Analysing Images and Human contrast sensitivity

?
When we view a visual scene what is formed?
An image of the world is formed on the retina at the back of each eye
1 of 63
What is the vision?
An information processing task... the process of discovering from images what is present in the world and where is it?
2 of 63
What is natural selection?
It has found a solution to this problem processing problems as evidenced by our own visual systems
3 of 63
If we are capable of knowing what is ?
Where in the world, our brains must somehow be able to represent or encode this information from images
4 of 63
What types of information do visual images contain?
Objects and surfaces in the world are composed of physical matter. the properties of the matter determines the manner in which it reflects
5 of 63
What do objects do?
Principally differ from each other, and their surroundings in terms of luminance of the luminance of the light that they each relect
6 of 63
Consequently images formed ?
images formed on the back of the eyes contain spatial variations in luminance that carry information about object boundaries, edges and contours in the world
7 of 63
Luminance
Slow changes in luminance reveal the coarse spatial structure of the world (E.G. Large objects, overall shapes), abrupt changes in luminance reveal the fine spatial structure of the world
8 of 63
How can we represent the information in the visual images?
- Could measure luminance at each spatial location in the image - This is the approach taken by digital cameras - The greater the number of points you measure the greater the quality of the resulting representation
9 of 63
Can images be broken down into more basic components?
Yeah, sinuosidal gratings are a very good choice
10 of 63
What are sinuosidal gratings?
Luminance varies across the pattern according to sinuosidal waveform, sinusoidal grating are simple 1 dimensional, periodic patterns in which luminance varies across space, luminance varies across the pattern according to a sinusoidal waveform
11 of 63
What is the four defining characteristics?
Spatial frequency, contrast, orientation, spatial phase
12 of 63
What is spatial freuqency?
Spatial scale of luminance variation (number of cycles in 1 degree of visual angle, because the width of the bars changes with viewing distance)
13 of 63
What is contrast?
Intensity difference between the light and dark bars (How visible it is)
14 of 63
What is orientation?
Axis of the grating's bar
15 of 63
What is the spatial phase?
Relative position of the bars (does it begin with a light or dark bar?)
16 of 63
Why are grating patterns so useful?
mathematically that it is possible to create any image, no matter how complex, from a set of sinusoidal grating patterns
17 of 63
What do you need to do?
find out what SF, contrast, orientation and phase are needed for each sinusoidal grating and then add them all together
18 of 63
WHat is low SF grating contribute to?
Coarse spatial information to resulting image, and high SFs contribute the fine detail
19 of 63
What is this process calleD?
Fourier synthesis
20 of 63
Importantly the reverse is what?
true. Yu can decompose or represent any complex image into a unique set of sinusoidal grating
21 of 63
What can happen?
2D pattern on the right can be made by simplying adding together the two sinusoidal grating patterns on the left, 2D on the right can be broken down into two simple sinusoidal gratings patterns on the left
22 of 63
What is the Einstein experiment?
Adding together more and more sinusoidal grating patterns with progressively higher SFs, we can make an image of his face
23 of 63
Coding images in this way is what?
is more economical
24 of 63
What does Sinusoidal gratings provide?
A universal language to precisely describe the input to the visual system, no matter how complex the visual scene
25 of 63
What can we measure?
the modulation transfer function of the system or cell – extent to which each grating gets through transmitted
26 of 63
What can we do?
Predict how it will respond to more complex images, because all images can be made from a sum of simple sinusoidal gratings
27 of 63
What do we assume?
The visual system obeys some simple rules. For example, its responses to a complex pattern is basically the sum of its responses to a set of simpler component grating patterns
28 of 63
How do you measure the MTF of the whole system?
- If our visual system does a good job of transferring some spatial frequencies, we should need very little contrast to see gratings of those spatial frequencies ( we would be relatively ‘sensitive’ to these gratings)
29 of 63
If vision is poor at transferring other spatial frequencies then we should what?
Be relatively insensitive to them and require a high contrast to see them
30 of 63
What should be measured?
Psychophysically contrast detection thresholds for a set of grating patterns of various spatial frequencies
31 of 63
What has been measured?
The CSF for human vision
32 of 63
What does it show?
contrast sensitivity (1/threshold contrast required to detect each sinusoidal grating pattern) Vs. The spatial frequency of the grating patterns. The higher the sensitivity, the less contrast is needed to see it
33 of 63
We are less sensitive to what?
less sensitive to very low and very high spatial frequency gratings than intermediate ones
34 of 63
What does the CSF define?
The visual system's window of visibility, under the curve patterns are visible
35 of 63
We are less sensitive to what?
Very low and very high spatial frequency gratings that intermediate ones, sensitivity is greatest for gratings with spatial frequencies of 2-6 c/deg
36 of 63
What does the CSF define?
They defines the visual system's window of visibility, under the curve patterns are visible
37 of 63
What happens under the curve?
Patterns are visible
38 of 63
What can we predict?
The CSF show that sinusoidal gratings that lie on either side of the optimum spatial frequenct
39 of 63
What does the CSF show?
sinusoidal gratings that lie on either side of the optimum spatial frequency are less visible
40 of 63
What can we do?
extrapolate from this to predict performance for more complex image - If the luminance variations in a scene occur at a spatial scale (Frequency) that is close to the region of optimum on the CSF it will be visible even when the contrast is low
41 of 63
If an image contains what?
contains only very coarse (Low spatial frequency) variations in luminance (large objects) or only very high spatial frequencies (small objects or fine details), it will be less visible and contrast will need to be high to see it
42 of 63
Why does the CSF have an inverted U shape?
Fall off in sensitivity to very high spatial frequency grating is due mainly to optical imperfections of the eye which ‘blur’ the fine detail in an image
43 of 63
What is optical imperfection?
The eye only degrade high spatial frequency information
44 of 63
What is insensitivity?
to very low spatial frequency gratings must be due to neural factors in the visual system
45 of 63
For example?
cells very early in the visual pathways (E.G. Retinal ganglion cells) have receptive fields that exhibit a concentric center-surround organisation
46 of 63
what do they respond to?
light spots on a dark background or Vice versa - When light covers the entire receptive field they show little or no change in firing rate
47 of 63
What do they respond to?
gratings but not when the spatial frequency is too low - On center cell: Light in the RF center excites the cell but light in the surround inhibits the cells
48 of 63
What does CSF look like at different light levels?
- Many researchers have measured the CSF for human vision as the overall luminance level is reduced from the phototopic range, to the mesopicrange and sceptic range (CF near darkness
49 of 63
Why does CSF look like a different light levels?
- As the overall luminance level is progressively reduced, the peak sensitivity shifts to gratings of lower and lower spatial frequency - Occurs mainly because sensitivity to high spatial frequency gratings becomes worse
50 of 63
What happens at night?
our rod system is most active, our ability to see fine detail is lost - Retinal ganglion cells are fed by rods from large areas of the reina and sacrifice acuity to operate at low light levels
51 of 63
What happens to the CSF when gratings flicker or move?
- Sinusoidal gratings can also have a temporal frequency which indicates how rapidly they move or flicker - When the CSF measured with moving/flickering gratings sensitivity to very low spatial frequencies improves when the temporal frequency is high
52 of 63
What is possible neural mechanism?
- Magnocellularcells (primate LGN) are 10 times more sensitive to low spatial frequency gratings than Parvocellular cells when patterns move/flicker at high rates
53 of 63
What are behaviour effects of lesions to cells in the LGn?
Behavioural effects of lesions in the parvocellular and magnocellular layers of primate LGN. Magno lesions cause loss in sensitivity to rapidly moving flickering low spatial frequency patterns
54 of 63
What do standard visual acuity test measure?
Historically the study of spatial vision emphasised the measurement of visual acuity – the finest spatial detail that an observer can disern Optometrists of ophthalmologists generally use a acuity measures to assess spatial vision, using HC stimuli
55 of 63
What is the Snellen eye chart?
Observer names the snellen letters
56 of 63
What is the landholt rings?
Observers reports the location of the gap
57 of 63
What is the parallel bars?
Observer reports the orientation of the bars
58 of 63
How are the CSF and standard acuity measure related?
Most visual acuity tests measure only the smallest spatial details that can be resolved by the visual systems. Contrast is fixed and maximal, so that size alone is the factor that limits performane
59 of 63
Are the two measures evers equivalent?
- The highest spatial frequency grating that you can detect, requires maximum contrast because sensitivity is so low - The single point on the CSF is also a measure of visual acuity
60 of 63
What is the advantages of using visual acuity tests?
- Standard acuity measures used in clinical settings by optometrists because many causes of loss in sensitivity to fine detail are optical in nature (e.g. ‘short sightedness’) - - These can usually be alleviated by the use of spectacles/contact lense
61 of 63
What are the advantages of measuring the CSF?
• Ginsburg et al. (1982) used the CSF to successfully predict how well pilots see objects in the air and on the ground. Under conditions when fine detail is lost (e.g. fog), visual acuity is a poor indicator of performance
62 of 63
Some patients suffering from alzheimers or Cataracts might what?
exhibit substantial contrast sensitivity deficits for both coarse (low frequency) and fine (high frequency) patterns. Visual acuity tests would not pick this up
63 of 63

Other cards in this set

Card 2

Front

What is the vision?

Back

An information processing task... the process of discovering from images what is present in the world and where is it?

Card 3

Front

What is natural selection?

Back

Preview of the front of card 3

Card 4

Front

If we are capable of knowing what is ?

Back

Preview of the front of card 4

Card 5

Front

What types of information do visual images contain?

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Analysing Images and Human contrast sensitivity resources »