An Automatic Calibration System Of Multiple Pointer Water Meter Based On Image Processing Technology

In traditional water meter calibration method, measurement results need to be read by people and water meter error should be calculated by people too. This manual calibration method has the low test efficiency and big work intensity. The test results are highly vulnerable to human factors. In order to overcome all the disadvantages brought by the artificial water meter calibration process, a multi pointer meter automatic calibration system based on image processing technology is designed in this paper. The machine vision technology and image processing technology are combined to use in the system. It instead of people to read and process the meter data and the water meter calibration is completed by it. It can save a lot of labor and increase productivity. So it can reduce the impact caused by human factors.The meter calibration systems are designed on VB.NET development platform. Through testing and analysis of this calibration system, it can prove that the system designed in the paper is feasible. It can achieve good calibration results. The main study of the paper is summarized as follows:1、Image reading module. The CMOS camera is connected to the computer with a USB cable. The meter image is photoed by the built-in dynamic link library of the camera. A particular meter image is acquired at a particular time. The image is converted to the RGB format image and saved to the specified folder. The corresponding parameters of the camera can be set for the actual needs.2、Image processing module. The acquired particular meter image is processed by Matlab. The image in RGB format is converted to the YCbCr format color space. Then the grayscale image is get by the grey level transformation. The grayscale image is denoised by erosion and dilation. Then binarization processing is done on the denoising image. Canny operator is used to extract the pointer edge of the binarization image. In order to coarsen the edge of the water meter pointer, the inflation is done on the edge extraction image. Hough transform is used to determine the only center coordinates on the every coarsening pointer edge of water meter image. Then it identifies the meter pointer tip coordinates. The actual meter pointer deflection angle and the corresponding value is identified. The water meter numerical reading is corrected by the corresponding relation of dial and the final meter reading is defined.3、Data processing module. The water meter calibration database is established through SQL Server. The obtained meter calibration data are inserted into the corresponding data column in the database. The meter eligibility is judged through meter error calculation. If the error is in the 1.5% and 1.5% respectively, the meter is qualified, otherwise it is unqualified. In order to inquire the data, the system should extract the meter calibration data from the database. Excel report is printed to facilitate the future data review and processing work. It can also clear the database according to the requirements of the test.4、Dynamic link library connection module. In this paper, VB.NET is used to establish development platform of calibration system. VB.NET calls for the camera MVCAMSDK.DLL to do secondary development of the application. It realizes the corresponding camera settings and the image of intake and store operation. VB.NET calls for the Matlab dynamic link library made in the paper. It realizes the meter image processing operations to determine the correlation meter readings. Through VB.NET and SQL Server connection, the corresponding meter image data processing operations are achieved.