Role of foveal vision in static and dynamic environments
Clayden, Adam Christopher
The visual field has multiple regions, with visual acuity being highest in the centre before declining rapidly outward toward the periphery. This central region, otherwise known as the fovea, is typically defined as the central 2.0o of vision. Although comparatively small with respect to other visual field regions, being able to discern objects of interest in fine detail is only possible in this region. Due to this, people make ballistic eye movements (saccades) towards the fine details and depending on the task, may stabilise their gaze in the form of a fixation to discriminate parts of this newly attended area. A typical everyday task that can exhibit such behaviours is visual search (scanning a visual environment for objects or features among distractors), and much of this thesis is concerned with the importance of foveal vision with respect to visual search tasks. Seven experiments are presented in this thesis, with chapters 1 – 4 containing a review of the literature, methodologies and a glossary. In Chapter 5, search experiments with targets of varying sizes were conducted to assess the role of foveal vision on search performance. This chapter builds on a debate on whether foveal vision was necessary for the successful completion of a visual search task. In addition, a novel algorithm was developed to embed targets at a specified change in local contrast to automate target placement. The presented results show that the time taken to find targets with and without foveal vision is nearly identical even when target size is small. Chapter 6 modulates search difficulty by manipulating salience to investigate the effect of target size and salience on search performance. Coined the Compensation Effect, the results of Chapter 6 show that the above-mentioned variables were able to compensate for one another, resulting in an improved search performance. This effect occurred regardless of visual field degradation. In the same chapter, another experiment revealed the stage in the search process where performance costs originated from, which was the final stage of search concerned with target verification. Finally, Chapter 7 transitions from static images to dynamic scenes which simulate self-motion. Additional algorithms were developed, including an extension to the existing gaze-based decomposition of search time (Malcolm & Henderson 2009). Chapter 7 investigated the role of foveal vision in visual search whilst optical flow was present. Unlike previous results, the final experiments of this thesis revealed the necessity of foveal vision for the attainment of a normal search performance. The results of this thesis demonstrate that the importance of foveal vision with respect to visual search is modulated by the stimulus environment, with it being relatively unimportant for static scenes, but important for dynamic scenes.