Perception
Perception
Perception
Pattern recognition
Scenes:
People are very fast at recognising scenes like cities or motorways because of the large
number of objects and because of occlusion. Scenes have very specific spatial organisation.
- Difficult to say which visual information belongs together and forms objects because
the environment is very complex and very confusing as objects overlap and hide
parts of objects.
1. The law or proximity – Visual elements close in space tend to be grouped together
2. The law of similarity – We group together the things that are similar
3. The law of good continuation – We group things together if they appear to form a
continuous pattern
4. The law of closure – We tend to complete figures with gaps in them by ignoring the
gaps and mentally filling in what we believe should be there
Gestalt Psychology:
- How we experience things that are not part of our perceptions. The mind adds
structure to the things we see.
- The notion that the simplest possible organisation of the visual environment is
perceived.
- Perceptual organisation of the visual field (objects of central interest) into a figure
and ground (less important background)
- Figure is perceived as being in front of the ground
Advantages: Describes perceptual organisation well, many principles have stood the test of
the time (applied to 3D images), Law of Pragnanz has proven useful.
Limitations: Underestimated the importance of knowledge, more descriptive than
explanatory, less applicable to detailed images, many further discoveries have been made,
too inflexible.
Simple perceptual principles to create meaning to our world, relevance to marketing and
art, film industry, useful tools of psychologists to study the mind
He proposed that object recognition was much more complex than previously thought.
Limitations – Maybe too much emphasis on bottom-up processes. He did admit that top-
down processes played a role. However, he ignored the role of expectations and prior
knowledge in visual perception.
If there’s enough information available to enable us to identify and object’s basic geons, we
will be able to identify objects.
Advantages – Good evidence for geons being important in object recognition. Provides
answer to how we identify objects even when there us substantial differences (shape, size,
orientation) between members of the same category, evidence that concavities and edges
are important in object recognition, many principles have stood the test of time
Limitations – de-emphasises importance of top-down influences from context, expectations
and previous knowledge, flails to account for most within-category discrimination, much
recognition is actually viewpoint dependend, Some classes do not have invariant geons yet
are still recognisable as members of a category (e.g., clouds)