An automatic interpretation method for LCD images of digital measuring instruments

A new automatic interpretation method for the Liquid Crystal Display (LCD) images of digital measuring instruments is proposed in this paper. The original image is down sampled to extract the edge of the LCD, which is utilized to locate the LCD region including the characters automatically. Subsequently, a group of multi-scale convolution masks are applied to deal with the LCD area in order to extract strokes of the characters. Finally, the characters are recognized according to the structural features, which are generated by using the lines to pass through the strokes of the characters. The experiment results prove the high robustness, high recognition accuracy and high running speed of the proposed method, which can meet the requirements in calibrating digital measuring instruments with LCD.