|
|
|
Eye Motion Tracking
Figure 1. Diagram of Eye
The vast majority of people with impaired vision retain some functional vision to perform everyday tasks and are characterized as having low vision. The ability to track eye movement in individuals suffering from low vision may provide some insight into how some are better able to adapt to this disability than others. This project utilizes correlation algorithms to track the movement of the vessels on the retina as a subject moves his/her eyes while performing a visual task. The images of the retina are from a scanning laser ophthalmoscope (SLO). There are three kinds of movements that are tracked: horizontal, vertical and rotational shift. The visual task utilizes a 5x5 grid of symbols. One symbol at a time is lit, starting at the upper right, and progresses across each row till the lower left one is lit. Subjects are asked to focus on the symbol that is currently lit as the SLO images their retina
Figure 2. A typical image of a retina from an SLO scan. Critical points are selected at the beginning of the tracking. After performing edge detection, a correlation algorithm is applied to subsequent images to determine the shift in the patient's retina from one frame to the next. Below are shown results from the first part of a patient test.
Figure 3. Progression of Lit Symbols in SLO experiment
(c) Rotational Movement Figure 4. The first 2000 frames of a patient test. (Starting at upper left symbol, and progressing to second row, second column)
Figure 5. A screen capture of the software interface used to track eyeball motion. |
|
|