TOFカメラによる3次元情報を用いた手話認識に関する検討 (福祉情報工学)  [in Japanese] A Study on Sign Language Recognition Using 3-D Information with TOF Camera  [in Japanese]

Search this Article

Author(s)

    • 森口 拓哉 MORIGUCHI Takuya
    • 名古屋工業大学工学部情報工学科 Department of Computer Science and Engineering Graduate School of Engineering Nagoya Institute of Technology
    • 酒向 慎司 SAKO Shinji
    • 名古屋工業大学工学部情報工学科 Department of Computer Science and Engineering Graduate School of Engineering Nagoya Institute of Technology
    • 北村 正 KITAMURA Tadashi
    • 名古屋工業大学工学部情報工学科 Department of Computer Science and Engineering Graduate School of Engineering Nagoya Institute of Technology

Abstract

手話の自動認識において,手話動作の抽出にはセンサを装着する方法や手話の動画像を用いる方法などがある.しかし,前者では身体的な拘束による話者への負担が,後者では画像処理の難しさから,手話の前後動作を正確に捉えることが困難となる.奥行き情報を容易に取得でき,なおかつ身体的な拘束のないものとして,TOF(Time-of-Flight)カメラを用いて手話動作を計測することが考えられる.TOFカメラによる手話認識の先行研究はそれほど多くなく,小規模な実験に留まっている.また,TOFカメラによる手話データも入手が困難な状況にある.本稿では,TOFカメラと一般のCCDカメラを併用した手話のデータベースを構築する.そして,このデータベースを用いて,これまで提案されてきたHMM(隠れマルコフモデル)に基づいた手話認識の枠組みに基づいて連続単語認識実験を試みた.

Feature extraction methods of hand movements for sign language, there are two approaches to use sensors and video. However, the sensor based approach would provide a physical burden to the signer. In the video based approach, it is difficult to accurately capture the 3-D information due to the difficulty of processing image. TOF(Time-of-Flight) camera can be easily obtained 3-D information without physical burden to the signer. But, there are few previous studies using the TOF camera. In addition, experiments have remained small. It is also difficult to obtain the sign data by the TOF camera. In this paper, we build a database of sign language combined with TOF camera and general CCD camera. Then, using this database and tried to experiment a continuous word recognition. The recognition method was used for the framework of HMM(Hidden Marcov Model) have proposed so far in sign language recognition.

Journal

  • IEICE technical report. Welfare Information technology

    IEICE technical report. Welfare Information technology 111(424), 1-6, 2012-01-27

    The Institute of Electronics, Information and Communication Engineers

Codes

  • NII Article ID (NAID)
    110009481696
  • NII NACSIS-CAT ID (NCID)
    AA1155230X
  • Text Lang
    JPN
  • ISSN
    0913-5685
  • NDL Article ID
    023424325
  • NDL Call No.
    Z16-940
  • Data Source
    NDL  NII-ELS 
Page Top