© Wesner, M. F.

Cue vs. Ecological Theory Cue Theories Feature extraction process (i.e., the computational construction of components) Feature extraction process (i.e., the computational construction of components) Ecological (James J. Gibson, 1950’s on) Gestalt-like approach - all items in environment are directly processed (i.e., the optic array of textures, shadows and edges are all processed directly to constitute depth and movement) Gestalt-like approach - all items in environment are directly processed (i.e., the optic array of textures, shadows and edges are all processed directly to constitute depth and movement) (1950) (1979)

Four Groups of Depth Cues 1. Pictorial 2. Oculomotor (accommodation & convergence*) 3. Kinetic (motion-produced cues) 4. Binocular Disparity* *Binocular Depth Cues

Monocular Depth Cues Pictorial Pictorial

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size b) Interposition (overlap)

A B Law of Prägnanz in operation here !

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size b) Interposition (overlap) c) Lighting & Shadowing

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size b) Interposition (overlap) c) Lighting & Shadowing d) Elevation -position of objects relative to the horizon

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size b) Interposition (overlap) c) Lighting & Shadowing d) Elevation -position of objects relative to the horizon e) Linear Perspective - “carpentered environment”

Linear Perspective - Geometric cues using converging lines and vanishing points Renaissance architect Leon Battista Alberti (1435) published an artist manual that systemized drawing in linear perspective. Renaissance architect Leon Battista Alberti (1435) published an artist manual that systemized drawing in linear perspective. Vanishing point Alberti was also involved with systemizing color layers, dark- to-full saturation-to-white, chiaroscuro. Alberti was also involved with systemizing color layers, dark- to-full saturation-to-white, chiaroscuro

Leonardo Da Vinci capitalized on the vanishing point concept..

Santa Maria delle Grazie in Milano.. at the end of a refectory hallway..

Monocular Depth Cues Pictorial Pictorial a) Size of image - usually involves a familiar size b) Interposition (overlap) c) Lighting & Shadowing d) Elevation -position of objects relative to the horizon e) Linear Perspective - “carpentered environment” f) Aerial Perspective - Rayleigh scatter

Titian - Vecellio Tiziano Bacchus & Ariadne (1520-3) Titian - Vecellio Tiziano (1488/1490 – 1576) Bacchus & Ariadne (1520-3)

Monocular Depth Cues Pictorial Pictorial a) *Size of image - usually involves a familiar size b) Interposition (overlap) c) Lighting & Shadowing d) Elevation -position of objects relative to the horizon e) Linear Perspective - “carpentered environment” f) Aerial Perspective - Rayleigh scatter g) Detail Perspective - Texture gradient (Size cue*)

Simple line textons:

Simple line textons with size cues:

Familiar size helps reduce the linear cue ambiguity

Familiar Size vs. linear perspective - “forced perspective” Size constancy gone awry..

Linear perspective conflicting with size Demo

Müeller Lyer Illusion

Linear Perspective vs. Size

Ponzo Illusion

Linear Perspective with Size, elevation....& texture

Monocular Depth Cues Pictorial Pictorial Accommodation (kinesthetic sense in ciliary muscles) Accommodation (kinesthetic sense in ciliary muscles) diverging light (within ∞)optical ∞

..although accommodation is considered a monocular depth cue, it does have a close link (or synkinesis) with binocular convergence or divergence (i.e., the accommodation/vergence ratio). convergencedivergence

Monocular Depth Cues Pictorial Pictorial Accommodation (kinesthetic sense in ciliary muscles) Accommodation (kinesthetic sense in ciliary muscles) Kinetic Cues (movement- produced cues Kinetic Cues (movement- produced cues

Slight move to the right ABAB

A A’BB’ AA’ > BB’..we perceive object A closer than B because object A moves faster than object B (i.e., equal time for image to sweep the arcs). With.. Nodal point

Nodal Point (~7 mm behind corneal apex) cornea Lens Thin lens system equivalent

Binocular Depth Cues Vergence Vergence divergenceconvergence

Binocular Depth Cues Vergence Vergence Binocular Disparity (true depth perception - stereopsis) Binocular Disparity (true depth perception - stereopsis)

Image correspondence - when the image is projected onto similar regions of the retinas in both eyes. foveas

Image noncorrespondence (disparity) - when the image is projected onto dissimilar regions of the retinas in both eyes. foveas Nodal points

Image disparity Left eye Right eye

Stereopsis makes use of image disparity:

RELE Front Screen Finger If objects are in line with one another: Crossed = image of nearer object projects to temporal side of fixation image; results in movement direction opposite that of open eye. Uncrossed = image of further object projects to nasal side of fixation image; results in movement direction that is the same as that of open eye.

Crossed vs. Uncrossed Disparity foveas Rays project to temporal hemiretinas

Crossed vs. Uncrossed Disparity foveas Rays project to nasal hemiretinas

Note: The magnitude of the disparity defines the distance from the fixation plane.. Greater crossed disparity

NOTE: Objects do NOT have to lie in visual alignment to establish a crossed or uncrossed disparity. Retinal image (non)correspondence has to do with a fixation point plane (or arc).

The geometry of stereoscopic depth. Begins with the Vieth-Müller (1862) horopter surface..

F Correspondence along the horopter surface

F Crossed disparity - fusion into one brain image

F Crossed diplopia - seeing double

F Uncrossed disparity - fusion into one brain image

F Uncrossed diplopia - seeing double

F Panum’s Fusion Area Crossed Diplopia Uncrossed Diplopia Crossed Diplopia Uncrossed Diplopia Panum’s Fusion Area

or anglyphs

F Panum’s Fusion Area Crossed Diplopia Uncrossed Diplopia Crossed Diplopia Uncrossed Diplopia Panum’s Fusion Area

A cortical binocular cell could be composed of disparate, monocular receptive fields that represent fusional space!

by definintion, within Panum’s Fusion Area

Size and Shape Constancy © Wesner, M. F.

Size constancy involves high-end processing. Yet, perceptual grouping (Prägnanz) can be achieved on a post-constancy representation (top-down influence).

A B

Notice that the central column of shapes is physically identical to those in the 3 left columns. However, using linear perspective, the central column appears as an in-depth surface. Thus, the perceived shape of these items are of a circle, not an ellipse, much like the circles in the 3 right-hand columns. Subjects will be more likely to group the central column with the circles ( post- constancy grouping). Post-constancy grouping grouping Size grouping

Size (shape) constancy are based on depth cues. Classic study by Boring done in the “hallowed halls” of Harvard University..

Monocular view Constant visual angle

Size (shape) constancy are often the culprits for many visual illusions..

Linear perspective in conflict with size..

Shape and size constancy in conflict..

Shallow bowl effect Sky is perceived as a flattened dome..

Problem: How is it that something that is registered far away (in the horizon) is judged to be closer because its apparent size is misperceived as bigger? A “top-down” nightmare. Shallow bowl effect Kaufman & Rock (1962) 1)The perceived difference in distance to the moon (registered distance) establishes the perceived size of the moon. 2)The judged distance to the moon is influenced by the perceived size of the moon (i.e., horizon is bigger than zenith.