| Popis: |
Identifying moving objects while we are moving is an important everyday skill. This ability allows us to monitor our surroundings, successfully interact with objects or people, and avoid potential hazards. Self-movement generates optical flow on the retina that complicates the recognition of moving objects from retinal motion alone. Rushton and Warren (2005) proposed a purely visual solution to this problem. They suggest that, in order to assess scene-relative object movement, the brain identifies and parses out (globally subtracts) patterns of visual flow that are consistent with self-movement. Existing research has demonstrated evidence of this flow parsing process in central vision (i.e. Warren & Rushton, 2008). This thesis aims to characterise the role of peripheral visual flow in this process. Research from the wider self-motion literature has often distinguished between central and peripheral vision. Some researchers have claimed that peripheral vision is specialised for self-motion perception, whilst more recent studies have challenged this assertion. This thesis investigates whether peripheral visual motion, traditionally considered to be a strong cue to self-movement, also contributes to flow parsing. The experimental work employed a simulated self-movement paradigm to isolate retinal motion from other non-visual cues. Thus, observers remained stationary whilst computer generated stimuli moved to produce patterns of retinal motion associated with actual self-movement. In the first set of experiments, I demonstrate that peripheral flow simulating forward or backward self-movement gives rise to characteristic flow parsing effects. This finding generalises to rotational observer motion (Chapter 3). Chapter 4 considers whether peripheral flow contributes to parsing for judgements of object size change. Finally, Chapter 5 investigates whether there is a benefit of peripheral information under conditions where central flow is potentially ambiguous. The results indicate that peripheral visual flow contributes to the flow parsing process. The contribution of flow in the near periphery appears to be maximally important. |
