A Recognition Method of RMB Numbers Based on Character Features

According to the features of RMB banknote images, combining digital image processing and pattern recognition, this paper proposes a recognition method of RMB word numbers based on character features. The recognition algorithms include: image preprocessing, number region location, feature extraction and character recognition. The experimental results show that the automatic number identification system based on this method achieves high recognition rate and processing speed. Keywords-RMB numbers; character segmentation; feature extraction ; character recognition I. INTRODUC TION RMB is the legal currency of China, and every RMB banknote has a unique crown word number, serving as the identity of it. Through the establishment of the banknotes' number sequence file, the circulation of the banknotes in the financial market can be monitored and tracked, and crimes like currency counterfeiting, robbery, stealing can be effectively controlled. Therefore, it is of great importance and application prospect to do research on the identification of RMB word numbers. Recently, many banks have applied this technology on their ATMs, identifying and recording the numbers of the RMB banknotes into and out of ATMs. In this way, banks can prevent receiving counterfeits and customers who get counterfeits from the ATMs would also be compensated. This paper studies on the identifying process of RMB numbers and proposes a method based on character features. By optimizing several algorithms, it achieves high recognition rate.