I. INTRODUCTION
With the development of society, the power departments
take more attention on its service quality and economic benefits.
On the basis of power system modernization, manual meter
reading and a high degree of automation of unattended are in
stark contrast, which arrest the modern progress of electricity
supply system. There are two ways to improve the degree of
automation of ammeter data acquisition. First, new automated
ammeter reading meter is used to replace old ammeter. Second
is the design of intelligent terminals with higher levels of handheld
meter reading. A method that makes use of image
processing technology for ammeter information acquisition is a
new ways proposed in recent years [1]. There is an important
information in ammeter image that is barcode, it represent the
user of the ammeter. When we record the ammeter, we should
know its barcode number first. So barcode image recognition in
the image acquisition system occupies an important position.