Digital Image Processing

V Unit notes

1. Color Models

Color models define a standard way to represent and describe colors mathematically, enabling
consistent processing and display of images. They are critical for tasks like rendering images on
devices, processing for analysis, and converting between formats.

1.1. RGB Model

• Components: Red, Green, and Blue.

• Applications: Widely used in:
o Electronic display systems (monitors, projectors, TVs).
o Cameras and scanners.
o Web and digital imaging.
• Properties:
o Additive Color Model:
▪ Colors are created by adding light of the three primary colors.
▪ Black is represented by (0,0,0), and white is represented by (255,255,255)
in 8-bit representation.
o Range:
▪ Each component typically ranges from 0 to 255 in 8-bit systems, allowing
for 16.7 million color combinations.

1.2. CMY and CMYK Model

• Components: Cyan (C), Magenta (M), Yellow (Y), and Black (K).
• Applications: Primarily used in:
o Printing (e.g., inkjet and offset printers).
o Color separation for publishing.
• Properties:
o Subtractive Color Model:
▪ Colors are produced by subtracting the reflected light from white.
▪ Black (K) is added to compensate for imperfect inks and to enhance depth
in printing.
o Conversion:
▪ Derived from the RGB model:
▪ C = 1 - R, M = 1 - G, Y = 1 - B (normalized RGB values).
 ▪ K is calculated to improve contrast.

1.3. HSV and HSI Models

• Components:
o HSV: Hue (H), Saturation (S), Value (V).
o HSI: Hue (H), Saturation (S), Intensity (I).
• Applications:
o Image editing (e.g., color adjustment in Photoshop).
o Feature detection and segmentation in computer vision.
o More intuitive for humans to interpret and manipulate colors.
• Properties:
o Hue (H): Represents the dominant color (e.g., red, green, blue) and is measured
in degrees (0° to 360°).
o Saturation (S): Indicates the purity or vividness of the color, ranging from 0
(gray) to 1 (pure color).
o Value (V) / Intensity (I): Defines the brightness, with 0 being black and 1 being
the brightest.

1.4. YUV and YCbCr Models

• Components:
o YUV: Luminance (Y), Chrominance-U (U), and Chrominance-V (V).
o YCbCr: Luminance (Y), Blue-difference Chrominance (Cb), Red-difference
Chrominance (Cr).
• Applications:
o Television broadcasting (analog and digital).
o Video compression formats (e.g., JPEG, MPEG, and H.264).
• Properties:
o Luminance (Y): Captures the brightness information, essential for grayscale
reproduction.
o Chrominance (U and V or Cb and Cr): Represent the color difference
information.
o Advantages:
▪ Separation of luminance and chrominance allows bandwidth optimization.
▪ Luminance can be displayed independently for grayscale devices or low-
bandwidth transmission.

Comparison of Color Models

 Model Components Application Properties
Displays and imaging Additive, intuitive for
RGB Red, Green, Blue
devices electronic displays
Cyan, Magenta, Yellow, Subtractive, better for
CMY/CMYK Printing and publishing
Black reflective mediums
Hue, Saturation, Editing and perception- User-friendly, aligns with
HSV/HSI
Value/Intensity based tasks human vision
Luminance, Broadcasting and Efficient for bandwidth,
YUV/YCbCr
Chrominance compression separates brightness

2. Pseudo-Color Image Processing

Definition

Pseudo-color image processing is the technique of mapping grayscale image intensities to colors
to enhance visual interpretation. It is particularly useful in visualizing subtle differences in
intensity that are difficult to distinguish in grayscale images.

Techniques

Two primary methods are used in pseudo-color processing:

1. Intensity Slicing
o Description:
▪ Divides the intensity range of the grayscale image into intervals or
"slices."
▪ Each interval is mapped to a specific color.
o Steps:
▪ Determine the intensity range of the grayscale image.
▪ Divide the range into distinct intervals.
▪ Assign a color to each interval.
o Visualization:
▪ Intensities corresponding to different ranges (e.g., low, medium, high) are
visualized with unique colors (e.g., blue for low, green for medium, red
for high).
o Applications:
▪ Meteorological data visualization (e.g., temperature or precipitation
maps).
 ▪ Highlighting regions in medical images like CT scans.

2. Gray Level to Color Transformation

o Description:
▪ Transforms the grayscale intensity values directly into corresponding
colors based on a predefined mapping function.
o Methods:
▪ Linear Transformation: Linearly maps intensity values to a color
spectrum (e.g., black to red, green, blue).
▪ Non-linear Transformation: Maps intensities using customized functions
or lookup tables for better differentiation in specific ranges.
o Advantages:
▪ Provides smoother transitions between colors compared to intensity
slicing.
o Applications:
▪ Thermal imaging for temperature mapping.
▪ Visualizing topographical elevation data.

Applications

Pseudo-color processing is widely used in diverse fields due to its ability to highlight subtle
intensity variations:

1. Medical Imaging
o Enhances the visualization of structures in X-rays, CT scans, or MRI images.
o Example: Assigning different colors to bone, muscle, and tissue densities for easy
interpretation.
2. Satellite Imagery
o Highlights specific geographical features such as vegetation health, water bodies,
or urban areas.
o Example: NDVI (Normalized Difference Vegetation Index) uses pseudo-colors to
differentiate healthy vegetation (green) from barren land (brown).
3. Scientific Data Visualization
o Analyzing weather patterns, ocean temperatures, or geological data.
o Example: Visualizing temperature variations in meteorological maps.
4. Industrial Applications
o Visualizing stress distributions in materials or temperature profiles in thermal
imaging.

Advantages of Pseudo-Color Processing

 • Enhances interpretability of images with subtle intensity variations.
• Helps non-experts understand complex grayscale data.
• Facilitates better feature detection and segmentation.

3. Color Transformation

Definition

Color transformations involve operations that modify the color attributes of an image to enhance
or analyze it for various applications. These transformations help in improving image quality,
extracting features, and preparing images for further processing.

Common Types

1. Color Corrections
o Description: Adjusts properties like brightness, contrast, and saturation to
improve the image's visual appeal.
o Examples:
▪ Brightness Adjustment: Adds or subtracts a constant value to the intensity
of all pixels.
▪ Contrast Adjustment: Scales the intensity values to widen or narrow the
intensity range.
▪ Saturation Adjustment: Enhances or reduces the vividness of colors.
o Applications: Photo editing, video enhancement.
2. Histogram Equalization in Color Images
o Description: Equalizes the intensity histogram to enhance contrast in color
images.
o Method:
▪ Convert the image to a different color space where luminance (Y or
intensity) is separable.
▪ Apply histogram equalization only to the luminance channel, keeping the
color information intact.
o Applications: Enhancing low-light images.
3. Color Space Conversion
o Description: Transforming images from one color model to another (e.g., RGB to
HSV, RGB to YCbCr).
o Applications:
▪ Simplifying image processing tasks like segmentation and object
detection.
▪ Compression and broadcasting.

Applications

• Image Enhancement: Improves the aesthetic or perceptual quality of images.

• Feature Detection: Highlights specific features for tasks like segmentation and pattern
recognition.
4. Smoothing in Color Images

Definition

Smoothing reduces noise and removes sharp transitions in color images, producing a cleaner and
more visually appealing image. It is a fundamental preprocessing step in image analysis.

Methods

1. Averaging Filters
o Description: Replaces each pixel value with the average of its surrounding pixel
values.
o Effect: Blurs the image, reducing random noise.
o Limitation: May result in loss of edge details.
2. Gaussian Filters
o Description: Applies a weighted average, where central pixels have a higher
weight.
o Effect: Smooths the image while preserving overall structure.
o Applications: Suitable for removing Gaussian noise.
3. Median Filters
o Description: Replaces each pixel value with the median value of its
neighborhood.
o Effect: Excellent at preserving edges while removing impulse (salt-and-pepper)
noise.
o Applications: Commonly used in real-world noisy environments.

Applications

• Noise Reduction: Improves the quality of images in photography.

• Preprocessing: Prepares images for segmentation and edge detection.

5. Sharpening in Color Images

Definition

Sharpening enhances edges and details in images, making them appear clearer and more defined.
It is often used to emphasize important features in images.

Methods
 1. Laplacian Filters
o Description: Computes the second derivative of intensity, highlighting regions of
rapid intensity change.
o Effect: Enhances edges but can amplify noise.
2. Unsharp Masking
o Description: Enhances edges by subtracting a blurred version of the image from
the original image.
o Effect: Produces a clearer and sharper image.
o Applications: Used in photography and printing.
3. High-Frequency Emphasis Filters
o Description: Combines a high-pass filter with the original image to enhance
high-frequency details (edges).
o Effect: Sharpens while retaining overall image content.

Applications

• Medical Imaging: Highlights important structures like tumors or blood vessels.

• Satellite Images: Enhances geographical features for analysis.

6. Image Representation in Color Processing

Bit Depth

• Definition: The number of bits used to represent each pixel determines the number of
possible colors.
• Examples:
o 8-bit: 256 colors.
o 24-bit: 16.7 million colors (8 bits each for R, G, and B).

Storage Formats

• JPEG: Compressed format, suitable for photographs.

• PNG: Lossless compression, ideal for graphics.
• BMP: Uncompressed, large file size.
• TIFF: High-quality format for professional use.

Pixel Composition

• Definition: Each pixel in a color image contains values for each color component in the
chosen color model (e.g., RGB, CMYK).
• Example: In a 24-bit RGB image, each pixel is represented by three 8-bit values.
Applications of Color Image Processing

1. Medical Imaging
o Detecting tumors and abnormalities in X-rays, CT scans, and MRI images.
o Enhancing image clarity for better diagnosis.
2. Remote Sensing
o Monitoring vegetation health, urban growth, and environmental changes.
o Analyzing satellite images to detect patterns and features.
3. Entertainment
o Editing and enhancing movies, video games, and animations.
o Creating realistic visual effects.
4. Surveillance
o Object tracking and recognition in security systems.
o Enhancing visibility in low-light conditions.
5. Image Compression
o Reducing file sizes for efficient storage and transmission.
o Used in formats like JPEG and MPEG for images and videos.