CN113762239A - Meter reflection identification method based on inspection robot - Google Patents
Meter reflection identification method based on inspection robot Download PDFInfo
- Publication number
- CN113762239A CN113762239A CN202110915177.4A CN202110915177A CN113762239A CN 113762239 A CN113762239 A CN 113762239A CN 202110915177 A CN202110915177 A CN 202110915177A CN 113762239 A CN113762239 A CN 113762239A
- Authority
- CN
- China
- Prior art keywords
- meter
- inspection robot
- area
- information
- measured
- 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
- 238000007689 inspection Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000008569 process Effects 0.000 claims abstract description 16
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 description 10
- 230000000007 visual effect Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000011514 reflex Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Manipulator (AREA)
Abstract
The invention relates to a meter light reflection identification method based on an inspection robot, which comprises the steps of determining the area where a meter to be detected is located to obtain meter area information; controlling the inspection robot to move to the area of the meter to be measured according to the meter area information; acquiring the reflection intensity of the area of the meter to be measured through the inspection robot; executing a judging process, wherein the judging process is to judge whether the reflection intensity is smaller than a preset threshold value: if not, acquiring image information of the meter to be measured through the inspection robot, correcting the image information to obtain correction information, and identifying the reading of the meter to be measured according to the correction information; if so, controlling the inspection robot to move according to the reflection intensity, acquiring the reflection intensity of the area of the meter to be measured in real time and executing a judgment process. The method can effectively reduce the influence of the reflection on the recognition algorithm, is simple and convenient, greatly reduces the interference of the reflection on the recognition algorithm, and ensures the recognition accuracy.
Description
Technical Field
The invention relates to the technical field of instrument and meter detection, in particular to a meter reflection identification method based on an inspection robot.
Background
The transformer substation is a place for converting voltage and current, receiving electric energy and distributing electric energy in an electric power system, the transformer substation in a power plant is a boosting transformer substation, the boosting transformer substation is used for boosting the electric energy generated by a generator and then feeding the electric energy into a high-voltage power grid, a patrol inspection system of the transformer substation is an effective measure for ensuring normal and safe operation of equipment, the equipment operation condition is known through regular patrol inspection of an attendant, operation abnormity is mastered, corresponding measures are taken timely, the method has important significance for reducing the occurrence of accidents and the influence range of the accidents, and a pointer instrument is an important tool for monitoring the operation state of the transformer substation.
At present, most of domestic transformer substations adopt a manual mode to record meters, but the monitoring of the meters is work which consumes time and has danger, so that manpower and material resources are wasted, and a quadruped robot is utilized to replace a person to patrol and examine, so that the method is a good alternative mode, but the identification precision of the meters is greatly influenced due to the fact that the quadruped robot is in a light reflection problem of the outdoor meters.
Disclosure of Invention
The invention aims to provide a meter reflection identification method based on an inspection robot, which can improve the accuracy of meter identification.
In order to achieve the purpose, the invention provides the following scheme:
a meter reflection identification method based on a patrol robot comprises the following steps:
determining the area of the meter to be measured to obtain meter area information;
controlling the inspection robot to move to the area of the meter to be measured according to the meter area information;
acquiring the reflection intensity of the region of the meter to be measured through the inspection robot;
executing a judging process, wherein the judging process is to judge whether the reflection intensity is smaller than a preset threshold value:
if not, acquiring image information of the meter to be measured through the inspection robot, correcting the image information to obtain correction information, and identifying the reading of the meter to be measured according to the correction information;
if yes, controlling the inspection robot to move according to the reflection intensity, acquiring the reflection intensity of the area of the meter to be detected in real time, and executing the judging process.
Preferably, the determining the region where the meter to be measured is located to obtain meter region information includes:
acquiring a position image of the meter to be measured; the position image comprises a bit area of the meter to be measured in the image;
and detecting the position image by using a YOLOv5 detection algorithm to obtain the meter region information.
Preferably, through patrol and examine robot and obtain the regional reflection intensity of the table meter that awaits measuring includes:
acquiring pixel information of an area of the meter to be measured by using the inspection robot; the pixel information comprises the number of pixel points of the region and the number of pixel points of the region;
and calculating the light reflection intensity according to the pixel information.
Preferably, the formula for calculating the reflection intensity includes:
wherein Np _ meter represents the number of the pixel points, Np _ reflect represents the number of the pixel points, and S represents the reflection intensity; the reflection intensity is the proportion of reflection points in the area of the meter to be measured.
Preferably, the correcting the image information to obtain correction information includes:
collecting a plurality of groups of image key points in the image information according to a G-RMI algorithm;
calculating an affine matrix of elevation transformation according to the image key points;
and determining the correction information according to the affine matrix.
Preferably, the identifying the reading of the meter under test according to the correction information includes:
identifying the correction information by using a depth regression algorithm, and regressing the direction of the instrument pointer in the correction information in an end-to-end mode;
and determining the reading of the meter to be measured according to the direction of the meter pointer.
Preferably, the identifying the correction information by using a depth regression algorithm, and regressing the meter pointer direction in the correction information in an end-to-end manner includes:
acquiring image characteristics of the correction information through a ResNet50 network;
and predicting the pointer direction of the image characteristics according to a direction regression module to obtain the direction of the instrument pointer.
Preferably, the controlling the inspection robot to move according to the reflection intensity includes:
and controlling the position of the inspection robot to be adjusted by utilizing a PID algorithm according to the reflection intensity.
Preferably, the method further comprises the following steps:
controlling the movement speed of the inspection robot according to the reflection intensity; the motion of the inspection robot rotates by taking the meter to be measured as a rotation center.
Preferably, the inspection robot is a quadruped robot; the quadruped robot comprises at least four mechanical legs.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides a meter light reflection identification method based on an inspection robot, which comprises the steps of determining the area where a meter to be detected is located to obtain meter area information; controlling the inspection robot to move to the area of the meter to be measured according to the meter area information; acquiring the reflection intensity of the region of the meter to be measured through the inspection robot; executing a judging process, wherein the judging process is to judge whether the reflection intensity is smaller than a preset threshold value: if not, acquiring image information of the meter to be measured through the inspection robot, correcting the image information to obtain correction information, and identifying the reading of the meter to be measured according to the correction information; if yes, controlling the inspection robot to move according to the reflection intensity, acquiring the reflection intensity of the area of the meter to be detected in real time, and executing the judging process. The method can effectively reduce the influence of the reflection on the recognition algorithm, is simple and convenient, greatly reduces the interference of the reflection on the recognition algorithm, and ensures the recognition accuracy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
Fig. 1 is a flowchart of a method for meter light reflection identification based on an inspection robot in an embodiment of the present invention.
Fig. 2 is an application schematic diagram of a meter light reflection identification method based on an inspection robot in an embodiment of the invention.
Detailed Description
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 is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method and a system for identifying meter reflection based on a quadruped robot, which can improve the accuracy of meter identification.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 and fig. 2 are a flowchart and an application schematic diagram of a method for meter light reflection identification based on an inspection robot in an embodiment of the present invention, respectively, and as shown in fig. 1 and fig. 2, the present embodiment provides a method for meter light reflection identification based on an inspection robot, including:
step 100: and determining the area of the meter to be measured to obtain meter area information.
Step 200: and controlling the inspection robot to move to the area of the meter to be measured according to the meter area information.
Step 300: and the inspection robot acquires the reflection intensity of the region of the meter to be measured.
Step 400: and executing a judging process, wherein the judging process is to judge whether the reflection intensity is smaller than a preset threshold value.
Step 401: if not, acquiring the image information of the meter to be measured through the inspection robot, correcting the image information to obtain correction information, and identifying the reading of the meter to be measured according to the correction information.
Step 402: if yes, controlling the inspection robot to move according to the reflection intensity, acquiring the reflection intensity of the area of the meter to be detected in real time, and executing the judging process.
The inspection robot in the embodiment is a quadruped robot; the quadruped robot comprises at least four mechanical legs.
Preferably, the determining the region where the meter to be measured is located to obtain meter region information includes:
acquiring a position image of the meter to be measured; the position image includes a bit region of the meter under test in the image.
And detecting the position image by using a YOLOv5 detection algorithm to obtain the meter region information.
The quadruped robot reaches the target point, the position area of the meter in the image is rapidly determined by using a YOLOv5 detection algorithm, reflected light intensity calculation is carried out in the area, YOLOv5 is an open-source target detection algorithm, the accurate position of the meter in the image can be located, the quadruped robot is accurately moved to the designated position, and the position of the meter is located based on the YOLOv5 algorithm, so that the position of the meter can be rapidly and accurately located. And accurately real-time through-reflection calculation formula
Calculate the regional reflection of light intensity of table meter in the image to according to the regional reflection of light intensity of table meter, control four-footed robot's position, and through PID control adjustment four-footed robot direction around the table meter, avoid the reflection of light to the interference of table meter discernment, improve the discernment precision, simultaneously, through PID control adjustment four-footed robotThe robot makes the four-footed robot can interact well with the user, makes the four-footed robot have good mobility, compares artifical meter, and effectual human resource of having practiced thrift and labour saving and time saving have avoided outdoor meter personnel's danger.
Optionally, in the meter region, when the gray value of a pixel point is greater than the illumination threshold, the pixel point is determined as a reflection point, and the reflection region of the region is detected, and the reflection calculation formula is used
The reflection intensity is calculated, the reflection area of the area can be detected according to the gray scale information of the area of the meter in the image, the identification precision of the meter is greatly influenced due to the reflection problem of the outdoor meter, and the interference of reflection on an identification algorithm can be avoided through the calculation of the reflection intensity.
Preferably, through patrol and examine robot and obtain the regional reflection intensity of the table meter that awaits measuring includes:
acquiring pixel information of an area of the meter to be measured by using the inspection robot; the pixel information includes the number of pixels of the region and the number of pixels of the region.
And calculating the light reflection intensity according to the pixel information.
Preferably, the formula for calculating the reflection intensity includes:
wherein Np _ meter represents the number of the pixel points, Np _ reflect represents the number of the pixel points, and S represents the reflection intensity; the reflection intensity is the proportion of reflection points in the area of the meter to be measured.
Specifically, the reflection intensity of the image is the proportion of the reflection points in the area, and a formula is calculated according to the reflection intensity
The reflection intensity of the meter area is calculated, the interference of reflection on the identification algorithm is reduced, the accuracy of the identification algorithm is improved, the problem of reflection of the meter in outdoor identification is avoided, and the identification accuracy of the meter is ensured.
Optionally, in calculating the formula of the reflected light intensity
Under the setting action of the light reflection intensity, the calculation result of the light reflection intensity is the proportion of the light reflection points in the area, the light reflection points are in the meter area, and when the gray value of the pixel point is greater than the illumination threshold value, the pixel point is determined to be the light reflection point.
Preferably, the correcting the image information to obtain correction information includes:
and collecting a plurality of groups of image key points in the image information according to a G-RMI algorithm.
And calculating an affine matrix of the front view transformation according to the image key points.
And determining the correction information according to the affine matrix.
Preferably, the identifying the reading of the meter under test according to the correction information includes:
identifying the correction information by using a depth regression algorithm, and regressing the direction of the instrument pointer in the correction information in an end-to-end mode;
and determining the reading of the meter to be measured according to the direction of the meter pointer.
Preferably, the identifying the correction information by using a depth regression algorithm, and regressing the meter pointer direction in the correction information in an end-to-end manner includes:
and acquiring the image characteristics of the correction information through a ResNet50 network.
And predicting the pointer direction of the image characteristics according to a direction regression module to obtain the direction of the instrument pointer.
Specifically, the meter image is corrected by using a G-RMI algorithm, the G-RMI algorithm can detect image key points, 3 groups of optimal image key points are selected, an affine matrix transformed with a front view is calculated, and since an imaging visual angle is not a front view visual angle generally when the meter collects images, after the meter is detected from a scene image, reading identification based on a pointer angle can not be directly performed generally, but the meter image needs to be corrected first to convert the meter image into the front view visual angle, so that accurate reading identification is performed, and the corrected image obtained by using the G-RMI algorithm can be used for conveniently identifying the meter number.
Optionally, the identification of the meter count is to identify the meter count by using a depth regression algorithm, obtain image features through ResNet50, predict a pointer direction through a direction regression module, finally calculate the meter reading from the pointer direction, extract the meter features through ResNet50, then regress the direction of the pointer, output a direction vector of the pointer, and finally calculate the meter count according to a linear relationship between the meter count and the pointer direction, thereby avoiding explicit detection of feature points and improving the identification performance.
Further, a depth regression algorithm regresses the direction of the pointer of the instrument in an end-to-end mode to obtain the reading of the instrument, explicit detection of feature points is avoided by using the depth regression algorithm, so that the recognition performance is improved, specifically, image features are obtained by ResNet50, the direction of the pointer is predicted by a direction regression module, and finally the reading of the instrument is calculated by the direction of the pointer.
Preferably, the controlling the inspection robot to move according to the reflection intensity includes:
and controlling the position of the inspection robot to be adjusted by utilizing a PID algorithm according to the reflection intensity.
Furthermore, the calculated reflection intensity is larger than a set threshold value, the quadruped robot can start to adjust the position of the quadruped robot, the quadruped robot rotates around the meter by taking the meter as a rotation center, the reflection intensity is calculated at the same time until the direction meeting the requirement of the reflection intensity is found, the movement direction of the quadruped robot is controlled by the position of the meter in the image, and the calculation is carried out through the reflection intensityFormula (II)
The direction of the quadruped robot is adjusted by calculating the reflection intensity, so that the influence of reflection on a recognition algorithm can be reduced, and the recognition accuracy of the quadruped robot on a meter is improved.
Preferably, the method further comprises the following steps:
controlling the movement speed of the inspection robot according to the reflection intensity; the motion of the inspection robot rotates by taking the meter to be measured as a rotation center.
Further, through calculating the direction of reflection of light intensity adjustment four-footed robot, can reduce the reflection of light to the influence of recognition algorithm, the precision of table meter discernment has been ensured, and to the direction adjustment of four-footed robot, the direction adjustment of four-footed robot adopts PID control, PID control constitutes the control deviation according to given value and actual output value, with the deviation in proportion, integral and differential pass through the linear combination and constitute the controlled quantity, control four-footed robot, it is simple to have the algorithm, the good and high advantage of reliability of robustness, can make four-footed robot reach ideal control effect, realize the accurate discernment of four-footed robot to the table meter.
As an optional implementation mode, the reflecting intensity controls the speed of the robot, the larger the reflecting intensity is, the faster the adjusting speed is, and the reflecting intensity is smaller than a set threshold value, the four-foot robot stops moving, compared with the traditional wheeled or tracked robot, the four-foot robot has better balance, has better passing capacity in rubble, steep slope and indoor complex environment, and can well interact with a user, meanwhile, the bearing capacity is good, the stability is high, the adaptability of the four-foot robot to the ground is strong, the four-foot robot has higher maneuverability in the complex environment, compared with an artificial meter, the manpower resource is effectively saved, the time and the labor are saved, the danger of outdoor meter personnel is avoided, and meanwhile, the reflecting intensity formula is calculated to achieve the purpose that the reflecting intensity is higher
Calculating the reflection intensity and adjusting the position of the quadruped robotAnd the interference of the reflected light on the recognition algorithm is greatly reduced.
As shown in fig. 2, the application process of meter reflection identification based on the inspection robot includes: the quadruped robot arrives at the target point → positions the meters → calculates the reflection intensity → judges the reflection intensity < the threshold value → corrects the meter image → identifies the meter degree.
The working principle of the embodiment is as follows: firstly, the quadruped robot moves to a specified position, a position area of a meter in an image is determined by using a YOLOv5 detection algorithm, and the reflected light intensity is calculated in the area, in the meter area, when the gray value of a pixel point is larger than an illumination threshold value, the pixel point is determined as a reflected light point, the reflected light intensity of the image is the proportion of the reflected light point in the area, and the calculated formula of the reflected light intensity is obtained as
The Np _ meter represents the number of pixel points in the area, the Np _ reflect represents the number of reflective points in the area, the reflective intensity of a meter in the current image can be calculated conveniently and rapidly, and when the reflective intensity calculation formula is used, the reflective intensity is calculated
The calculated reflection intensity is larger than a set threshold value, the quadruped robot can start to adjust the position of the quadruped robot through PID control, the PID control forms a control deviation according to a given value and an actual output value, the deviation forms a control quantity through linear combination according to proportion, integral and differential, the quadruped robot is controlled, and the quadruped robot has the advantages of simple algorithm, good robustness and high reliability, wherein the reflection intensity controls the speed of the robot, the larger the reflection intensity is, the higher the adjustment speed is, the meter is taken as a rotation center, the quadruped robot can rotate around the meter, and meanwhile, the four-legd robot can rotate around the meter according to a reflection intensity calculation formula
Calculating the reflection intensity until finding the direction meeting the requirement of the reflection intensity, stopping the quadruped robot when the calculation result of the reflection intensity is smaller than the set threshold value, and collecting the image of the instrumentSince the imaging viewing angle is not generally the front viewing angle when the instrument is captured, after the instrument is detected from the scene image, reading identification based on pointer angle can not be directly carried out, but the instrument image needs to be corrected firstly to convert the instrument image into a front view angle, reading identification is thus performed, whereby image keypoints are detected by using the G-RMI algorithm, and by selecting the optimal three sets of image keypoints, then an affine matrix transformed with the front view is calculated, so that a corrected meter image is calculated, finally, a depth regression algorithm is utilized to identify the meter degree, then, the ResNet50 is used for extracting instrument and meter characteristics, then the direction of the pointer is regressed, the direction vector of the pointer is output, and finally the meter degree is accurately calculated according to the linear relation between the meter degree and the pointer direction.
The invention has the following beneficial effects:
1. in the invention, the meter position is positioned based on YOLOv5, the meter position can be quickly positioned and accurately positioned, meanwhile, the number of pixel points in the area is represented by an Np _ meter, the number of reflection points in the area is represented by an Np _ reflex, and the formula is used for
Calculating the reflection intensity of the meter in the current image, judging the calculation result and a set threshold value, adjusting the direction of the quadruped robot according to the reflection intensity, and rotating the quadruped robot around the meter through a formula
The reflection intensity is calculated until the direction meeting the requirement of the reflection intensity is found, through the arrangement, the influence of reflection on the recognition algorithm can be effectively reduced, simplicity and convenience are realized, the interference of reflection on the recognition algorithm is greatly reduced, and the recognition accuracy is ensured.
2. The system uses the quadruped robot to replace a human to patrol, the quadruped robot has better balance and better passing capacity in rubble, steep slope and indoor complex environment compared with the traditional wheeled or tracked robot, can well interact with a user, has good bearing capacity, high stability and strong adaptability to the ground, has higher maneuverability in the complex environment compared with an artificial meter, effectively saves human resources, saves time and labor, avoids the danger of personnel in the outdoor meter, controls the adjustment of the accurate position of the quadruped robot by adopting PID (proportion integration differentiation), can form control deviation according to a given value and an actual output value, linearly combines the deviation, the integral and the differential to form a control quantity, controls the quadruped robot, and has simple algorithm, The quadruped robot has the advantages of good robustness and high reliability, can achieve an ideal control effect, and is convenient for the quadruped robot to accurately identify the meter.
3. In the invention, when the instrument acquires an image, the imaging visual angle is not a front-view visual angle generally, so that after the instrument is detected from a scene image, reading identification based on a pointer angle can not be directly performed generally, the instrument image needs to be corrected to be a front-view image so as to perform accurate identification of the reading, 3 groups of optimal key points are selected by detecting the image through a G-RMI algorithm, an affine matrix transformed with a front view is calculated, the corrected instrument image can be efficiently calculated, and the counting degree can be rapidly and conveniently identified.
4. According to the invention, the meter reading is obtained by adopting a depth regression algorithm to identify the meter degree, the direction of the pointer of the meter is regressed in an end-to-end mode, the characteristics of the meter are extracted by utilizing ResNet50, then the direction of the pointer is regressed, the direction vector of the pointer is output, and finally the meter degree is calculated according to the linear relation between the meter degree and the direction of the pointer.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. The utility model provides a method of table meter reflection of light discernment based on robot patrols and examines, its characterized in that includes:
determining the area of the meter to be measured to obtain meter area information;
controlling the inspection robot to move to the area of the meter to be measured according to the meter area information;
acquiring the reflection intensity of the region of the meter to be measured through the inspection robot;
executing a judging process, wherein the judging process is to judge whether the reflection intensity is smaller than a preset threshold value:
if not, acquiring image information of the meter to be measured through the inspection robot, correcting the image information to obtain correction information, and identifying the reading of the meter to be measured according to the correction information;
if yes, controlling the inspection robot to move according to the reflection intensity, acquiring the reflection intensity of the area of the meter to be detected in real time, and executing the judging process.
2. The method for meter light reflection identification based on the inspection robot according to claim 1, wherein the step of determining the area where the meter to be detected is located to obtain meter area information comprises the steps of:
acquiring a position image of the meter to be measured; the position image comprises a bit area of the meter to be measured in the image;
and detecting the position image by using a YOLOv5 detection algorithm to obtain the meter region information.
3. The inspection robot-based meter reflection identification method according to claim 1, wherein the acquiring of the reflection intensity of the area of the meter to be detected by the inspection robot comprises:
acquiring pixel information of an area of the meter to be measured by using the inspection robot; the pixel information comprises the number of pixel points of the region and the number of pixel points of the region;
and calculating the light reflection intensity according to the pixel information.
4. The inspection robot based meter light reflection identification method according to claim 1, wherein the formula for calculating the light reflection intensity comprises:
wherein Np _ meter represents the number of the pixel points, Np _ reflect represents the number of the pixel points, and S represents the reflection intensity; the reflection intensity is the proportion of reflection points in the area of the meter to be measured.
5. The inspection robot based meter light reflection identification method according to claim 1, wherein the correcting the image information to obtain corrected information comprises:
collecting a plurality of groups of image key points in the image information according to a G-RMI algorithm;
calculating an affine matrix of elevation transformation according to the image key points;
and determining the correction information according to the affine matrix.
6. The inspection robot-based meter light reflection identification method according to claim 1, wherein the identifying the reading of the meter to be tested according to the correction information comprises:
identifying the correction information by using a depth regression algorithm, and regressing the direction of the instrument pointer in the correction information in an end-to-end mode;
and determining the reading of the meter to be measured according to the direction of the meter pointer.
7. The inspection robot based meter light reflection identification method according to claim 6, wherein the identifying the correction information by using a depth regression algorithm, and the regressing the meter pointer direction in the correction information in an end-to-end manner comprises:
acquiring image characteristics of the correction information through a ResNet50 network;
and predicting the pointer direction of the image characteristics according to a direction regression module to obtain the direction of the instrument pointer.
8. The inspection robot based meter light reflection identification method according to claim 1, wherein the controlling of the inspection robot according to the light reflection intensity comprises:
and controlling the position of the inspection robot to be adjusted by utilizing a PID algorithm according to the reflection intensity.
9. The inspection robot based meter light reflection identification method according to claim 8, further comprising:
controlling the movement speed of the inspection robot according to the reflection intensity; the motion of the inspection robot rotates by taking the meter to be measured as a rotation center.
10. The inspection robot based meter light reflection identification method according to claim 1, wherein the inspection robot is a four-legged robot; the quadruped robot comprises at least four mechanical legs.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110915177.4A CN113762239B (en) | 2021-08-10 | 2021-08-10 | Meter light reflection identification method based on inspection robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110915177.4A CN113762239B (en) | 2021-08-10 | 2021-08-10 | Meter light reflection identification method based on inspection robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113762239A true CN113762239A (en) | 2021-12-07 |
| CN113762239B CN113762239B (en) | 2024-07-19 |
Family
ID=78789026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110915177.4A Active CN113762239B (en) | 2021-08-10 | 2021-08-10 | Meter light reflection identification method based on inspection robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113762239B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114764821A (en) * | 2022-04-19 | 2022-07-19 | 成都极米科技股份有限公司 | Moving object detection method, moving object detection device, electronic apparatus, and storage medium |
| CN115082922A (en) * | 2022-08-24 | 2022-09-20 | 济南瑞泉电子有限公司 | Water meter digital picture processing method and system based on deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106570436A (en) * | 2016-11-14 | 2017-04-19 | 三峡大学 | Substation equipment meter reading system and meter reading method based on two-dimensional code |
| CN110781731A (en) * | 2019-09-17 | 2020-02-11 | 上海电机学院 | Inspection robot instrument identification method based on specular reflection removal |
| CN112149441A (en) * | 2020-09-04 | 2020-12-29 | 北京布科思科技有限公司 | Two-dimensional code positioning control method based on reflector |
| CN112419185A (en) * | 2020-11-20 | 2021-02-26 | 湖北工业大学 | Accurate high-reflectivity removing method based on light field iteration |
-
2021
- 2021-08-10 CN CN202110915177.4A patent/CN113762239B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106570436A (en) * | 2016-11-14 | 2017-04-19 | 三峡大学 | Substation equipment meter reading system and meter reading method based on two-dimensional code |
| CN110781731A (en) * | 2019-09-17 | 2020-02-11 | 上海电机学院 | Inspection robot instrument identification method based on specular reflection removal |
| CN112149441A (en) * | 2020-09-04 | 2020-12-29 | 北京布科思科技有限公司 | Two-dimensional code positioning control method based on reflector |
| CN112419185A (en) * | 2020-11-20 | 2021-02-26 | 湖北工业大学 | Accurate high-reflectivity removing method based on light field iteration |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114764821A (en) * | 2022-04-19 | 2022-07-19 | 成都极米科技股份有限公司 | Moving object detection method, moving object detection device, electronic apparatus, and storage medium |
| CN115082922A (en) * | 2022-08-24 | 2022-09-20 | 济南瑞泉电子有限公司 | Water meter digital picture processing method and system based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113762239B (en) | 2024-07-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113762239A (en) | Meter reflection identification method based on inspection robot | |
| CN107229930B (en) | Intelligent identification method for numerical value of pointer instrument | |
| CN108111125B (en) | IV characteristic curve scanning and parameter identification system and method for photovoltaic array | |
| CN112766103B (en) | Machine room inspection method and device | |
| WO2021063062A1 (en) | Live-line testing system for power grid apparatus, thermal infrared imager, and method | |
| CN104006757A (en) | Welding deformation real-time detection system and detection method thereof | |
| CN106125744A (en) | The Intelligent Mobile Robot cloud platform control method of view-based access control model servo | |
| CN107393270A (en) | A kind of portable vision inspection device and method for electric meter detection | |
| CN207215806U (en) | A kind of intelligent robot monitored automatically with form and aspect for river water quality | |
| CN109268928A (en) | A kind of heat supply monitoring system | |
| CN102778900A (en) | Counterglow tracking method and tracking system for disc type solar thermal power generation system | |
| CN115018872B (en) | Intelligent control method of dust collection equipment for municipal construction | |
| CN114166147A (en) | Portable detector and method for steel rail profile based on three-dimensional laser scanning technology | |
| CN103323216B (en) | A kind of device and method detecting parallax error of collimator | |
| CN112906602B (en) | Automatic identification device and identification method for electric quantity meter of power distribution cabinet based on image processing | |
| CN108180871B (en) | method for quantitatively evaluating composite insulator surface pulverization roughness | |
| CN106119856A (en) | A kind of Intelligent region cathodic protection monitoring system and control method thereof | |
| CN106197683B (en) | A kind of portable intelligent infrared temperature measurement system | |
| Liu et al. | Research on intelligent detection of concrete aggregate level based on monocular imaging | |
| JP6806325B2 (en) | Photovoltaic system evaluation device, evaluation method, and program for evaluation device | |
| Hu et al. | A detecting algorithm for occlusion on the surface of photovoltaic modules based on improved YOLOv5 | |
| CN205901685U (en) | Solar photovoltaic detects and data system | |
| CN107063403A (en) | Mechanical water meter measuring accuracy automatic detection device and method | |
| CN113074797A (en) | Water level measuring device without water gauge and water level calibration method | |
| CN207540507U (en) | A kind of one camera real-time measurement system of groove type solar paraboloid surface shape |
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 |