| |
Abstract:
The aim of the present study is to investigate how the visual
system takes into account shadows and lighting for object
recognition. Participants were instructed to recognise 3D objects
(geometrical shapes and familiar objects) as quickly as possible.
The objects were accompanied by a congruent shadow (e.g. a cylinder
with a cylinder shadow) or by an incongruent shadow (e.g. a
cylinder with a sphere shadow) and by a congruent lighting (e.g.
lighting from the right, shadow to the left) or by an incongruent
lighting (e.g. lighting from the right, shadow to the right).
Control conditions included 3D objects without shadows and 2D
silhouette of objects with shadows. Response times (RTs) for 3D
objects without shadows were longer than RTs for 3D objects
presented with shadows. RTs for objects presented with incongruent
shadows and incongruent lighting were greater than those for
objects presented with congruent shadows and congruent lighting.
RTs for 2D silhouettes of objects presented with cast shadows were
the longest. These findings suggest that an incongruent cast shadow
and incongruent lighting were sufficient to delay the recognition
of familiar stationary objects. Further that unusual conditions
such as a 2D silhouette accompanied by a shadow produces longer
processing time. Together these results suggest that the coding of
the direction of lighting and cast shadows serve the function of
improving object recognition.
|
