| 摘要 |
PROBLEM TO BE SOLVED: To correctly hear out the signals sent from a sound source that is noticed by a viewer by detecting the line of sight of the viewer shown on a monitor screen and giving the weighting to the voices after selecting the voice signals sent from the sound source positioned in the direction of the line of sight of the viewer. SOLUTION: A specific part that is noticed by a viewer 13 is detected on a display screen which is projected on a monitor device 7. For instance, a compact video camera 11 is fixed at a position near the device 7 where an image is projected and then photographs the face of the viewer 13. Then the eye areas are detected out of an entire face image and then the eyelid and eyeball areas are separated from the eye areas. A line-of-sight detection circuit 12 detects the direction of the line of sight of the viewer 13 based on the eyelid shape changes and the position relation changes of eyeballs. Then the position that is caught by the line of sight is specified on the display screen. If a sound source is included in the specified part, the voice signal generated by the sound source is specified. The weight of this specified signal is increased and the weight of other signals are reduced. Then the synthetic stereo signals are outputted. |
