Pattern recognition (Physiological Psychology)

From Wikipedia, the free encyclopedia

Pattern recognition involves identification of faces, objects, words, melodies, etc. The visual system does more than just interpret forms, contours and colours.Pattern recognition refers to the process of recognising a set of stimuli arranged in a certain pattern that is cahracteristic of that set of stimuli. Pattern recognition is an innate ability of animals.

Contents

[edit] Theories of pattern recognition

  • Template matching theory
  • Prototype matching theory
  • Feature analysis theory
  • Recognition by components
  • Fourier analysis
  • Bottom-up and top-down processing

[edit] Template matching theory

The incoming sensory information is compared directly to copies (templates) stored in the long term memory. These copies are stored in the process of our past experiences and learning.

Eg. A A A are all recognised as the letter A

[edit] Prototype matching theory

Prototype means a concept of average characteristics of a particular subject. For instance a concept of small animal with feathers, beak, two wings that can fly is a prototype concept of a crow, sparrow, hen, eagle, etc. Prototype matching unlike template matching does not emphasize a perfect match between the incoming stimuli and the stored concept in the brain.

[edit] Feature analysis theory

According this theory, the visual system breaks down the incoming stimuli into its features and process the information. Some features may be more important for recognition than others. All stimuli have a set of distinctive features. Feature analysis proceeds through 4 stages.

  1. Detection
  2. Pattern dissection
  3. Feature comparison in memory
  4. Recognition

[edit] Theory of recognition of components (Irving Biederman)

According this theory every object is made up of Geons - the building blocks of all objects (cylinders, cones, are combined in many ways (on top of, to the side, etc.).

Hierarchy of detectors:

  1. feature detectors - lowest; respond to curves, edges, etc.
  2. geon detectors - activated by feature detectors
  3. higher level detectors - recognize combinations of features and geons

According Biederman an individual on an average is familiar with about 30,000 objects and recognising them requires no more than 36 geons.