||Time-segmentation and position-free recognition of air-drawn gestures and characters in videos
Time-segmentation and position-free recognition of air-drawn gestures and characters in videos
新妻, 佑記 ,
鳥居, 駿平 ,
矢口, 勇一岡, 嶐一
Multimedia tools and applications = Multimedia tools and applications
11639 , 2016-10 , Springer
We report the recognition in video streams of isolated alphabeticcharacters and connected cursive textual characters, such as alphabetic, hi-ragana and kanji characters, that are drawn in the air. This topic involves anumber of difficult problems in computer vision, such as the segmentation andrecognition of complex motion on videos. We use an algorithm called time-space continuous dynamic programming (TSCDP), which can realize bothtime- and location-free (spotting) recognition. Spotting means that the priorsegmentation of input video is not required. Each reference (model) characteris represented by a single stroke that is composed of pixels. We conductedtwo experiments involving the recognition of 26 isolated alphabetic charactersand 23 Japanese hiragana and kanji air-drawn characters. We also conductedgesture recognition experiments based on TSCDP, which showed that TSCDPwas free from many of the restrictions imposed by conventional methods.