Abstract
It is desired to enhance traditional mechanical water meters with automatic meter reading (AMR) in the era of Internet of Things (IoT), In this paper, we propose and implement an add-on module using optical character recognition (OCR) to achieve that. The proposed module will be attached to a conventional mechanical water meter, capture and translate the register readings into plain text, which can be sent wirelessly to a database server. In addition, we propose a novel light-weight optical character recognition algorithm and test it in a prototype with embedded digital system. Experimental results show that the recognition accuracy can be as high as 98.1%, which indicates a promising candidate technology for AMR to revolutionise water meters. Finally, this technique can be applied to other mechanical meters with analogue number displays, paving the way for ubiquitous IoT.
