Image Model – Notes
1. What is an Image?
An image is represented as a two-dimensional function f(x, y), where (x, y) are spatial
coordinates and f(x, y) is the intensity (brightness) at that point. In digital form, this
becomes a matrix of pixel values.
2. Illumination–Reflectance Model
The image intensity at each point is determined by two components:
f(x, y) = i(x, y) × r(x, y)
1. Illumination, i(x, y): The amount of light falling on the scene. Measured in lux (lm/m²).
Range: 0 ≤ i(x, y) < ∞.
2. Reflectance, r(x, y): The proportion of incident light reflected by objects. Unitless. Range:
0 ≤ r(x, y) ≤ 1.
3. Typical Values
Illumination (i(x, y)) examples:
- Clear sunlight: ~90,000 lux
- Cloudy day: < 10,000 lux
- Full moon: ~0.1 lux
- Commercial office: ~1,000 lux
Reflectance (r(x, y)) examples:
- Black velvet: 0.01
- Stainless steel: 0.65
- Flat white paint: 0.80
- Silver metal: 0.90
- Snow: 0.93
4. Intensity (Gray Level) Scale
The image intensity I = f(x, y) lies within a range:
L_min ≤ I ≤ L_max
• L_min: The minimum intensity. Theoretically 0, but practically determined by the smallest
illumination × reflectance combination detectable by the sensor.
• L_max: The maximum intensity, limited by maximum illumination × reflectance, and
practically by sensor saturation.
Example: In an office with ~1000 lux illumination:
- L_min ≈ 1000 × 0.01 = 10
- L_max ≈ 1000 × 1 = 1000
In practice, intensity values are normalized to [0, 1] or scaled to [0, 255] for 8-bit digital
images.
