| To reduce mercury pollution, the United Nations adopted the "Minamata Convention" international conventions in January 2013. Because of mercury pollution,mercury glass thermometers are being phased out gradually, and electronic thermometers are used widely. To guarantee the accuracy of electronic thermometer temperature characteristics, production companies need to verify electronic thermometer before they are sold. However, manufacturers mainly rely on manual for verification, which has low efficiency and high labor costs.At present, China Jiliang University has developed a set of single-station electronic thermometer verification device, which can verify multivessel electronic thermometers in a once, but when verification at multiple temperature points in the single-station, it requires a lot of time for constant temperature water bath to reach each verification point and keep temperature stable. And it can not complete the batch electronic thermometer verification efficiently. In this paper, according to medical electronic thermometer verification specification, through theoretical research, design and develop a set of multi-position rod type electronic thermometers automatic verification system to meet the manufacturer’s verification requirements. The research work of this article is as follows.(1) Design a multi-station electronic thermometer automatic verification equipment device. Each station corresponds to a verification point of the bath temperature to avoid insufficient of single-station on multi-point verification.Verification platform is designed as circular rotating structure in order to achieve the transform from single station to multi-station, and each station installs electric pusher device, by controlling pusher device to achieve electronic thermometer clamp frame automatically stretch and shrink from constant temperature water bath. Build a machine vision system to do readings recognition automatically.(2) Design motion control system for multi-point verification control. By controlling pusher device to achieve electronic thermometer clamp frame automatically stretch and shrink from constant temperature water bath; By controllingindustrial cameras to get electronic thermometer image; By controlling verification equipment platforms to move electronic thermometer clamp frame from one station to another position.(3) Put forward K-nearest neighbor method to recognize readings of multivessel electronic thermometers. The process is as follows: Do gray-scale transformation of the image; Use Sobel operator to do edge detection; Get reading area based on feature extraction; Use Hough transform to do tilt correction; Numeric character segmentation; Character normalization process; Use K-nearest neighbor method to recognize readings.Finally, do performance testing on the bath which is used on the verification, and test the readings recognition rate of multivessel electronic thermometers. Summarized in this article, then propose the problems and further improvement. |
