A thermometer verification reading automatic identification system and method

Abstract

The invention discloses a thermometer verification reading automatic identification system and method. The system comprises a camera (1), a computer (2), a stepping motor (3) and a temperature controlgroove (4). The computer (2) is connected with the camera (1) and the stepping motor (3) through wires, and an image acquisition card and a motion control card are installed in the computer (2). Thestepping motor (3) is fixed at the top of the temperature control groove (4); A constant-temperature liquid environment is arranged in the temperature control groove (4) and used for providing the temperature of a verification point, a thermometer reading window (5) is formed in the surface of the temperature control groove (4), a hole is formed in the top end of the temperature control groove (4), and a thermometer clamp capable of being taken out independently is arranged inside the temperature control groove. The method comprises the steps of image display and acquisition, image format conversion, image grayscale conversion, image segmentation, image binaryzation, image morphological operation, image feature extraction and reading recognition. Human eyes are replaced with computer recognition software, automation of thermometer reading is achieved, and therefore the manual labor amount is reduced, and the reading precision is improved.

Description

technical field [0001] The invention belongs to the fields of thermometer measurement, image processing and identification, and in particular relates to a system and method for automatic identification of thermometer verification readings. Background technique [0002] Temperature is an important measurement parameter that people use in production and life. As an important tool for temperature measurement, liquid glass thermometers are widely used in many important fields such as industry, agriculture, environment, construction, and energy. The reading error of the thermometer itself will affect the quality of temperature measurement, and larger errors may cause irreparable losses. Therefore, when the thermometer has been used for a period of time, it needs to be verified to calibrate its reading error. [0003] At present, most of the methods used in the verification of thermometers are relatively traditional. The working process is as follows: according to the verificatio...

AI Technical Summary

Problems solved by technology

It can be seen from the verification process that the reading of the thermometer is realized by the operator's eye observation, which may lead to two problems: first, the accuracy of the reading of the thermometer depends on the skill level of the reader, and there is a certain degree of subjectivity in temperature interpretation The second is that when performing batch verification readings on multiple thermometers or multiple verification points, the amount of manual labor is large, and visual fatigue is prone to occur. These problems will all lead to a decline in the accuracy of thermometer verification readings

Generally speaking, these methods are relatively cumbersome to operate, and the process is not optimized enough; when the image quality of the original acquisition is not high, the reading recognition effect is not good; some steps are not highly automated, and the reading recognition results are affected by certain human subjective factors

[0005] In the article "Image Processing Technology in Automatic Verification of Liquid Thermometer" in the 6th issue of "Measurement and Testing Technology" in 2002 and "Image Processing Technology in Automatic Verification of Liquid Thermometer" in the 4th issue of "Measurement Technology" in 2003 "In the article, a fixed threshold is used in the binarization of the grayscale image, which has poor adaptability. When determining the threshold, the amount of manual labor is large, and the selection of the threshold is sometimes difficult to achieve the desired effect; in the feature extraction of the grayscale image of the thermometer The projection method used at the time did not consider the impact of the thermometer tilt on the final identification reading, and the reading accuracy and reliability were not high.

[0006] In the article "Design and Implementation of a Thermometer Automatic Verification System" in the 7th issue of "Measurement Technology" magazine in 2004, although the image processing and recognition technology was used to realize the automatic reading of the thermometer verification, the operation was cumbersome and the process was not enough Optimization, there are many human and subjective factors in the processing process, and the accuracy of feature extraction needs to be improved. For example, binarization is performed before the grayscale image of the thermometer is segmented, and then according to the results of binarization and median filtering. Segmentation of the original gray-scale image is cumbersome; the threshold used for binarization of the gray-scale image is selected based on experience. On the one hand, the selection of the threshold needs to be adjusted and verified repeatedly. It is also necessary to select and experiment the threshold again, and the operability is poor; the feature extraction of the segmented thermometer image is based on the grayscale image, and the projection method used cannot be independent and difficult to extract the thermometer accurately. Mercury column and scale line features, the feature extraction effect is greatly affected by image quality and noise, which in turn affects the final recognition reading results, and the reading reliability is not high

[0007] In the article "Image Segmentation Technology for Automatic Verification of Thermometers" in the 6th issue of "Ordnance Automation" in 2005, although an adaptive threshold method was used for the binarization of grayscale images, the amount of calculation was large; The feature extraction of the binarized image is carried out on the premise that the thermometer is completely in a vertical ideal state. In fact, this ideal state is difficult to guarantee. If there is a tilt, the reading accuracy will be reduced; the mercury column and the mercury column are not included in the feature extraction. The scale lines are separated independently, which will lead to inaccurate feature extraction of scale lines and mercury columns, resulting in poor reading reliability

[0008] The above problems have affected the application of the researched method in the actual verification of thermometers to a certain extent.

Although the automatic reading technology based on image processing and recognition can theoretically improve the reading accuracy of thermometers and the efficiency of verification, due to the imperfection of the method itself and the low reliability of readings, it has not been obtained in the verification of thermometer readings. to promote

Images