CN108460331A - A kind of readings of pointer type meters automatic identification equipment of robust and its recognition methods - Google Patents
A kind of readings of pointer type meters automatic identification equipment of robust and its recognition methods Download PDFInfo
- Publication number
- CN108460331A CN108460331A CN201810046784.XA CN201810046784A CN108460331A CN 108460331 A CN108460331 A CN 108460331A CN 201810046784 A CN201810046784 A CN 201810046784A CN 108460331 A CN108460331 A CN 108460331A
- Authority
- CN
- China
- Prior art keywords
- cylinder
- instrument
- pointer
- readings
- camera
- 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
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of readings of pointer type meters automatic identification equipment of robust and its recognition methods, belong to instrument and meter and field of machine vision, described device includes a cylinder, a camera and the circuit board for controlling camera, the connection shaft of the instrument disk center of the cylinder and pointer instrument is integrated, the pointer tail portion of pointer instrument is set on cylinder, the camera is set to the instrument plate edge of pointer instrument, and the outer surface of the cylinder is around a circle characteristic pattern.In the case that the present invention is by changing gauge internal structure without changing instrument function, in the way of machine vision, the image of instrument other parts is associated with total indicator reading, achieve the purpose that identify meter reading, be influenced by dial plate environment to eliminate when pointer instrument is read from outside；Use simple image processing algorithm, you can the automatic identification function for completing readings of pointer type meters improves the accuracy of identification of readings of pointer type meters.
Description
Technical field
The present invention relates to instrument and meter and field of machine vision more particularly to a kind of pointer instruments of robust
Read automatic identification equipment and its recognition methods.
Background technology
Pointer instrument is widely used in the fields, such as locomotive instrumentation such as machinery, medical treatment, military affairs, building, chemical industry, water meter,
Pressure gauge etc. is capable of the operating status of real time indication system.
Pointer instrument has the advantages that digital instrument is irreplaceable, i.e. intuitively the trend of reactions change amount, such as vapour
Vehicle instrument directly can determine speed change procedure according to throw of the pointer and direction, very clear, so pointer instrument will be grown
Phase exists without being eliminated by industry.
Pointer instrument is generally read by two ways：(1) direct labor reads the data on dial plate；(2) table is shot
Disk image carries out processing and obtains data.Wherein, mode (1) is time-consuming and laborious, is influenced by human factor, error is larger.Mode (2)
Shooting the image of dial plate, factors are influenced by video camera placement position and angle, illumination, dial plate be not clean etc., image procossing difficulty
Greatly, bad adaptability, algorithm versatility are low.
For Most scholars from the angle of image procossing, the image processing algorithm by improving dial plate obtains higher at present
Accuracy of identification, as number of patent application is：201510066208.8 having invented a kind of readings of pointer type meters of robust automatically to know
Other method, it proposes shade and iteration of the comentropy algorithm based on probability distribution image for eliminating the generation of dial plate edge
Optimization algorithm determines the center of dial plate, and meter reading accuracy of identification is improved with this.
Pointer instrument is had a problem that by improving image processing algorithm raising accuracy of identification, if dial plate is not done
Only, circular scale region can not be clearly captured, then will produce prodigious error by image reading or even can not read.Institute
It can be handled with the image presence shot from gauge external and two kinds of situations can not be handled, similarly, there is also this for artificial reading
Problem.Therefore, only how being not is more accurately identified to readings of pointer type meters from the point of view of improving image processing algorithm
Foot, it is also necessary to go to consider from the angle of instrument itself.
For above two mode read from gauge external, there are shortcomings, therefore, it may be considered that out of instrument
Portion shoots image.But space workable for gauge internal is less, and the image of meter dial is included to shooting, camera is general
There are two types of laying methods：(1) it is placed on the ring edge for the opposite end that instrument range mid-scale line faces, and camera lens surface
It has a certain degree with dial plate；(2) middle of instrument lid is placed, camera lens surface is towards dial plate and parallel with dial plate；For side
Method (1), adjustable height is smaller above and below camera, and focal distance too short-range missile causes the dial plate of shooting very fuzzy, and it is big to have an X-rayed variation.
For method (2), the instrument that camera is located at right over dial plate covers, and the circuit for controlling camera just needs to connect up from dial plate, no
The appearance of instrument is only affected, is also unfavorable for reading.Therefore, it is also infeasible in gauge internal shooting dial scale image.
Invention content
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of readings of pointer type meters of robust is automatic
Identification device and its recognition methods read existing shortcomings, it is intended to design one to solve existing pointer instrument from outside
Dial reading can not all can be read by dial plate such environmental effects, any situation in kind device.
The present invention is achieved by the following technical solutions：
The present invention provides a kind of readings of pointer type meters automatic identification equipments of robust, including a cylinder, a camera
And the circuit board of camera is controlled, the connection shaft of the instrument disk center of the cylinder and pointer instrument is integrated, pointer-type
The pointer tail portion of instrument is set on cylinder, and the camera is set to the instrument plate edge of pointer instrument, the cylinder it is outer
Surface loop is around a circle characteristic pattern, the characteristic pattern of the camera lens face cylinder of the camera；It, will be special by shooting characteristic pattern
Sign pattern is associated with total indicator reading, achievees the purpose that identify meter reading.
Further, the optical axis of the camera perpendicular to cylinder shaft, and by cylinder center so that circle
The heart is imaged cylinder in the picture, convenient subsequently to the processing of image.
Further, the size of the cylinder and the accuracy of identification of meter reading are related, determine that the size of cylinder is
One of the key point of the present invention, can account in terms of accuracy of instrument and shaft step size two, wherein：
If from the aspect of accuracy of instrument, the diameter D of the cylinder meets：
In formula, A is accuracy of instrument, and C is the corresponding angle of range (range angle), and FOV is camera fields of view, and M is camera
Horizontal pixel；
If from the aspect of shaft step size, the diameter D of the cylinder meets：
In formula, H is the step angle of stepper motor, and S is the high-subdividing number of stepper motor, and FOV is camera fields of view, and M is camera
Horizontal pixel.
Further, the diameter needs of the cylinder are designed according to actual demand, can neither too greatly can not be too
It is small, both influence appearance greatly very much, weight also influences performance, too small not meet design principle, usually the body diameter be 4~
9mm is proper.
Further, the characteristic pattern is the right angled triangle and one group of parallel lines around cylinder periphery, described
One right-angle side of right angled triangle is parallel with the height of cylinder, and the bottom/top circumference of another right-angle side and cylinder is flat
Row, the parallel lines are parallel or vertical with the bottom/top circumference of cylinder.
Further, the characteristic pattern is the bottom surface around the saw-tooth like pattern of cylinder periphery and with saw-tooth like pattern
The width of parallel parallel lines, the sawtooth corresponds to the high range of instrument, includes the character for distinguishing other sawtooth in each sawtooth.
Further, the bottom/top circle with cylinder of a parallel lines in the parallel lines and right angled triangle
A parallel lines in the right-angle side coincidence of Zhou Pinghang or the parallel lines are overlapped with the line where the bottom surface of saw-tooth like pattern.
The present invention also provides a kind of reading knowledge is carried out using the readings of pointer type meters automatic identification equipment of above-mentioned robust
Method for distinguishing, step include：
Step S1：Camera is demarcated, takes some fixed point in image as datum mark, the pixel for obtaining the point is sat
Mark, image when shooting 0 scale and current scale respectively, and image is corrected；
Step S2：The edge pixel for extracting the characteristic pattern in image, does a vertical line with the datum mark of calibration, calculates
Line segment pixels value when to 0 scale and current scale between corresponding characteristic pattern edge and vertical line intersection point；
Step S3：The edge line segment calculated for pixel values cylinder of corresponding characteristic pattern when according to 0 scale and current scale
Rotational angle；
Step S4：Since the angle of cylinder rotation is equal to indicators turn angle, the angle that can be rotated according to cylinder
Degree converts, and obtains current scale value.
Further, in the step S1, correction includes level correction and the perspective distortion correction of image.
Further, in the step S1, point on the basis of the central point of image is selected.
The present invention prior art that compares has the following advantages：The present invention lacking from outside reading for pointer instrument
Point provides readings of pointer type meters automatic identification equipment and its recognition methods of a kind of robust, and this method, which passes through, changes instrument
In the case that internal structure is without changing instrument function, in the way of machine vision, by the image and finger of instrument other parts
Needle reading associates, and achievees the purpose that identify meter reading, to eliminate when pointer instrument is read from outside by dial plate
The influence of environment；Use simple image processing algorithm, you can the automatic identification function for completing readings of pointer type meters improves
The accuracy of identification of readings of pointer type meters.
Description of the drawings
Fig. 1 is the three-dimensional structure diagram of the readings of pointer type meters automatic identification equipment of robust；
Fig. 2 is the side structure view of the readings of pointer type meters automatic identification equipment of robust；
Fig. 3 is the step flow chart of the readings of pointer type meters automatic identification equipment recognition methods of robust；
Fig. 4 is the characteristic pattern expanded view of embodiment 1 and its intersection point with datum mark straight line excessively；
Fig. 5 is the characteristic image edge intersection point calculated for pixel values method schematic diagram of embodiment 1；
Fig. 6 is the characteristic image expanded view of embodiment 2；
Fig. 7 is the characteristic image expanded view of embodiment 3；
Fig. 8 is grid chart；
Fig. 9, which is that pointer is practical, rotates 2 °, 4 °, 10 °, 15 °, 20 °, recognition result figure when 60 ° (from left to right)；
Figure 10, which is that pointer is practical, rotates to be 10 °, 30 °, 50 °, 70 °, 90 °, 110 °, 130 °, 150 °, 170 ° (from left to right)
When recognition result figure.
Specific implementation mode
The present invention provides a kind of readings of pointer type meters automatic identification equipments of robust, as shown in Figure 1, including a cylinder
Body 300, a camera 200 and the circuit board 210 for controlling camera 200, the instrument board 100 of the cylinder 300 and pointer instrument
The connection shaft at center is integrated, and the pointer tail portion of pointer instrument is set on cylinder 300, is rotated synchronously with cylinder 300,
The camera 200 is set to the instrument board frontside edge of pointer instrument, and the outer surface of the cylinder 300 is around a circle characteristic pattern
Case, the characteristic pattern of the camera lens face cylinder 300 of the camera 200；The circuit board 210 is hidden set on pointer instrument
Under instrument board 100, other positions of pointer instrument can also be set to.
As shown in Fig. 2, shaft of the optical axis of the camera 200 perpendicular to cylinder 300, and by cylinder 300
Center so that the heart is imaged cylinder 300 in the picture, convenient subsequently to the processing of image.The optical axis of the camera 200 can also
Non-perpendicular to the shaft installation of cylinder 300, or without the center by cylinder 300, at this point, being had an X-rayed in the image of shooting
Distortion is serious, is corrected to 300 image of cylinder using grid image, to measure again after eliminating perspective distortion.
The size of the cylinder 300 is related with the accuracy of identification of meter reading, determines that the size of cylinder 300 is this hair
One of bright key point accounts in terms of accuracy of instrument and shaft step size two, determines cylinder 300 separately below
Size：
A, from the aspect of accuracy of instrument：
Assuming that accuracy of instrument is A, range B, the corresponding angle of range is known as range angle C.General accuracy of identification is
3 times of accuracy of instrument, i.e. accuracy of identification are A/3, then identify the calculation formula of angle R：
It derives：
R=C*A/3 (2)
By image knowledge it is found that identifiable angle R at least needs a pixel (not consider sub- picture to reflect on the image
The situation of element), it is assumed that 300 a diameter of D of cylinder, 200 visual field of camera are FOV, and 200 resolution ratio of camera is M*N, then cylinder
The relationship of 300 diameters and accuracy of instrument is：
Formula (2) is substituted into formula (3) to derive：
Usually industrial table accuracy class is 0.5~4 grade, has made table 1 according to formula (4), wherein table 1 is enumerated to measure
Journey is the body diameter value of 0.5~2.5 grade of the speedometer of 210km/h.As can be seen from Table 1, when range, range angle, camera
200 resolution ratio, one timing, accuracy of instrument is higher, and required 300 diameter of cylinder is bigger, and 300 diameter change of cylinder is got over
Greatly；When range, range angle, one timing of accuracy of instrument, 200 resolution ratio of camera is bigger, and required 300 diameter of cylinder is smaller；
Therefore, cylinder 300 diameter needs be designed according to actual demand, can neither too greatly can not be too small, both influenced greatly very much
Appearance, weight also influence performance, too small not meet design principle, are designed according to some common locomotive instrumentation pointer tail portions,
Usually cylinder diameter is that 14-20mm is proper, and diameter is bigger, and meter reading accuracy of identification is higher.
Table 1：Range is the body diameter value of 0.5~2.5 grade of the speedometer of 210km/h
B, consider from shaft step size
Instrument central rotating shaft is directly related to the rotation of 300 structure of cylinder, minimum angles, that is, pointer of shaft rotation
The minimum angles of rotation at least need a pixel to reflect (the case where not considering sub-pix) on the image.Assuming that pointer rotates
Minimum angles be E, 300 a diameter of D of cylinder, 200 visual field of camera is FOV, and 200 resolution ratio of camera is M*N, then cylinder
300 diameters and the relationship of pointer rotation angle are：
It derives：
The rotation angle of shaft is related to stepper motor, and usually, stepper motor is rotated by 360 °, and pointer rotates 1 degree, refers to
The minimum angles of needle rotation are related with the minimum angles of stepper motor energy stepping, it is assumed that the step angle of stepper motor is H, high-subdividing number
S, then the minimum step angle H/S of stepper motor, it can thus be concluded that：
Formula (7) is brought into formula (6) to derive：
Table 2 is made according to formula (8), as can be seen from Table 2, when one timing of 200 resolution ratio of camera and high-subdividing number, step angle
Smaller, required 300 diameter of cylinder is bigger；When one timing of step angle, 200 resolution ratio of camera is bigger, required cylinder 300
Diameter is smaller.
Table 2：The body diameter value determined according to instrument central rotating shaft step size
| Step angle ° | High-subdividing number | 200 resolution ratio of camera | Visual field mm | 300 diameter mm of cylinder |
| H | S | M*N | FOV | D |
| 1.8 | 4 | 640*480 | 26 | 10.35 |
| 1.5 | 4 | 640*480 | 26 | 12.42 |
| 1.2 | 4 | 640*480 | 26 | 15.53 |
| 1.8 | 4 | 1280*960 | 35 | 6.97 |
| 1.5 | 4 | 1280*960 | 35 | 8.36 |
| 1.2 | 4 | 1280*960 | 35 | 10.45 |
Formula (7) shows that the ratio of step angle and high-subdividing number indicates that the minimum angles that pointer rotates, Tables 1 and 2 compare, table 2
Middle step angle and the ratio of high-subdividing number are equivalent to the identification angle in table 1,300 diameter of cylinder that Tables 1 and 2 calculates be
Under corresponding parameter, for identifiable angle theoretically attainable maximal accuracy.In fact, due to being indicated with a pixel
The change of angle is difficult to differentiate, so generally require a little can just recognize corresponding angle greatly than the value of theoretical calculation.
By above-mentioned analysis, 300 diameter of cylinder can be designed from accuracy of instrument and the consideration of shaft step size, but
Be from shaft step size consider in the case of, have a problem that, when shaft stepping minimum angles naked eyes be difficult recognize
When, 300 diameter of cylinder is designed according to it and is had little significance, and error in reading is generally all related with accuracy of instrument, so below
Embodiment designs 300 diameter of cylinder with accuracy of instrument.
A known cylinder 300, if whole is all identical color and without any other feature, human eye can not all be distinguished
The angle not rotated, is still more identified using image, so how cylinder 300 to be mapped with total indicator reading is also
The key point of the present invention, the present invention propose the design feature pattern on cylinder 300, when by identifying 0 scale and current scale
Characteristic pattern, by between characteristic pattern and total indicator reading establish be associated with, achieve the purpose that identification meter reading, the identification side
The flow of method is as shown in figure 3, include the following steps：
Step S1：Camera 200 is demarcated, picture centre pixel coordinate is obtained as datum mark, shoots 0 scale respectively
With image when current scale, and image is corrected；
Step S2：The edge pixel for extracting the characteristic pattern in image, does a vertical line with the datum mark of calibration, calculates
Line segment pixels value when to 0 scale and current scale between corresponding characteristic pattern edge and vertical line intersection point；
Step S3：The edge line segment calculated for pixel values cylinder of corresponding characteristic pattern when according to 0 scale and current scale
300 rotational angles；
Step S4：Since the angle that cylinder 300 rotates is equal to indicators turn angle, it can be 300 turns according to cylinder
Dynamic angle converts, and obtains current scale value.
With reference to specific embodiment, technical scheme of the present invention is described in more detail.
Embodiment 1
Present embodiments provide readings of pointer type meters automatic identification equipment and its recognition methods of a kind of robust, the knowledge
300 a diameter of D of cylinder of other device, the optical axis of camera 200 and pass through cylinder 300 perpendicular to the shaft of cylinder 300
Center.
By laser scoring or the modes such as label printing or film printing, characteristic pattern is set in the periphery of cylinder 300,
After the characteristic pattern expansion of the present embodiment as shown in figure 4, for a right angled triangle and one group of parallel lines combination, wherein described straight
The a line of angle triangle is parallel with the height of cylinder 300, and another a line is parallel with the base circumference of cylinder 300, right angle three
Angular slope is K, and the parallel lines are set to below right angled triangle, is two parallel with the base circumference of cylinder 300
Parallel lines, the distance between two parallel lines are G.
Based on features described above pattern, the recognition methods of the present embodiment includes the following steps：
Step S1：The pretreatment of image：
Camera 200 is demarcated, obtains the pixel coordinate of image center, right angled triangle with 300 bottom of cylinder
The one end at the angle of the parallel right-angle side of face circumference and bevel edge composition is directed at 300 side of cylinder towards 200 camera lens of camera, picture centre
The upper bottom surface at face center, cylinder 300 is parallel to image level direction；When indicators turn, cylinder 300 rotates therewith,
Right angled triangle thereon is rotated also along the high parallel direction of cylinder 300；Shoot pointer be directed toward 0 scale when and current scale
When image；
Image is corrected, so that the parallel lines in image is parallel to the horizontal direction of image, to avoid due to machining accuracy or dress
With on cylinder caused by error 300 parallel lines and image level direction it is not parallel.
Step S2：Picture edge characteristic extracts：
Point generates the vertical line of the mistake datum mark, this vertical line and right angled triangle on the basis of the image center of calibration
Two intersection points a and b are intersected at, intersect at two intersection points c and d with parallel lines, line segment pixels value, cd between calculating two intersection points of ab
Line segment pixels value between two intersection points,
The computational methods of pixel value between two line segment of ab, cd, specifically include following steps：
1. known intersection point a is in hypotenuse, intersection point b, c and d are on three parallel lines, centered on a intersection points,
Make the rectangle perpendicular to bevel edge, be the long axis of rectangle perpendicular to the direction of bevel edge, centered on intersection point c, makees perpendicular to parallel lines
Rectangle, which contains tri- intersection points of b, c and d, is the long axis of rectangle perpendicular to the direction of parallel lines, as shown in Figure 5；
2. rectangle corrects：When mapping, rectangle may be not orthogonal to edge, need to be corrected.In rectangle, carry out first
Edge detection obtains edge contour, each pixel and the tangent orientation angle of profile on profile is then calculated, finally to institute
The angled deflection for being averaged to obtain edge corrects the rectangular true vertical that obtains of rectangle in the square at edge using direction angle
Shape；
3. after the completion of rectangle correction, in rectangle, along rectangular centre line on the direction at edge, to each picture
Vegetarian refreshments position, all gray value of images summation that will be perpendicular in the rectangle width of long axis are averaged, obtain a width and be similar to side
The figure of wave.Figure horizontal axis is the pixel coordinate along rectangular centre long axis, and the longitudinal axis is the rectangle width perpendicular to rectangle long axis
Interior gray value mean value.Wherein, the low gray value in low ebb correspondence image, the high gray value in wave crest correspondence image, by low ebb
It is the position at edge to wave crest or by the position of wave crest to low ebb, since lines have certain width, so there are two
Marginal position is averaged two edge point coordinates, obtains the exact position at lines center, according to above method, respectively obtains
A, the exact pixels coordinate value of tetra- intersection points of b, c, d；
4. the line segment pixels value of 2 points of ab is calculated according to the exact pixels coordinate value of tetra- intersection points of a, b, c, d, 2 points of cd's
Line segment pixels value.
Step S3：Calculate 300 rotational angle of cylinder：
Assuming that when 0 scale, the pixel value in image between two intersection points of ab is Ia, and the pixel value between two intersection points of cd is Ib；
When current scale, the pixel value in image between two intersection points of ab is Ic, and the pixel value between two intersection points of cd is Ib '；Calculate cd two
The average value of pixel value between intersection pointThe distance between the slope K of known triangle and two parallel lines G, then each picture
Element size beThen the distance value of two intersection points of ab is P*Ia when 0 scale, and the distance value of two intersection points of ab is after rotation
P*Ic, then the arc length L that cylinder 300 rotates are：
L=(P*Ic-P*Ia)/K (9)
Cylinder 300 rotate angle F be：
Step S4：Calculate dial scale：
Since the relationship of dial scale and angle is known, what cylinder 300 rotated carves per the corresponding dial plate in once angle
Degree is B/C, then cylinder 300 and the relationship of dial scale Z are：
Z=F*B/C (11)
Similarly, when being designed to characteristic pattern, two parallel lines can pass through triangle interior, and two flat
Known to the distance between line；Also can a line segment parallel with triangle base, triangle directly be designed in 300 periphery of cylinder
Known to the distance between shape bottom edge and the line segment (one i.e. in parallel lines overlaps with the bottom surface of triangle).
Embodiment 2
The readings of pointer type meters automatic identification equipment of the robust of the present embodiment and its recognition methods, the identification device
Cylinder 300 a diameter of D, as shown in fig. 6, being a right angled triangle and two perpendicular to right angle three after the characteristic pattern expansion
The parallel segment on angular bottom edge forms, wherein the slope of the right angled triangle is K, one right-angle side and cylinder 300
High parallel, another a line (bottom edge) is parallel with the base circumference of cylinder 300, and two parallel segments are located at right angled triangle
Interior, endpoint is located on the edge of right angled triangle.
The optical axis of the camera 200 perpendicular to cylinder 300 shaft, and by cylinder 300 center.Work as pointer
In 0 scale, picture centre corresponds to short line segment, and rotary indicator makes pointer picture centre at a certain position correspond to long line
Section obtains the angle T rotated through between two lines section.Again with the calculating of pixel value between the line segment of 1 step S2 offers of embodiment
Method, the number of pixels for obtaining two parallel segments are respectively Na, Nb, then the angle Q that each pixel turns over meets：
With embodiment 1, when indicators turn is to current location, 1 step S1 image preprocessings of embodiment and step S2 are utilized
After picture edge characteristic extraction, the point on the basis of the picture centre of calibration counted the line of the datum mark being located in triangle
The number of pixels of section, it is assumed that the line segment pixels number that current time obtains is N 'b, then the angle F rotated through is：
Finally, F is converted into dial scale Z to get to the reading of instrument further according to formula (11).
Embodiment 3
The readings of pointer type meters automatic identification equipment of robust provided in this embodiment and its recognition methods, the identification dress
The 300 a diameter of D of cylinder set, the optical axis of camera 200 perpendicular to cylinder 300 shaft, and by cylinder 300
The heart.As shown in fig. 7, for around the saw-tooth like pattern and and saw-tooth like pattern of 300 periphery of cylinder after the characteristic pattern expansion
Bottom surface where straight line parallel parallel lines, wherein straight line where the bottom surface of sawtooth is parallel with the base circumference of cylinder 300,
The width of each sawtooth corresponds to a high range of instrument, and the bottom surface of each sawtooth is equipped with the character for distinguishing other sawtooth, such as word
Female or number etc., is distinguished in the present embodiment using letter.
When identification, it is known that the correspondence of letter and instrument high range, the slope K of each sawtooth, width under each sawtooth
W, corresponding scale size Y, the distance G between parallel lines, pointer picture centre in 0 scale are directed at 0 scale of sawtooth.When measurement,
It is still using picture centre as datum mark, wherein the size P of each pixel is as described in Example 1, meets：
After shooting piece image, identify the character under the corresponding sawtooth of picture centre first, it is known that pointer be directed toward it is a wide range of (Zl,
Zr).Then being accurately directed to for pointer is calculated, the pixel value Ic between the line segment intersected with sawtooth is obtained, then line segment physical length is
P*Ic, the corresponding scale Z calculation formula of pointer such as formula (14)：
Embodiment 4
The readings of pointer type meters automatic identification equipment of robust provided in this embodiment and its recognition methods, wherein described
The optical axis of the camera 200 of identification device and the non-perpendicular installation of shaft of cylinder 300 (300), or without passing through cylinder 300
Center.Perspective distortion is serious in the image shot at this time, can not directly acquire a certain section in triangle of distance.
The present embodiment is corrected to eliminate perspective distortion, so 300 image of cylinder using grid image (gridding method)
It measures again afterwards.
The grid is as shown in Figure 8, it is known that grid image generated hints obliquely at figure, using the image to subsequent all video cameras
300 image of cylinder of shooting is corrected, and the image after correction is equivalent to the flat image of the expansion of cylinder 300, in plane
Have four line segments on image, two intersecting lenses, two parallel lines, the corresponding angle of two intersecting lenses be triangle slope anyway
Cut value, then according to the angle Plane of rotation image, so that flat image is looked at and is shot with face cylinder 300, finally, choosing
Some point for selecting triangle interior is used as datum mark, generates the vertical line of the mistake datum mark.Cylinder 300 rotates preceding rotation
Afterwards, the pixel coordinate of the datum mark is consistent, and each image procossing is all the line segment distance for calculating the datum mark.After correction
Pattern eliminate perspective distortion, be equal to the optical axis of camera 200 perpendicular to the shaft of cylinder 300 and pass through cylinder 300
Center when the image that shoots, subsequent calculation and 300 triangle of cylinder of previous embodiment 1-3 refer to gauge pointer
Correspondence to scale is identical.
Two groups of experiments are carried out with the scheme of embodiment 1 below, to verify accuracy of identification, concrete operation step is as follows：
Known experiment material：Manual precision rotation platform, the camera 200OV5640 that resolution ratio is 2592 × 1944, instrument refer to
Needle, 300 characteristic pattern of cylinder.270 ° of known instrument range 210, range angle, accuracy of instrument 1.5%.According to the essence of instrument
Degree, accuracy of identification are 3 times of accuracy of instrument, the worst error ranging from 1.05 of permission.
300 characteristic pattern of cylinder has been printed by the way of printing, has been affixed on 300 pointer of cylinder, wherein
Parallel lines are located at the top of right angled triangle, are one group of heavy line, and the lower edges of the heavy line form one group of parallel lines, slightly
The width of solid line is 1mm, the as spacing between parallel lines, for demarcating Pixel Dimensions.Pointer is fixed on accurate rotation manually
On turntable, when test, manual rotation precision rotation platform shoots the image after 0 scale image and rotation angle, is surveyed with using image
The angle of amount is compared.
Experiment 1：As shown in Figure 9：The present image and 0 scale image rotation of the first row red numerical expression measurement in figure
Angle, two lines of numbers indicate that the current scale value measured, the absolute error that the third line digital representation measures, absolute error are just
Number indicates to be more than actual range, is that negative number representation is less than actual range, the relative error that fourth line digital representation measures.In figure,
Compared with 0 scale, the angle that actually rotates is respectively 2 °, 4 °, 10 °, 15 °, 20 °, 60 °, in terms of the measured value in figure, absolutely
Accidentally absolute value of the difference is both less than 1.05, meets the requirement that accuracy of identification is 3 times of accuracy of instrument.
Experiment 2：With the mode of experiment 1, the angle actually rotated is respectively 10 °, 30 °, 50 °, 70 °, 90 °, 110 °,
130 °, 150 °, 170 °, the results are shown in Figure 10, and in terms of the measured value in figure, the absolute value of absolute error is both less than 1.05,
Meet the requirement that accuracy of identification is 3 times of accuracy of instrument.
It is a kind of detailed embodiment and specific operating process of the present invention above, is before being with technical solution of the present invention
It puts and is implemented, but protection scope of the present invention is not limited to the above embodiments.
Claims (10)
1. a kind of readings of pointer type meters automatic identification equipment of robust, which is characterized in that including a cylinder (300), a phase
The circuit board (210) of machine (200) and control camera (200), the instrument board (100) of the cylinder (300) and pointer instrument
The connection shaft at center is integrated, and the pointer tail portion of pointer instrument is set on cylinder (300), and the camera (200) is set to
Instrument board (100) edge of pointer instrument, the outer surface of the cylinder (300) is around a circle characteristic pattern, the camera
(200) characteristic pattern of the camera lens towards cylinder (300).
2. a kind of readings of pointer type meters automatic identification equipment of robust according to claim 1, which is characterized in that described
The diameter D of cylinder (300) meets：
In formula, A is accuracy of instrument, and C is the corresponding angle of range (range angle), and FOV is camera (200) visual field, and M is camera
(200) horizontal pixel.
3. a kind of readings of pointer type meters automatic identification equipment of robust according to claim 1, which is characterized in that described
The diameter D of cylinder (300) meets：
In formula, H is the step angle of stepper motor, and S is the high-subdividing number of stepper motor, and FOV is camera (200) visual field, and M is camera
(200) horizontal pixel.
4. a kind of readings of pointer type meters automatic identification equipment of robust according to claim 1, which is characterized in that described
A diameter of 14-20mm of cylinder (300).
5. a kind of readings of pointer type meters automatic identification equipment of robust according to claim 1, which is characterized in that described
Characteristic pattern is the right angled triangle and one group of parallel lines around cylinder (300) periphery, one of the right angled triangle
Right-angle side is parallel with the height of cylinder (300), and the parallel lines and the bottom/top circumference of cylinder (300) are parallel or vertical.
6. a kind of readings of pointer type meters automatic identification equipment of robust according to claim 1, which is characterized in that described
Characteristic pattern is the saw-tooth like pattern and the parallel lines parallel with the bottom surface of saw-tooth like pattern around cylinder (300) periphery, institute
The width for stating sawtooth corresponds to the high range of instrument, and each sawtooth includes the character for distinguishing other sawtooth.
7. according to a kind of readings of pointer type meters automatic identification equipment of any robusts of claim 5-6, feature exists
In straight parallel with the bottom/top circumference of cylinder (300) with right angled triangle of, parallel lines in the parallel lines
A parallel lines in arm of angle coincidence or the parallel lines are overlapped with the straight line where the bottom surface of saw-tooth like pattern.
8. a kind of side of the identification of the readings of pointer type meters automatic identification equipment of robust as described in claim 1-7 is any
Method, which is characterized in that step includes：
Step S1：Camera (200) is demarcated, takes some fixed point in image as datum mark, obtains the pixel of the point
Coordinate, image when shooting 0 scale and current scale respectively, and image is corrected；
Step S2：The edge pixel for extracting the characteristic pattern in image, does a vertical line with the datum mark of calibration, is calculated 0
Line segment pixels value when scale and current scale between corresponding characteristic pattern edge and vertical line intersection point；
Step S3：The edge line segment calculated for pixel values cylinder (300) of corresponding characteristic pattern when according to 0 scale and current scale
Rotational angle；
Step S4：The angle that cylinder (300) rotates is converted into current scale value.
9. a kind of knowledge method for distinguishing of the readings of pointer type meters automatic identification equipment of robust according to claim 7,
It is characterized in that, in the step S1, correction includes level correction and the perspective distortion correction of image.
10. a kind of knowledge method for distinguishing of the readings of pointer type meters automatic identification equipment of robust according to claim 7,
It is characterized in that, in the step S1, selects point on the basis of the central point of image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810046784.XA CN108460331B (en) | 2018-01-18 | 2018-01-18 | Robust pointer instrument reading automatic identification device and identification method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810046784.XA CN108460331B (en) | 2018-01-18 | 2018-01-18 | Robust pointer instrument reading automatic identification device and identification method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108460331A true CN108460331A (en) | 2018-08-28 |
| CN108460331B CN108460331B (en) | 2022-12-23 |
Family
ID=63220634
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810046784.XA Active CN108460331B (en) | 2018-01-18 | 2018-01-18 | Robust pointer instrument reading automatic identification device and identification method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460331B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543682A (en) * | 2018-11-23 | 2019-03-29 | 电子科技大学 | A kind of readings of pointer type meters method based on deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2064926U (en) * | 1989-12-09 | 1990-10-31 | 薛宏斌 | Automatic range selector |
| CN104657711A (en) * | 2015-02-09 | 2015-05-27 | 中国石油大学(华东) | Robust pointer instrument reading automatic identification method |
| CN205090948U (en) * | 2015-11-02 | 2016-03-16 | 长春师范大学 | High accuracy corner locating support of straight line laser instrument |
| JP2017129531A (en) * | 2016-01-22 | 2017-07-27 | ファナック株式会社 | Automatic reading device for micrometers |
-
2018
- 2018-01-18 CN CN201810046784.XA patent/CN108460331B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2064926U (en) * | 1989-12-09 | 1990-10-31 | 薛宏斌 | Automatic range selector |
| CN104657711A (en) * | 2015-02-09 | 2015-05-27 | 中国石油大学(华东) | Robust pointer instrument reading automatic identification method |
| CN205090948U (en) * | 2015-11-02 | 2016-03-16 | 长春师范大学 | High accuracy corner locating support of straight line laser instrument |
| JP2017129531A (en) * | 2016-01-22 | 2017-07-27 | ファナック株式会社 | Automatic reading device for micrometers |
Non-Patent Citations (3)
| Title |
|---|
| LI YANWEN 等: ""Detection system of meter pointer based on computer vision"", 《IEEE》 * |
| 刘丽君: ""基于机器视觉的双排刻线器具自动读数研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 罗鑫 等: ""基于ADSP-2185m的嵌入式汉语语音合成系统"", 《信号处理》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543682A (en) * | 2018-11-23 | 2019-03-29 | 电子科技大学 | A kind of readings of pointer type meters method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108460331B (en) | 2022-12-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107167169B (en) | Readings of pointer type meters based on NI Vision Builder for Automated Inspection identifies measurement method | |
| CN108426701B (en) | Method and system for testing orientation elements in discontinuous star camera of CCD (Charge coupled device) | |
| CN101311963B (en) | Round mark point center picture projection point position acquiring method for positioning video camera | |
| CN104657711B (en) | A kind of readings of pointer type meters automatic identifying method of robust | |
| CN107101582A (en) | Axial workpiece run-out error On-line Measuring Method based on structure light vision | |
| CN105043259A (en) | Numerical control machine tool rotating shaft error detection method based on binocular vision | |
| CN110163918A (en) | A kind of line-structured light scaling method based on projective geometry | |
| CN107356202A (en) | A kind of laser scanning measurement system target sights method automatically | |
| CN109341601A (en) | A kind of Space locality establishment method of revolving platform central axis in vision measurement device | |
| CN110455222A (en) | A kind of high-precision rotary angle measuring method, device and equipment | |
| CN109712157B (en) | Gravity field method accelerometer calibration method based on monocular vision | |
| WO2024011981A1 (en) | Method for verifying light spot of cylindrical target | |
| CN111637851A (en) | Aruco code-based visual measurement method and device for plane rotation angle | |
| CN108830899A (en) | A kind of center of circle detection method that the quick self-adapted angle point based on geometrical characteristic combines | |
| JPH0914965A (en) | Target for surveying | |
| CN108460331A (en) | A kind of readings of pointer type meters automatic identification equipment of robust and its recognition methods | |
| CN208435805U (en) | Space positioning apparatus and rectifier | |
| CN114061472B (en) | Method for correcting measurement coordinate error based on target | |
| CN108662980B (en) | Welding stud position detection device and method based on silhouette technology | |
| CN107945228B (en) | Method for extracting height of oil tank based on single satellite image | |
| WO2023060717A1 (en) | High-precision positioning method and system for object surface | |
| CN112432612B (en) | High-precision micro rotation angle measuring method based on monocular vision | |
| CN108489515A (en) | A kind of big field stereo vision system calibration method based on out-of-focus image | |
| CN108286960A (en) | Focusing light pipe array apparatus and photography detection method | |
| CN114088088A (en) | Angular rate and angular acceleration measuring 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |