Binocular depth processing: neural basis or stereopsis or solid 3D vision Flashcards
which two types of pictorial cues allows to achieve stereopsis with only one eye
- interposition - in real 3D world, can tell e.g. the zebra is standing in front of the striped wall as the zebra is masking the stripes on the wall i.e. it is interposed between the observer and the background
- shadowing - demonstrated when the sun is shining from one direction, it creates shadows which gives off a strong 3D piece of information
what is the origin of the ability to use 2 eyes & extract a third dimension/binocular stereopsis
optic geometry
- horizontal separation of 2 eyes - each eye has slightly different view on the world due to separation which…
- generates disparities between 2 retinal images
what is the perception of binocular stereopsis achieved from
visual cortex
- primary v1 cortex + numerous extra striate areas
- single neurons tuned to different retinal image disparities = disparity detectors (neurons in the extra striate cortex)
what part of binocular stereopsis enhances depth discrimination
- stereo acuity (clinical grades & tests)
- /~100 arc secs = coarse/reduced (defect in stereo acuity), tests are usually placed at 40cm where the two eyes can tell how far something is which can’t do with one eye
what is the vieth-muller circle also known as
the horopter
when the foveas of both eyes are looking at the same place, what property will the images have
they will have no disparity = 0 i.e. the two images are in retinal correspondence as there is no difference in how far away from you that you can perceive that fixation point to be
why is the fovea not the only part in the retina where there in retinal correspondence/0 disparity
for every point on your right retina, there will also be a part on your left retina which is looking at the same point in space = object with 0 disparity & there is correspondence
what does the horopter represent
in reality there is a hemi sphere all around the head, on this hemi sphere, every point will be the exact same distance away from you, as the fixation point is
what is non correspondence in relation to the horopter
if object is beyond or before the horopter, it will form a disparity between both eyes
what are the two points of non correspondence/properties of disparity
- sign
nearer than the horopter causes crossed (-ve) disparity
further than the horopter causes uncrossed (+ve) disparity - size
disparities increase with distance from the horopter
what is the sign of disparity when the image is to the right in the right eye and to the left in the left eye
uncrossed +ve disparity
what is the sign of disparity when the object is slightly to the left in the right eye and slightly to the right in the left eye
crossed -ve disparity
what does a large disparity cause and why
diplopia due to narrow range wither side of the horopter
what is referred to as panum’s fusional area
fixation plane in which you can fuse two images, even though they are slightly different in two eyes, that region of space is the panum’s fusional area
where is the width of panum’s fusional area narrow
around the fovea = narrow region of 3D space
where does panum’s fusional area get broader and why
further out in the eccentricity, because the brain is able to tolerate the slop is disparities in a more eccentric location, compared to a point in fixation i.e. stereo acuity is better if closer to the fovea & worse out in periphery
what are the nasal axons of RGCs that cross over at the optic chiasm and have neurons that go to certain layers of the LGN on the other side, look at the same point in space as
temporal axons of RGCs that do not cross over at the optic chiasm and have neurons that go to certain layers on the LGN on the same side
name the steps in which the neurons take after they leave the LGN
- dive into posterior limb of the internal capsule
- up the optic radiations
- end up in primary v1 cortex
- layer 4c called granule cells (little local circuit neurons) which only receive information from one eye from same point in space
- granule cells have neurons which come out & make connections with granule cells in layers above and below layer 4c
- these connections coverage from left and right eyes onto single neurons of layers above and below e.g. pyramidal cells in layer 3
- which are interested in information from both left and right eyes
which cells are only interested in information from one eye/monocular
connections from LGN to primary v1 cortex to layer 4c
which cells are interested in information from both eyes/binocularly
layers above and below layer 4c
how much % of v1 cells beyond 4c are binocularly driven (excited by stimuli presented to both eyes)
75%
what do all binocular driven cortical cells have in similarity
receptive fields in the left and right eyes
what do all binocular driven cortical cells have similar RFs in respect to
- spatial location (come from same point in space)
- preferred orientation (e.g. if likes vertical lines as seen from left eye, will like vertical lines as seen from right eye)
- preferred spatial frequency
- organisation i.e. simple, complex, hypercomplex
e. g. if complex cells will be the same in right and left eye as those binocular cells are looking at same point in space with right and left eye, so interested in the same stimulus properties
what do binocular driven cortical cells differ in
their
- preferences for horizontal stimulus disparity
- small spatial offsets in their left vs right eye RF positions (known as positional disparities) i.e. Rsf are not completely super imposed in space, sometimes they are slightly crossed in left and right eyes or uncrossed = RF offsets
what are the responses of 75% of binocularly driven cells in v1 cortex influenced by
the horizontal disparity between stimuli presented to the left and right eyes
how many % of binocularly driven cells in v1 cortex are not influenced by horizontal disparity between stimuli presented to the left and right eyes
25% do not care about disparity at all
