A digital image-based water meter calibration system
The water meter verification system based on digital images has achieved automated water meter testing, solving the problem of high manpower and material costs associated with traditional verification methods, improving testing efficiency and accuracy, and meeting national standards.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-10-24
- Publication Date
- 2026-03-20
AI Technical Summary
Traditional water meter calibration methods consume a lot of manpower and resources, and have large reading errors, making it difficult to meet the needs of efficient and accurate testing.
Design a water meter calibration system based on digital images, including a water meter calibration installation platform, a constant water flow supply module, a digital image recognition module, and a computer data acquisition and processing module, to realize automated testing and intelligent data processing. The system acquires images through a camera and performs digital image processing, and combines computer algorithms to obtain water meter error and repeatability results.
It realizes the intelligent and automated verification of water meters, reduces the reliance on professional personnel, improves testing efficiency and accuracy, complies with national metrological verification regulations, and reduces errors introduced by manual reading.
Smart Images
Figure CN115540978B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of intelligent detection of water meters, in particular to a water meter calibration system based on digital images. BACKGROUND
[0002] According to the requirements of JJG 162-2019 "Drinking Cold Water Meter", for mechanical water meters, the value error test should be carried out, and the test procedure is: the water meter is installed on the water meter calibration device individually or in batches, the air in the water meter and the experimental device pipeline is discharged by water, then the water pressure is slowly raised under the static state of the water meter, and the pressure is observed for 1 min without leakage or damage, then the flow is adjusted to the normal flow, the boundary flow and the minimum flow, and the calibration of all flow points is completed by starting and stopping method.
[0003] At present, as one of the three civil meters, the use of water meters is very large, and the update iteration is very rapid. The unit of measurement of water meters is unified, and the accurate and reliable data plays a very important role in national economy and people's livelihood. The traditional calibration method produces a huge workload, and in addition to the professional requirements of the calibration personnel, it is also a consumption in physical and mental aspects. Therefore, a mechanical water meter dial reading system with automatic calibration function is designed, which can effectively improve the detection efficiency and accuracy and reduce the dependence on professional detection personnel. SUMMARY
[0004] In view of the above problems, the present application provides a water meter calibration system based on digital images, which can automatically obtain the automatic water meter detection device of the water meter dial reading, realize automatic test and intelligent data processing. The water meter calibration original record can be issued combined with the water meter information (manufacturer, model specification, factory number, test parameter).
[0005] In order to realize the above functions, the following technical solutions and installation entities are adopted:
[0006] A water meter calibration system based on digital images, comprising: a water meter calibration installation table module, a constant water flow supply module, a digital image recognition module and a computer data acquisition and processing module.
[0007] The water meter calibration installation platform module is responsible for the series installation of three water meters of various diameters and specifications. The constant water flow supply module is the water supply source of the overall system and is fixedly connected with the water meter calibration installation platform through a mechanical structure. The digital image recognition module acquires digital images through a camera placed at the upper end of the water meter calibration installation platform and calls a digital image processing algorithm to process the acquired digital images to obtain real-time reading data of the water meter. The computer data acquisition and processing module obtains output data through the algorithm interface of the digital image recognition module and performs formula calculation on the corresponding data to obtain the error and repeatability results of the water meter calibration and generate an electronic original calibration record.
[0008] The water meter calibration installation platform module controls the installation distance between the water meters through a mechanical telescopic transmission meter clamping structure to ensure seamless connection of the water meters being tested and complete the series connection of three water meters of different specifications and diameters through different diameter water pipes.
[0009] The constant water flow supply module includes a water pump, a pressure stabilizing tank, a flow switch valve, a flow regulating valve, and an indicating flowmeter. The water pump draws circulating water from an underground pool to the pressure stabilizing tank for pressure stabilization and regulation. Then the flow switch valve is opened to input the stabilized circulating water into the water meter calibration installation platform module, through the sealed mechanical water meter, to the flow regulating valve for circulating water flow rate regulation. The circulating water flow rate is obtained through the indicating flowmeter. The circulating water with stable pressure and fixed flow rate enters the standard device through the mechanical connection pipeline to complete one cycle of constant water flow.
[0010] The digital image recognition module includes a camera, a data transmission link, and a digital image packager for intelligent reading of real-time accumulated data of the mechanical water meter. The digital image recognition module captures digital images of the dial of the mechanical water meter through the camera, transmits the digital images to the industrial computer through the data transmission link, and extracts the features of the water meter dial through the digital image packager embedded in the industrial computer to obtain real-time accumulated data of the water meter.
[0011] The computer data acquisition and processing module includes a pressure sensor installed in the standard device, a temperature sensor installed on the indicating flowmeter, a database, an experimental data processor, and a report template. The temperature sensor and the pressure sensor are connected with the experimental data processor through a communication line to collect, arrange, and calculate the pressure and medium temperature during various tests. First, the computer data acquisition and processing module converts the temperature and pressure sensor data of the standard device to obtain the circulating water quality data collected during the test. Then, the accumulated value data of the mechanical water meter output by the digital image recognition module through the algorithm interface are obtained. Both are input into the experimental data processor to generate a Word version of the water meter calibration original record.
[0012] Further, the computer data acquisition and processing module obtains the real-time mechanical readings of the water meter dial before and after the test, substitutes the readings into the volume algorithm formula to calculate the actual flow accumulation of the water meter during the test, and then compares the volume accumulation of the standard device during the same test time, i.e., the standard value, substitutes the error formula to calculate the water meter indication error of this test. Through repeated tests at different detection flow points, complete experimental data of the water meter calibration are obtained, the water meter calibration original record is formed, the water meter calibration is completed, and the water meter calibration original record is printed.
[0013] The effective results of the present application are:
[0014] 1. The water meter calibration can be realized intelligently. After the mechanical water meter is installed, the corresponding parameters are set in the water meter calibration system according to the basic information of the water meter, and the corresponding calibration test can be completed. The device can automatically obtain the test data required for calibration according to the adjusted detection flow point, reducing the dependence on professional calibration personnel, improving the calibration efficiency, and reducing the work pressure of the experimental personnel.
[0015] 2. The test data of the calibration is accurate and reliable. On the basis of the start-stop method, the error and uncertainty introduced by manual reading can be effectively reduced by digital image recognition instead of manual reading. The related operation fully meets the related requirements in the national metrological calibration regulation JJG 162-2019.
[0016] 3. It has strong data integration and processing function. The previous water meter calibration test needs to be completed by manual reading and handwritten original record to process the experimental results. The intelligent mechanical water meter dial display reading system can realize the recording, query, statistics and printing of the calibration data, completely eliminating the manual repetitive labor. BRIEF DESCRIPTION OF DRAWINGS
[0017] Figure 1 is a structural schematic view of the present application;
[0018] Figure 2 is a water meter panel.
[0019] In the figure: 1, experimental data processor, 2, industrial computer, 3, water meter, 4, flow regulating valve, 5, camera, 6, standard device, 7, indicating flowmeter, 8, water pump, 9, pressure stabilizing tank, 10, flow switch valve, 11, mechanical telescopic transmission clamp meter structure. DETAILED DESCRIPTION
[0020] The present application will be further described below in conjunction with the drawings.
[0021] As Figure 1As shown, it is a structural schematic diagram of the application, which includes an experimental data processor 1, an industrial computer 2, a water meter 3, a flow regulating valve 4, a camera 5, a standard device 6, an indicating flowmeter 7, a water pump 8, a pressure stabilizing tank 9, a flow switch valve 10 and a mechanical telescopic transmission clamp meter structure 11. The industrial computer 2 sets specific detection processes and experimental requirements according to the actual model specifications of the water meter 3, including: verification flow point, precision control, verification time, etc. The industrial computer 2 sends control instructions through the data transmission line to adjust the detection medium flowing through each water meter in turn.
[0022] A water meter verification system based on digital image, comprising: a water meter verification installation platform module, a constant water flow supply module, a digital image recognition module and a computer data acquisition and processing module.
[0023] The water meter verification installation platform module is responsible for the three-in-line installation of water meters of various diameters and specifications, achieving the functions of overall sealing and smooth internal drainage. The constant water flow supply module is the water supply source of the overall system, which plays the functions of adjusting water flow pressure and flow rate, and is fixedly connected with the water meter verification installation platform through a mechanical structure. The digital image recognition module acquires digital images through a camera placed at the upper end of the water meter verification installation platform, and processes the acquired digital images by calling a digital image processing algorithm to obtain real-time reading data of the water meter. The computer data acquisition and processing module obtains output data through the algorithm interface of the digital image recognition module and performs complex formula calculation on the corresponding data to obtain the error and repeatability results of the water meter verification, and generates an electronic verification original record. The modules in the system are organically combined, cooperate and collaboratively complete the indication error test of mechanical water meters.
[0024] The mechanical water meter dial indication reading system realizes the three-in-line connection of mechanical water meters 3 of different specifications and diameters through the water meter verification installation platform module, and the water meters are connected through five groups of water meter short connection pipe sections of different diameters. Through a mechanical telescopic transmission clamp meter structure 11, the installation distance between the water meters is controlled, thereby ensuring seamless connection of the mechanical water meters being tested, achieving overall sealing and smooth internal water flow.
[0025] The constant water flow supply module includes a water pump 8, a pressure stabilizing tank 9, a flow switch valve 10, a flow regulating valve 4 and an indicating flowmeter 7, for the circulation of water flow in the system. The circulating water is first pumped from the underground pool to the pressure stabilizing tank 9 by the water pump 8, and the water pressure of the input circulating water is regulated by the pressure stabilizing tank 9. Then the flow switch valve 10 is manually opened, and the stabilized circulating water is input into the water meter calibration installation platform module 3 through the flow switch valve, and the flow rate of the circulating water is regulated by the flow regulating valve 4. The flow rate of the circulating water is obtained by the indicating flowmeter 7, which includes three electromagnetic flowmeters with different diameters and flow ranges, which can be intelligently selected according to the flow points to be detected, and the real-time flow rate through the mechanical water meter is indicated in time. Finally, the circulating water with stable pressure and fixed flow rate enters the standard device 6 through the mechanical connection pipeline, completing a cycle of constant water flow. The standard device 6 is a water flow collecting device, which can obtain the cumulative amount of water flow quality within the cumulative working time.
[0026] The digital image recognition module includes a camera, a data transmission link and a digital image encapsulator, for the system to intelligently read the real-time cumulative data of the mechanical water meter. The digital image recognition module captures the digital image of the dial of the mechanical water meter through the camera 5, and the digital image is transmitted to the industrial computer through the data transmission link, and then the digital image encapsulator embedded in the industrial computer 2 extracts the features of the feedback dial of the mechanical water meter, and finally obtains the real-time cumulative data of the water meter.
[0027] The digital image encapsulation module includes an independently written digital image algorithm. The specific implementation principle is: intelligently reading the pointer wheel combination type counting mechanism of the water meter dial. The structure of the water meter dial is as follows Figure 2As shown, its structure is generally composed of a one-bit star pointer a1, four-bit red pointer a2 and a five-bit word wheel a3. The star pointer a1 is used to indicate the initial flow, the four-bit red pointer a2 is arranged in order from left to right, indicating the cumulative amount of the corresponding unit division, and the five-bit word wheel directly displays the cumulative amount, with the unit being cubic meters. The digital image algorithm mainly completes the intelligent recognition of the water meter reading by identifying the reading pointed by the four-bit red pointer and the reading displayed by the five-bit word wheel, and accumulating the two. The identification of the four-bit red pointer: first, a normalized digital image operator is established, then the color region where the red pointer is located is determined through the linear relationship between the digital image operators, and finally the pixel point coordinates where the red pointer is located are directly grabbed in the RGB color space and manually assigned, completing the image segmentation of the four-bit pointer and the water meter dial. Secondly, the image is preprocessed through grayscale conversion and adaptive binaryzation, the preprocessed image is scanned horizontally and vertically, once the pixel point with abnormal background value is identified, its two-dimensional coordinates are marked, then the scanning is stopped and the next column or row is continued to be searched. The marked coordinates are reproduced in another image. After this processing, the boundary and coordinate data of the binary image can be obtained. The boundary coordinates obtained are analyzed by connected domain, the coordinates of the closed loop interval formed by mutual adhesion are classified, and the four pointers are classified into four categories. For the pointer coordinates of a single category, the smallest x-axis coordinate point x1, the smallest y-axis coordinate point y1, the largest x-axis coordinate point x2 and the largest y-axis coordinate point y2 are obtained, and then the center point coordinates of the pointer are: Then the distances between the edge points and the center point of the pointer are calculated, the maximum value is taken, and the edge point coordinates F edge (x, y) are obtained. Connecting the center point and the edge point is the direction of the pointer. The digital area pointed by the pointer is determined by the angle between the direction of the pointer and the vertical direction, and then a true value table of angle and number is established for matching. Then the product operation is performed according to the flow unit where the pointer is located to obtain the cumulative value A of the pointer. The identification of the five-bit word wheel: the rectangular frame in the dial is identified through Hough transform, and then the coordinates of the rectangular frame are obtained to complete the positioning of the image area where the word wheel is located. Since the word wheel display is composed of a string of numbers from 0 to 9, the reading of the word wheel is first processed through grayscale processing, adaptive binaryzation and other image preprocessing in the image area where the five-bit word wheel is located. Then the connected domain detection is performed on the pixel points, the independent digital regions are cropped one by one to obtain five separate numbers, and then the numbers are sorted according to the left and right coordinates and matched with the general number template prepared in advance to complete the identification. The cumulative amount B directly displayed by the five-bit word wheel is obtained. The actual cumulative amount displayed by the water meter is the cumulative value of A and B.
[0028] The computer data acquisition and processing module includes a pressure sensor installed in the standard device, a temperature sensor installed on the indicating flowmeter, a database, an experimental data processor 1 and a report template. The sensors are connected with the experimental data processor through a communication line, and can collect, arrange and calculate various test data. First, the computer data acquisition and processing module obtains the circulating water quality data collected during the test by converting the temperature and pressure sensor data. Second, the digital image recognition module outputs the mechanical water meter cumulative value data through the algorithm interface. Both of them are input into the experimental data processor 1, and the water temperature, density, pressure and other environmental parameters during the test are comprehensively considered to finally process and output the indication error and repeatability of the water meter calibration, and generate the water meter calibration original record.
[0029] The water meter calibration system realizes the connection of each sub-module through the data transmission line, realizes the command sending and data acquisition between the pressure stabilizing tank 9, the flow regulating valve 4, the standard device 6 and the indicating flowmeter 7. After the water meter 3, the constant water flow supply module, the flow regulating valve 4, the indicating flowmeter 7 and the standard device reach the working state, the industrial computer issues a calibration instruction to adjust the medium flow to the required flow point. The standard device obtains the real-time standard quality through the pressure sensor therein.
[0030] Further, the digital image recognition module captures the digital image of the water meter dial, inputs the digital image packager, and obtains the real-time reading of the water meter dial after calculation. This process is realized through the independent development of digital image algorithm. Specifically, by combining C++ and OPENCV, the obtained digital image is subjected to red pointer extraction, gray scale processing, binary processing, edge contour extraction and template recognition matching in sequence, and the real-time mechanical cumulative reading of the water meter dial is obtained.
[0031] Further, the camera 5 is connected with the industrial computer 2, and is responsible for capturing the digital image of the mechanical dial of the water meter 3 before and after the start of the calibration test. The captured image is converted into data information by the digital image recognition module, and is sent to the computer data acquisition and processing module through the algorithm interface. The computer data acquisition and processing module obtains the real-time mechanical reading of the water meter dial before and after the test, substitutes it into the volume algorithm formula to calculate the actual flow accumulation of the water meter during the test, compares it with the volume accumulation of the standard device during the same test time, and substitutes it into the error formula to calculate the indication error of the water meter during the test. Through repeated tests at different detection flow points, complete experimental data of the water meter calibration are obtained, the water meter calibration original record is formed, the water meter calibration is completed, and the water meter calibration original record is printed.
Claims
1. A water meter calibration system based on digital images, characterized in that... include: The system includes a water meter calibration and installation platform module, a constant water flow supply module, a digital image recognition module, and a computer data acquisition and processing module. The water meter calibration and installation platform module is responsible for connecting three water meters of various diameters and specifications in series; the constant water flow supply module is the water supply source for the entire system and is connected to the water meter calibration and installation platform fixing module through a mechanical structure; the digital image recognition module uses a camera placed on top of the water meter calibration and installation platform module to collect digital images, and calls a digital image processing algorithm to process the collected digital images to obtain real-time reading data of the water meters; the computer data acquisition and processing module obtains output data through the algorithm interface of the digital image recognition module, calculates the corresponding data, obtains the error and repeatability results of the water meter calibration, and generates an electronic calibration original record; The water meter calibration installation platform module uses a mechanical telescopic transmission meter clamping structure to control the installation distance between each water meter, ensuring seamless connection of the water meters being tested, and completing the series connection of three water meters of different specifications and diameters through water pipes of different diameters. The constant water flow supply module includes a water pump, a pressure stabilizing tank, a flow switch valve, a flow regulating valve, and an indicating flow meter. The water pump draws circulating water from the underground water tank into the pressure stabilizing tank for pressure regulation. Then, the flow switch valve is opened to input the pressure-stabilized circulating water into the water meter calibration installation platform module. Through the sealed mechanical water meter, the water flows into the flow regulating valve to regulate the flow rate of the circulating water. The circulating water flow rate is obtained through the indicating flow meter. The circulating water with stable pressure and fixed flow rate enters the standard device through the mechanical connection pipeline, completing one cycle of constant water flow. The digital image recognition module includes a camera, a data transmission link, and a digital image encapsulator. The digital image recognition module captures digital images of the dial of the mechanical water meter through the camera. The digital images are transmitted to the industrial control computer through the data transmission link. The digital image encapsulator embedded in the industrial control computer then extracts the features of the mechanical water meter dial to obtain real-time water meter cumulative data. The computer data acquisition and processing module includes a pressure sensor installed in the standard, a temperature sensor installed on the flow indicator, a database, an experimental data processor, and report templates. The temperature and pressure sensors are connected to the experimental data processor via communication lines to collect, organize, and calculate the pressure and medium temperature during various tests. First, the computer data acquisition and processing module converts the temperature and pressure sensor data to obtain the circulating water quality data collected during the test. Second, it obtains the mechanical water meter cumulative value data output by the digital image recognition module through the algorithm interface. Finally, both are input into the experimental data processor to generate a Word version of the water meter calibration original record. The computer data acquisition and processing module obtains the real-time mechanical readings of the water meter dial at two points before and after the test, substitutes them into the volumetric algorithm formula to calculate the actual cumulative flow of the water meter during the test time, and then compares it with the cumulative volume of the standard instrument during the same test time, i.e., the standard value, and substitutes it into the error formula to calculate the water meter reading error of this test; through multiple repeated tests at different detection flow points, complete experimental data for water meter verification is obtained, forming the original record of water meter verification, completing the water meter verification, and printing the original record of water meter verification; The digital image encapsulation module includes an independently developed digital image algorithm that intelligently reads the water meter's dial pointer-wheel combination counting mechanism. The digital image algorithm identifies the readings pointed to by the four red pointers and the readings displayed by the five-digit wheel, and adds the two together to complete the intelligent recognition of the water meter reading. The specific process for identifying the direction pointed to by the four red pointers is as follows: First, a normalized digital image operator is established. The color region where the red pointer is located is determined by the linear relationship between the digital image operators. The pixel coordinates where the red pointer is located are directly captured by the RGB color space and manually assigned to complete the image segmentation of the four pointers and the water meter dial. Secondly, grayscale conversion and adaptive binarization image preprocessing are performed. The preprocessed image is scanned horizontally and vertically to identify pixels with abnormal values compared to the background and mark their two-dimensional coordinates. Then the process stops and continues to search in the next column or row. The marked coordinate pixels are reproduced in another image to obtain the boundary and coordinate data of the binary image; connected component analysis is performed on the obtained boundary coordinates, and the coordinates that are connected to form closed loop intervals are classified. The four pointers are classified into four categories in total. To find the smallest x-axis coordinate x1, the smallest y-axis coordinate y1, the largest x-axis coordinate x2, and the largest y-axis coordinate y2 for a single type of pointer, the coordinates of the pointer center point are F_center((x1+x2) / 2,(y1+y2) / 2). Then, calculate the distance between each edge point of the pointer and the center point, take the maximum value, and obtain the coordinates of the edge point: F_edge(x,y). Connecting the center point and the edge point gives the pointer's direction. Finally, the digital region pointed to by the pointer is determined by the angle between the pointer and the vertical direction. Then, a truth table of angle and number is established for matching, and the pointer accumulation value A is obtained by multiplying according to the flow unit where the pointer is located. The specific process for reading the five-digit character wheel is as follows: The Hough transform is used to identify rectangular boxes in the dial, and the coordinates of the rectangular boxes are obtained to locate the image area where the dial wheel is located. The reading of the character wheel is first processed by grayscale and adaptive binarization of the image area where the five-digit character wheel is located. Then, connected component detection is performed on the pixels. The independent digit regions are cropped one by one to obtain five individual digits. The digits are then sorted according to their coordinates and matched one by one with a pre-prepared general digit template to complete the reading. The cumulative amount B of the five-digit character wheel is then directly displayed.
2. The water meter verification system based on digital images according to claim 1, characterized in that: The indicating flow meter includes three electromagnetic flow meters with different diameters and flow ranges, which promptly indicate the real-time flow rate of water passing through the mechanical water meter.
3. The water meter verification system based on digital images according to claim 1, characterized in that: A standard is a water flow collection device that obtains the cumulative water flow quality over a cumulative working time.
4. The water meter verification system based on digital images according to claim 1, characterized in that: The water meter calibration system described above enables the transmission of commands and data acquisition between the pressure stabilizing tank, flow regulating valve, standard device, and indicating flow meter via a data transmission line. After the water meter, constant water flow supply module, flow regulating valve, indicating flow meter and standard reach the working state, the industrial control computer issues a verification command to adjust the medium flow to the required flow point; the standard obtains the real-time standard quality through the pressure sensor in it.
5. The water meter verification system based on digital images according to claim 1, characterized in that: The experimental data processor integrates the calculation method for the indication error of water meter verification, stores the original record template of water meter verification in EXCEL format, calculates the experimental data and puts it into the original record template of water meter verification, and then processes and outputs the indication error and repeatability of water meter verification by taking into account the environmental parameters of the comprehensive test process.
6. The water meter verification system based on digital images according to claim 1, characterized in that: The digital image recognition module acquires digital images of the water meter dial, inputs them into a digital image encapsulator, and obtains the real-time reading of the water meter dial after calculation. This process is implemented through digital image algorithms, specifically by combining C++ and OpenCV. The acquired digital image is subjected to red pointer extraction, grayscale processing, binarization processing, edge contour extraction, and template recognition and matching to obtain the real-time mechanical cumulative reading of the water meter dial.
Citation Information
Patent Citations
Full-automatic water meter production inspecting method
CN101408452A
Dial generation method, terminal device and computer readable storage medium
WO2022078054A1