2009 Fiscal Year Final Research Report
Development of automatic fatigued bill classification system based on acoustic signal information
Project/Area Number |
19560407
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
System engineering
|
Research Institution | Hiroshima Institute of Technology (2008-2009) Nara University of Education (2007) |
Principal Investigator |
TERANISHI Masaru Hiroshima Institute of Technology, 情報学部, 准教授 (50237004)
|
Project Period (FY) |
2007 – 2009
|
Keywords | 計算機システム / 信号処理 / 計測工学 / 自己組織化 / ソフトコンピューティング |
Research Abstract |
The study has developed and improved an automatic fatigued bill classification system for banking machines, based on acoustic signal information of passing bills. The classification resolution of fatigued bills is improved by introducing the fatigue level estimation method with the supervised self-organizing map. Furthermore, feature selection methods based on correlation information between fatigue levels and acoustic signal information of banking machines led the performance of the proposed method to practical level, by reducing classification time and improving classification accuracies of fatigue levels of bills.
|