Summary
International Technical Conference on Circuits/Systems, Computers and Communications
2008
Session Number:C3
Session:
Number:C3-4
Morphometrics for Shape Analysis in Kansei Engineering
Shigekazu Ishihara, Keiko Ishihara,
pp.-
Publication Date:2008/7/7
Online ISSN:2188-5079
Summary:
In Kansei engineering, we have been treated sample shapes as categorical variable (nominal scale). For example, categories like wide / tall. These categories were assigned as x variables and evaluation values on a Kansei word was assigned as a y variable of linear equation. This equation has been computationally solved by Quantification theory type 1 or similar regression methods. Although qualitative analysis of shapes is relatively robust and commonly used, but shapes are not directly treated. In this study, we attempted to treat shapes as statistical values with the various methods of Morphometrics those have been developed between paleontology, biology and statistics. By treating shapes as statistical values, we can apply various statistical methods from basic statistics such as testing distribution to multivariate analysis techniques (i.e., classification, projection onto lower numbers of dimension).