Skip to Main Content
Presents an on-line bilingual handwritten character recognition to classify both Thai and English characters using distinctive feature extraction. In this paper, we analytically derive distinctive features for Thai-English language classification. Decision tree diagram based on distinctive features derived is then used in practical applications. In addition, a language classifier has been used as the front-end recognizer in order to improve the performance of the recognition in terms of complexity reduction and recognition accuracy. From the experimental results, our implemented system can recognize an input as Thai or English character with an accuracy of 86.34% and 95.42% respectively. These numbers can be translated into an overall language classification accuracy of 90.21%.