Container Code Recognition Using Isolated Character Recognition and Continuous Characters Recognition Combined Technique
FU Guang-rong · Journal of Sichuan University · 2011
Container code characters can be arranged in many ways,and the distances between neighbor characters in different images vary largely.Sometimes the code characters are closely arranged,and are difficult to be segmented apart.To address this problem,a new method for container code recognition,which combined both isolated character recognition technique and continuous character recognition technique,was proposed.Firstly,in the character segmentation phase,the characters arranged with some distance would be extracted as isolated characters,and the characters closely arranged would be extracted as a character region.Secondly,in the character recognition phase,the isolated characters were recognized by a RBF neutral network based isolated character recognition technique,and those closely arranged characters were recognized by an HMM based continuous characters recognition technique.The recognition rate of 1 040 container images reached 92.5%.The experiment results showed that the container code images under various situations can be effectively recognized by the present method.