Area Intelligent Analysis System for Diaphragm Gas Meters
Through the deep neural network model combined with image processing technology, the direction sharpening image of the membrane gas meter is analyzed, which solves the problem of the membrane gas meter size not meeting the standards, realizes intelligent detection, improves the accuracy and reliability of the detection, and simplifies the Operation and reduce costs.
Patent Information
- Application Number
- CN202410546966.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-06
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-05-06
AI Technical Summary
During the production process of membrane gas meter, due to component production dimensional errors and assembly deviations, the actual dial area data varies greatly, making it difficult to ensure that the size of membrane gas meter meets the standards, which increases the detection difficulty and cost.
Deep neural network model combined with direction sharpening image analysis, detect the pixel coordinates, contrast and imaging focal length of membrane gas meter, and filter the effective data and judge the 膜式燃气表的实际面积是否在预设范围内,发出误差可控或超限信号。
It realizes intelligent analysis of the front-end surface area of the membrane gas meter, improves the accuracy and reliability of detection, simplifies operation, and can promptly determine whether the membrane gas meter meets the standards, reducing manual inspection costs.
Abstract
Description
Technical Field
[0001] The present invention relates to the field of gas meters, and more specifically, to an area intelligent analysis system for a membrane gas meter. Background Art
[0002] There are two mainstream gas meters on the market, one is the traditional mechanical diaphragm gas meter; the other is the prepaid diaphragm gas meter. The mechanical diaphragm gas meter is measured by a mechanical roller. The mechanical roller is added according to the amount of gas used. Every time a unit is used, the roller counts one, and finally the gas volume measurement record is realized. The advantages of mechanical gas meters are reliable measurement and stable quality. The disadvantages are that meter reading is troublesome. Manual door-to-door meter reading is required, and gas companies need to invest a lot of financial and human resources. Prepaid diaphragm gas meters are improved on the basis of traditional mechanical meters. On the basis of the original, electronic metering methods, prepayment and display functions are added, which can display the current working status of the gas meter. The prepaid method of buying gas first and then using gas can be realized, which greatly reduces the cost of gas companies and labor.
[0003] Generally, the surface area corresponding to the front end surfaces of multiple diaphragm gas meters of the same model should be within the preset area value range corresponding to the standard diaphragm gas meter of the model in order to meet the layout requirements of the dial assembly of the diaphragm gas meter of the model. However, in the production process of diaphragm gas meters, due to errors in the production dimensions of each component and assembly deviations when assembling each diaphragm gas meter, the actual dial area data of each diaphragm gas meter actually produced may vary greatly. Summary of the invention
[0004] In order to solve the technical problems in the related fields, the present invention provides an area intelligent analysis system for a diaphragm gas meter, which can detect the horizontal coordinate values corresponding to the pixels occupied by the front end surface of the diaphragm gas meter in the direction-sharpened image and the vertical coordinate values corresponding to the pixels, detect the contrast and imaging focal length of the direction-sharpened image, so as to screen effective data for subsequent intelligent analysis, input the horizontal coordinate values, the vertical coordinate values, the contrast and imaging focal length of the direction-sharpened image into a deep neural network model in parallel, and run the deep neural network model to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model, and send an error controllable signal when the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model is within the preset area value interval corresponding to the front end surface of the standard diaphragm gas meter, otherwise, send an error limit signal, so as to complete the intelligent analysis result based on the actual area value of the front end surface of the diaphragm gas meter that is currently offline, and judge whether the size of the front end surface of the diaphragm gas meter that is currently offline meets the standard.
[0005] According to the present invention, there is provided an area intelligent analysis system for a membrane gas meter, the system comprising:
[0006] A trigger camera is used to perform a camera operation on the front end surface of the latest offline diaphragm gas meter when receiving an offline trigger instruction corresponding to the offline diaphragm gas meter, so as to obtain and output a corresponding front end camera image;
[0007] A content filtering device, connected to the trigger camera device, for performing Butterworth low-pass filtering on the received front-end camera image to obtain and output a corresponding content filtering image;
[0008] an affine transformation device, connected to the content filtering device, for performing affine transformation processing on the received content filtering image to obtain and output a corresponding affine transformation image;
[0009] A directional sharpening device, connected to the affine transformation device, for sequentially performing horizontal direction sharpening processing and vertical direction sharpening processing on the received affine transformation image to obtain and output a corresponding directional sharpened image;
[0010] A numerical mapping mechanism is connected to the direction sharpening device and includes a pixel point analysis device, a model building device and a model use device, wherein the model use device is respectively connected to the model building device and the pixel point analysis device, wherein the numerical mapping mechanism is used to detect the horizontal coordinate values corresponding to the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image and the vertical coordinate values corresponding to the pixels, detect the contrast and imaging focal length of the direction sharpened image, input the horizontal coordinate values, the vertical coordinate values, the contrast and imaging focal length of the direction sharpened image into a deep neural network model in parallel, and run the deep neural network model to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model, wherein the number of hidden layers of the deep neural network model is positively correlated with the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image;
[0011] an error identification mechanism, connected to the value mapping mechanism, for issuing an error controllable signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is within a preset area value interval corresponding to the front end surface of the standard diaphragm gas meter, and for issuing an error exceeding limit signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is outside the preset area value interval corresponding to the front end surface of the standard diaphragm gas meter;
[0012] Among them, the model building device is used to establish a deep neural network model, and the number of hidden layers of the deep neural network model is positively correlated with each pixel point occupied by the front end surface of the membrane gas meter in the directionally sharpened image.
[0013] It can be seen that the present invention has at least the following three key invention points:
[0014] First, detect the horizontal coordinate values corresponding to the pixels of the front end surface of the membrane gas meter in the direction-sharpened image and the vertical coordinate values corresponding to the pixels, and detect the contrast and imaging focal length of the direction-sharpened image, so as to select effective data for subsequent intelligent analysis;
[0015] Secondly, each horizontal coordinate value, each vertical coordinate value, the contrast of the directionally sharpened image, and the imaging focal length are input into the deep neural network model in parallel, and the deep neural network model is run to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model, and when the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model is within the preset area value interval corresponding to the front end surface of the standard diaphragm gas meter, an error controllable signal is issued, otherwise, an error exceeding limit signal is issued, thereby completing the intelligent analysis result based on the actual area value of the front end surface of the diaphragm gas meter that has just come off the production line to determine whether the size of the front end surface of the diaphragm gas meter that has just come off the production line meets the standard;
[0016] Third, the number of hidden layers of the deep neural network model selected for intelligent analysis is positively correlated with each pixel point occupied by the front end surface of the diaphragm gas meter in the directionally sharpened image, thereby ensuring the reliability and stability of the intelligent analysis results.
[0017] The area intelligent analysis system for diaphragm gas meters of the present invention is intelligently designed and easy to operate. Since a deep neural network model can be used to perform intelligent analysis of the actual area value of the front end surface of each diaphragm gas meter, and when the actual area value obtained is outside the preset area value range corresponding to the front end surface of the standard diaphragm gas meter, an error limit signal is issued, thereby completing the intelligent judgment of whether the size of the front end surface of the latest diaphragm gas meter off the production line meets the standard. DETAILED DESCRIPTION
[0018] The area intelligent analysis system for a membrane gas meter according to the primary embodiment of the present invention comprises:
[0019] A trigger camera is used to perform a camera operation on the front end surface of the latest offline diaphragm gas meter when receiving an offline trigger instruction corresponding to the offline diaphragm gas meter, so as to obtain and output a corresponding front end camera image;
[0020] A content filtering device, connected to the trigger camera device, for performing Butterworth low-pass filtering on the received front-end camera image to obtain and output a corresponding content filtering image;
[0021] For example, the content filtering device is connected to the trigger camera device and is used to perform Butterworth low-pass filtering on the received front-end camera picture to obtain and output the corresponding content filtering image, including: an FPGA device can be selected to implement the content filtering device, which is connected to the trigger camera device and is used to perform Butterworth low-pass filtering on the received front-end camera picture to obtain and output the corresponding content filtering image;
[0022] an affine transformation device, connected to the content filtering device, for performing affine transformation processing on the received content filtering image to obtain and output a corresponding affine transformation image;
[0023] A directional sharpening device, connected to the affine transformation device, for sequentially performing horizontal direction sharpening processing and vertical direction sharpening processing on the received affine transformation image to obtain and output a corresponding directional sharpened image;
[0024] A numerical mapping mechanism is connected to the direction sharpening device and includes a pixel point analysis device, a model building device and a model use device, wherein the model use device is respectively connected to the model building device and the pixel point analysis device, wherein the numerical mapping mechanism is used to detect the horizontal coordinate values corresponding to the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image and the vertical coordinate values corresponding to the pixels, detect the contrast and imaging focal length of the direction sharpened image, input the horizontal coordinate values, the vertical coordinate values, the contrast and imaging focal length of the direction sharpened image into a deep neural network model in parallel, and run the deep neural network model to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model, wherein the number of hidden layers of the deep neural network model is positively correlated with the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image;
[0025] an error identification mechanism, connected to the value mapping mechanism, for issuing an error controllable signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is within a preset area value interval corresponding to the front end surface of the standard diaphragm gas meter, and for issuing an error exceeding limit signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is outside the preset area value interval corresponding to the front end surface of the standard diaphragm gas meter;
[0026] The model building device is used to establish a deep neural network model, and the number of hidden layers of the deep neural network model is positively correlated with each pixel point occupied by the front end surface of the membrane gas meter in the direction-sharpened image;
[0027] And wherein, the pixel point analysis device is used to detect the horizontal coordinate values corresponding to each pixel point occupied by the front end surface of the membrane gas meter in the direction-sharpened image and the vertical coordinate values corresponding to each pixel point, and detect the contrast and imaging focal length of the direction-sharpened image.
[0028] Compared with the primary embodiment of the present invention, the area intelligent analysis system for a membrane gas meter according to the secondary embodiment of the present invention may further include the following components:
[0029] A temperature measurement mechanism, comprising a plurality of temperature measurement units, for respectively measuring current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism;
[0030] The temperature measurement mechanism includes a plurality of temperature measurement units for respectively measuring the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, including: the plurality of temperature measurement units respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism are a plurality of non-contact temperature sensors;
[0031] Wherein, the temperature measurement mechanism includes a plurality of temperature measurement units for respectively measuring the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, and further includes: the internal structures of a plurality of non-contact temperature sensors respectively used for the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism are the same;
[0032] The temperature measurement mechanism includes a plurality of temperature measurement units for respectively measuring the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, and further includes: a plurality of non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism have the same temperature measurement upper limit threshold and temperature measurement lower limit threshold;
[0033] And wherein, the temperature measurement mechanism includes multiple temperature measurement units, which are used to respectively measure the current surface temperature values of the content filtering device, the affine transformation device, the directional sharpening device and the numerical mapping mechanism, and also includes: multiple non-contact temperature sensors respectively used for the content filtering device, the affine transformation device, the directional sharpening device and the numerical mapping mechanism are at equal distances from the content filtering device, the affine transformation device, the directional sharpening device and the numerical mapping mechanism.
[0034] Compared with the primary embodiment of the present invention, the area intelligent analysis system for a membrane gas meter according to the secondary embodiment of the present invention may further include the following components:
[0035] a photoelectric alarm mechanism, connected to the plurality of non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, for performing corresponding temperature alarm actions based on temperature measurement results of the plurality of non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism;
[0036] Wherein, the photoelectric alarm mechanism is respectively connected to a plurality of non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, and is used to perform corresponding temperature alarm actions based on the temperature measurement results of the plurality of non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism, including: the photoelectric alarm mechanism includes an optical alarm unit and an electrical alarm unit;
[0037] And wherein, the photoelectric alarm mechanism is respectively connected to multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism, and is used to perform corresponding temperature alarm actions based on the temperature measurement results of the multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism, and also includes: the optical alarm unit flashes at a set frequency to perform the temperature alarm action.
[0038] In addition, in the area intelligent analysis system for diaphragm gas meters, the model uses a device for inputting the various horizontal coordinate values, the various vertical coordinate values, the contrast of the directionally sharpened image, and the imaging focal length into a deep neural network model in parallel, and running the deep neural network model to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model.
[0039] It will be appreciated by those skilled in the art that various modifications and changes may be made without departing from the scope and spirit of the present invention. It should therefore be understood that the above embodiments are for illustration only and are not intended to be limiting. Because the scope of the present invention is defined by the claims rather than the foregoing description, any changes and modifications that fall within the scope and boundaries of the claims or the equivalents of these scopes and boundaries are subject to the claims.
Claims
1. An area intelligent analysis system for a membrane gas meter, characterized in that: The system comprises: A trigger camera is used to perform a camera operation on the front end surface of the latest offline diaphragm gas meter when receiving an offline trigger instruction corresponding to the offline diaphragm gas meter, so as to obtain and output a corresponding front end camera image; A content filtering device, connected to the trigger camera device, for performing Butterworth low-pass filtering on the received front-end camera image to obtain and output a corresponding content filtering image; an affine transformation device, connected to the content filtering device, for performing affine transformation processing on the received content filtering image to obtain and output a corresponding affine transformation image; A directional sharpening device, connected to the affine transformation device, for sequentially performing horizontal direction sharpening processing and vertical direction sharpening processing on the received affine transformation image to obtain and output a corresponding directional sharpened image; A numerical mapping mechanism is connected to the direction sharpening device and includes a pixel point analysis device, a model building device and a model use device, wherein the model use device is respectively connected to the model building device and the pixel point analysis device, wherein the numerical mapping mechanism is used to detect the horizontal coordinate values corresponding to the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image and the vertical coordinate values corresponding to the pixels, detect the contrast and imaging focal length of the direction sharpened image, input the horizontal coordinate values, the vertical coordinate values, the contrast and imaging focal length of the direction sharpened image into a deep neural network model in parallel, and run the deep neural network model to obtain the actual area value of the front end surface of the diaphragm gas meter output by the deep neural network model, wherein the number of hidden layers of the deep neural network model is positively correlated with the pixels occupied by the front end surface of the diaphragm gas meter in the direction sharpened image; an error identification mechanism, connected to the value mapping mechanism, for issuing an error controllable signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is within a preset area value interval corresponding to the front end surface of the standard diaphragm gas meter, and for issuing an error exceeding limit signal when the actual area value of the front end surface of the diaphragm gas meter outputted by the received deep neural network model is outside the preset area value interval corresponding to the front end surface of the standard diaphragm gas meter; Wherein, the model building device is used to establish a deep neural network model; A temperature measurement mechanism, comprising a plurality of temperature measurement units, for respectively measuring current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism; Wherein, the multiple temperature measurement units respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the value mapping mechanism are multiple non-contact temperature sensors; The photoelectric alarm mechanism is respectively connected to multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism, and is used to perform corresponding temperature alarm actions based on the temperature measurement results of the multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism.
2. The area intelligent analysis system for a membrane gas meter according to claim 1, characterized in that: The pixel point analysis device is used to detect the horizontal coordinate values corresponding to each pixel point occupied by the front end surface of the membrane gas meter in the direction-sharpened image and the vertical coordinate values corresponding to each pixel point, and detect the contrast and imaging focal length of the direction-sharpened image.
3. The area intelligent analysis system for a membrane gas meter according to claim 1, characterized in that: The temperature measurement mechanism includes multiple temperature measurement units, which are used to respectively measure the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism. It also includes: the internal structures of multiple non-contact temperature sensors respectively used for the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism are the same.
4. The area intelligent analysis system for a membrane gas meter according to claim 1, characterized in that: The temperature measurement mechanism includes multiple temperature measurement units, which are used to respectively measure the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism. It also includes: multiple non-contact temperature sensors respectively used for the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism have the same temperature measurement upper limit threshold and temperature measurement lower limit threshold.
5. The area intelligent analysis system for a membrane gas meter according to claim 4, characterized in that: The temperature measurement mechanism includes multiple temperature measurement units, which are used to respectively measure the current surface temperature values of the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism. It also includes: multiple non-contact temperature sensors respectively used for the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism are at equal distances from the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism.
6. The area intelligent analysis system for a membrane gas meter according to claim 1, characterized in that: The photoelectric alarm mechanism is respectively connected to multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism, and is used to perform corresponding temperature alarm actions based on the temperature measurement results of the multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism. The photoelectric alarm mechanism includes an optical alarm unit and an electrical alarm unit.
7. The area intelligent analysis system for a membrane gas meter according to claim 6, characterized in that: The photoelectric alarm mechanism is respectively connected to multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism, and is used to perform corresponding temperature alarm actions based on the temperature measurement results of the multiple non-contact temperature sensors respectively used by the content filtering device, the affine transformation device, the direction sharpening device and the numerical mapping mechanism. The optical alarm unit also flashes at a set frequency to perform the temperature alarm action.
Citation Information
Patent Citations
Aluminum profile packaging length detection method and system
CN111951324A
Diaphragm gas meter reading identification method and device based on projection segmentation method
CN114863415A
