2019 Volume 58 Issue 3 Pages 340-344
Japanese Finger-spelled Sign Language is a visual language and this represents Japanese syllabary characters called “Hiragana”. To assist the communication with hearing-impaired people, many researchers have already been proposed automatic finger-spelled sign language recognition systems. In this paper, we introduce an example of that recognition system which includes a recognition method with hand shape information that is extracted from a hand region by using a RGB-D camera. In addition, we introduce a spotting technique with SVM and rule-based approaches and show a prototype of typing system with Japanese finger-spelled sign language by converting a Hiragana word into its Kanji representation.