The recognition of RMB serial number based on CNN

A method of using the two levels convolution neural network (CNN) is proposed to recognize the RMB serial number (SN), it is a more efficient, more high-speed, more precision way than before, and it based on the CNN which can directly extract characters from the training sample. And in the process character segmentation, the window moving method is proposed for the contaminated image. Through the DSP platform, it can effectively reach the robustness of the recognition algorithm under some circumstances such as the damaged, smirched identified object, the recognition accuracy can reach at least 99.99% and the recognition time can be controlled within 5ms.