Perceiving the Visual World: Recognizing Patterns

We perceive ourselves as moving through a stationary environment, rather then ourselves being stationary and the surrounding environment as moving.

We can recognize and actually perceive common objects as having an expected shape, even though we view them from different distances, orientations and under unknown lighting conditions.

We can adjust to more than ten orders of magnitude of light intensity variation without conscious awareness and be bothered by shadows or partial occlusions.

It is obvious that perception is not the inevitable result of a set of stimulus patterns but rather a best interpretation of sensory data based on the past experience of both the organism and its ancestors.

Perhaps the most concise way of summing up our visual capability is that, except in the case of physical injury, it appears to operate flawlessly, spontaneously and without surprises.

At the highest levels of performance, shape recognition involves the ability to ignore variations in size, brightness, position and orientation, e.g. if a person learns to identify a shape using one part of his retina he is able to identify the same shape when it is presented to other parts of the retina, or even to the other eye. However when the brightness of an object and its background are inverted, humans are sometimes unable to recognize the object. Moreover humans have great difficulty in recognizing faces presented upsidedown. Considering pattern vision as a single integrated function is an oversimplification, biologically important visual tasks are often handled by special mechanisms (e.g. it appears that human brain has distinct procedures for processing visual information about faces)

References


back.gif (279 bytes) Main Vision Page
HyperVis Table of Contents

Last modified on April 05, 1999, G. Scott Owen, owen@siggraph.org