Skip to Main Content
Most methods of the tradition character reorganization are based on JPEG and the classifier and then matching with the templates. Due to the environmental of plate with large noise plate and interference factors, the accuracy of this method is relatively low. In this paper, based on the analysis of the traditional method of the character recognition, it uses wavelet processing for image pre-processing, and then carry out template matching under different multi-resolutions. Experiments show that this method can effectively solve the shortcomings of traditional methods to improve the accuracy of character recognition.
Date of Conference: 11-13 May 2010