Module 4: Chapter 6 - Color

Image Processing
6.1 COLOR FUNDAMENTALS
Concept:
Color is the visual effect that happens when light reflects off an object and
enters our eyes. When light of different wavelengths hits our eyes, specialized
cells called cones respond to specific ranges of light wavelengths. Humans
have three main types of cones that detect red, green, and blue light, which we
often abbreviate as RGB.
Example:
Imagine looking at a rainbow. It has a sequence of colors from violet to red.
Each color in the rainbow corresponds to a different wavelength of light: violet
has a shorter wavelength, while red has a longer wavelength. When we see
red, it’s because the light has a wavelength that stimulates the “red” cones in
our eyes. The visible spectrum ranges approximately from 380 to 780 nm.

Primary Colors of Light

Three primary colors form the basis for color mixing:

Blue: ~435.8 nm

Green: ~546.1 nm

Red: ~700 nm

Module 4: Chapter 6 - Color Image Processing 1

 Fig: Color spectrum seen by passing white light through a prism.

Fig: Wavelengths comprising the visible range of the

electromagnetic spectrum.

Fig: Absorption of light by the red, green, and blue cones in the human eye as
a function of wavelength.

Characteristics:

Module 4: Chapter 6 - Color Image Processing 2

 The way we perceive color is based on three main characteristics:

1. Brightness: Subjective descriptor that embodies the achromatic notion of

intensity. Think of adjusting your phone's screen brightness - that's similar
to how we perceive the intensity of color

2. Hue: Attribute associated with the dominant wavelength in a mixture of light

waves. This is what we typically mean when we say "color" - like when you
say "that car is red"

3. Saturation: Refers to relative purity or the amount of white light mixed with
a hue. Imagine adding white paint to red paint - as you add more white, the
red becomes less saturated

Key Points:

RGB Model: On digital screens like TVs, computers, or smartphones, colors

are displayed by combining red, green, and blue light at various intensities.
This is why we talk about screens in terms of the RGB color model.

Achromatic Light: Achromatic colors are shades of gray, which have no

hue—just different levels of intensity. Think of black and white photography,
which is made up only of achromatic colors.

Chromatic Light: Colors like red, green, blue, etc., are called chromatic
because they have hue and saturation. For example, a red apple reflects red
wavelengths, which is why it appears red to us.

6.2 COLOR MODELS

Concept:
A color model is a mathematical way of describing colors using numbers. Each
model has its own system of coordinates (or values) that represent colors
based on different primary elements.

Examples of Different Color Models:

1. RGB Model:

This model uses three numbers to represent colors: one for red, one for
green, and one for blue. Each number ranges from 0 to 255. For instance:

Red is represented as (255, 0, 0), meaning maximum red, no green, and

no blue.

Module 4: Chapter 6 - Color Image Processing 3

 Green is represented as (0, 255, 0).

Blue is represented as (0, 0, 255).

Real-life example: Any digital screen you use—whether it’s a smartphone,

computer, or TV—produces colors by combining red, green, and blue light
in various intensities.

2. CMYK Model:
CMYK stands for Cyan, Magenta, Yellow, and Key (Black). It’s used in
printing because it represents the colors of inks and dyes. Unlike screens
that emit light, printed colors are produced by layering inks.
Real-life example: When you print a color photo, the printer combines
different amounts of cyan, magenta, yellow, and black inks to produce the
full range of colors.

3. HSI Model:

The HSI (Hue, Saturation, Intensity) model represents colors more like the
way we naturally see them. Hue is the type of color (like red or blue),
Saturation is the color’s vividness, and Intensity is the brightness.

Real-life example: The HSI model is often used in image editing software to
separate color from brightness. For instance, if you want to adjust only the
brightness of a photo without changing its colors, you can use the HSI
model to do this more easily.

6.5 COLOR TRANSFORMATIONS

Concept:

Module 4: Chapter 6 - Color Image Processing 4

 Color transformations are adjustments to the colors in an image to enhance or
change its appearance. These transformations can involve altering the
brightness, contrast, or color balance of an image.

Example:
If a photo appears too dark, you can increase the brightness to make it look
clearer. Similarly, if it lacks contrast (making it look flat), adjusting the contrast
can help make the details stand out.

Key Points:

Brightness/Contrast Adjustments: Brightness increases or decreases the

light level of an image, while contrast increases the difference between
light and dark areas. Think of a sunset photo that looks too dim; increasing
brightness can make the scene look more vivid.

Histogram Equalization: This is a technique that balances the brightness

across an image to make details more visible. It’s often used in medical
imaging or satellite images to bring out details in low-contrast areas. For

Module 4: Chapter 6 - Color Image Processing 5

 example, in a dimly lit X-ray image, histogram equalization can help doctors
see details that were previously hard to distinguish.

6.6 COLOR IMAGE SMOOTHING AND SHARPENING

Concept:
Image smoothing and sharpening are methods to refine an image's appearance
by reducing noise (unwanted variations in color or brightness) or enhancing
edges to make objects in the image clearer.
Examples:

1. Smoothing:

Smoothing reduces “graininess” or “noise” in an image by averaging out the

colors in small areas. This makes the image look softer.

Two main approaches:

1. Component-wise:

Like smoothing each R, G, and B channel separately

Like applying a blur filter in photo editing apps

Common filters: averaging, Gaussian smoothing

Module 4: Chapter 6 - Color Image Processing 6

 2. Vector approach:

Treats each pixel's RGB values as a single unit

Generally gives better results but more complex

Real-life example: When you take a photo in low light, it might appear
grainy. Applying a smoothing filter can help make the photo look smoother
by reducing this graininess.

2. Sharpening:
Sharpening increases the contrast at the edges within an image, making
details clearer. It’s especially useful for photos that are slightly blurry.
Laplacian is commonly used for sharpening

Common technique: Unsharp masking

1. Create a blurred copy of the image

Module 4: Chapter 6 - Color Image Processing 7

 2. Subtract it from the original

3. Add the difference back to original

Real-life example: If you take a picture of a building and it looks slightly out
of focus, sharpening can bring out the lines and edges of the building,
making it look clearer.

6.7 USING COLOR IN IMAGE SEGMENTATION

Concept:
Image segmentation is a process that divides an image into different regions or
segments based on color. It’s widely used in image processing to analyze or
classify different parts of an image.
Example:
In a photo of a fruit bowl, image segmentation can help separate different fruits
based on color—for instance, separating the red apples from the green ones.
Key Points:

HSI Model:

Module 4: Chapter 6 - Color Image Processing 8

 The HSI model is particularly useful in segmentation because the Hue
(color) component can be used to differentiate between regions more
easily. For example, in a forest photo, green leaves can be separated from
red flowers using the Hue value alone.

HSI Model separates color into Hue (type of color, 0°–360°), Saturation
(vibrancy), and Intensity (brightness), aligning with human color perception.

Advantages:

Closer to human color perception.

Separates color (Hue) from brightness (Intensity), making it less

sensitive to lighting changes.

Ideal for tasks like image segmentation and color-based object

detection.

Disadvantages:

Requires conversion from RGB, which adds computational

overhead.

Hue becomes ambiguous in low-saturation areas (grays).

Limited compatibility with some software and hardware.

RGB Model:

Module 4: Chapter 6 - Color Image Processing 9

 The RGB model can also be used for segmentation by calculating the
differences in red, green, and blue values. For instance, if you’re trying to
find blue-colored objects in a photo, you can look for pixels where the blue
value is much higher than red and green.

RGB Model represents color by mixing Red, Green, and Blue intensities,
commonly used in digital systems.

Advantages:

Simple and fast since most images are already in RGB format.

Universally supported by digital devices.

Performs consistently even in low-saturation areas.

Disadvantages:

Sensitive to lighting changes as color and brightness are not

separated.

Less intuitive for color differentiation tasks.

Real-life Applications:

Medical Imaging: Doctors use segmentation to identify different tissues or

organs in MRI or CT scans, where certain colors or intensities can signify
different types of tissue.

Module 4: Chapter 6 - Color Image Processing 10

 Self-Driving Cars: Image segmentation helps self-driving cars detect
objects like pedestrians, vehicles, and road signs based on color
differences.

Agriculture: Farmers use drones equipped with color segmentation

software to monitor crop health. Healthy plants often appear in shades of
green, while stressed plants may have different color tones that can be
segmented and analyzed.

Key Learnings:
1. Color Perception: Human eyes perceive color through the interaction of
light wavelengths with cone cells, detecting red, green, and blue light.

2. Primary Colors of Light: Red (~700 nm), green (~546 nm), and violet-blue
(~436 nm) form the basis of color mixing.

3. Color Models:

RGB Model: Represents colors via combinations of red, green, and blue
intensities, widely used in digital screens.

HSI Model: Separates color into Hue (type of color), Saturation

(vividness), and Intensity (brightness), aligning with human perception.

CMYK Model: Used in printing, layering Cyan, Magenta, Yellow, and

Black inks.

4. Color Transformations: Enhance image appearance by adjusting

brightness, contrast, or color balance.

5. Image Smoothing and Sharpening: Smoothing reduces noise, while

sharpening enhances edge details.

6. Image Segmentation: Divides images into regions based on color (e.g., HSI
for separating hues, RGB for direct intensity comparison), useful in fields
like medical imaging, self-driving cars, and agriculture.

Module 4: Chapter 6 - Color Image Processing 11