Abstract :
In this paper, we present the ATM (Awesome Translation Machine), which translates handwriting texts in English into Chinese, and then provides its pronunciations in both the two languages. Specifically, two types of the databases that contain characters and sentences for training the ATM are constructed. Various signal processing techniques are employed sequentially for processing and analyzing the image raw data. After all the preparation stages, we apply multiple pattern recognition techniques, i.e., Principle component analysis, linear discriminant analysis, and support vector machines, for the purpose of character recognition. The identified characters are thereby automatically linked to their actual meanings stored. Extensive experiments are conducted to gauge the performance for different techniques.
Keywords :
handwritten character recognition; language translation; natural language processing; principal component analysis; support vector machines; text analysis; ATM; English language translation; awesome translation machine; character recognition; handwriting text translation; image raw data analysis; linear discriminant analysis; pattern recognition techniques; principle component analysis; signal processing techniques; support vector machines; Character recognition; Databases; Noise; Support vector machines; Training; Vectors; Writing; Character recognition; Image capture; Image processing; Pattern recognition methods;
