One major field in cognitive psychology is the study of how we recognize objects in the world around us. How do we look at a chair that we have never seen before, and yet know that it's a chair? Object recognition encompasses the process of perception from the preliminary identification of shapes to the placement of the object into a particular category.

There is mounting evidence that the brain contains a large number of feature detectors. These detectors are neurons that fire a signal when a particular feature hits the retina. Features are simple parts that make up all the objects that we see. Some examples of features include orientation (straight or tilted), primary colors (red, yellow, and blue), and line enclosure (C versus O). The mind is wired to group together objects with similar features, making dissimilar objects "pop out." The grouping together of similarly featured objects is called the Gestalt principle. Look at the figure below. All of the horizontal lines can be seen as one object with multiple parts, while the vertical line clearly stands out on its own.

A researcher named Triesman used pop-out studies to identify which parts of an object were features. Subjects were told to search for the one straight (or yellow, or enclosed) object among many other tilted (or blue, or open) objects. The sought-after object was the target and the others were the distracters. If people needed to look at each object, process it, and decide whether it had the sought-after quality, then the time taken to find the correct object should increase as the number of distracters increased. This was the case when subjects had to search for some complex quality, such as a letter among numbers. However, Triesman found that, for simpler qualities, which she called features, the number of distracters was irrelevant; subjects always found the target equally quickly, suggesting that they didn't need to examine each object, but rather that the target "popped out" at them. Features, therefore, seem to be processed without much effort or attention devoted to their recognition.

Biederman's theory of object recognition proposed that all objects are made up of a limited number of simple, three-dimensional shapes called geons. These geons act like features to make up structural descriptions of objects. The geons and the structural descriptions are supposedly view-independent, meaning that people are able to recognize them equally well from any angle. For example, the geons making up a table are a flattened cube and four cylinders, and the structural description is the combination of those five geons to form a "table."

By contrast, Tarr (another cognitive researcher) argues that we have a large warehouse of templates that represent past objects in memory, one for each object or type of object that we have encountered. These templates are used to identify objects when they are seen again. You might have separate templates for the chair in your room, your father's armchair, and your desk chair at school. However, if we truly have a template for each object we have ever encountered, and we use it to recognize the object when we see it again, how can we identify new objects? Even though you may have never seen my red chair before, and therefore have no template for it, you can still easily identify it as a chair. We will study the process of categorization further in the section on concepts and categories.