| Project/Area Number |
16330141
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Experimental psychology
|
|---|
| Research Institution | Kyoto Institute of Technology |
|---|
Principal Investigator |
OOTANI Yoshio Kyoto Institute of Technology, Graduate School of Science and Technology, Professor, 工芸科学研究科, 教授 (00192518)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
VAN TONDER G, J Kyoto Institute of Technology, Graduate School of Science and Technology, Associate Professor, 工芸科学研究科, 助教授 (30362586)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥15,100,000 (Direct Cost: ¥15,100,000)
Fiscal Year 2006: ¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 2005: ¥5,200,000 (Direct Cost: ¥5,200,000)
Fiscal Year 2004: ¥7,900,000 (Direct Cost: ¥7,900,000)
|
|---|
| Keywords | Pattern recognition / Medial axis / Fractal / Masking / Eye movement / Analysis of visual structure / 配置 / 方位 |
|---|
| Research Abstract |
Pattern recognition process in the human visual system is composed of the sub-processes which segment input image into 'meaningful parts', and then integrate the parts into a unified perceptual 'whole'. It has been suggested that, to make the sub-processes operate effectively, the human visual system compresses the input retinal image thereby obtaining an efficient representation of the input image. To explore the processing strategy for the efficient representation, the present study focused the following three research themes:
(1) Psychophysical experiments aiming at the visualization (i.e. making perceptible) of the medial axis of the image
(2) Computational study for the medial-axis structure of the naturalistic landscape (e.g. Japanese gardens)
(3) Psychophysical analysis of the relation between the fractal property of the image, and eye-movement and visual preference.
With respect to the theme (1), we failed to obtain positive results in the masking experiments initially planned, but we got promising quantitative data from the 'dot-placement' experiments introduced in the latter half of the research term, which suggest that medial-axis representation does exists in the human visual system.
With respect to the theme (2), we elucidate the visual structures of many highly-artistic Japanese gardens by using a computational model of medial-axis transformation.
With respect to the theme (3), we constructed an experimental system composed of an eye-movement tracking device and image-presentation device, and analyzed the relation between eye-movement/pupil-size and visual preference.
The results obtained in the present research remain preliminary, but it is expected that they will be useful in promoting further our knowledge on the human pattern recognition process.
|
