Research On Automatic Recognition System Of Glass Thermometer Based On Machine Vision

The glass liquid thermometer is widely used in various fields because of its high temperature measurement accuracy,easy to carry and so on.Glass thermometers need regular verification,the local department in charge of the metrological verification work,leading to the Metrology Department tasks,through previous studies about the measurement department,found that most of the measuring department for verification of glass thermometer readings by artificial way.Artificial reading way leads to the following results: the number of the thermometer calibrating personnel fatigue caused verification result is not accurate;the subjective factors influence the thermometer verification personnel to test results is not accurate;not timely storage data leads to arbitrary thermometer verification speed.Therefore,it is necessary to study the automatic calibration of the glass liquid thermometer.With the development of science and technology,especially the development of intelligence and automation,machine vision has been widely used by many researchers.Based on the machine vision,the use of thermostatic bath to maintain the temperature stability of the thermometer,the thermometer for images obtained by a certain resolution of industrial cameras,an important part of the system is analyzed,and the feature of target extraction by image processing technique,so as to further study.The liquid column,the scale line and the number of the thermometer are extracted from the image of the thermometer.The main contents of this paper include:The thermometer automatic recognition system based on machine vision is analyzed and studied,the processing of the collected images and the ways to achieve constant temperature device of the system,analyses the related parameters of industrial camera,and the camera calibration of Zhang Zhengyou industrial calibration,interference makes the collected images without distortion.In accordance with the previous design of the automatic identification system for the temperature of the field simulation experiment,adjust the constant temperature water bath device set different temperatures,the temperature is stable after the use of industrial cameras to capture images.By using the MATLAB software platform for image acquisition to processing,processing procedures include: image noise reduction to remove noise generated in the process of acquiring the image;the image contrast stretching to highlight the image of the thermometer;segmentation,image feature extraction to extract the thermometer liquid column and the scale line image;in order to improve the recognition accuracy of the tilt correction thermometer;and the thermometer on the number of segmentation and recognition,template matching.Through the image processing to calculate the number of thermometer,and the use of Visual Studio combined with HALCON machine vision software to design the corresponding user interface to display the image and the results of information.