| 摘要 |
<p>An object of the present invention is, when determining a line-of-sight vector of a subject, to identify a three-dimensional position of a pupil center or corneal reflection center thereof. The line-of-sight vector is determined by measuring a distance from a camera to the pupil center or corneal reflection center of the subject using a two-dimensional CMOS distance image sensor by a light time-of-flight method and obtaining position information of the pupil center and corneal reflection center of the subject using the distance image sensor. If a position and a direction of the camera are identified, a camera-pupil center vector is determined from image information of the pupil center of the subject. Since the camera-pupil center distance has been measured, a pupil center position in a three-dimensional space is determined from the above information. Combining this with two line-of-sight vectors determines a position where line-of-sight vectors intersect, or a view-point.</p> |
