An AI vision-based industrial instrument reading automatic identification system

Abstract

The application discloses an industrial instrument reading automatic identification system based on AI vision, relates to the technical field of automatic identification of instrument reading, and comprises an industrial instrument visual monitoring module, an industrial instrument visual analysis module, an industrial instrument reading analysis module, an early warning terminal and a database. The visual conditions corresponding to the monitoring of the industrial instrument are analyzed, the monitoring corresponding to the industrial instrument is then performed, the operation state of the industrial operation corresponding to the industrial instrument is identified, the abnormal body of the industrial operation abnormality corresponding to the industrial instrument is analyzed, and finally the maintenance measures of the abnormal body of the industrial operation abnormality corresponding to the industrial instrument are analyzed, so that omnibearing and efficient visual monitoring of industrial production is realized, an AI vision technology based on intelligence is used, full-coverage industrial instrument views are presented, efficient reading evaluation is carried out, accurate evaluation results are obtained, targeted maintenance measures are taken, and the equipment failure rate and safety risk during industrial operation are reduced.

Description

Technical Field

[0001] This invention relates to the field of automatic instrument reading recognition technology, and specifically to an automatic industrial instrument reading recognition system based on AI vision. Background Technology

[0002] Industrial instruments are crucial for monitoring equipment operating status and controlling process parameters in industrial production. Therefore, this paper proposes an automatic identification system for industrial instrument readings based on AI vision. Through AI vision technology, this system enables efficient monitoring and real-time understanding of industrial production, thereby ensuring safe and stable operation, reducing equipment failure rates, and minimizing safety issues.

[0003] Current technology suffers from the following problems: Current industrial instrument readings are mainly based on manual inspection. This manual inspection method is inefficient and difficult to operate. In industrial production, equipment is scattered in various locations, and it often requires a great deal of effort for manual inspection to achieve full coverage. As a result, abnormalities in equipment cannot be identified in a timely manner, which increases the failure rate of equipment, reduces its service life, increases safety risks in industrial production, and also affects the products produced, reducing product quality and efficiency. Furthermore, the lack of timely response and identification of industrial faults increases the risk to inspection personnel. Summary of the Invention

[0004] In view of the above-mentioned technical shortcomings, the purpose of this invention is to provide an automatic identification system for industrial instrument readings based on AI vision.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides an automatic identification system for industrial instrument readings based on AI vision, including: an industrial instrument vision monitoring module, used to acquire basic information of industrial instruments corresponding to the industrial site, then analyze the set of visual conditions for monitoring corresponding to the industrial instruments, and collect images of industrial instruments, acquire image data, and analyze whether the visual monitoring corresponding to the industrial instruments meets the standards.

[0006] The industrial instrument vision analysis module is used to analyze the secondary control of the vision of the industrial instrument when the vision monitoring of the industrial instrument is not up to standard, obtain the visual image of the industrial instrument, and identify the operating status of the industrial operation corresponding to the industrial instrument.

[0007] The industrial instrument reading analysis module is used to analyze the anomalies of industrial instruments corresponding to industrial operation anomalies, and to analyze the maintenance measures for the industrial instruments corresponding to industrial operation anomalies.

[0008] The early warning terminal is used to issue early warnings when the visual monitoring corresponding to the industrial instrument fails to meet the standards or when the operation of the industrial instrument corresponding to the industrial operation is abnormal.

[0009] The beneficial effects of this invention are as follows: 1. This invention provides an automatic identification system for industrial instrument readings based on AI vision. By analyzing the visual conditions corresponding to the monitoring of industrial instruments, the system performs the corresponding monitoring of industrial instruments, identifies the operational status of the industrial operation corresponding to the industrial instruments, analyzes the abnormal entities of the industrial operation corresponding to the industrial instruments, and finally analyzes the maintenance measures for the abnormal entities of the industrial operation corresponding to the industrial instruments. This achieves comprehensive and efficient visual monitoring of industrial production. Based on intelligent AI vision technology, a full-coverage industrial instrument scene display is obtained, thereby enabling efficient reading evaluation and obtaining timely and accurate evaluation results. This allows for targeted maintenance measures to be taken, reducing equipment failure rates and safety risks during industrial operations, and intelligently ensuring the quality of industrial operations.

[0010] 2. Obtain basic information of industrial instruments corresponding to the industrial site, analyze the visual condition set of industrial instruments, collect industrial instrument images, and analyze whether the visual monitoring of industrial instruments meets the standards. This will achieve full coverage monitoring of industrial visual instruments, provide comprehensive monitoring and protection for industrial production, and facilitate data support for subsequent automatic identification of industrial instrument readings.

[0011] 3. When visual monitoring fails to meet standards, analyze the secondary control of the corresponding vision of the industrial instruments and identify the operating status of the corresponding industrial operation. This enables timely control of the corresponding visual detection in the industry, facilitating the acquisition of more accurate and effective industrial instrument images, thereby ensuring the correctness of the automatic recognition results of industrial instrument readings.

[0012] 4. Based on the analysis of the abnormal entities corresponding to industrial operation anomalies of industrial instruments, and the analysis of the maintenance measures for the abnormal entities corresponding to industrial operation anomalies of industrial instruments, in order to achieve timely understanding and control of industrial production problems, further curb the deterioration of fault problems, reduce the safety risks of industrial production, and ensure the safety and stability of industrial production. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Please see Figure 1 As shown, an AI vision-based automatic identification system for industrial instrument readings includes an industrial instrument visual monitoring module, an industrial instrument visual analysis module, an industrial instrument reading analysis module, an early warning terminal, and a database.

[0017] The industrial instrument visual monitoring module is connected to the industrial instrument visual analysis module and the early warning terminal, respectively. The industrial instrument visual analysis module is connected to the industrial instrument reading analysis module and the early warning terminal, respectively. The database is connected to the industrial instrument visual monitoring module, the industrial instrument visual analysis module and the industrial instrument reading analysis module, respectively.

[0018] The industrial instrument visual monitoring module is used to acquire basic information of industrial instruments in the industrial field, analyze the set of visual conditions for monitoring industrial instruments, collect images of industrial instruments, obtain image data, and analyze whether the visual monitoring of industrial instruments meets the standards.

[0019] The system acquires basic information about industrial instruments in the industrial field, analyzes the visual conditions of the industrial instruments, collects images of the industrial instruments, and analyzes whether the visual monitoring of the industrial instruments meets the standards. This achieves full coverage monitoring of industrial visual instruments, providing comprehensive monitoring and protection for industrial production and facilitating data support for the automatic identification of subsequent industrial instrument readings.

[0020] As an optional implementation, the analysis of the set of visual conditions corresponding to the industrial instruments is carried out in the following specific process: the distribution location and spatial location of the industrial instruments in the industrial site are obtained, and the distribution location and spatial location of the industrial instruments are compared with the location data range of each placement measure in the visual library stored in the database. If the distribution location and spatial location of the industrial instruments are included in the location data range of a certain placement measure in the visual library stored in the database, then the placement measure is taken as the visual condition of the industrial instruments in the industrial site, including the placement angle, placement location and acquisition frequency.

[0021] The set of visual conditions for monitoring corresponding to industrial instruments is constructed by the placement angle, placement position, and acquisition frequency.

[0022] It should be noted that, based on the acquisition of historical visual monitoring images of industrial instruments corresponding to various industrial sites, and the extraction of each historical visual monitoring image that meets the standards, the placement conditions corresponding to each historical visual monitoring image during monitoring are obtained. A visual library is constructed based on the historical visual monitoring images of industrial instruments that meet the standards and the placement conditions. Basic images of industrial instruments corresponding to industrial sites are captured by industrial cameras, and the distribution location and spatial location occupied by the instruments are extracted from the basic images. The distribution location and spatial location occupied by the instruments are the basic information.

[0023] As an optional implementation, the specific analysis process for acquiring industrial instrument images is as follows: Industrial instruments in the industrial field are monitored according to a set of visual conditions, and industrial instrument images are acquired. The acquired industrial instrument images undergo image preprocessing, including denoising, illumination correction, and angle correction. Noise is denoised using Gaussian filtering and Messian filtering, and illumination correction is performed based on a histogram equalization algorithm. Then, the tilt angle of the industrial instrument image is calculated using the circular contour of the instrument panel detected by Hough transform, and angle correction is performed to obtain a standardized industrial instrument image.

[0024] It should be noted that the industrial instrument image is denoised by replacing the center pixel value with the median of the neighboring pixels corresponding to the Gaussian kernel in the Gaussian kernel convolution kernel. The image's grayscale level is dynamically expanded using a histogram equalization algorithm, and the grayscale values ​​of the image pixels are redistributed to complete the illumination correction. The core of the Hough circle transform is to convert the circle of the image into a three-dimensional accumulator in the parameter space. By extracting the edge contour of the denoised and illumination-corrected image, a circle is obtained. Several feature points are selected to establish a local coordinate system, and ellipse fitting is performed on the contour points in the local coordinate system to obtain the major and minor axes of the ellipse. The major axis of the ellipse corresponds to the actual horizontal direction of the dial, and the angle between the major axis and the X-axis of the image coordinate system is the tilt angle, the formula of which is... , , and These are the coefficients corresponding to the ellipse fitting, and their formula is: Gaussian filtering, Messian filtering, histogram equalization algorithm, and Hough transform are all existing technologies for obtaining the tilt angle of the circular contour of the instrument panel, and will not be elaborated on further here.

[0025] As an alternative implementation, the image data includes a visual sharpness parameter and a visual pointer length.

[0026] It should be noted that visual clarity parameters include image grayscale and high-frequency energy ratio, etc.; image grayscale is extracted from industrial instrument images based on AI vision; fast Fourier transform is performed on industrial instrument images to obtain frequency domain spectrum, and high-frequency energy and total full-frequency energy are obtained. Substitute them into the high-frequency energy ratio formula, divide the high-frequency energy by the total full-frequency energy and then multiply by the percentage to obtain the high-frequency energy ratio.

[0027] As an optional implementation, the analysis process for determining whether the visual monitoring corresponding to the industrial instrument meets the standards is as follows: The visual clarity parameter and visual pointer length corresponding to the industrial instrument image are compared with preset reference visual clarity parameters and reference visual pointer lengths, respectively. If the visual clarity parameter corresponding to the industrial instrument image is less than the preset reference visual clarity parameter, or the visual pointer length is less than the reference visual pointer length, then the visual monitoring corresponding to the industrial instrument is determined to be non-compliant. If both the visual clarity parameter and the visual pointer length corresponding to the industrial instrument image are greater than the preset reference visual clarity parameter and reference visual pointer length, then the visual monitoring corresponding to the industrial instrument is determined to be compliant.

[0028] It should be noted that the reference visual clarity parameters are set by professional industrial managers; based on the preset parameters, the most accurate image is obtained, which facilitates more accurate instrument reading recognition, obtains accurate recognition results, and allows for targeted measures to be taken; the reference visual pointer length is the actual length corresponding to the instrument.

[0029] The industrial instrument vision analysis module is used to analyze the secondary control of the vision of the industrial instrument when the vision monitoring of the industrial instrument is not up to standard, obtain the visual image of the industrial instrument, and identify the operating status of the industrial operation corresponding to the industrial instrument.

[0030] When visual monitoring fails to meet standards, the secondary control of the vision corresponding to the industrial instruments is analyzed, and the operating status of the industrial operation corresponding to the industrial instruments is identified. This enables timely control of the visual detection of the industrial instruments, making it easier to obtain a more accurate and effective picture of the industrial instruments, thereby ensuring the correctness of the automatic recognition results of the industrial instrument readings.

[0031] As an optional implementation, the secondary control of the vision corresponding to the industrial instrument is analyzed as follows: Based on the image data pointed to when the vision image of the industrial instrument is non-compliant, a reference control parameter range corresponding to each historical image data with the same non-compliant phenomenon as the current image is extracted from the database. The reference control parameter range is used as the data control reference range under the current non-compliant vision image of the industrial instrument. Vision control is performed. When the image data of the vision image corresponding to the industrial instrument after vision control is greater than the preset reference value, the secondary control of the vision corresponding to the industrial instrument is completed.

[0032] It should be noted that, based on the acquisition of historical images of industrial instruments when their visual images were non-compliant and historical image data, and by recording the corresponding current control measures, a set of control measures for historical industrial instruments when their visual images were non-compliant is obtained. This set consists of reference control parameter ranges corresponding to each historical image data, and serves as a control reference item for the current non-compliant visual image of the industrial instrument.

[0033] As an optional implementation, the specific analysis process for identifying the operational status of the industrial instruments corresponding to the industrial operations is as follows: The instrument range is extracted from the visual image of the industrial instrument, and the instrument range readings are converted to obtain the actual operating parameters corresponding to the industrial instrument. The actual operating parameters corresponding to the industrial instrument are compared with a preset reference operating parameter threshold. If the actual operating parameters corresponding to the industrial instrument are greater than the preset reference operating parameter threshold, it is determined that the current industrial operation exceeds the normal load and the operation is abnormal; otherwise, it is determined that the current industrial operation is normal. This is how the operational status of the industrial instruments corresponding to the industrial operations is identified.

[0034] It should be noted that the reading conversion is as follows: The instrument range includes a maximum value and a minimum value. The range span is obtained by subtracting the minimum value from the maximum value. Based on the visual image of the industrial instrument, the included angle and the total angle from the minimum value of the pointer to the maximum value are obtained. The range span is divided by the total angle to obtain the angle. The minimum value plus the angle is multiplied by the included angle to obtain the actual operating parameters. The actual operating parameters can cover reading conversion directions including pressure values, temperature values, etc. Reference operating parameter thresholds are set by professional industrial managers. Based on the preset reference operating parameter thresholds that ensure normal operation of industrial operations, abnormal industrial operation conditions can be identified more efficiently, timely over-operation maintenance can be performed, and industrial operating pressure can be reduced.

[0035] The industrial instrument reading analysis module is used to analyze the anomalies of industrial instruments corresponding to industrial operation anomalies, and to analyze the maintenance measures for the industrial instruments corresponding to industrial operation anomalies.

[0036] Based on the analysis of the abnormal entities corresponding to industrial operation anomalies of industrial instruments, and the analysis of the maintenance measures for the abnormal entities corresponding to industrial operation anomalies of industrial instruments, the goal is to achieve timely understanding and control of industrial production problems, further curb the deterioration of fault problems, reduce the safety risks of industrial production, and ensure the safety and stability of industrial production.

[0037] As an optional implementation, the analysis of the abnormal body corresponding to the industrial operation anomaly of the industrial instrument is specifically performed as follows: extract the operation direction characteristics of each device corresponding to the industrial operation from the database, and based on the direction type of the actual operation parameters corresponding to the industrial instrument, filter out each device with the same direction type and record it as each target device, monitor each target device, obtain the actual working value, compare the working value corresponding to each target device with the actual operating parameters, if the working value corresponding to a target device is greater than or equal to the actual operating parameters, then the device is recorded as an abnormal body, thereby analyzing the abnormal body corresponding to the industrial operation anomaly of the industrial instrument.

[0038] It should be noted that the operational indication characteristics refer to the specific performance of the equipment in industrial operation. For example, the specific manifestation of a certain equipment is the presentation of pressure, temperature, etc. The indication type of the actual operating parameters corresponding to industrial instruments refers to pressure values ​​or temperature values, etc.; the corresponding working values ​​of each target equipment are obtained based on temperature sensors and pressure sensors.

[0039] As an optional implementation, the analysis of maintenance measures for industrial instruments corresponding to industrial operation anomalies is specifically carried out as follows: The actual working value of the industrial operation anomaly is compared with the reference problem value range of the operation and maintenance measures for each piece of equipment corresponding to the anomaly stored in the database. If the actual working value of the industrial operation anomaly is included in the reference problem value range of the operation and maintenance measures for a certain piece of equipment corresponding to the anomaly stored in the database, then the operation and maintenance measures for that piece of equipment are taken as maintenance measures for the industrial instruments corresponding to industrial operation anomalies.

[0040] It should be noted that, based on the recorded historical industrial operations, the problem images, problem parameters, and measures taken when each piece of equipment was in an abnormal state, a problem database for each piece of equipment is constructed. This database serves as a reference for the maintenance measures taken when the corresponding equipment in the current industrial operation has a problem, facilitating efficient and accurate problem resolution in industrial operations.

[0041] The database is used to store basic information, location data ranges of each installation measure, image data, reference control parameter ranges corresponding to each historical image data, instrument ranges, operating direction characteristics, and reference problem value ranges for each equipment operation and maintenance measure.

[0042] The early warning terminal is used to issue early warnings when the visual monitoring corresponding to the industrial instrument fails to meet the standards or when the operation of the industrial instrument corresponding to the industrial operation is abnormal.

[0043] This invention, through analyzing the visual conditions monitored by industrial instruments, performs corresponding monitoring, identifies the operational status of the industrial operations mapped by the industrial instruments, analyzes the abnormal entities of the industrial operations corresponding to the industrial instruments, and finally analyzes the maintenance measures for the abnormal entities of the industrial operations corresponding to the industrial instruments. This achieves comprehensive and efficient visual monitoring of industrial production. Based on intelligent AI vision technology, it obtains a full-coverage industrial instrument scene display, thereby performing efficient reading evaluation and obtaining timely and accurate evaluation results. This allows for targeted maintenance measures to be taken, reducing equipment failure rates and safety risks during industrial operations and intelligently ensuring the quality of industrial operations.

[0044] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0045] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. An automatic identification system for industrial instrument readings based on AI vision, characterized in that, Includes the following modules: The industrial instrument visual monitoring module is used to acquire basic information of industrial instruments in the industrial field, analyze the set of visual conditions for monitoring industrial instruments, collect images of industrial instruments, acquire image data, and analyze whether the visual monitoring of industrial instruments meets the standards. The industrial instrument vision analysis module is used to analyze the secondary control of the vision of the industrial instrument when the vision monitoring of the industrial instrument is not up to standard, obtain the visual image of the industrial instrument, and identify the operating status of the industrial operation corresponding to the industrial instrument. The industrial instrument reading analysis module is used to analyze the anomalies of industrial instruments corresponding to industrial operation anomalies, and to analyze the maintenance measures for the industrial instruments corresponding to industrial operation anomalies. The early warning terminal is used to issue early warnings when the visual monitoring corresponding to the industrial instrument fails to meet the standards or when the operation of the industrial instrument corresponding to the industrial operation is abnormal.

2. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 1, characterized in that, The analysis process is as follows, based on the set of visual conditions monitored by the industrial instruments: The distribution location and spatial location of the industrial instruments in the industrial field are obtained. The distribution location and spatial location of the industrial instruments are compared with the location data range of each placement measure in the vision library stored in the database. If the distribution location and spatial location of the industrial instruments are included in the location data range of a certain placement measure in the vision library stored in the database, then the placement measure is used as the vision condition of the industrial instruments in the industrial field, including placement angle, placement location and acquisition frequency. The set of visual conditions for monitoring corresponding to industrial instruments is constructed by the placement angle, placement position, and acquisition frequency.

3. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 2, characterized in that, The specific analysis process for acquiring industrial instrument images is as follows: The industrial instruments in the industrial field are monitored according to the visual condition set. The images of the industrial instruments are collected and preprocessed, including noise reduction, illumination correction and angle correction. The industrial instrument images are denoised by Gaussian filtering and Messian filtering, and illumination correction is performed on the industrial instrument images based on the histogram equalization algorithm. Then, the tilt angle of the industrial instrument images is calculated and corrected by using the circular contour of the instrument panel of the Hough transform detector, so as to obtain standardized industrial instrument images.

4. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 3, characterized in that, The image data includes visual clarity parameters and visual pointer length.

5. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 4, characterized in that, The analysis process for determining whether the visual monitoring corresponding to the industrial instruments meets the standards is as follows: The visual clarity parameter and visual pointer length corresponding to the industrial instrument image are compared with the preset reference visual clarity parameter and reference visual pointer length, respectively. If the visual clarity parameter corresponding to the industrial instrument image is less than the preset reference visual clarity parameter, or the visual pointer length is less than the reference visual pointer length, the visual monitoring corresponding to the industrial instrument is determined to be non-compliant. If both the visual clarity parameter and visual pointer length corresponding to the industrial instrument image are greater than the preset reference visual clarity parameter and reference visual pointer length, the visual monitoring corresponding to the industrial instrument is determined to be compliant.

6. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 5, characterized in that, The analysis process for the secondary control of the industrial instrument corresponding to vision is as follows: Based on the image data pointed to when the visual image of the industrial instrument is non-compliant, and extracting the reference control parameter range corresponding to each historical image data with the same non-compliant phenomenon as the current image from the database, the reference control parameter range is used as the data control reference range under the current non-compliant visual image of the industrial instrument. Visual control is performed, and when the image data of the corresponding visual image of the industrial instrument after visual control is greater than the preset reference value, the secondary control of the visual image of the industrial instrument is completed.

7. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 1, characterized in that, The identification of industrial instruments corresponds to the operational status of industrial operations, and the specific analysis process is as follows: The instrument range is extracted from the visual image of the industrial instrument, and the instrument range readings are converted to obtain the actual operating parameters of the industrial instrument. The actual operating parameters of the industrial instrument are compared with the preset reference operating parameter threshold. If the actual operating parameters of the industrial instrument are greater than the preset reference operating parameter threshold, it is determined that the current industrial operation exceeds the normal load and the operation is abnormal. Otherwise, it is determined that the current industrial operation is normal. In this way, the operating status of the industrial operation corresponding to the industrial instrument is identified.

8. The automatic identification system for industrial instrument readings based on AI vision according to claim 1, characterized in that, The analysis process for the anomaly body corresponding to the industrial operation abnormality of the analyzed industrial instrument is as follows: The system extracts the operational characteristics of each device corresponding to industrial operations from the database. Based on the orientation type of the actual operating parameters of the industrial instruments, it filters out devices with the same orientation type and records them as target devices. It then monitors each target device to obtain the actual operating values. The system compares the operating values ​​of each target device with the actual operating parameters. If the operating value of a target device is greater than or equal to the actual operating parameters, the device is recorded as an anomaly. This process is used to analyze the anomalies of industrial operations corresponding to industrial instruments.

9. The automatic identification system for industrial instrument readings based on AI vision according to claim 1, characterized in that, The analysis of maintenance measures for industrial instruments corresponding to industrial operation anomalies is as follows: The actual working value of the industrial operation corresponding to the anomaly is compared with the reference problem value range of the operation and maintenance measures of each equipment corresponding to the anomaly stored in the database. If the actual working value of the industrial operation corresponding to the anomaly is included in the reference problem value range of the operation and maintenance measures of a certain equipment corresponding to the anomaly stored in the database, then the operation and maintenance measures of that equipment are taken as the maintenance measures of the industrial instrument corresponding to the industrial operation anomaly.

10. The AI ​​vision-based automatic identification system for industrial instrument readings according to claim 1, characterized in that, It also includes a database, which stores basic information, location data ranges for each installation measure, image data, reference control parameter ranges corresponding to each historical image data, instrument ranges, operating direction characteristics, and reference problem value ranges for each equipment operation and maintenance measure.

Images