Bulletin of the Graduate School of Education, Hiroshima University. Part. II, Arts and science education Issue 54
2006-03-28 発行

布の手触り感とニューロネット

Neural Networks and Fabric Touch
Kinoshita, Mizuho
file
abstract
By the hand sensory method Shin-gosen fabrics were classified into four groups (New Silky, Peach Skin, Rayon Dry and New Worsted). Neural networks on a personal computer were build by trainnng with the results of mechanical measurements. The results of sensory test were compared with the output of neural networks. The comparison were good in the case of Rayon Dry and Peach Skin type samples. In the case of New Silky and New Worsted type, the results of sensory test were rather dispersed and the dispersion patterns were fairly similar to those of the output figures of the networks. So we are able to expect neural networks to be effective tools to analyze the relation betweeen fabric touch and mechanical properties.
subject
手触り
官能検査
ニューロネット
新合繊
摩擦特性
Fablic Touch
Hand Sensory Evaluation
Neural Networks
Shin-gosen
Friction
SelfDOI