Chinese Finger Alphabet Flow Recognition System Based on Multi-Features and Multi-Classifier
-
摘要: 手指语是用手指指式进行交流,一个指式代表一个汉语拼音字母,按照汉语拼音方案 拼成普通话.文中提出了一种基于多特征多分类器的汉语手指语识别方法,并利用该方法建造 了手指字母流识别系统.实验表明,该方法的识别效率明显优于基于单分类器的识别方法.Abstract: Finger-spelling language is the language that uses finger patterns to communicate. One finger pattern stands for one Chinese spelling alphabet, which is spelled into mandarin according to Chinese spelling scheme. In this paper a learning and recognition algorithm of Chinese finger alphabet flow based on multi-features and multi-classifiers is proposed, and a Chinese finger alphabet flow recognition system is built on the algorithm. Experiment shows that the system apparently outgoes the recognition system based on a single classifier.
计量
- 文章访问数: 2716
- HTML全文浏览量: 73
- PDF下载量: 1013
- 被引次数: 0