作者: Chung-Lin Huang , Ming-Shan Wu
DOI: 10.1109/ICSMC.1996.569746
关键词:
摘要: In this paper, a hand gesture recognition system is introduced to recognize the gestures of sequential digital number in complex background. We present method which uses flexible models describe shapes and its variability. First, we training set by manually labeled points located on boundary objects. Next, align all for examining statistics coordinates model over set. This can only fit new examples are similar corresponding From extracted features, judge state it belong to, entire continuous sequence. experiments, illustrate that both single number.