Human eyes see the same object from slightly different viewpoints. Unless the object is placed at infinity, our eyes have to verge towards each other in order to fixate on the same object. Such vergence eye movements are also coupled to appropriate accommodation changes of our eyes in order to focus on the object. Unfortunately, this coupling is disturbed when viewing current binocular displays. On binocular displays, images with correct stereoscopic depth cues can be presented to induce appropriate vergence eye movements. However, the lens focus of such displays is typically fixed regardless of changing depth cues. This posts an un-natural demand on the viewer’s eyes since the viewing condition demands vergence movement in the absence of appropriate accommodation changes. This thesis s...[ Read more ]

Human eyes see the same object from slightly different viewpoints. Unless the object is placed at infinity, our eyes have to verge towards each other in order to fixate on the same object. Such vergence eye movements are also coupled to appropriate accommodation changes of our eyes in order to focus on the object. Unfortunately, this coupling is disturbed when viewing current binocular displays. On binocular displays, images with correct stereoscopic depth cues can be presented to induce appropriate vergence eye movements. However, the lens focus of such displays is typically fixed regardless of changing depth cues. This posts an un-natural demand on the viewer’s eyes since the viewing condition demands vergence movement in the absence of appropriate accommodation changes. This thesis studies, for the first time, the effects of matching lens focus with stereoscopic depth cues on the time taken to form a single stereoscopic image when viewing a binocular micro-display. A micro-display system with dynamically adjustable lens focus was designed and prototyped for this study. The design and prototyping work forms part of the contribution of the thesis. An experiment was then conducted to study the time taken to merge a pair of left and right binocular images into a single stereoscopic image under four viewing conditions that exhaust the combinations of two lens focus (40cm and 200cm) and two object depths (40cm and 200cm).

Results indicate that viewers took significantly shorter time to form a single overlaid stereoscopic image when the lens focus matched with the object depth (p<0.05, paired t test). Further examinations of the data suggest that unnatural demand for the eyes to diverge (e.g., lens focus was smaller than the object depth) are associated with significantly longer periods of double images.

Among the ten viewers, seven consistently and significantly benefited from the effect of matching lens focus with stereoscopic depth cues but three viewers did not. The relationships between individuals’ performance in the study and their visual parameters have been studied. Significant correlations are found between the time taken to form a single stereoscopic image and the negative fusional reserve of the participants. The work of this thesis evaluates the potential benefits of applying dynamically adjustable lens focus to the future design of binocular micro-displays.