Basics of vision
Eyes have 2 functions:
Forming an image of an object
Transducing Images into electrochemical signals which are sent to the cortex via neural pathways.
Image forming system
1) the cornea
Light enters through the cornea and is bent inward through the pupil to the lens which focuses light onto the retina.
1) the retina..
-at the back of the eye
- outward growth of the brain
- consists of a thin and complete network of photo receptor cells
3 types of photoreceptors:
1) rods; low light density - night time, 1 eye has 120 million rods
2) cones; high light density - daylight, leads to the perception of colour, 8 million cones
3) photosensitive ganglion cells; moderates circadian rhythms and pupil reaction (zaidi,2007)
Basics of vision
The optic nerve.
- collections of axons leading to the visual centres of the brain from the eye.
- optic fibres transmit neural messages to the optic chiasms.
- optic nerves from each eye join here.
- allows messages from any eye to arrive at a common destination in the visual cortex.
- lies under the cerebrum and helps guide visual attention.
- if an object appears in the extremity of the visual field, this guides eye movements so the object can be observed.
The visual cortex .
- located at the back of the cerebral hemispheres in the occipital lobe.
Cells in the visuaL cortex
1. Simple cells.
- respond to particular stimuli
- only fires if an image falls exactly on a cells receptive field
- as responses are so specific, simple cells can signal the orientation of a stimulus falling within a particular region of the receptive field.
2. Complex cells.
- also respond to particular orientations.
- do not have well defined on and off areas; it's harder to predict what stimulus will cause activation.
- respond Most vigorously to a beam of light moving across the visual field.
3. Hypercomplex cells.
- these cells detect lines of specific length
- recent research indicates these are subclasses of simple and complex cells.
Gestalt principles of perception
- focuses on 'form'.
- argues perception cannot be comprehended by merely looking at individual components.
- form; dependant on relationship between individual elements, rather tan the elements themselves.
- human perceptual systems use a range of principles to detect form.
- argues principles are 'innate'.
Gestalt principles of organisation.
- developed by Wertheimer, further developed by Rock and Palmer (1990).
1) pragnaz - simplist and most stable interpretations are favoured : elements of this law can be seen in all laws.
2) proximity - group projects that are close to one another as a perceptual unit.
3) similarity - aliked items grouped
4) closure - contours close to one another are likely to be united
5) good continuation - 2 or more neibouring components of a scene are grouped when thy appear to be connected by a straight or smooth line.
6) common fate - objects moving in the same direction will be perceived as related.
7) common region - elements will be grouped together when they are enclosed by a defined boundary.
8). Uniform connectedness - elements that appear to be connected are more likely to be grouped
9) meaningfulness or familiarity - objects grouped when meaningful.
Problems with gestalt approach
Problems with innate view;
- some appear to have innate components.
- lots of evidence for learned component (segall, Campbell and herskovils, 1966).
- found people from environments where straight lines are rare (rural south Africa) were less susceptible to muller-lyer illusion than people from western cultures.
- bottom up
Difficulties applying laws;
- pragnaz not easy to demonstrate
- people may view the same figure in very different ways or perceptions may change over time.
Descriptive rather than explanatory;
- describes rather than explains
- lack of support
- hard to generate predictions
- fails to say why the rules are necessary
- uses vague language I.e. 'good' or 'simple' shape ?
- ecological approach to space perception
- argued attention should be on cues in present in the environment rather than taking the retinal image as the starting point for perceptual processing
- information is detected rather than processed
- the visual environment provides sufficient information to allow interaction with environment without the need for further internal processing
- stimuli are not passive images but actively sampled information termed optic array.
- bottom - up theory.
Several cues that give observer information about
Optic flow pattern:
- the apparent motion of objects in the environment caused by relative motion between the observer and the scene.
- provides information about distance, elements become densely packed, and decrease in size as distance decreases.
- allows observers to judge the relative size of objects
Direct perception good and bad..
- Gibson argues we should examine perception in the real world.
- his theory makes strong links between perception and actions
- the individual and the environment interact
- end result - not an internal representation but an awareness of possibilities to interact
- EXAMPLE - a branch is graspable and fruit can be eaten
- radical and controversial
- emphasis of environmental cues on perception
- made future research on interaction between individuals and their environments
- avoiding the growing tendency to examine perception out if context - bars and grids
- Fodor and Pylyshyn (1981)
- argued some terms Gibson used such as 'directly detected' and 'invariant', are so broadly defined as to be almost meaningless
- the Idea we require no further processing implies we have invariants specifying friends faces that do not require store memory - is this really possible?
- Marr (1982), Ullman (1980) - pointed out detection of physical invariants such as image surfaces and optic flow is a problem.
The constructivist approach
- perceiver has an internal constructive process which transforms an incoming stimulus into a percent.
- therefor it is an active approach
- perception is an end product of a series of interactions between 'initial stimuli', 'internal representations' 'memory' and 'expectations'.
- if expectations are incorrect visual illusions will occur
Depth perception and constructivism : illusions
- retinal image is 2d
- perceiver has to add depth to create 3D
- accomplished by a range of cues including size constancy, perspective, and stereopsis
- Gregory (1997) investigated how such cues aid perception by examining conditions in which which they broke down to give perceptual errors
- they found the Muller-Lyer illusion occurs when there is a conflict between the cues in the scene and retinal image
- children are less susceptible to some illusions supporting the constructivist approach as they have less experience using cues
Critics of the constructivist approach
- led to the discovery of a range of visual cues which influence perception
- explanations are often unsatisfactory and contradictory
- Day (1972) argues that the presence of conflicting cues rather than depth cues specifically cause the Muller-Lyer illusion .
- others criticise artificial nature of the tasks used to produce perceptual errors.
The computational approach
- based on information-processing perspective which divides processes into a series of steps
- based on information from the environment and a calculation of these inputs
- based in the following steps
1) image on the retina
2) grey level descriptor - intensity of light at each point in the visual field
3) Raw primal sketch - recognises edges and differences in light intensity.
3) full primal sketch - raw elements assembled. Simple grouping process to link the features
4)2 1/2 D - layout of the structures, depth and object motion. Objects are viewer centred and superficial.
5) 3D model representation - object recognition and computation, information about objects identity and stored knowledge.
Critics of computation theory
- relatively simple with each stage involving significant computation
- Marr and HIldreth (1980) developed a computer programme to produce a primal sketch but this bears little relation to a humans vision
- bottom up processing