CN113029220A - State recognition system and method for industrial instrument panel - Google Patents
State recognition system and method for industrial instrument panel Download PDFInfo
- Publication number
- CN113029220A CN113029220A CN202110120310.7A CN202110120310A CN113029220A CN 113029220 A CN113029220 A CN 113029220A CN 202110120310 A CN202110120310 A CN 202110120310A CN 113029220 A CN113029220 A CN 113029220A
- Authority
- CN
- China
- Prior art keywords
- instrument panel
- industrial
- picture
- industrial instrument
- processing
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 56
- 238000012805 post-processing Methods 0.000 claims abstract description 14
- 238000005520 cutting process Methods 0.000 claims abstract description 8
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 239000011800 void material Substances 0.000 claims description 3
- 238000009776 industrial production Methods 0.000 abstract description 7
- 238000012544 monitoring process Methods 0.000 abstract description 6
- 238000001514 detection method Methods 0.000 abstract description 5
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000012549 training Methods 0.000 description 14
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012935 Averaging Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D13/00—Component parts of indicators for measuring arrangements not specially adapted for a specific variable
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a state recognition system and method for an industrial instrument panel, and belongs to the technical field of industrial equipment. The system comprises an acquisition module, a recognition module and a recognition module, wherein the acquisition module is used for acquiring a first picture to be recognized; the type identification module is used for identifying and obtaining the type of the industrial instrument panel by adopting a first identification model; the characteristic area positioning module is used for identifying and obtaining a dial area of the industrial instrument panel by adopting a second identification model and cutting to obtain a second picture to be identified; the characteristic point positioning module is used for selecting a third identification model according to the type of the industrial instrument panel to identify and obtain the current state of the industrial instrument panel; and the post-processing module is used for processing the current reading of the industrial instrument panel according to the current state of the industrial instrument panel. The beneficial effects of the above technical scheme are: the identification method is simpler, has higher processing speed and detection speed, and can be suitable for reading monitoring of an industrial instrument panel in a complex industrial production environment.
Description
Technical Field
The invention relates to the technical field of industrial equipment, in particular to a state recognition system and method of an industrial instrument panel.
Background
With the further development of economy, no matter in the power industry, the industry or other industries, the operation state of important equipment needs to be monitored in real time to ensure the safe operation of the whole industry, so that the monitoring of the reading of an industrial instrument panel is a very important link in the industrial production process.
In the prior art, the industrial instrument panel is usually processed in a manual reading mode, so that the operation is very troublesome, the risk factors in the industrial production process are many, and production accidents are easy to happen in the manual reading mode. Based on this, some manufacturers try to read the industrial dashboard by means of image recognition, and specifically, usually, a picture of the industrial dashboard is detected by means of line detection, circle detection, template matching, and the like, and then the readings of the dashboard are obtained by processing the pictures. However, the image algorithm has extremely high requirements on the quality of the picture, and the environment in the industrial production process is usually not suitable for shooting the picture, so that the picture suitable for processing cannot be acquired. Moreover, the processing speed of the image algorithm is low, the detection speed is influenced, and the accuracy is low.
Disclosure of Invention
The invention provides a state recognition system and method for an industrial instrument panel, aiming at solving the problems that reading processing of the industrial instrument panel is slow, the speed of real-time monitoring is influenced, the reading accuracy rate is low and the like in the prior art, and specifically comprises the following steps:
a state identification system of an industrial instrument panel specifically comprises:
the acquisition module is used for acquiring a first to-be-identified picture comprising an industrial instrument panel;
the type recognition module is connected with the acquisition module and used for recognizing and obtaining the type of the industrial instrument panel in the first picture to be recognized by adopting a pre-trained first recognition model and outputting the type;
the characteristic area positioning module is connected with the acquisition module and used for identifying and obtaining a dial area of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial area and outputting the second picture to be identified;
the feature point positioning module is respectively connected with the type recognition module and the feature area positioning module, and is preset with a plurality of pre-trained third recognition models, each third recognition model corresponds to the type of one industrial instrument panel, and the feature point positioning module is used for selecting one corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further adopting the third recognition model to recognize the current state of the industrial instrument panel in the second picture to be recognized and output the current state;
and the post-processing module is connected with the characteristic point positioning module and used for processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain and output the current reading of the industrial instrument panel.
Preferably, in the state identification system, the acquisition module specifically includes:
the acquisition unit is used for acquiring and obtaining an original picture comprising the industrial instrument panel;
and the adjusting unit is connected with the acquisition unit and used for performing brightness equalization processing on the original picture so as to obtain and output the first picture to be identified.
Preferably, in the state identification system, the adjusting unit performs averaging processing on the brightness of the original picture by using a histogram equalization method, so as to obtain and output the first to-be-identified picture.
Preferably, in the state recognition system, the first recognition model is implemented by using an xception network, and all convolution layers in the xception network are replaced by void convolution layers.
Preferably, in the state recognition system, the second recognition model is implemented by using an SSD network, and the feature extraction network in the second recognition model is implemented by using a reduced version of mobilenetv2 network.
Preferably, in the state recognition system, the feature point locating module recognizes and obtains a plurality of feature points of the dial area, and outputs the plurality of feature points as the current state of the industrial instrument panel;
and the post-processing module processes the characteristic points to obtain the current reading of the industrial instrument panel and outputs the reading.
Preferably, the state recognition system, wherein the plurality of feature points include a start position of a meter pointer within the industrial instrument panel, an end position of the meter pointer, a center position of the industrial instrument panel, and a start position of a meter scale within the industrial instrument panel;
the post-processing module specifically includes:
the first processing unit is used for processing to obtain a first straight line according to the coordinates of the starting position and the end position of the instrument pointer;
the second processing unit is used for processing to obtain a second straight line according to the central position of the industrial instrument panel and the initial position of the instrument scale;
the third processing unit is respectively connected with the first processing unit and the second processing unit and used for calculating to obtain an included angle between the first straight line and the second straight line;
and the conversion unit is connected with the third processing unit and used for processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading and outputting the current reading.
A state recognition method of an industrial instrument panel is applied to the state recognition system of the industrial instrument panel and comprises the following steps:
step S1, acquiring a first to-be-identified picture comprising an industrial instrument panel;
step S2, identifying and obtaining the type of the industrial instrument panel in the first picture to be identified by adopting a pre-trained first identification model and outputting the type;
step S3, identifying and obtaining a dial area of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial area, and outputting the second picture to be identified;
step S4, selecting a corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further recognizing by using the third recognition model to obtain and output the current state of the industrial instrument panel in the second picture to be recognized;
and step S5, processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain the current reading of the industrial instrument panel and outputting the current reading.
Preferably, in the state identification method, the step S1 specifically includes:
step S11, acquiring an original picture comprising the industrial instrument panel;
and step 12, performing brightness equalization processing on the original picture to obtain and output the first picture to be recognized.
Preferably, in the state recognition method, the current state of the industrial dashboard output in step S4 includes a start position of a dashboard pointer in the industrial dashboard, an end position of the dashboard pointer, a center position of the industrial dashboard, and a start position of a dashboard scale in the industrial dashboard;
the step S5 specifically includes:
step S51, processing to obtain a first straight line according to the coordinates of the starting position and the end position of the instrument pointer, and processing to obtain a second straight line according to the central position of the industrial instrument panel and the starting position of the instrument scale;
step S52, calculating an included angle between the first straight line and the second straight line;
and step S53, processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading, and outputting the current reading.
The invention has the beneficial effects that:
the method and the device completely abandon the traditional algorithm with extremely high requirements on the picture quality, have simpler identification method, have higher processing speed and higher detection speed compared with the traditional method, and can be suitable for reading monitoring of the industrial instrument panel in the complex industrial production environment.
Drawings
FIG. 1 is a schematic diagram of a general configuration of a status recognition system for an industrial dashboard in accordance with a preferred embodiment of the present invention;
FIG. 2 is a schematic diagram of a specific structure of an acquisition module according to a preferred embodiment of the present invention;
FIG. 3 is a diagram illustrating a label file of training data of a first recognition model according to a preferred embodiment of the present invention;
FIG. 4 is a schematic diagram of a post-processing module according to a preferred embodiment of the present invention;
FIG. 5 is a schematic flow chart of a method for identifying the status of an industrial dashboard according to a preferred embodiment of the present invention;
FIG. 6 is a flowchart illustrating the detailed process of step S1 according to the preferred embodiment of the present invention;
fig. 7 is a flowchart illustrating the step S5 in accordance with the preferred embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described in detail 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 real-time embodiments of the invention, and not all embodiments. 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.
The invention is further described with reference to the following figures and specific examples.
The invention provides a state recognition system of an industrial instrument panel, which is specially used for reading monitoring of the industrial instrument panel in an industrial production process, such as industrial thermometers, pressure gauges, flow meters and the like. The state recognition system is specifically shown in fig. 1, and includes:
the acquisition module 1 is used for acquiring a first to-be-identified picture comprising an industrial instrument panel;
the type recognition module 2 is connected with the acquisition module 1 and is used for recognizing and obtaining the type of the industrial instrument panel in the first picture to be recognized by adopting a pre-trained first recognition model and outputting the type;
the characteristic region positioning module 3 is connected with the acquisition module 1 and is used for identifying and obtaining a dial region of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial region and outputting the second picture to be identified;
the feature point positioning module 4 is respectively connected with the type recognition module 2 and the feature area positioning module 3, and is preset with a plurality of pre-trained third recognition models, each third recognition model corresponds to the type of an industrial instrument panel, and the feature point positioning module is used for selecting one corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further adopting the third recognition model to recognize to obtain and output the current state of the industrial instrument panel in the second picture to be recognized;
and the post-processing module 5 is connected with the characteristic point positioning module 4 and used for processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain and output the current reading of the industrial instrument panel.
Specifically, in view of the complexity of the image processing algorithm for reading the industrial dashboard in the prior art, in the present embodiment, the acquired image including the industrial dashboard is first processed once to separate the image portion including the industrial dashboard from the whole image, and at the same time, the type of the industrial dashboard included in the image is identified.
In this embodiment, before the above process, specific recognition models for different types of industrial instrument panels are obtained by training in advance, and the reason for the respective training is that the different types of instrument panels may be different in terms of dials, scales, and instrument pointers, so that the settings of the key feature points may also be different.
After the image including the industrial instrument panel part separated from the whole image is obtained, a recognition model formed by pre-training is selected according to the recognized type of the industrial instrument panel, and the dial plate part of the industrial instrument panel is further recognized to recognize the current state of the instrument panel. The state of the dashboard is in fact the reading of the pointer of the dashboard, the identification of which is described in detail below.
And finally, the reading of the industrial instrument panel is processed and output through the post-processing module, so that the real-time monitoring of the industrial instrument panel is realized.
In a preferred embodiment of the present invention, as shown in fig. 2, the acquisition module 1 specifically includes:
the acquisition unit 11 is used for acquiring an original picture comprising an industrial instrument panel;
and the adjusting unit 12 is connected with the acquisition unit 11 and is used for performing brightness equalization processing on the original picture so as to obtain and output a first picture to be identified.
Further, the adjusting unit 12 performs an averaging process on the brightness of the original picture by using a histogram equalization method, so as to obtain and output the first to-be-identified picture.
In particular, in this embodiment, because the lighting conditions at the industrial production site are not suitable for taking a picture, the picture judged under such conditions may have uneven brightness, so that the object in the picture cannot be effectively identified. Therefore, after the original picture is obtained by shooting, the brightness of the original picture needs to be adjusted firstly, so that the original picture can clearly show the industrial instrument panel in the picture, and the subsequent identification is convenient. The adjusting method may be averaging the brightness of the original picture by using a histogram equalization method, so that the brightness of the picture is adjusted to a uniform numerical level, thereby forming a first to-be-identified picture for subsequent identification.
In a preferred embodiment of the present invention, the first recognition model is implemented by using an xception network, and all convolutional layers (conv) in the xception network are replaced by void convolutional layers (scaled conv). Specifically, the xception classification module is realized by adopting a hole convolution layer, the hole convolution layer has the advantage that the receptive field can be greatly improved, and compared with a common convolution layer, most values in a convolution kernel of the hole convolution layer are 0, so that the whole network is sparse, and the identification accuracy can be improved.
In this embodiment, the training data of the first recognition model is prepared as follows:
firstly, the pictures of various different types of instrument panels are acquired in advance, the pictures of the different types of instrument panels are placed in different folders, and each folder only stores the pictures of one type of instrument panel. After storing the pictures of the instrument panel, generating a label file by using the script. As shown in FIG. 3, the sequence of each label file has two columns, the first column for indicating the instrument type and the second column for indicating the training class number. The prepared label file is then added to the training data to train the first recognition model. And finally, taking the trained model as a first recognition model for use.
In a preferred embodiment of the present invention, the second recognition model is implemented by using an SSD (Single Shot multi box Detector) network, and the feature extraction network in the second recognition model is implemented by using a reduced version of mobilenetv2 network, so that the model compression is small, and because the task of positioning the dial region is relatively simple, it is not necessary to add a customized network such as fpn module.
In a preferred embodiment of the present invention, the characteristic point positioning module 4 identifies a plurality of characteristic points in the dial area, and outputs the plurality of characteristic points as the current state of the industrial instrument panel;
the post-processing module 5 processes the current reading of the industrial instrument panel according to the plurality of feature points and outputs the reading.
Further, the plurality of feature points comprise a starting position of a meter pointer in the industrial instrument panel, an end position of the meter pointer, a central position of the industrial instrument panel and a starting position of a meter scale in the industrial instrument panel;
the post-processing module 5 is specifically shown in fig. 4, and includes:
the first processing unit 51 is configured to process the coordinates of the start position and the end position of the meter pointer to obtain a first straight line;
the second processing unit 52 is used for processing the central position of the industrial instrument panel and the initial position of the instrument scale to obtain a second straight line;
the third processing unit 53 is respectively connected to the first processing unit 51 and the second processing unit 52, and is used for calculating an included angle between the first straight line and the second straight line;
and the conversion unit 54 is connected with the third processing unit 53, and is used for processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading and outputting the current reading.
Specifically, in this embodiment, since the feature points of the surface of different instrument panels are different, the formed training data are also different, and therefore, a classification step of the industrial instrument panel (i.e., the function of the type identification module 2) needs to be performed in advance. After the types of the industrial instrument panels are identified, the corresponding third identification model is selected according to the types, and further the instrument pointer in the dial area is identified.
In this embodiment, when training data is prepared, feature points of a dial area in a drawing are respectively marked for different types of industrial instrument panels, specifically, a start position and an end position of an instrument pointer of the dial area, a center position of the dial area, and a start position where scales start in the dial area are marked. Since the center position of the dial usually coincides with the starting position of the meter hand in the dial area, only three feature points need to be marked in the actual marking.
In this embodiment, the third recognition model extracts modules s1, s2, and s3 from layers 5, 8, and 13 of the Shuffle Net based on the Shuffle Net, superimposes the outputs of the extracted modules together using embedding, and finally accesses a convolution layer of 1x1, where the output channel is 6 channels, and the values represented by each channel are x1, y1, x2, y2, x3, and y 3.
In this embodiment, the learning rate is reduced in stages during the training process, each stage being reduced by 0.1, the stages being [10,50,100,120,130 ]]. The initial learning rate is 0.1, and the final learning rate is 10-6。
Aiming at the training processes of the first recognition model, the second recognition model and the third recognition model, image enhancement, brightness/darkness/saturation adjustment, random cutting/stretching/rotation, affine transformation, perspective transformation and other adjustment steps need to be embedded into the training process, so that training data are enriched, and the recognition accuracy of the recognition model obtained through final training can be improved.
Further, in this embodiment, the feature point positioning module 4 finally identifies and outputs coordinates of a plurality of feature points in the instrument panel area through the third identification model, and further outputs coordinates of four feature points, i.e., a start position of the instrument pointer, an end position of the instrument pointer, a center position of the industrial instrument panel, and a start position of the instrument scale in the industrial instrument panel.
The post-processing module 5 then processes the current reading of the instrument panel according to the four characteristic points. Specifically, the post-processing module respectively processes a first straight line (used for representing the current position of the instrument pointer) between the starting position and the end position of the instrument pointer and a second straight line (used for representing the reference position of the zero point of the scale of the instrument panel) between the central position of the instrument panel and the starting position of the scale of the instrument according to the four characteristic points, and calculates an included angle between the first straight line and the second straight line. After the included angle is obtained through calculation, the current reading of the instrument panel is obtained through the following formula:
Result＝currentAngle*perValue；
wherein Result is used to represent the current reading of the industrial dashboard;
currentAngle is used for representing the angle of a current included angle between a first straight line and a second straight line in the industrial instrument panel;
perValue is used for representing a scale value corresponding to a unit angle on the industrial instrument panel obtained through a priori knowledge.
In a preferred embodiment of the present invention, based on the above-mentioned state recognition system for an industrial dashboard, a state recognition method for an industrial dashboard is now provided, which is specifically shown in fig. 5, and includes:
step S1, acquiring a first to-be-identified picture comprising an industrial instrument panel;
step S2, identifying and obtaining the type of the industrial instrument panel in the first picture to be identified by adopting a first identification model trained in advance and outputting the type;
step S3, identifying and obtaining the dial area of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial area, and outputting the second picture to be identified;
step S4, selecting a corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further recognizing by adopting the third recognition model to obtain and output the current state of the industrial instrument panel in the second picture to be recognized;
and step S5, processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain the current reading of the industrial instrument panel and outputting the reading.
Further, in a preferred embodiment of the present invention, as shown in fig. 6, step S1 specifically includes:
step S11, acquiring an original picture comprising an industrial instrument panel;
and step 12, performing brightness equalization processing on the original picture to obtain and output a first picture to be recognized.
Further, in a preferred embodiment of the present invention, the current state of the industrial dashboard output in step S4 includes a starting position of a dashboard pointer in the industrial dashboard, an ending position of the dashboard pointer, a central position of the industrial dashboard, and a starting position of a dashboard scale in the industrial dashboard;
as shown in fig. 7, step S5 specifically includes:
step S51, processing to obtain a first straight line according to the coordinates of the starting position and the end position of the instrument pointer, and processing to obtain a second straight line according to the central position of the industrial instrument panel and the starting position of the instrument scale;
step S52, calculating an included angle between the first straight line and the second straight line;
and step S53, processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading, and outputting the current reading.
Claims (10)
1. A state identification system of an industrial instrument panel is characterized by specifically comprising:
the acquisition module is used for acquiring a first to-be-identified picture comprising an industrial instrument panel;
the type recognition module is connected with the acquisition module and used for recognizing and obtaining the type of the industrial instrument panel in the first picture to be recognized by adopting a pre-trained first recognition model and outputting the type;
the characteristic area positioning module is connected with the acquisition module and used for identifying and obtaining a dial area of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial area and outputting the second picture to be identified;
the feature point positioning module is respectively connected with the type recognition module and the feature area positioning module, and is preset with a plurality of pre-trained third recognition models, each third recognition model corresponds to the type of one industrial instrument panel, and the feature point positioning module is used for selecting one corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further adopting the third recognition model to recognize the current state of the industrial instrument panel in the second picture to be recognized and output the current state;
and the post-processing module is connected with the characteristic point positioning module and used for processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain and output the current reading of the industrial instrument panel.
2. The state recognition system of claim 1, wherein the acquisition module specifically comprises:
the acquisition unit is used for acquiring and obtaining an original picture comprising the industrial instrument panel;
and the adjusting unit is connected with the acquisition unit and used for performing brightness equalization processing on the original picture so as to obtain and output the first picture to be identified.
3. The state recognition system according to claim 2, wherein the adjusting unit averages the brightness of the original picture by histogram equalization, so as to obtain and output the first picture to be recognized.
4. The state recognition system of claim 1, wherein the first recognition model is implemented using an xception network and replaces all convolutional layers in the xception network with void convolutional layers.
5. The status recognition system according to claim 1, wherein the second recognition model is implemented by an SSD network, and the feature extraction network in the second recognition model is implemented by a reduced version of a mobilenetv2 network.
6. The status recognition system of claim 1, wherein the feature point location module identifies a plurality of feature points of the dial area and outputs a plurality of the feature points as the current status of the industrial dashboard;
and the post-processing module processes the characteristic points to obtain the current reading of the industrial instrument panel and outputs the reading.
7. The status recognition system of claim 6, wherein the plurality of feature points comprises a starting position of a meter pointer within the industrial dashboard, an ending position of the meter pointer, a center position of the industrial dashboard, and a starting position of a meter scale within the industrial dashboard;
the post-processing module specifically includes:
the first processing unit is used for processing to obtain a first straight line according to the coordinates of the starting position and the end position of the instrument pointer;
the second processing unit is used for processing to obtain a second straight line according to the central position of the industrial instrument panel and the initial position of the instrument scale;
the third processing unit is respectively connected with the first processing unit and the second processing unit and used for calculating to obtain an included angle between the first straight line and the second straight line;
and the conversion unit is connected with the third processing unit and used for processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading and outputting the current reading.
8. A state recognition method of an industrial dashboard, applied to the state recognition system of the industrial dashboard according to any one of claims 1 to 7, and comprising:
step S1, acquiring a first to-be-identified picture comprising an industrial instrument panel;
step S2, identifying and obtaining the type of the industrial instrument panel in the first picture to be identified by adopting a pre-trained first identification model and outputting the type;
step S3, identifying and obtaining a dial area of the industrial instrument panel in the first picture to be identified by adopting a pre-trained second identification model, cutting the first picture to be identified to obtain a second picture to be identified only comprising the dial area, and outputting the second picture to be identified;
step S4, selecting a corresponding third recognition model according to the type of the industrial instrument panel output by the type recognition module, and further recognizing by using the third recognition model to obtain and output the current state of the industrial instrument panel in the second picture to be recognized;
and step S5, processing the current state of the industrial instrument panel output by the characteristic point positioning module to obtain the current reading of the industrial instrument panel and outputting the current reading.
9. The state identification method according to claim 8, wherein the step S1 specifically includes:
step S11, acquiring an original picture comprising the industrial instrument panel;
and step 12, performing brightness equalization processing on the original picture to obtain and output the first picture to be recognized.
10. The status recognition method according to claim 8, wherein the current status of the industrial dashboard output in the step S4 includes a start position of a dashboard pointer within the industrial dashboard, an end position of the dashboard pointer, a center position of the industrial dashboard, and a start position of a dashboard scale within the industrial dashboard;
the step S5 specifically includes:
step S51, processing to obtain a first straight line according to the coordinates of the starting position and the end position of the instrument pointer, and processing to obtain a second straight line according to the central position of the industrial instrument panel and the starting position of the instrument scale;
step S52, calculating an included angle between the first straight line and the second straight line;
and step S53, processing the current reading of the industrial instrument panel according to the included angle and the preset unit reading, and outputting the current reading.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110120310.7A CN113029220A (en) | 2021-01-28 | 2021-01-28 | State recognition system and method for industrial instrument panel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110120310.7A CN113029220A (en) | 2021-01-28 | 2021-01-28 | State recognition system and method for industrial instrument panel |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113029220A true CN113029220A (en) | 2021-06-25 |
Family
ID=76459418
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110120310.7A Pending CN113029220A (en) | 2021-01-28 | 2021-01-28 | State recognition system and method for industrial instrument panel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113029220A (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002188939A (en) * | 2000-12-21 | 2002-07-05 | Nec Corp | Pointer-type meter measuring system by image recognition and pointer-type meter measuring method |
| CN101498592A (en) * | 2009-02-26 | 2009-08-05 | 北京中星微电子有限公司 | Reading method and apparatus for pointer instrument |
| CN103034838A (en) * | 2012-12-03 | 2013-04-10 | 中国人民解放军63963部队 | Special vehicle instrument type identification and calibration method based on image characteristics |
| CN103364023A (en) * | 2013-06-21 | 2013-10-23 | 广东电网公司电力科学研究院 | Total least square-based extraction method for central point of instrument image |
| CN104913797A (en) * | 2015-05-29 | 2015-09-16 | 广州供电局有限公司 | Pointer type instrument number reading and recognition method and system |
| CN105678344A (en) * | 2016-02-29 | 2016-06-15 | 浙江群力电气有限公司 | Intelligent classification method for power instrument equipment |
| CN108549840A (en) * | 2018-03-19 | 2018-09-18 | 广东核电合营有限公司 | A kind of method and device that analog instrument simultaneously and rapidly identifies |
| CN109948469A (en) * | 2019-03-01 | 2019-06-28 | 吉林大学 | The automatic detection recognition method of crusing robot instrument based on deep learning |
| CN110378225A (en) * | 2019-06-14 | 2019-10-25 | 南京泛函智能技术研究院有限公司 | The recognition methods of pointer instrument automatic reading and identification device |
| CN110543878A (en) * | 2019-08-07 | 2019-12-06 | 华南理工大学 | pointer instrument reading identification method based on neural network |
| CN110659636A (en) * | 2019-09-20 | 2020-01-07 | 随锐科技集团股份有限公司 | Pointer instrument reading identification method based on deep learning |
| CN110929710A (en) * | 2019-10-29 | 2020-03-27 | 湖南大学 | Method and system for automatically identifying meter pointer reading based on vision |
| CN111291691A (en) * | 2020-02-17 | 2020-06-16 | 合肥工业大学 | Deep learning-based substation secondary equipment instrument panel reading detection method |
| CN111950330A (en) * | 2019-05-16 | 2020-11-17 | 杭州测质成科技有限公司 | Pointer instrument indicating number detection method based on target detection |
-
2021
- 2021-01-28 CN CN202110120310.7A patent/CN113029220A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002188939A (en) * | 2000-12-21 | 2002-07-05 | Nec Corp | Pointer-type meter measuring system by image recognition and pointer-type meter measuring method |
| CN101498592A (en) * | 2009-02-26 | 2009-08-05 | 北京中星微电子有限公司 | Reading method and apparatus for pointer instrument |
| CN103034838A (en) * | 2012-12-03 | 2013-04-10 | 中国人民解放军63963部队 | Special vehicle instrument type identification and calibration method based on image characteristics |
| CN103364023A (en) * | 2013-06-21 | 2013-10-23 | 广东电网公司电力科学研究院 | Total least square-based extraction method for central point of instrument image |
| CN104913797A (en) * | 2015-05-29 | 2015-09-16 | 广州供电局有限公司 | Pointer type instrument number reading and recognition method and system |
| CN105678344A (en) * | 2016-02-29 | 2016-06-15 | 浙江群力电气有限公司 | Intelligent classification method for power instrument equipment |
| CN108549840A (en) * | 2018-03-19 | 2018-09-18 | 广东核电合营有限公司 | A kind of method and device that analog instrument simultaneously and rapidly identifies |
| CN109948469A (en) * | 2019-03-01 | 2019-06-28 | 吉林大学 | The automatic detection recognition method of crusing robot instrument based on deep learning |
| CN111950330A (en) * | 2019-05-16 | 2020-11-17 | 杭州测质成科技有限公司 | Pointer instrument indicating number detection method based on target detection |
| CN110378225A (en) * | 2019-06-14 | 2019-10-25 | 南京泛函智能技术研究院有限公司 | The recognition methods of pointer instrument automatic reading and identification device |
| CN110543878A (en) * | 2019-08-07 | 2019-12-06 | 华南理工大学 | pointer instrument reading identification method based on neural network |
| CN110659636A (en) * | 2019-09-20 | 2020-01-07 | 随锐科技集团股份有限公司 | Pointer instrument reading identification method based on deep learning |
| CN110929710A (en) * | 2019-10-29 | 2020-03-27 | 湖南大学 | Method and system for automatically identifying meter pointer reading based on vision |
| CN111291691A (en) * | 2020-02-17 | 2020-06-16 | 合肥工业大学 | Deep learning-based substation secondary equipment instrument panel reading detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109948469B (en) | Automatic inspection robot instrument detection and identification method based on deep learning | |
| CN112949564B (en) | Pointer type instrument automatic reading method based on deep learning | |
| CN103955694B (en) | Image recognition meter reading system and method | |
| CN110659636B (en) | Pointer instrument reading identification method based on deep learning | |
| CN106599830B (en) | Face key point positioning method and device | |
| CN112818988B (en) | Automatic identification reading method and system for pointer instrument | |
| CN105975979B (en) | A kind of instrument detecting method based on machine vision | |
| CN110543878A (en) | pointer instrument reading identification method based on neural network | |
| CN108182433A (en) | A kind of meter reading recognition methods and system | |
| CN109508709B (en) | Single pointer instrument reading method based on machine vision | |
| CN105894002A (en) | Instrument reading identification method based on machine vision | |
| CN106326887A (en) | Method and device for checking optical character recognition result | |
| CN108709500B (en) | Circuit board element positioning and matching method | |
| CN111353502A (en) | Digital table identification method and device and electronic equipment | |
| CN112381177A (en) | Dial digital character recognition method and system based on deep learning | |
| CN109034151A (en) | A kind of localization method for the identification of multiple pointer instruments | |
| CN111062448B (en) | Equipment type recognition model training method, equipment type recognition method and device | |
| CN113029220A (en) | State recognition system and method for industrial instrument panel | |
| CN108897781B (en) | Duplicate checking system for paper graphs | |
| CN116664817A (en) | Power device state change detection method based on image difference | |
| CN115564734A (en) | Product detection method, electronic equipment and storage medium | |
| CN113111849B (en) | Human body key point detection method, device, system and computer readable medium | |
| CN115393838A (en) | Pointer instrument reading identification method and device, electronic equipment and storage medium | |
| CN108898632B (en) | Method and device for determining rotation angle of instrument pointer | |
| CN114646563B (en) | Method for detecting surface abrasion resistance of polyester film with metal coating |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210625 |
|
| RJ01 | Rejection of invention patent application after publication |