ABSTRACT
Introductory textbooks on visual perception often state that the two main goals of the visual system are the visual guidance of action and the identification of objects (see Chapter 14). This chapter focuses specifically on the latter goal, though of course the two are related to one another. The discussion begins by outlining the problems inherent in a visual system that attempts to recognize environmental objects, and then covers several possible solutions proposed for these problems. It is important to note that these solutions take into account many of the visual processes covered in previous chapters, such as frames of reference and viewpoint, as well as determination of contours and occlusion. Overall, object recognition is a complex process (or set of processes) that builds on those lower-or middle-level visual processes that precede them. However, which processes are included in recognition and how they operate are still under debate. Much of the research on object recognition has been done by the computer vision community, as will become apparent throughout this chapter. While this has had significant influence on psychology research, many of the computer-vision-based theories of object recognition have not yet been evaluated as to their ability to account for human performance.
