Return to top English Japanese
Kansei and image features

This research is based on an initiative by Dr. Chie Muraki-Asano of Yasuda Women's University, and collaborated by Professor Fujimoto and her laboratory staff of Hokkaido University of Education in charge of experiment works, Dr. C. Muraki-Asano in charge of statistical analysis, and Dr. Akira Asano, Hiroshima University, in charge of image processing. The following text is written by Dr. C. Muraki-Asano.

Human kansei, which is a widely used Japanese word for sensibility or emotion, for "colors" has been actively investigated in the areas of color psycology, etc. However, it seems to be treated trivial and its factors have not been investigated that the human impressions from different materials are different even if the color is the same. For example, human impressions of metallic materials and of soft materials like cloths are completely different even if their color is commonly black. Metallic materials cause an impression of coldness, and soft ones cause that of warmness. It is suggested that this difference of impressions are caused by the human perceptivity that is the integration of tactual and visual senses.

This is the same even if we restrict ourselves to cloths only. The quality of a high-quality cloth is recognized only by visual sense without touching it, since the visual sense is related to the experience of touching high-quality cloths in the human perceptivity. This suggests that the information of visual sense contains information about tactual feels of materials. We recently showed the relationship between physical properties as well as human impressions of black cloths for formal wears and their image features. Extending this research, we aim to describe feels of materials by image features./p>

We currently extend our researches on kanseito clothes, and investigate kansei to synthesized textures based on our research about the morphological texture modification. Since this method can generate a new texture synthesized by mixing features of different textures, we can investigate what feature is effective to kansei by measuring human kansei to the generated textures.

Selected publications

English Japanese