyze
The performance of instrument.Resolution capability yet with human eye is limited, when pointer be located at two graduation marks between when, can only rough estimate refer to
The position of pin, it is impossible to which the accurate registration for reading instrument, this will directly affects the order of accuarcy of calibrating.Furthermore the eyes of people are big
Visual fatigue occurs after the visual tasks of amount, carelessness and mistake occurs in verification test work, if can find in time
These mistakes are corrected, then to add the duplication of labour.If can not find in time, it is likely that bring serious consequence.Using
Manual type works, and time-consuming, labor intensity is big, calibrating efficiency is low, gauging error is big, poor reliability.For manual type calibrating
Shortcoming, develop as early as possible can quickly, accurate, stable full-automatic pointer instrument detecting system, by the instrument of China
Huge impetus is played in production and calibrating.
The content of the invention
There is provided a kind of gauge pointer scale value calculating method for defect and deficiency of the invention for prior art.
The technical solution adopted by the present invention is as follows：
A kind of gauge pointer position accurate detecting method, it is characterised in that comprise the following steps：
1) Background, is set up：All image pixel by pixel when each scale is in pointer analyze brightness, by position of sorting
The brightness value of the respective pixel position in middle brightness value is as Background is put, and then sets up Background；
When carrying out automatic detection to pointer, the position of instrument and camera is all static, dial, number in instrument
Word, icon etc., are all actionless；But when being detected, pointer is motion；This characteristic can be utilized, is improved
The effect of pointer detection；Therefore before testing, it is necessary to Background be set up, for later use image difference extraction pointer, background
The specific Making programme of figure is as follows：
1.1), change the position of pointer, shoot the image that pointer is in all graduation positions of dial plate, dial plate is carved for 17 altogether
To ensure the absolute rest of camera in degree, therefore 17 images of shooting altogether, shooting process；
1.2), the image that all graduation positions of dial plate are in pointer carries out comprehensive analysis, sets up Background；
By analyzing the corresponding brightness value of pointer location of pixels in every image, and it is ranked up, by sorting position in
Between brightness value as the brightness value of pointer location of pixels in Background, 17 images are analyzed here；
2) instrument disc template figure, is selected from Background, it is real by currently scheming the matching with instrument dial Prototype drawing
The now positioning of current figure；
During the use of system, it is very common phenomenon that camera, which is moved, in order to ensure subsequent step, Neng Gouzheng
The operation of true execution image subtraction is, it is necessary to first complete the positioning of image, and detailed process is as follows：
2.1), Prototype drawing and the circle fitting of the outline of current figure center disk；
2.2), image translation so that align in the center of circle；
2.3), image scaling so that current figure disc radius is consistent with Prototype drawing radius；
2.4), Prototype drawing and current figure center disk bottom lines detection, rectilinear direction are calculated；
2.5), image rotation so that current figure disk bottom rectilinear direction is consistent with Prototype drawing；
3) color training, is carried out to pointer pixel in Prototype drawing, and carries out the checking of color training result；
In order to improve the quality of pointer extracting, it is necessary to which on Prototype drawing, with reference to artificial method, extraction gets pointer color
Color approximate range, the step is referred to as " color training ", comprised the following steps that：
3.1) scope of pointer in Prototype drawing, is selected；
It is artificial to draw a circle to approve a scope on Prototype drawing for pointer using mouse；
3.2) distribution of color, is automatically extracted；
Because COLOR COMPOSITION THROUGH DISTRIBUTION is not single on pointer, single Gauss model is set up, color can not be described exactly
Distribution, it is therefore desirable to set up three Gauss models for the distribution of color of pointer area, entered using HSV color spaces to distribution of color
Row modeling, carries out statistical analysis and cluster to the HSV value of all colour elements in pointer range, obtains each Gauss model
Average and variance；
3.3), displaying pointer automatically extracts effect；
Using three Gauss models, the pointer on Prototype drawing is detected, each pixel is calculated and belongs to three models
Highest confidence level, if highest confidence level is higher than 0.2, then it is assumed that be pointer pixel, according to three kinds of different colors, refer to show
Pin pixel, each color one Gauss model of correspondence；
3.4), manual repair burst pixel, increases new Gauss model；
The pointer pixel do not included by three Gauss models is, it is necessary to by way of manually clicking on, by burst
Pixel is included into, the pixel newly included into, continues to increase new Gauss model, the process is continued for, included as far as possible more
Many pointer pixels, come while ensureing that no non-pointer pixel is preferentially absorbed into；
4), preceding image is compared with Background and primarily determines that pointer pixel, then extracts current according to color training result
Doubtful pointer pixel in figure；
In Background, storage is all actionless element in instrument；By with Background comparative analysis, can obtain
The candidate pixel of pointer；Certainly, in result now, the key element of some other variations is also included, such as the numeral changed changes
Scale strip, or the icon, etc. changed, but pass through the step greatly reduces the model of the potential site of pointer
Enclose, idiographic flow is as follows：
4.1), each pixel in traversal present image, calculates its highest confidence level for belonging to color Gauss model；
4.2), confidence level is higher than 0.2, is come out as pointer pixel extraction；
4.3), medium filtering, filters out isolated noise spot；
5) connected domain extraction, splicing, size and dimension analyzing and processing, are carried out to doubtful pointer pixel, pointer area is determined；
The pointer candidate pixel result figure obtained in step 4 is further analyzed, to filter out the interference of noise,
Comprise the following steps that：
5.1), connected domain is extracted and contours extract；
5.2), the closely splicing of connected domain；
Due to the influence of camera actual photographed effect, the pointer pixel extracted in step 4, it is likely to fracture,
Pointer is possible to be divided into some, now needs to analyze all connected domain profiles, calculates its minimum distance,
Minimum distance is less than the connected domain of 5 pixels, it is necessary to be spliced into an entirety；
5.3) small size connected domain, is filtered：The boundary rectangle of connected domain is calculated, the wide and high of boundary rectangle is both less than 50
The connected domain of pixel is, it is necessary to filter out, and they are unlikely to be pointer area；
5.4), connected domain shape analysis, filters out the connected domain not substantially being inconsistent：The company of pointer pixel formation in step 4
, even if being influenceed by background pixel, or after splicing, still there is phenomenon of rupture, its profile is also close to rectangle in logical domain；
The minimum enclosed rectangle of connected domain profile is calculated, its rectangular degree is then calculated, rectangular degree is less than 0.6 connected domain, directly filters；
6), pointer edge extracting in present image；
, it is necessary to calculate the edge in image in advance, subsequent step is in edge graph before the straight line of detection pointer both sides
On basis, fitting a straight line is further carried out, is comprised the following steps that：
6.1) canny operator extraction edge graphs, are utilized in current color figure；
6.2), mask artwork is generated using the result of step 5；
Left connected domain after step 5 is filtered, is plotted in new image, forms black white binarization image, is used for
As mask artwork, the white pixel in mask artwork represents to need processed pixel；
6.3), mask artwork expansion
When connected domain is extracted, its border is probably the border of pointer；In order to prevent pointer edge to be missed, it is necessary to right
Mask artwork is expanded, and expands the scope of each connected domain；
6.4), mask artwork and edge graph ask with
Ask and afterwards, the edge graph of only pointer area is retained, remaining background area or non-pointer region
All filtered at edge；
6.5), sub-line section in edge is interrupted
All marginal points of traversal, analyze the number of marginal point in its 8 neighborhood, are T if marginal point number is more than 2
Type crosspoint or right-angled intersection point remove the marginal point, it is necessary to interrupt；
The curvature of marginal point is analyzed, curvature exceedes threshold value, it is also desirable to interrupts, removes the marginal point；
6.6), edge shape is analyzed, filtration curve
The shape at each edge is analyzed, calculates and is put most on the center of gravity of all marginal points on edge, centroidal distance edge
Big distance is more than 10 pixels, it is believed that is curve, is filtered out, finally remained, entirely straight line sub-line section；
7), according to step 6) the edge sub-line section extracted, enter line pointer edge lines fitting；
In the curve map that step 6 is obtained, fitting a straight line is carried out, while calculating the equation of line correspondences；
8), according to step 7) obtained fitted straight lines of edges calculates pointer center line equation；
The shape of pointer is generally not the rectangle of rule, but up-small and down-big or up big and down small shape, portion is presented
It is even more non-regular shape occur to divide pointer；But, most pointers is all symmetrical, and both sides are all in the presence of straight
Line boundary, using the pointer boundary straight line fitting result obtained in step 7, can obtain the center line equation of pointer, idiographic flow
It is as follows：
8.1), straight line is to search
The straight line of pointer both sides, with it is following the characteristics of：A), the distance to disc centre is less than 30 pixels；B), both sides
Straight line is symmetrical on disc centre；According to the two features, all straight lines in traversing graph picture, it is possible to find and meet the two
The straight line pair of feature, is used as the linear equation of pointer both sides；
8.2), straight line is to equation amendment
Because edge is in extraction process, always disturbed by various factors, cause linear equation deflection occur, therefore
Need according to a certain degree of amendment is carried out, method is as follows：A) it is straight apart from this in search edge graph, for current linear equation
The nearer marginal point of line；B), in rectilinear periphery, the parameter value of edge extracting is reduced, expands number of edge points；C), by distance compared with
Near marginal point, re-starts fitting a straight line, is used as new revised linear equation；
8.3), straight line antinode is calculated：Using both sides linear equation, intersecting point coordinate is obtained；
8.4), angular bisector is calculated：Using the corresponding angle of both sides linear equation, the angle of center line is obtained, center line is located at
On angular bisector；
8.5), center line equation is calculated：Using intersection point and angle, center line linear equation is obtained；
8.6), center line equation is verified：Pointer is sent from the center of disk, can calculate disc centre into pointer
The distance of line, distance is more than 5 pixels, it is believed that equation calculates mistake, carries out early warning；
9), according to the center line equation of the center position of each graduation mark, corresponding actual scale value and pointer, meter
Calculate the actual scale value of pointer；Idiographic flow is as follows：
9.1) scale central point, is traveled through one by one, calculates its that side for being located at pointer center line：Left side is labeled as just, right side
Labeled as negative；
9.2) two adjacent graduation marks, are traveled through one by one, find that a pair of the graduation mark A for being located at pointer center line both sides respectively
And B：All graduation marks on disk, only a pair of scales are separations in the both sides of center line, it is only necessary to find symbol different
Scale is to just；
9.3), distances of the computing scale line A and B to center line；
9.4), according to distance than the actual scale value with A, B, the corresponding actual scale value of pointer is calculated：Using linear
The method of interpolation, obtains the real scale value of pointer.
The present invention by the pixel extraction and connected component analysis of the gauge pointer in present image, edge extracting,
Fitting a straight line, center line equation are calculated, and realize the quick calculating of the pointer scale value on the dial for known scale value, tool
The characteristics of having efficiently and accurately, overcomes the deficiency of artificial reading；The present invention, which will be produced and examined and determine to instrument, has huge promotion
Effect.
Brief description of the drawings
Accompanying drawing is used for providing a further understanding of the present invention, and constitutes a part for specification, the reality with the present invention
Applying example is used to explain the present invention together, is not construed as limiting the invention.
Fig. 1 is flow chart of the invention.
Embodiment
A kind of gauge pointer scale value calculating method, it is characterised in that comprise the following steps：
1) Background, is set up：All image pixel by pixel when each scale is in pointer analyze brightness, by position of sorting
The brightness value of the respective pixel position in middle brightness value is as Background is put, and then sets up Background；
When carrying out automatic detection to pointer, the position of instrument and camera is all static, dial, number in instrument
Word, icon etc., are all actionless；But when being detected, pointer is motion；This characteristic can be utilized, is improved
The effect of pointer detection；Therefore before testing, it is necessary to Background be set up, for later use image difference extraction pointer, background
The specific Making programme of figure is as follows：
1.1), change the position of pointer, shoot the image that pointer is in all graduation positions of dial plate, dial plate is carved for 17 altogether
To ensure the absolute rest of camera in degree, therefore 17 images of shooting altogether, shooting process；
1.2), the image that all graduation positions of dial plate are in pointer carries out comprehensive analysis, sets up Background；
By analyzing the corresponding brightness value of pointer location of pixels in every image, and it is ranked up, by sorting position in
Between brightness value as the brightness value of pointer location of pixels in Background, 17 images are analyzed here；
2) instrument dial Prototype drawing, is selected from Background, by currently scheming the matching with instrument dial Prototype drawing,
Realize the positioning of current figure；
During the use of system, it is very common phenomenon that camera, which is moved, in order to ensure subsequent step, Neng Gouzheng
The operation of true execution image subtraction is, it is necessary to first complete the positioning of image, and detailed process is as follows：
2.1), Prototype drawing and the circle fitting of the outline of current figure center disk；
2.2), image translation so that align in the center of circle；
2.3), image scaling so that current figure disc radius is consistent with Prototype drawing radius；
2.4), Prototype drawing and current figure center disk bottom lines detection, rectilinear direction are calculated；
2.5), image rotation so that current figure disk bottom rectilinear direction is consistent with Prototype drawing；
3) color training, is carried out to pointer pixel in Prototype drawing, and carries out the checking of color training result；
In order to improve the quality of pointer extracting, it is necessary to which on Prototype drawing, with reference to artificial method, extraction gets pointer color
Color approximate range, the step is referred to as " color training ", comprised the following steps that：
3.1) scope of pointer in Prototype drawing, is selected；
It is artificial to draw a circle to approve a scope on Prototype drawing for pointer using mouse；
3.2) distribution of color, is automatically extracted；
Because COLOR COMPOSITION THROUGH DISTRIBUTION is not single on pointer, single Gauss model is set up, color can not be described exactly
Distribution, it is therefore desirable to set up three Gauss models for the distribution of color of pointer area, entered using HSV color spaces to distribution of color
Row modeling, carries out statistical analysis and cluster to the HSV value of all colour elements in pointer range, obtains each Gauss model
Average and variance；
3.3), displaying pointer automatically extracts effect；
Using three Gauss models, the pointer on Prototype drawing is detected, each pixel is calculated and belongs to three models
Highest confidence level, if highest confidence level is higher than 0.2, then it is assumed that be pointer pixel, according to three kinds of different colors, refer to show
Pin pixel, each color one Gauss model of correspondence；
3.4), manual repair burst pixel, increases new Gauss model；
The pointer pixel do not included by three Gauss models is, it is necessary to by way of manually clicking on, by burst
Pixel is included into, the pixel newly included into, continues to increase new Gauss model, the process is continued for, included as far as possible more
Many pointer pixels, come while ensureing that no non-pointer pixel is preferentially absorbed into；
4), preceding image is compared with Background and primarily determines that pointer pixel, then extracts current according to color training result
Doubtful pointer pixel in figure；
In Background, storage is all actionless element in instrument；By with Background comparative analysis, can obtain
The candidate pixel of pointer；Certainly, in result now, the key element of some other variations is also included, such as the numeral changed changes
Scale strip, or the icon, etc. changed, but pass through the step greatly reduces the model of the potential site of pointer
Enclose, idiographic flow is as follows：
4.1), each pixel in traversal present image, calculates its highest confidence level for belonging to color Gauss model；
4.2), confidence level is higher than 0.2, is come out as pointer pixel extraction；
4.3), medium filtering, filters out isolated noise spot；
5) connected domain extraction, splicing, size and dimension analyzing and processing, are carried out to doubtful pointer pixel, pointer area is determined；
The pointer candidate pixel result figure obtained in step 4 is further analyzed, to filter out the interference of noise,
Comprise the following steps that：
5.1), connected domain is extracted and contours extract；
5.2), the closely splicing of connected domain；
Due to the influence of camera actual photographed effect, the pointer pixel extracted in step 4, it is likely to fracture,
Pointer is possible to be divided into some, now needs to analyze all connected domain profiles, calculates its minimum distance,
Minimum distance is less than the connected domain of 5 pixels, it is necessary to be spliced into an entirety；
5.3) small size connected domain, is filtered：The boundary rectangle of connected domain is calculated, the wide and high of boundary rectangle is both less than 50
The connected domain of pixel is, it is necessary to filter out, and they are unlikely to be pointer area；
5.4), connected domain shape analysis, filters out the connected domain not substantially being inconsistent：The company of pointer pixel formation in step 4
, even if being influenceed by background pixel, or after splicing, still there is phenomenon of rupture, its profile is also close to rectangle in logical domain；
The minimum enclosed rectangle of connected domain profile is calculated, its rectangular degree is then calculated, rectangular degree is less than 0.6 connected domain, directly filters；
6), pointer edge extracting in present image；
, it is necessary to calculate the edge in image in advance, subsequent step is in edge graph before the straight line of detection pointer both sides
On basis, fitting a straight line is further carried out, is comprised the following steps that：
6.1) canny operator extraction edge graphs, are utilized in current color figure；
6.2), mask artwork is generated using the result of step 5；
Left connected domain after step 5 is filtered, is plotted in new image, forms black white binarization image, is used for
As mask artwork, the white pixel in mask artwork represents to need processed pixel；
6.3), mask artwork expands
When connected domain is extracted, its border is probably the border of pointer；In order to prevent pointer edge to be missed, it is necessary to right
Mask artwork is expanded, and expands the scope of each connected domain；
6.4), mask artwork and edge graph ask with
Ask and afterwards, the edge graph of only pointer area is retained, remaining background area or non-pointer region
All filtered at edge；
6.5), sub-line section in edge is interrupted
All marginal points of traversal, analyze the number of marginal point in its 8 neighborhood, are T if marginal point number is more than 2
Type crosspoint or right-angled intersection point remove the marginal point, it is necessary to interrupt；
The curvature of marginal point is analyzed, curvature exceedes threshold value, it is also desirable to interrupts, removes the marginal point；
6.6), edge shape is analyzed, filtration curve
The shape at each edge is analyzed, calculates and is put most on the center of gravity of all marginal points on edge, centroidal distance edge
Big distance is more than 10 pixels, it is believed that is curve, is filtered out, finally remained, entirely straight line sub-line section；
7), according to step 6) the edge sub-line section extracted, enter line pointer edge lines fitting；
In the curve map that step 6 is obtained, fitting a straight line is carried out, while calculating the equation of line correspondences；
8), according to step 7) obtained fitted straight lines of edges calculates pointer center line equation；
The shape of pointer is generally not the rectangle of rule, but up-small and down-big or up big and down small shape, portion is presented
It is even more non-regular shape occur to divide pointer；But, most pointers is all symmetrical, and both sides are all in the presence of straight
Line boundary, using the pointer boundary straight line fitting result obtained in step 7, can obtain the center line equation of pointer, idiographic flow
It is as follows：
8.1), straight line is to search
The straight line of pointer both sides, with it is following the characteristics of：A), the distance to disc centre is less than 30 pixels；B), both sides
Straight line is symmetrical on disc centre；According to the two features, all straight lines in traversing graph picture, it is possible to find and meet the two
The straight line pair of feature, is used as the linear equation of pointer both sides；
8.2), straight line is to equation amendment
Because edge is in extraction process, always disturbed by various factors, cause linear equation deflection occur, therefore
Need according to a certain degree of amendment is carried out, method is as follows：A) it is straight apart from this in search edge graph, for current linear equation
The nearer marginal point of line；B), in rectilinear periphery, the parameter value of edge extracting is reduced, expands number of edge points；C), by distance compared with
Near marginal point, re-starts fitting a straight line, is used as new revised linear equation；
8.3), straight line antinode is calculated：Using both sides linear equation, intersecting point coordinate is obtained；
8.4), angular bisector is calculated：Using the corresponding angle of both sides linear equation, the angle of center line is obtained, center line is located at
On angular bisector；
8.5), center line equation is calculated：Using intersection point and angle, center line linear equation is obtained；
8.6), center line equation is verified：Pointer is sent from the center of disk, can calculate disc centre into pointer
The distance of line, distance is more than 5 pixels, it is believed that equation calculates mistake, carries out early warning；
9), according to the center line equation of the center position of each graduation mark, corresponding actual scale value and pointer, meter
Calculate the actual scale value of pointer；Idiographic flow is as follows：
9.1) scale central point, is traveled through one by one, calculates its that side for being located at pointer center line：Left side is labeled as just, right side
Labeled as negative；
9.2) two adjacent graduation marks, are traveled through one by one, find that a pair of the graduation mark A for being located at pointer center line both sides respectively
And B：All graduation marks on disk, only a pair of scales are separations in the both sides of center line, it is only necessary to find symbol different
Scale is to just；
9.3), distances of the computing scale line A and B to center line；
9.4), according to distance than the actual scale value with A, B, the corresponding actual scale value of pointer is calculated：Using linear
The method of interpolation, obtains the real scale value of pointer.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, although with reference to foregoing reality
Apply example the present invention is described in detail, for those skilled in the art, it still can be to foregoing each implementation
Technical scheme described in example is modified, or carries out equivalent substitution to which part technical characteristic；All essences in the present invention
God is with principle, and any modification, equivalent substitution and improvements made etc. should be included in the scope of the protection.
Claims (8)
1. a kind of gauge pointer scale value calculating method, it is characterised in that comprise the following steps：
1) Background, is set up：All image pixel by pixel when each scale is in pointer analyze brightness, and sorting position is existed
Middle brightness value sets up Background as the brightness value of respective pixel position in Background；
2) instrument dial Prototype drawing, is selected from Background, by currently scheming the matching with instrument dial Prototype drawing, positioning
Current figure；
3) color training, is carried out to pointer pixel in Prototype drawing, and carries out the checking of color training result；
4), preceding image is compared with Background and primarily determines that pointer pixel, is then extracted according to color training result in current figure
Doubtful pointer pixel；
5) connected domain extraction, splicing, size and dimension analyzing and processing, are carried out to doubtful pointer pixel, pointer area is determined；
6), pointer edge extracting in present image；
7), according to step 6）The edge sub-line section of extraction, enters line pointer edge lines fitting；
8), according to step 7）Obtained fitted straight lines of edges calculates pointer center line equation；
9), according to the center line equation of the center position of each graduation mark, corresponding actual scale value and pointer, line is utilized
Property interpolation method calculate pointer actual scale value.
2. the gauge pointer scale value calculating method according to claim 1, it is characterised in that the step 2）In pass through
The matching of current figure and instrument dial Prototype drawing, the process for positioning current figure is as follows：
A), Prototype drawing and the circle fitting of the outline of current figure center disk；
B), image translation so that align in the center of circle；
C), image scaling so that current figure disc radius is consistent with Prototype drawing radius；
D), Prototype drawing and current figure center disk bottom lines detection and rectilinear direction are calculated；
E), image rotation so that current figure disk bottom rectilinear direction is consistent with Prototype drawing.
3. the gauge pointer scale value calculating method according to claim 1, it is characterised in that the step 3）In in mould
Color training is carried out in plate figure to pointer pixel, and carries out also including manual repair burst picture after the checking of color training result
Element, increases the process of new color training Gauss model.
4. the gauge pointer scale value calculating method according to claim 1, it is characterised in that the step 6) in it is current
The process of pointer edge extracting is as follows in image：
A) canny operator extraction edge graphs, are utilized in current color figure；
B) step 5, is utilized）Pointer area generation mask artwork；
C), mask artwork expands, and expands the scope of each connected domain；
D), mask artwork and edge graph are asked and determine pointer edge；
E), T-junction point or right-angled intersection point edge sub-line section are interrupted and filtered out；
F) edge shape, is analyzed, filtration curve retains straight line sub-line section.
5. gauge pointer scale value calculating method according to claim 1, it is characterised in that the step 8）Pointer in
Line equation calculating process is as follows：
A), search on the symmetrical straight line of disc centre to the linear equation as pointer both sides；
B), straight line is modified to equation；
C) straight line antinode calculating, is carried out to equation according to revised straight line, intersecting point coordinate is obtained；
D) angle of angular bisector, is calculated to the corresponding angle of equation using straight line；
E), using intersecting point coordinate and bisector angle, center line linear equation is calculated.
6. gauge pointer scale value calculating method according to claim 5, it is characterised in that also testing including center line equation
Card：Disc centre is calculated to the distance of pointer center line, distance is more than threshold value L1, then center line equation mistake, carries out early warning.
7. gauge pointer scale value calculating method according to claim 6, it is characterised in that described 5 pictures in threshold value L1 positions
Element.
8. the gauge pointer scale value calculating method according to claim 1, it is characterised in that the step 9) according to every
The center line equation of the center position of individual graduation mark, corresponding actual scale value and pointer, calculates the actual scale value of pointer
Idiographic flow it is as follows：
A) scale central point, is traveled through one by one, calculates its that side for being located at pointer center line：Left side is labeled as just, and right side is labeled as
It is negative；
B) two adjacent graduation marks, are traveled through one by one, find that a pair of the graduation marks A and B for being located at pointer center line both sides respectively；
C), distances of the computing scale line A and B to center line；
D), according to distance than the actual scale value with A, B, using the method for linear interpolation, pointer corresponding actual quarter is calculated
Angle value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710331133.0A CN107239742B (en) | 2017-05-11 | 2017-05-11 | Method for calculating scale value of instrument pointer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710331133.0A CN107239742B (en) | 2017-05-11 | 2017-05-11 | Method for calculating scale value of instrument pointer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107239742A true CN107239742A (en) | 2017-10-10 |
| CN107239742B CN107239742B (en) | 2020-06-09 |
Family
ID=59985574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710331133.0A Active CN107239742B (en) | 2017-05-11 | 2017-05-11 | Method for calculating scale value of instrument pointer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107239742B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109086763A (en) * | 2018-07-23 | 2018-12-25 | 国网辽宁省电力有限公司信息通信分公司 | A kind of pointer instrument read method and device |
| CN109145899A (en) * | 2018-07-30 | 2019-01-04 | 安徽慧视金瞳科技有限公司 | A kind of automobile meter pointer extracting method |
| CN109470290A (en) * | 2018-10-31 | 2019-03-15 | 惠州华阳通用电子有限公司 | A kind of gauge pointer automatic calibrating method and device |
| CN110335282A (en) * | 2018-12-25 | 2019-10-15 | 广州启明星机器人有限公司 | A kind of profile line segment feature extraction algorithm based on grid |
| CN110852333A (en) * | 2019-11-13 | 2020-02-28 | 广东电科院能源技术有限责任公司 | Automatic reading method and device for pointer instrument |
| CN111298238A (en) * | 2020-02-21 | 2020-06-19 | 济南欣格信息科技有限公司 | Infusion early warning method based on image recognition |
| CN111325164A (en) * | 2020-02-25 | 2020-06-23 | 北京眸视科技有限公司 | Pointer indication number identification method and device and electronic equipment |
| CN112288049A (en) * | 2020-12-29 | 2021-01-29 | 广东科凯达智能机器人有限公司 | Pointer meter reading method based on template matching and limitation of connected domain pointer positioning |
| CN112640415A (en) * | 2018-08-30 | 2021-04-09 | 欧吉斯信息技术有限公司 | Meter reading device, meter reading system and computer program |
| CN115655576A (en) * | 2022-12-13 | 2023-01-31 | 成都千嘉科技股份有限公司 | Automatic sensing method for displacement abnormity of pointer type pressure gauge |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521560A (en) * | 2011-11-14 | 2012-06-27 | 上海交通大学 | Instrument pointer image identification method of high-robustness rod |
| CN103927507A (en) * | 2013-01-12 | 2014-07-16 | 山东鲁能智能技术有限公司 | Improved multi-instrument reading identification method of transformer station inspection robot |
| CN104050446A (en) * | 2014-05-30 | 2014-09-17 | 浙江大学 | Meter pointer image identification method based on pointer width character |
| CN104392206A (en) * | 2014-10-24 | 2015-03-04 | 南京航空航天大学 | Image processing method for automatic pointer-type instrument reading recognition |
| CN105783969A (en) * | 2016-04-01 | 2016-07-20 | 湖北工程学院 | Stepper motor pointer meter video reading method for state-IV commercial automobile instruments |
| CN106339707A (en) * | 2016-08-19 | 2017-01-18 | 亿嘉和科技股份有限公司 | Instrument pointer image recognition method based on symmetrical characteristics |
| CN106529519A (en) * | 2016-09-19 | 2017-03-22 | 国家电网公司 | Automatic number identification method and system of power pointer type instrument |
-
2017
- 2017-05-11 CN CN201710331133.0A patent/CN107239742B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102521560A (en) * | 2011-11-14 | 2012-06-27 | 上海交通大学 | Instrument pointer image identification method of high-robustness rod |
| CN103927507A (en) * | 2013-01-12 | 2014-07-16 | 山东鲁能智能技术有限公司 | Improved multi-instrument reading identification method of transformer station inspection robot |
| CN104050446A (en) * | 2014-05-30 | 2014-09-17 | 浙江大学 | Meter pointer image identification method based on pointer width character |
| CN104392206A (en) * | 2014-10-24 | 2015-03-04 | 南京航空航天大学 | Image processing method for automatic pointer-type instrument reading recognition |
| CN105783969A (en) * | 2016-04-01 | 2016-07-20 | 湖北工程学院 | Stepper motor pointer meter video reading method for state-IV commercial automobile instruments |
| CN106339707A (en) * | 2016-08-19 | 2017-01-18 | 亿嘉和科技股份有限公司 | Instrument pointer image recognition method based on symmetrical characteristics |
| CN106529519A (en) * | 2016-09-19 | 2017-03-22 | 国家电网公司 | Automatic number identification method and system of power pointer type instrument |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109086763A (en) * | 2018-07-23 | 2018-12-25 | 国网辽宁省电力有限公司信息通信分公司 | A kind of pointer instrument read method and device |
| CN109145899A (en) * | 2018-07-30 | 2019-01-04 | 安徽慧视金瞳科技有限公司 | A kind of automobile meter pointer extracting method |
| CN112640415A (en) * | 2018-08-30 | 2021-04-09 | 欧吉斯信息技术有限公司 | Meter reading device, meter reading system and computer program |
| CN112640415B (en) * | 2018-08-30 | 2022-09-13 | 欧吉斯信息技术有限公司 | Meter reading device, meter reading system and storage medium |
| CN109470290A (en) * | 2018-10-31 | 2019-03-15 | 惠州华阳通用电子有限公司 | A kind of gauge pointer automatic calibrating method and device |
| CN109470290B (en) * | 2018-10-31 | 2022-02-18 | 惠州华阳通用电子有限公司 | Automatic calibration method and device for instrument pointer |
| CN110335282A (en) * | 2018-12-25 | 2019-10-15 | 广州启明星机器人有限公司 | A kind of profile line segment feature extraction algorithm based on grid |
| CN110335282B (en) * | 2018-12-25 | 2023-04-18 | 广州启明星机器人有限公司 | Contour line segment feature extraction method based on grids |
| CN110852333B (en) * | 2019-11-13 | 2021-07-23 | 南方电网电力科技股份有限公司 | Automatic reading method and device for pointer instrument |
| CN110852333A (en) * | 2019-11-13 | 2020-02-28 | 广东电科院能源技术有限责任公司 | Automatic reading method and device for pointer instrument |
| CN111298238A (en) * | 2020-02-21 | 2020-06-19 | 济南欣格信息科技有限公司 | Infusion early warning method based on image recognition |
| CN111325164A (en) * | 2020-02-25 | 2020-06-23 | 北京眸视科技有限公司 | Pointer indication number identification method and device and electronic equipment |
| CN111325164B (en) * | 2020-02-25 | 2023-11-21 | 北京眸视科技有限公司 | Pointer representation number identification method and device and electronic equipment |
| CN112288049A (en) * | 2020-12-29 | 2021-01-29 | 广东科凯达智能机器人有限公司 | Pointer meter reading method based on template matching and limitation of connected domain pointer positioning |
| CN115655576A (en) * | 2022-12-13 | 2023-01-31 | 成都千嘉科技股份有限公司 | Automatic sensing method for displacement abnormity of pointer type pressure gauge |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107239742B (en) | 2020-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107239742A (en) | A kind of gauge pointer scale value calculating method | |
| CN110659636B (en) | Pointer instrument reading identification method based on deep learning | |
| US20220148213A1 (en) | Method for fully automatically detecting chessboard corner points | |
| CN113688807B (en) | Self-adaptive defect detection method, device, recognition system and storage medium | |
| CN107633516B (en) | Method and device for identifying road surface deformation diseases | |
| WO2021143343A1 (en) | Method and device for testing product quality | |
| CN109900711A (en) | Workpiece, defect detection method based on machine vision | |
| CN105069790A (en) | Rapid imaging detection method for gear appearance defect | |
| WO2017181724A1 (en) | Inspection method and system for missing electronic component | |
| CN103345755A (en) | Chessboard angular point sub-pixel extraction method based on Harris operator | |
| WO2017177717A1 (en) | Element positioning method and system based on color and gradient | |
| CN112257676A (en) | Pointer instrument reading method and system and inspection robot | |
| CN102496161A (en) | Method for extracting contour of image of printed circuit board (PCB) | |
| CN111222507B (en) | Automatic identification method for digital meter reading and computer readable storage medium | |
| CN115937203B (en) | Visual detection method, device, equipment and medium based on template matching | |
| CN115014248B (en) | Laser projection line identification and flatness judgment method | |
| CN110211178A (en) | A kind of pointer instrument recognition methods calculated using projection | |
| CN115597494A (en) | Precision detection method and system for prefabricated part preformed hole based on point cloud | |
| CN113705564B (en) | Pointer type instrument identification reading method | |
| CN105374045B (en) | One kind is based on morphologic image given shape size objectives fast partition method | |
| CN104655041B (en) | A kind of industrial part contour line multi-feature extraction method of additional constraint condition | |
| CN110288040A (en) | A kind of similar evaluation method of image based on validating topology and equipment | |
| CN115719326A (en) | PCB defect detection method and device | |
| CN115375608A (en) | Detection method and device, detection equipment and storage medium | |
| CN113554688B (en) | O-shaped sealing ring size measurement method based on monocular vision |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 230000 Yafu Park, Juchao Economic Development Zone, Chaohu City, Hefei City, Anhui Province Applicant after: ANHUI HUISHI JINTONG TECHNOLOGY Co.,Ltd. Address before: 102, room 602, C District, Hefei National University, Mount Huangshan Road, 230000 Hefei Road, Anhui, China Applicant before: ANHUI HUISHI JINTONG TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |