Reading identification Algorithms of mercury thermometer based on machine vision

Based on machine vision techniques, the reading identification algorithms of mercury thermometer was presented. According to the Huffman transformation methods, three-time expansion operation combined with three-time erosion operation is used to locate the mercury, mercury pole and the scales. On the basis of performing spatial moment subpixel localization, the temperature value indicated by thermometer is calculated using location information between edge pixels. Experiment results show that these algorithms with accuracy of 0.1degC can meet the practical application.