Central Visual Processes
-
Upload
charity-carlson -
Category
Documents
-
view
38 -
download
2
description
Transcript of Central Visual Processes
Central Visual Processes
Anthony J Greene 2
Central Visual Pathways
I. Primary Visual CortexReceptive Field
• Columns• Hypercolumns
II. Spatial Frequency
III. Nerve or Cortical Damage
IV. Higher Visual Areas
Anthony J Greene 3
Occipital Lobe
Anthony J Greene 4
Occipital Lobe: Calcarine Sulcus -- V1 -- Striate Cortex
Anthony J Greene 5
Cells in V1
Anthony J Greene 6
Single-Cell Recording (Hubel & Weisel, 1962)
• Attempted to discover what sorts of information cells in (cat) V1 respond to
• Accidentally discovered orientation specific cells organized into columns and hypercolumns
V1
Anthony J Greene 7
Anthony J Greene 8
Cells In V1
Cells in V1 receive messages from certain ganglion cells such that they respond to stimuli of a certain orientation from a small portion of the retina - Orientation Specific
~ 200 Million Cells in V1
Inputs from Ganglion Cells
Anthony J Greene 9
Cells In V1
• One V1 cell receives inputs from many ganglion cells
• One ganglion cell may send inputs to numerous V1 cells
• Stimuli from every possible orientation, and from every position in the visual field are detected by different cells in V1
• Simple Cells detect only orientation -- Complex Cells detect orientation and motion
Anthony J Greene 10
How to Make a Complex Cell
• Orientation specific inputs from ganglion cells is similar to simple cells
• However, the receptive field is much larger and is designed to respond maximally when inputs from sub-fields are sequential
Anthony J Greene 11
Cells in V1
Occular Dominance
Anthony J Greene 12
Columns in V1
Anthony J Greene 13
Organization of Cells in V1• Columns are sections of cortex which all respond to
the same orientation from approximately the same region of cortex
Anthony J Greene 14
Organization of Cells in V1• Hypercolumns are groups of columns, from both
eyes, which are influenced by the same minute portion of the visual field
Anthony J Greene 15
Organization of Cells in V1• What sort of information are these cells detecting? • Is the information from any single cell in V1
informative?
Anthony J Greene 16
Response Properties of Cells in V1The extent to which columns will respond to stimuli with no interactions from other columns
Orientation Column Position on Occular Dominance Slab
Cel
lula
r A
ctiv
ity
Anthony J Greene 17
Lateral InhibitionThe column with the strongest response to a given stimuli will suppress the respondse of neighboring columns
+
-
Anthony J Greene 18
Response Properties of Cells in V1The extent to which columns will respond to stimuli with lateral inhibition from other columns
Orientation Column Position on Occular Dominance Slab
Cel
lula
r A
ctiv
ity
Anthony J Greene 19
Processing at V1 Is Edge Detection
Anthony J Greene 20
Edge Detection
• While lateral inhibition normally improves the accuracy of edge detection, in this case it creates the “Deli Wall Illusion”
Anthony J Greene 21
Anthony J Greene 22
Understanding Acuity: Spatial Frequency Analysis
• Measuring visual acuity:
– Eye doctors use distance (e.g., 20/20)
– Vision scientists use visual angle
Anthony J Greene 23
Understanding Acuity: Spatial Frequency
Anthony J Greene 24
Understanding Acuity: Spatial Frequency
Anthony J Greene 25
Describing Processes in V1: Spatial Frequency Analysis
Anthony J Greene 26
Describing Processes in V1: Spatial Frequency Analysis (cont.)
• Orientation
• Frequency
• Contrast
Orientation
Decreasing Contrast
Anthony J Greene 27
Spatial Frequency Analysis (cont.)
• Fourier - French mathematician, came up with theory that one can create any complex wave through a summation of Sinusoids (or sub-parts, sub-waves)
• Fourier Analysis divides all orientation specific cells in V1 according to the width of their receptive fields or Spatial Frequency
• 1) Low
• 2) Medium
• 3) High
Anthony J Greene 28
Spatial Frequency Analysis (cont.)• Neurons can then be viewed as Spatial Filters which
separately analyze differing levels of detail or scale
• Any scene can then be decomposed into images with varying spatial frequencies - low frequency images are blurry and only the most prominent features are represented - high frequency images exaggerate the fine details
• Construing form vision in terms of an emergent property of these different scales of receptors is referred to as the Multichannel Model
Anthony J Greene 29
Spatial frequency Analysis (cont.)
Anthony J Greene 30
Spatial frequency Analysis (cont.)
Anthony J Greene 31
Spatial frequency Analysis (cont.)
• Once divided by width, cells can further be grouped according to their orientation specificity
• This allows a vastly simplified organization of neural activity - 3 major variables - Spatial Frequency, Orientation & Contrast
• Additionally, Fourier analysis helps explain how individual cells may contribute information to the aggregate
Anthony J Greene 32
Spatial frequency Analysis (cont.)
Anthony J Greene 33
Spatial frequency Analysis (cont.)
Anthony J Greene 34
Spatial Frequency Analysis (cont.)
• 1f gives the fundamental waveform
• 2f ... xf : are called harmonics - increasing details
Anthony J Greene 35
Spatial Frequency Analysis (cont.)
Anthony J Greene 36
Spatial Frequency Illusions
Anthony J Greene 37
Spatial Frequency Illusions
Anthony J Greene 38
Spatial Frequency Illusions
Anthony J Greene 39
Spatial Frequency Illusions
Anthony J Greene 40
Color at V1
• Among cells selective for orientation are patches of cells selective for color (and not orientation), which are known as Blobs.
• Other cell (orientation specific cells) regions are known as interblobs.
Anthony J Greene 41
Organization of V2
• Thin Stripes receive information from Blobs and pass it to V4
• Thick Stripes recieve information from complex cells and send it to V5 and V3
• Interstripes recieve information from simple cells and send it to V3 and V4
• Information at V2 is 3-D
Anthony J Greene 42
Nerve or Cortical Damage1) Retina / Optic Nerve2) Optic Chiasm3) Optic Tract4) V1/V2
Anthony J Greene 43
Nerve or Cortical DamageRetina/Optic Nerve: Monocular blindness
Anthony J Greene 44
Nerve or Cortical DamageOptic Chiasm: Nasal field blindness
Anthony J Greene 45
Nerve or Cortical DamageOptic Chiasm: Bitemporal field blindness
Anthony J Greene 46
Nerve or Cortical Damage
Optic Tract/LGN/Radiations: Homonymous Blindness
Anthony J Greene 47
Nerve or Cortical DamageV1: Quadrantic blindness
Anthony J Greene 48
Nerve or Cortical Damage
V1/V2:
• Scotoma
• Complete blindness
• case of Blindsight
Anthony J Greene 49
Higher Visual Areas
• V3: Form & Dynamic Form
• V4: Color
• V5: Motion
• IT: What System: Object Recognition– Lingual Gyrus of IT: Face Recognition
• PP: Where System: Object Location and Navigation
Anthony J Greene 50
Simplified Functional Visual Anatomy
Type Lobe Area Function
Primary Occipital V1 (17) Initial Processing
Secondary Occipital V2 (18) 3-D Form
Secondary Occipital V3 (19) Dynamic Form
Secondary Occipital V4 (19) Color & Form
Secondary Occipital V5/MT (37) Motion
Tertiary Temporal IT (20, 21, 22) "What"
Tertiary Parietal PP (7) "Where"
Anthony J Greene 51
Simplified Flow Diagram of the Visual System
LGN
Parvo
Magno
Thalamus Occipital Lobe
Temporal Lobe
Parietal Lobe
V1 V2
V3
V4
V5 or MT IT
Optic Nerve
PP
Anthony J Greene 52
Summary
Color Form Motion
Eye / LGN P P M
V1 Blobs Simple Complex
V2 Thin Stripes Interstripes Thick Stripes
V4Form &Color
V3Dynamic
Form
V5Motion
} } }
Anthony J Greene 53