| Project/Area Number |
12680826
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Biomedical engineering/Biological material science
|
|---|
| Research Institution | The University of Electro-Communications |
|---|
Principal Investigator |
ITAKURA Naoaki The University of Electro-Communications, Faculty of Electro-Communications, Associate Professor, 電気通信学部, 助教授 (30223069)
|
|---|
| Project Period (FY) |
2000 – 2002
|
| Project Status |
Completed (Fiscal Year 2002)
|
| Budget Amount *help |
¥2,800,000 (Direct Cost: ¥2,800,000)
Fiscal Year 2002: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 2001: ¥900,000 (Direct Cost: ¥900,000)
Fiscal Year 2000: ¥1,200,000 (Direct Cost: ¥1,200,000)
|
|---|
| Keywords | Interface / Evoked Potentials / Eye-gaze Input / Electro-oculogram / Discriminant Analysis / ニューラルネットワーク |
|---|
| Research Abstract |
For the purpose of developing an eye-gaze input interface with a head mounted display utilizing information of EEG signals, we have focused on the following three aims : (1) Whether the user gazes at the specification position or not is decided by the EEG. (2) The suitable visual stimuli to decide the eye-gaze point by the EEG are investigated. (3) The simple eye-gaze point measuring method is developed.
For the aim (1), we tried to distinguish between a correct EEG and an incorrect EEG. The correct EEG is an EEG measured when the background of the character actually gazed at changes to red. The incorrect EEG is an EEG measured when the background of the character next to the character actually gazed at changes to red. Consequently, we got the following results : 1) The rate of the distinction between the two EEGs does not improve with using only amplitude information at the single time point of the EEG signal. 2) The method of extracting the typical waveform with time reproducibility from the EEG signal is effective. 3) Unless removing the background EEG effectively, the typical waveform is not extracted. 4) Unless using a suitable stimulus, the typical waveform of the EEG itself may not occur.
For the aim (2) and (3), we got the following results : 1) The decision of eye-gaze point utilizing photic driving phenomenon that need optical blink stimulus for 10 seconds or more is not practical because the user's burden is large. 2) The decision of eye-gaze point with utilizing visual evoked potentials may be able to be carried out by using the optical blink stimulus for about 5 seconds and by removing the background EEG. 3) The algorithm that can decide the eye-gaze point from an AC amplified electro-oculogram signal was developed. 4) The eye-gaze point can be decided at a rate of about 90% in the horizontal direction. 5) The accuracy of the decision of eye-gaze point in the vertical direction falls according to the influence of eye blinking.
|
