Abstract
Optic flow can be used by humans to determine their direction of heading as well as to control steering. Dot-flow displays have been widely used to investigate heading perception due to the potential they provide for various quantifiable manipulations (e.g. flow density and luminance) and findings from these experiments have been used to better understand the real-world control of steering. However, it is still unclear whether textures more akin to those in the real world are more informative than dot-flow when judging heading and controlling steering. It is also unclear whether the visual conditions that enable accurate heading judgments will necessarily support successful steering. To investigate this the accuracy of heading judgements were compared across dot-flow displays of different densities and luminance as well as to a photorealistic textured ground plane. In a separate experiment we measured steering performance towards an offset target in order to have a direct comparison between heading and steering performance under the same visual conditions. Our findings suggest that the bright, dense, dot-flow displays led to equivalent performance as the realistic ground texture both when judging heading and actively steering. Very sparse displays (with few low luminance dots) led to poor performance for both heading and steering tasks. Interestingly the intermediate dot-flow quality revealed conditions that led to accurate heading judgements but this did not translate to accurate steering performance. We conclude that heading perception should not necessarily be considered synonymous with successful steering control, since some visual conditions that support heading judgements that will not necessarily support steering.
Meeting abstract presented at VSS 2013