Brightness/Lightness and Edge Detection terms

Question 1

Radiance

Answer 1

adiant power from a light source, in order to see we need light
units: lumen (light produced by a standard candle), candela = 1 lumen per steradian (squared radians, unit of solid angle that is conical in shape)

Question 2

Illuminance

Answer 2

amount of light falling on a surface, unit= lux (1 lumen per square meter of area)

eg. daylight = 10 000 lux, full moon - 0.1 lux

Question 3

Luminance

Answer 3

amount of light reflected from a surface. Unit = nit (1 candela per square metre of area)

eg. typical active-matrix LCD panel = 200-300 nits, typical CRT monitor 50-125 nits

Question 4

Reflectance

Answer 4

proportion of light reflected from a matter surface. Unit: percent (%) or ‘albedo’ (luminance/illuminance x 100)

eg. white paper = 90%, black paper = 10%

Question 5

Brightness

Answer 5

perceptual impression of intensity of light source

Question 6

Lightness

Answer 6

perceptual impression of surface ‘greyness’, psychological counterpart to reflectance of a surface

eg. white paper = high lightness, black paper = low lightness

Question 7

Ratio Principle

Answer 7

Percentage of light reflected determines perception of lightness

Question 8

Lightness Constancy

Answer 8

ability to perceive the true reflectance properties of an object - no matter what the illumination is

eg. looking at white and black paper on a partially lit desktop, the black paper in the light has greater luminance than the white paper in the dark yet we still maintain lightness constancy, see this results for unfamiliar stimuli so not just memory colour

Question 9

Retinex Theory

Answer 9

retina and cortex involved, based on two natural constraints:
1) variations in illumination are gradual within an object,
2) changes in reflectance are abrupt between objects - this defines an edge

eg. Adelson’s Checkerboard: see one square as lighter due to shadows even thought they are the same colour!

Question 10

Convolution

Answer 10

matching a certain receptive field (or feature detector) with an image - matching stimulus with thing thats detecting it.
Does two things: 1) smooths out any noise or small differences in illumination and 2) detects edges

Question 11

Reconstitution

Answer 11

to construct anew, via deconvolution. Activity is spead from an edge smoothing out all small changes in illumination (or reflectance). Process distorts things a little bit can result in incorrect lightness perception.

Creates Contrast illusions (simultaneous contrast, Chevreul Illusion, COCE Effect) - incorrect lightness perception due to breakdown in lightness constancy - from errors in reconstitution

eg. canned orange juice + water = orange juice. Water is taking out to make it canned than added back in to make it orange juice again - not the exact same

Question 12

Raw Primal Sketch

Answer 12

detects edge segments, representation of an image so that intensity differences are made explicit, global structures, represents primitives (features) and place tokens (their locations) are determined (contours that may belong together). Ignore colour

Question 13

Full Primal Sketch

Answer 13

represents higher-level boundaries, boundaries and regions found by grouping primitive elements together

uses gestalt groupings / principles

Question 14

Zero-Crossing

Answer 14

function that crosses the baseline, produced by an edge, how we know that edge exists

Second derivative

Laplacian Operater