A New Fast Method for RMB Paper Currency Number Recognition
Yingming Hao · Microcomputer Information · 2008
In the engineering application,recognizing paper currency number from the skew RMB image requires high processing speed and good identification effect,so we utilize smart camera to establish paper currency number recognition system and present a new fast RMB number recognition method based on many features. To avoid rectifying the incline and save the locating time,this method uses the improved project method to locate the numbers. Then it recognizes the located numbers by features,like gap feature,the ratio of the character’s width to height,crossing feature,and so on. In a good lighting condition,we test 50 pieces of 100 Yuan denomination,whose skew angle is in the range of ±13. The results are that the time of recognizing one image (only including Ara-bic numerals) is not more than 4ms and recognition rate is 100%. Experimental results show that the method has a high speed,al-lows the image to have a wide skew angle and meets the engineering requires.