CN113610094A - Distribution room pointer instrument reading method based on rotation projection calibration - Google Patents

Distribution room pointer instrument reading method based on rotation projection calibration Download PDF

Info

Publication number
CN113610094A
CN113610094A CN202110998184.5A CN202110998184A CN113610094A CN 113610094 A CN113610094 A CN 113610094A CN 202110998184 A CN202110998184 A CN 202110998184A CN 113610094 A CN113610094 A CN 113610094A
Authority
CN
China
Prior art keywords
pointer instrument
pointer
distribution room
instrument
angle
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
Application number
CN202110998184.5A
Other languages
Chinese (zh)
Other versions
CN113610094B (en
Inventor
唐冬来
李玉
宋卫平
胡州明
黄璞
王赢方
叶鸿飞
杨俏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan Zhongdian Aostar Information Technologies Co ltd
Original Assignee
Sichuan Zhongdian Aostar Information Technologies Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Sichuan Zhongdian Aostar Information Technologies Co ltd filed Critical Sichuan Zhongdian Aostar Information Technologies Co ltd
Priority to CN202110998184.5A priority Critical patent/CN113610094B/en
Publication of CN113610094A publication Critical patent/CN113610094A/en
Application granted granted Critical
Publication of CN113610094B publication Critical patent/CN113610094B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computational Linguistics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Evolutionary Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a switchboard pointer instrument reading method based on rotation projection calibration. The influence on the identification precision caused by the angle, the dial plate pollution and the like is avoided, and more accurate automatic identification is realized.

Description

Distribution room pointer instrument reading method based on rotation projection calibration
Technical Field
The invention belongs to the technical field of computer image acquisition, identification and intelligent processing, and particularly relates to a distribution room pointer instrument reading method based on rotation projection calibration.
Background
The distribution room is an indoor distribution place with low-voltage load, mainly distributes electric energy for low-voltage users, and is provided with a medium-voltage incoming line (a small amount of outgoing lines can be provided), a distribution transformer and a low-voltage distribution device. Facilities of equipment with voltage class of 10kV or below are divided into a high-voltage distribution room and a low-voltage distribution room. The high-voltage distribution room generally refers to a 6 kV-10 kV high-voltage switch room; the low-voltage distribution room generally refers to a 400V distribution room for a transformer line of a 10kV or 35kV station, and a large number of pointer instruments are arranged in the distribution room.
The instrument pointer refers to a part used for indicating data on the instrument. The function of the pointer is to indicate complex data structures in a relatively objective and direct way. The pointer is generally elongated and pointed at one end to facilitate pointing and reading. Since accuracy is one of the main properties of the instrument, this design and construction of the pointer is advantageous for improving the accuracy of the instrument and reducing errors.
The distribution room is used as a key link for monitoring the carbon emission of the energy consumption unit, and has great significance for carrying out online monitoring on the energy consumption. Usually, the distribution room has a large number of pointer instruments such as current, oil temperature, etc., and these pointer instruments do not have data acquisition function. If the pointer instrument is read in a manual mode, the problems of waste of human resources, low data acquisition frequency and the like can be caused, and the requirement of the government on the carbon emission monitoring frequency of key energy consumption units can not be met. In recent years, with the rapid development of information technology, a distribution room inspection robot gradually replaces manual work to read images of a pointer instrument, but how to accurately read data of the pointer instrument is still a problem. When the robot is patrolled and examined in the electricity distribution room, the field of vision angle that shoots the pointer panel board at every turn has the change, and interference fringe influence on the pointer panel board has caused electricity distribution room pointer instrument recognition accuracy poor.
The current prior art for taking a meter pointer mainly comprises the following steps:
(1) and eliminating the influence of illumination on the identification of the pointer instrument through bilateral filtering, and sampling the data of the pointer instrument in the full convolution network identification image.
(2) And acquiring the central position of the pointer instrument panel through the mutation positions of the color pixel points, and acquiring the reading of the pointer instrument through color segmentation.
(3) And analyzing the characteristics of the image window through a linear classifier of a support vector machine, and detecting pointer instrument data through image gray scale information.
(4) The interference of external light on image reading is eliminated through a pointer instrument panel distribution histogram, and the positioning of the instrument pointer is realized by detecting an image corner point through a FAST rotating and rotating BRIEF (ORB) method.
(5) And detecting the edge of the current pointer instrument image through a convolutional neural network, and extracting the reading of the pointer instrument after the image quality is corrected by zooming the camera.
However, when the inspection robot of the distribution room inspects the power distribution room, the visual field angle of the pointer instrument panel shot each time changes, and interference stripes on the pointer instrument panel affect the change, so that the identification accuracy of the pointer instrument of the distribution room is poor. Such as:
(1) the invention has the following patents: an automatic reading method of a pointer instrument based on an intelligent inspection robot,
the key point of the method is that the dial plate of the pointer instrument is used as the center to rotate, and information such as scales and pointers is obtained. The key premise of this method is that the center point of the pointer meter can be confirmed by the pointer center position. But in practice, many mechanical pointer mechanisms of the pointer instrument are arranged at the lower right corner or other positions, and the position of the central point cannot be acquired by using the central point for confirmation. There is also no reference to the angle of the pointer instrument offset obtained. Meanwhile, the instrument panel is affected by interference fringes such as scratches, so that the identification accuracy is reduced, and no processing technology exists.
(2) The invention has the following patents: automatic reading method for scale of pointer dial
According to the method, only scale and pointer reading is considered, the problem of obtaining the angle of the instrument panel and the influence of interference scratches, dirt and seasonal fogging on the surface of the instrument are not considered, and if the angle has inclination and the interference scratches on the surface of the instrument, the error rate of the read data is high.
(3) The invention has the following patents: method for detecting key points of instrument pointer
According to the method, only scale and pointer reading is considered, the problem of obtaining the angle of the instrument panel and the influence of interference scratches, dirt and seasonal fogging on the surface of the instrument are not considered, and if the angle has inclination and the interference scratches on the surface of the instrument, the error rate of the read data is high.
Disclosure of Invention
Based on the defects of the prior art, the invention provides a switchboard pointer instrument reading method based on rotation projection calibration. The influence on the identification precision caused by the angle, the dial plate pollution and the like is avoided, and more accurate automatic identification is realized.
The specific implementation content of the invention is as follows:
the invention provides a distribution room pointer instrument reading method based on rotation projection calibration, which comprises the following steps of:
step 1: calibrating the shooting angle of a pointer instrument of a distribution room;
step 2: dividing the dial blocks, and then extracting the scale characteristics and the pointer color characteristics to finish the extraction of the scale data of the pointer instrument of the power distribution room;
and step 3: and identifying the scale data of the pointer instrument of the distribution room by adopting Hough transform.
In order to better implement the present invention, further, the step 1 specifically includes the following steps:
step 1.1: acquiring image edge characteristics of a pointer instrument of a power distribution room by adopting a canny edge detection algorithm;
step 1.2: acquiring a central point and 4 edge points of a pointer instrument of a power distribution room by adopting a line detection algorithm;
step 1.3: calculating the inclination angle of the pointer instrument of the power distribution room through a cosine trigonometric function by using the coordinate data of the central point and the 4 edge points of the pointer instrument of the power distribution room;
step 1.4: pooling image edge areas of the pointer instrument of the power distribution room by adopting an area convolution neural network to obtain a characteristic diagram of a fixed area of a pointer table of the pointer instrument;
step 1.5: training the pointer rotation angle label by adopting a regional convolution neural network;
step 1.6: and performing angle rotation on the fixed area of the pointer table on the pointer instrument picture according to the training result to finish angle calibration.
In order to better implement the method, an image angle threshold value is further set, and the shot images of the distribution room pointer instrument image with the angle outside the angle threshold value range are discarded and shot again.
In order to better implement the present invention, further, the method for calculating the image angle of the pointer instrument of the electricity distribution room comprises: setting the distance length from the center point of the pointer instrument of the electricity distribution room to one vertex as LaAcquiring the distance length from the instrument central point to the corresponding vertex through the instrument file data of the distribution room to be LbThe distribution room pointer instrument image angle delta theta shot by the distribution room inspection robot is as follows:
Figure BDA0003234536110000031
in order to better implement the present invention, further, the step 2 specifically includes the following steps:
step 2.1: the image data of the pointer instrument of the power distribution room after angle correction is set as F (x)b,yb) Fixing the space scale of the standard feature transformation algorithm as sigma, and adopting Gaussian filtering to obtain image information L (x)b,yb,σ)；
Step 2.2: setting u and v at the edge of the pointer instrument of the power distribution room after angle calibration, and calculating to obtain a spatial scale extreme value P (x) of the pointer instrument of the power distribution roomb,ybσ), the specific calculation formula is:
P(xb,yb,σ)＝L(xbu,ybu,σu)-L(xbv,ybv,σv)；
step 2.3: after the spatial scale extreme value of the pointer instrument in the power distribution room is detected, scale conversion is carried out on the image data of the pointer instrument to obtain a conversion value EaThe specific calculation formula is as follows:
Ea＝P(xb,yb,σ)×H(xb,yb)；
step 2.4: obtaining scale characteristic values of a pointer instrument of a power distribution room through Red, Green and Blue (RGB) gray level detection;
step 2.5: let the coordinate of any point of the scale of the pointer instrument be (x)c,yc) The vertex coordinate is (x)d,yd) The image integral function is phi, and the sum V (x) of the gray values is obtained through calculationc,yc) The specific calculation formula is as follows:
Figure BDA0003234536110000041
step 2.6: the coordinates of four corner edge points of the scale of the pointer instrument are set as V (x)c1,yc1)、V(xc2,yc2)、V(xc3,yc3)、V(xc4,yc4) Calculating to obtain the region gray value PkThe specific calculation formula is as follows:
Pk＝V(xc4,yc4)-V(xc2,yc2)-V(xc3,yc3)+V(xc1,yc1)；
step 2.7: establishing a scale matrix of a pointer instrument of a power distribution room, and setting coordinates of scales and pointers as xz、yzThe detection point is JcAnd calculating to obtain a detection matrix JsThe specific calculation formula is as follows:
Figure BDA0003234536110000042
step 2.8: and extracting the scale data of the pointer instrument of the power distribution room according to RGB color detection.
In order to better implement the present invention, further, the specific operations of step 3 are:
step 3.1: constructing an image according to the extracted scale data of the switchboard pointer instrument, and setting the pointer independent variable of the switchboard pointer instrument as R in the constructed imagee(ii) a The slope of the straight line of the pointer is PeIntercept of FjCalculating the dependent variable S of the pointerkThe specific calculation formula is as follows:
Sk＝PeRe+Fj；
step 3.2: the sine angle of the pointer visual angle is set as
Figure BDA0003234536110000043
Cosine angle of
Figure BDA0003234536110000044
Expressing a straight line Z of a pointer instrument of a distribution room by using a parameter equation through Hough transformationtThe concrete formula is as follows:
Figure BDA0003234536110000045
step 3.3: and after the straight-line pointer characteristic of the pointer instrument of the power distribution room is obtained, the reading of the pointer instrument is obtained according to the scale of the instrument panel.
Compared with the prior art, the invention has the following advantages and beneficial effects:
(1) the angle calibration is realized through a self-grinding rotation calibration algorithm, so that the measurement result is more accurate;
(2) the fixed standard feature transformation algorithm is improved, the dial plate block division, the scale, the pointer color feature extraction and the scale splicing are added, and the influence of interference fringe data of the instrument dial plate is reduced;
(3) the pointer coordinate data of the pointer instrument of the electricity distribution room is converted into the parameter space data through Hough transform, so that the accuracy of straight line identification of the pointer instrument of the electricity distribution room can be improved, and the influence of a strip interference source on the reading of the pointer instrument of the electricity distribution room is reduced.
Drawings
FIG. 1 is a schematic view of a complete flow chart of an embodiment of the present invention.
Detailed Description
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments, and therefore should not be considered as a limitation to the scope of protection. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
In the description of the present invention, it is to be noted that, unless otherwise explicitly specified or limited, the terms "disposed," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Example 1:
the embodiment provides a switchboard pointer instrument reading method based on rotation projection calibration, as shown in fig. 1, including the following steps:
step 1: calibrating the shooting angle of a pointer instrument of a distribution room; the step 1 specifically comprises the following steps:
step 1.1: acquiring image edge characteristics of a pointer instrument of a power distribution room by adopting a canny edge detection algorithm;
step 1.2: acquiring a central point and 4 edge points of a pointer instrument of a power distribution room by adopting a line detection algorithm;
step 1.3: calculating the inclination angle of the pointer instrument of the power distribution room through a cosine trigonometric function by using the coordinate data of the central point and the 4 edge points of the pointer instrument of the power distribution room;
step 1.4: pooling image edge areas of the pointer instrument of the power distribution room by adopting an area convolution neural network to obtain a characteristic diagram of a fixed area of a pointer table of the pointer instrument;
step 1.5: training the pointer rotation angle label by adopting a regional convolution neural network;
step 1.6: and performing angle rotation on the fixed area of the pointer table on the pointer instrument picture according to the training result to finish angle calibration.
Step 2: dividing the dial blocks, and then extracting the scale characteristics and the pointer color characteristics to finish the extraction of the scale data of the pointer instrument of the power distribution room; the step 2 specifically comprises the following steps:
step 2.1: the image data of the pointer instrument of the power distribution room after angle correction is set as F (x)b,yb) Fixing the space scale of the standard feature transformation algorithm as sigma, and adopting Gaussian filtering to obtain image information L (x)b,yb,σ)；
Step 2.2: setting u and v at the edge of the pointer instrument of the power distribution room after angle calibration, and calculating to obtain a spatial scale extreme value P (x) of the pointer instrument of the power distribution roomb,ybσ), the specific calculation formula is:
P(xb,yb,σ)＝L(xbu,ybu,σu)-L(xbv,ybv,σv)；
step 2.3: after the spatial scale extreme value of the pointer instrument in the power distribution room is detected, scale conversion is carried out on the image data of the pointer instrument to obtain a conversion value EaThe specific calculation formula is as follows:
Ea＝P(xb,yb,σ)×H(xb,yb)；
step 2.4: obtaining scale characteristic values of a pointer instrument of a power distribution room through Red, Green and Blue (RGB) gray level detection;
step 2.5: let the coordinate of any point of the scale of the pointer instrument be (x)c,yc) The vertex coordinate is (x)d,yd) The image integral function is phi, and the sum V (x) of the gray values is obtained through calculationc,yc) The specific calculation formula is as follows:
Figure BDA0003234536110000061
step 2.6: the coordinates of four corner edge points of the scale of the pointer instrument are set as V (x)c1,yc1)、V(xc2,yc2)、V(xc3,yc3)、V(xc4,yc4) Calculating to obtain the region gray value PkThe specific calculation formula is as follows:
Pk＝V(xc4,yc4)-V(xc2,yc2)-V(xc3,yc3)+V(xc1,yc1)；
step 2.7: establishing a scale matrix of a pointer instrument of a power distribution room, and setting coordinates of scales and pointers as xz、yzThe detection point is JcAnd calculating to obtain a detection matrix JsThe specific calculation formula is as follows:
Figure BDA0003234536110000071
step 2.8: and extracting the scale data of the pointer instrument of the power distribution room according to RGB color detection.
And step 3: and identifying the scale data of the pointer instrument of the distribution room by adopting Hough transform. The specific operation of the step 3 is as follows:
step 3.1: constructing an image according to the extracted scale data of the switchboard pointer instrument, and setting the pointer independent variable of the switchboard pointer instrument as R in the constructed imagee(ii) a The slope of the straight line of the pointer is PeIntercept of FjCalculating the dependent variable S of the pointerkThe specific calculation formula is as follows:
Sk＝PeRe+Fj；
step 3.2: the sine angle of the pointer visual angle is set as
Figure BDA0003234536110000072
Cosine angle of
Figure BDA0003234536110000073
Expressing a straight line Z of a pointer instrument of a distribution room by using a parameter equation through Hough transformationtThe concrete formula is as follows:
Figure BDA0003234536110000074
step 3.3: and after the straight-line pointer characteristic of the pointer instrument of the power distribution room is obtained, the reading of the pointer instrument is obtained according to the scale of the instrument panel.
The working principle is as follows: the flow of the distribution room pointer instrument automatic reading method based on Hough transform is shown in fig. 1, and the pointer instrument automatic reading branch pointer instrument shooting angle calibration, pointer instrument scale data extraction and pointer instrument data identification are carried out.
In the pointer instrument shooting angle calibration link, firstly, the shooting image angle of the inspection robot in the power distribution room is identified, if the shooting image angle exceeds a threshold value, a picture is shot again, and then, the regional convolution neural network is adopted to carry out rotation calibration on the pointer instrument picture data. In the pointer instrument scale data extraction link, pointer instrument data of a power distribution room are segmented through a fixed standard feature transformation algorithm, and then pointer instrument scale data extraction is carried out based on color feature data. And in a pointer instrument data identification link, extracting the pointer reading of the pointer instrument through Hough transform.
Firstly, calibrating the shooting angle of a pointer instrument:
the distribution room inspection robot has a view angle difference when shooting a distribution room pointer instrument every time, and the identification accuracy of the distribution room pointer instrument is reduced. Therefore, the angle calibration of the shot image is needed before the pointer instrument of the power distribution room is identified.
The Region-based connected Neural Network (R-CNN) has the capability of image feature learning, and can be used for shooting angle offset rotation calibration of the pointer instrument through model training.
And if the angle of the pointer instrument image is out of the threshold range, controlling the power distribution room inspection robot to shoot the pointer instrument image again.
Firstly, acquiring the image edge characteristics of the pointer instrument of the power distribution room by adopting canny edge detection,
and then, obtaining the central point and 4 edge points of the pointer instrument of the power distribution room by adopting line detection, and calculating the inclination angle of the pointer instrument of the power distribution room through a cosine trigonometric function after obtaining the coordinate data of the central point and 4 edge points of the pointer instrument of the power distribution room.
Setting the distance length from the center point of the pointer instrument of the electricity distribution room to one vertex as LaAcquiring the distance length from the instrument central point to the corresponding vertex through the instrument file data of the distribution room to be LbThe pointer instrument angle delta theta shot by the distribution room inspection robot is as follows:
Figure BDA0003234536110000081
and if the angle of the pointer instrument image is out of the threshold range, controlling the inspection robot of the power distribution room to shoot the pointer instrument image again until the angle of the pointer instrument image meets the image recognition requirement.
Then, pooling image edge areas of the pointer instrument by using an area convolution neural network to obtain a characteristic map of a fixed area of the pointer instrument. Let the length of the characteristic region image of the pointer instrument be haWidth of laThe length of the horizontal axis of the position of the center point of the pointer instrument is JzLength of longitudinal axis yaHorizontal line G of the longest side pointer instrument imageuComprises the following steps:
Gu＝max(Jz) (2)
secondly, training the pointer rotation label by using a regional convolution neural network, setting the offset angle of a pointer instrument as delta beta, setting the region displayed in the image as K, setting the rotation angle times as s, and setting a regression target B (v) as:
Figure BDA0003234536110000082
and finally, performing angle rotation on the pointer instrument picture area to realize angle calibration of the pointer instrument picture.
Second, pointer instrument scale data extraction:
the fixed standard feature transformation algorithm is a local feature analysis algorithm aiming at image scale space identification. The pointer instrument data block analysis is realized by a method of searching an extreme point in a spatial scale so as to improve the identification precision [16-17], and the method can be used for pointer instrument data identification. The fixed standard feature transformation algorithm is used in the text, dial block segmentation, scale extraction, pointer color feature extraction and scale splicing are added, and the influence of interference fringe data of the instrument dial can be reduced.
The image data of the pointer instrument of the power distribution room after angle correction is set as F (x)b,yb) Fixing the space scale of the standard feature transformation algorithm as sigma, and adopting Gaussian filtering to obtain image information L (x)b,yb,σ)，
Setting u and v edges of the distribution room pointer instrument after angle calibration, and setting the distribution room pointer instrumentSpatial scale extremum P (x)b,ybσ) is:
P(xb,yb,σ)＝L(xbu,ybu,σu)-L(xbv,ybv,σv) (4)
after the spatial scale of the pointer instrument in the power distribution room is detected, the scale conversion is carried out on the image data of the pointer instrument, and the conversion value EaComprises the following steps:
Ea＝P(xb,yb,σ)×H(xb,yb) (5)
on the basis, the dial image data after the dimension conversion of the distribution room pointer instrument is segmented to eliminate the influence of interference fringes on the distribution room pointer instrument. And then, obtaining the scale characteristic value of the pointer instrument of the power distribution room through Red, Green and Blue (RGB) gray detection.
Let the coordinate of any point of the scale of the pointer instrument be (x)c,yc) The vertex coordinate is (x)d,yd) The image integration function is phi, and the sum of the gray values is V (x)c,yc)：
Figure BDA0003234536110000091
The coordinates of four corner edge points of the scale of the pointer instrument are set as V (x)c1,yc1)、V(xc2,yc2)、V(xc3,yc3)、V(xc4,yc4) Calculating the region gray-scale value Pk：
Pk＝V(xc4,yc4)-V(xc2,yc2)-V(xc3,yc3)+V(xc1,yc1) (7)
Finally, a scale matrix of the pointer instrument of the power distribution room is established, and coordinates of scales and pointers are set as xz、yzThe detection point is JcDetection matrix JsComprises the following steps:
Figure BDA0003234536110000101
through RGB color detection, the scale data of the pointer instrument of the power distribution room can be extracted.
Thirdly, identifying data of the pointer instrument:
the Hough transform is adopted in the text to convert pointer coordinate data of the pointer instrument of the electricity distribution room into parameter space data [18-19], so that the accuracy of straight line identification of the pointer instrument of the electricity distribution room can be improved, and the influence of a strip interference source on the reading of the pointer instrument of the electricity distribution room is reduced.
Set in image space, the pointer independent variable of the pointer instrument is ReThe slope of the straight line of the pointer is PeIntercept of FjDependent variable S of pointerkComprises the following steps:
Sk＝PeRe+Fj (9)
the sine angle of the pointer visual angle is set as
Figure BDA0003234536110000102
Cosine angle of
Figure BDA0003234536110000103
Expressing a straight line Z of a pointer instrument of a distribution room by using a parameter equation through Hough transformationt：
Figure BDA0003234536110000104
And after the straight-line pointer characteristic of the pointer instrument of the power distribution room is obtained, the reading of the pointer instrument is obtained according to the scale of the instrument panel.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and all simple modifications and equivalent variations of the above embodiments according to the technical spirit of the present invention are included in the scope of the present invention.

Claims (6)

1. A distribution room pointer instrument reading method based on rotation projection calibration is characterized by comprising the following steps:
step 1: calibrating the shooting angle of a pointer instrument of a distribution room;
step 2: dividing the dial blocks, and then extracting the scale characteristics and the pointer color characteristics to finish the extraction of the scale data of the pointer instrument of the power distribution room;
and step 3: and identifying the scale data of the pointer instrument of the distribution room by adopting Hough transform.
2. The distribution room pointer instrument reading method based on the rotating projection calibration as claimed in claim 1, wherein the step 1 specifically comprises the following steps:
step 1.1: acquiring image edge characteristics of a pointer instrument of a power distribution room by adopting a canny edge detection algorithm;
step 1.2: acquiring a central point and 4 edge points of a pointer instrument of a power distribution room by adopting a line detection algorithm;
step 1.3: calculating the inclination angle of the pointer instrument of the power distribution room through a cosine trigonometric function by using the coordinate data of the central point and the 4 edge points of the pointer instrument of the power distribution room;
step 1.4: pooling image edge areas of the pointer instrument of the power distribution room by adopting an area convolution neural network to obtain a characteristic diagram of a fixed area of a pointer table of the pointer instrument;
step 1.5: training the pointer rotation angle label by adopting a regional convolution neural network;
step 1.6: and performing angle rotation on the fixed area of the pointer table on the pointer instrument picture according to the training result to finish angle calibration.
3. A method as claimed in claim 2, wherein an image angle threshold is set, and images taken of the switchhouse pointer instrument image angle outside the angle threshold are discarded and re-taken.
4. The method of claim 3A switchboard pointer instrument reading method based on rotation projection calibration is characterized in that a switchboard pointer instrument image angle calculation method comprises the following steps: setting the distance length from the center point of the pointer instrument of the electricity distribution room to one vertex as LaAcquiring the distance length from the instrument central point to the corresponding vertex through the instrument file data of the distribution room to be LbThe distribution room pointer instrument image angle delta theta shot by the distribution room inspection robot is as follows:
Figure FDA0003234536100000011
5. a cubicle pointer instrument reading method based on rotation projection calibration according to claim 1, 2, 3 or 4, wherein said step 2 comprises the following steps:
step 2.1: the image data of the pointer instrument of the power distribution room after angle correction is set as F (x)b,yb) Fixing the space scale of the standard feature transformation algorithm as sigma, and adopting Gaussian filtering to obtain image information L (x)b,yb,σ)；
Step 2.2: setting u and v at the edge of the pointer instrument of the power distribution room after angle calibration, and calculating to obtain a spatial scale extreme value P (x) of the pointer instrument of the power distribution roomb,ybσ), the specific calculation formula is:
P(xb,yb,σ)＝L(xbu,ybu,σu)-L(xbv,ybv,σv)；
step 2.3: after the spatial scale extreme value of the pointer instrument in the power distribution room is detected, scale conversion is carried out on the image data of the pointer instrument to obtain a conversion value EaThe specific calculation formula is as follows:
Ea＝P(xb,yb,σ)×H(xb,yb)；
step 2.4: obtaining scale characteristic values of a pointer instrument of a power distribution room through Red, Green and Blue (RGB) gray level detection;
step 2.5: let the coordinate of any point of the scale of the pointer instrument be (x)c,yc) The vertex coordinate is (x)d,yd) The image integral function is phi, and the sum V (x) of the gray values is obtained through calculationc,yc) The specific calculation formula is as follows:
Figure FDA0003234536100000021
step 2.6: the coordinates of four corner edge points of the scale of the pointer instrument are set as V (x)c1,yc1)、V(xc2,yc2)、V(xc3,yc3)、V(xc4,yc4) Calculating to obtain the region gray value PkThe specific calculation formula is as follows:
Pk＝V(xc4,yc4)-V(xc2,yc2)
-V(xc3,yc3)+V(xc1,yc1)；
step 2.7: establishing a scale matrix of a pointer instrument of a power distribution room, and setting coordinates of scales and pointers as xz、yzThe detection point is JcAnd calculating to obtain a detection matrix JsThe specific calculation formula is as follows:
Figure FDA0003234536100000022
step 2.8: and extracting the scale data of the pointer instrument of the power distribution room according to RGB color detection.
6. A distribution room pointer instrument reading method based on rotation projection calibration as claimed in claim 1, characterized in that the specific operation of step 3 is:
step 3.1: constructing an image according to the extracted scale data of the switchboard pointer instrument, and setting the pointer independent variable of the switchboard pointer instrument as R in the constructed imagee(ii) a The slope of the straight line of the pointer is PeCutting offDistance of FjCalculating the dependent variable S of the pointerkThe specific calculation formula is as follows:
Sk＝PeRe+Fj；
step 3.2: setting the sine angle and the cosine angle of the pointer visual angle to be sin theta and cos theta, and representing the straight line Z of the pointer instrument of the distribution room by a parameter equation through Hough transformationtThe concrete formula is as follows:
Zt＝Recosθ+Sksinθ；
step 3.3: and after the straight-line pointer characteristic of the pointer instrument of the power distribution room is obtained, the reading of the pointer instrument is obtained according to the scale of the instrument panel.
CN202110998184.5A 2021-08-27 2021-08-27 Distribution room pointer instrument reading method based on rotation projection calibration Active CN113610094B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110998184.5A CN113610094B (en) 2021-08-27 2021-08-27 Distribution room pointer instrument reading method based on rotation projection calibration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110998184.5A CN113610094B (en) 2021-08-27 2021-08-27 Distribution room pointer instrument reading method based on rotation projection calibration

Publications (2)

Publication Number Publication Date
CN113610094A true CN113610094A (en) 2021-11-05
CN113610094B CN113610094B (en) 2023-06-02

Family

ID=78309626

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110998184.5A Active CN113610094B (en) 2021-08-27 2021-08-27 Distribution room pointer instrument reading method based on rotation projection calibration

Country Status (1)

Country Link
CN (1) CN113610094B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114339179A (en) * 2021-12-23 2022-04-12 深圳市火乐科技发展有限公司 Projection correction method, projection correction device, storage medium and projection equipment

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711400A (en) * 2018-11-06 2019-05-03 华中科技大学 A kind of electric inspection process method and apparatus identifying simulated pointer formula meter reading
US20210217129A1 (en) * 2016-09-26 2021-07-15 Digimarc Corporation Detection of encoded signals and icons

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20210217129A1 (en) * 2016-09-26 2021-07-15 Digimarc Corporation Detection of encoded signals and icons
CN109711400A (en) * 2018-11-06 2019-05-03 华中科技大学 A kind of electric inspection process method and apparatus identifying simulated pointer formula meter reading

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
JIAN-XUN MI 等: "A Robust and Automatic Recognition Method of Pointer Instruments in Power System", INTELLIGENT COMPUTING THEORIES AND APPLICATION, pages 223 *
徐林加: "基于图像识别的多指针仪表自动读数方法研究与实现", 中国优秀硕士学位论文全文数据库 (信息科技辑), pages 138 - 949 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114339179A (en) * 2021-12-23 2022-04-12 深圳市火乐科技发展有限公司 Projection correction method, projection correction device, storage medium and projection equipment
CN114339179B (en) * 2021-12-23 2024-05-28 深圳市火乐科技发展有限公司 Projection correction method, apparatus, storage medium and projection device

Also Published As

Publication number Publication date
CN113610094B (en) 2023-06-02

Similar Documents

Publication Publication Date Title
CN106951900B (en) A kind of automatic identifying method of arrester meter reading
CN112906694B (en) Reading correction system and method for transformer substation inclined pointer instrument image
CN104050446B (en) Meter pointer image identification method based on pointer width character
CN109635806B (en) Ammeter value identification method based on residual error network
CN110580480A (en) Meter reading identification method based on image processing
CN109284718B (en) Inspection robot-oriented variable-view-angle multi-instrument simultaneous identification method
CN110807355A (en) Pointer instrument detection and reading identification method based on mobile robot
CN104236498B (en) Method for measuring end size of switch rail
CN108764134A (en) A kind of automatic positioning of polymorphic type instrument and recognition methods suitable for crusing robot
CN111598942A (en) Method and system for automatically positioning electric power facility instrument
CN110852213B (en) Pointer instrument multi-condition automatic reading method based on template matching
CN111368906A (en) Pointer type oil level indicator reading identification method based on deep learning
CN111401357A (en) Pointer instrument reading method based on text detection
CN113469178B (en) Power meter identification method based on deep learning
CN112560837A (en) Reading method, device, equipment and storage medium of pointer instrument
CN113610094B (en) Distribution room pointer instrument reading method based on rotation projection calibration
CN113902894B (en) Automatic reading identification method for strip level based on image processing
CN105423975A (en) Calibration system and method of large-size workpiece
CN113705564B (en) Pointer type instrument identification reading method
CN113705350A (en) Pointer instrument reading identification method and device for transformer substation, medium and electronic equipment
CN109359637B (en) Pointer instrument value reading method based on machine vision
Wu et al. Intelligent instrument recognition scheme based on unattended substation inspection
CN115661446A (en) Pointer instrument indication automatic reading system and method based on deep learning
CN114037993A (en) Substation pointer instrument reading method and device, storage medium and electronic equipment
CN111652055B (en) Intelligent switch instrument identification method based on two-stage positioning

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