ABSTRACT

Over the past 15 years or so, one of the most influential distinctions in the field of visual information processing has been that between object and spatial coding. This dichotomy, first introduced by Ungerleider and Mishkin, holds that the neural areas supporting object recognition are separate from those supporting location coding, as are the computational processes involved in object and spatial coding in the brain. Indeed, to recognize the same object across different viewing positions and angles, it can be argued that the varying spatial information needs to be discarded so that recognition is based on the invariant object information present. This chapter examines the nature of object and spatial coding in the brain. It argues for the existence of at least two forms of visual representation: one in which elements is coded as parts of a single object and one in which elements are coded independently.