品質工学
Online ISSN : 2189-9320
Print ISSN : 2189-633X
ISSN-L : 2189-633X
25 巻, 6 号
選択された号の論文の4件中1~4を表示しています
解説
事例研究
  • 五味 伸之, 北崎 友哉, 根本 真孝, 矢野 耕也
    2017 年 25 巻 6 号 p. 33-40
    発行日: 2017/12/01
    公開日: 2019/03/20
    ジャーナル フリー

    The present study was carried out with the aim of constructing a signature recognition system using the MT system. A pen tablet was used to configure a signature data input system, the data obtained were used to construct the signature recognition system, and the recognition system was then analyzed to find whether static data or dynamic data, which have different sampling methods, would be more advantageous for signature recognition. The analysis was carried out by the N1 method, i.e., with a single unit space, by comparing results from the use of dynamic data with results from the use of static data. As a result of the analysis, it became possible to reduce the number of false recognitions by using dynamic data and employing a weighting method to adjust the influence of the various statistical quantities.

feedback
Top