In vision sensation and perception are rarely equivalent. Most notably, humans perceive the world with three-dimensional depth in a manner consistent with having a single, cyclopean eye centered above the nose even though the inputs to the visual system are a pair of binocular inputs each of which is two-dimensional. In this dissertation we address three aspects of cyclopean perception and how each is influenced by its underlying binocular inputs. In Chapter 1 we study how viewing distance affects retinal disparity and how the visual system is able to maintain a depth percept that is invariant to these changes. In Chapter 2 we examine how binocular disparity and monocular luminance cues are integrated to perceive edges. Finally, in Chapter 3 we consider how monocular images from the left and right eyes are combined to optimize the cyclopean image. Together these studies provide insights into how the visual system interprets retinal inputs and transforms them into a single perception of the world.

Three scene properties determine the image of a 3D object: the material reflectance, the illumination, and the object's shape. Because all three properties determine the image, one cannot solve for any one property without knowing the other two. Nevertheless, people often are able to perceive these properties consistently and relatively accurately. We explore the relationship between these properties, the sources of image information the visual system can use to recover these properties, and the assumptions the visual system tends to make. We first conducted a shape perception experiment in which we investigate whether the visual system assumes the angle between the lighting direction and the viewing direction. Observer errors were minimized when the light was 20-30 degrees above the viewing direction, confirming the light-from-above prior. In a second study, we conducted two psychophysical experiments to determine how viewers use shape information to estimate the lighting direction from shaded images. We found that observers can accurately determine lighting direction when a host of shape cues specify the objects. When shading the is the only cue, observers always set lighting direction to be from above. We modeled the results in a Bayesian framework that include a prior distribution describing the assumed lighting direction. Finally, we explore how disparity and defocus information may be useful in material perception to distinguish glossy and matte surfaces. We describe the types of images needed to investigate this question, and introduce a method to render them.

Determining the three-dimensional (3D) layout of the natural environment is an important but difficult visual function. It is difficult because the 3D layout of a scene and the shapes of 3D objects are not explicit in the 2D retinal images. The visual system uses both binocular (two eyes) and monocular (one eye) cues to determine 3D layout from these images. The same cues that are used to determine the layout of real scenes are also present in photographs and realistic pictures of 3D scenes. The work described in this dissertation examines how people perceive 3D layout in the natural environment and in pictures. Chapter 1 provides an introduction to the specific 3D cues that will be studied. Chapter 2 describes a series of experiments examining how the visual system's use of binocular cues might be adaptive to improve depth perception in the natural environment. Chapter 3 describes a series of experiments examining how the visual system interprets monocular linear perspective cues, and how this can lead to perceptual distortions in perceived 3D from pictures. Chapter 4, the Conclusion, discusses how the work described in this dissertation might be applied to creating better displays and pictures.

Typical stereoscopic displays produce a vivid impression of depth by presenting each eye with its own image on a flat display screen. This technique produces many depth signals (including disparity) that are consistent with a 3-dimensional (3d) scene; however, by using a flat display panel, focus information will be incorrect. The accommodation distance to the simulated objects will be at the screen distance, and blur will be inconsistent with the simulated depth.

In this thesis I will described several studies investigating the importance of focus cues for binocular vision. These studies reveal that there are a number of benefits to presenting correct focus information in a stereoscopic display, such as making it easier to fuse a binocular image, reducing visual fatigue, mitigating disparity scaling errors, and helping to resolve the binocular correspondence problem. For each of these problems, I discuss the theory for how focus cues could be an important factor, and then present psychophysical data showing that indeed focus cues do make a difference.

Next, I describe a new approach to construct a stereoscopic display that presents these signals correctly. This new display uses a custom lens system to present correct focus information in a time-multiplexed manner. This approach has significant advantages over other volumetric display approaches, and is a significant advancement in stereoscopic display technology.

In addition to the technical development of the volumetric display, I have also developed a way of analyzing volumetric displays to assess the retinal image quality of objects in various positions within the display. This model of image formation in a volumetric display allows us to test how various properties of the viewer, display and stereoscopic content will change the image quality. The main conclusion of this analysis is that a volumetric display requires only a coarse depth resolution to create stereoscopic images that are a close approximation of real-world images.