Department of Computer Science

Introduction to computer vision and image

processing

Chapter One
Topic Coverage
 Computer Vision
Every picture tells a story
⮚ Image carries vast amount of
information.
⮚ We, humans, are selective of what we
consume through visual sense.
⮚ The goal of computer vision is to write
computer programs that can interpret
images
⮚ Can computers match human
perception?
Yes and no (but mostly no!)
humans are much better at “hard”
things
computers can be better at “easy”
things
 Cont...
Computer vision involves enabling computers to interpret and make decisions
based on visual data from the world.
Computer vision seeks to replicate and exceed human vision capabilities by
leveraging computational algorithms, image processing, and machine learning
techniques.
It involves the development of systems and methods that allow computers to
acquire, process, analyze, and interpret visual data from the surrounding
environment.
 What is an image?
An image is a visual representation or depiction of an object, scene, concept, or
phenomenon.
In the context of computer science and digital technology, an image is typically a
two-dimensional array of pixels, where each pixel represents a small unit of color or
intensity.
Images can be captured using various devices such as cameras, scanners, or
generated digitally using computer graphics software.
 Cont…
Digital images can be in different formats, including raster and vector formats.
Raster images are composed of a grid of pixels, where each pixel contains
color information.
Common raster image formats include JPEG, PNG, and GIF.
Vector images, on the other hand, are composed of geometric shapes defined by
mathematical equations, making them resolution-independent and scalable.
Common vector image formats include SVG (Scalable Vector Graphics), EPS
(Encapsulated PostScript), and PDF (Portable Document Format).
 Cont…
• It is a 2D representation of a 3D reality
•The pattern is defined is a coordinate system whose origin is
conventionally defined as the upper-left corner of the image.
We can describe the pattern by a function f(x,y).
Image formation in the Eye
The principal difference between the lens of the eye and an ordinary optical lens is that lens of the
eye is flexible.
The process of creating images in our eyes and minds involves several complex steps that occur
within our visual system.
1. Light Enters the Eye: When we look at something, light enters our eyes through the
cornea, which is the transparent outer covering of the eye.
2. Light Refraction: The light then passes through the pupil, which is the black circular
opening in the center of the iris.
The iris adjusts the size of the pupil to control the amount of light entering the eye.
After passing through the pupil, the light is refracted by the lens, which focuses the light onto
the retina at the back of the eye.
 Cont…
3.Image Formation on the Retina: The retina contains millions of light-sensitive cells
called photoreceptors, namely rods and cones.
These photoreceptors convert light into electrical signals, which are then transmitted to the
brain via the optic nerve.
4.Signal Processing in the Retina: The photoreceptors in the retina respond differently
to different wavelengths of light and are responsible for detecting light and color.
Rods are more sensitive to low light levels and are responsible for night vision, while
cones are
responsible for color vision and function best in bright light.
 Cont…
5.Transmission to the Brain: The electrical signals generated by the photoreceptors
travel through intermediate neurons in the retina before reaching the ganglion cells.
The ganglion cells then bundle the signals and transmit them along the optic nerve to the
brain's visual processing centers.
6.Visual Processing in the Brain: The optic nerve carries these electrical signals to the
brain's visual cortex, located at the back of the brain.
Here, the brain processes the signals received from both eyes, interprets them, and constructs
a visual representation of the scene.
This process involves various areas of the brain, including the primary visual cortex and
higher- order visual areas.
 Cont…
7. Perception and Interpretation: Finally, the brain interprets the visual information
received and constructs our perception of the world around us.
This includes recognizing objects, shapes, colors, and depth, as well as forming mental
images and memories.
Image formation in the Eye
The distance between the center of the lens and the retina (called the focal length)
varies from approximately 17 mm to about 14 mm, as the refractive power of the
lens increases from its minimum to its maximum.
If h is the height in mm of that object in the retinal image, the geometry of
visualization yields 15/100=h/17 and hence h=2.55mm
Issue of Contrast
Objects appear to the eye to become darker as the background gets lighter.
The example below is a piece of paper that seems white when lying on a desk, but
can appear totally black in a lighter background
 Related fields in CV?
Computer Vision (CV) is an interdisciplinary field that intersects with various
related fields.
These fields contribute to and benefit from the advancements and applications of
computer vision.
Image Processing:
Image processing is closely related to computer vision,
providing foundational techniques for manipulating and enhancing images
before higher-level analysis.

Machine Learning:
Machine learning, particularly deep learning, plays a crucial role in computer
vision by providing algorithms and models for tasks such as image classification,
object detection, and segmentation.
 Related fields in CV?
Artificial Intelligence (AI):
AI is the broader field encompassing computer vision.
Computer vision is a specialized application of AI that focuses on visual data
interpretation.

Robotics:
Computer vision is essential in robotics for tasks such as navigation, object
manipulation, and scene understanding.
Robots equipped with vision systems can perceive and
interact with their environment.

Augmented Reality (AR) and Virtual Reality (VR):

Computer vision contributes to AR and VR by enabling real-time tracking,
recognition, and interaction with the physical world or virtual environments.
 Computer Vision Vs image
Image processing isprocessing
a field of computer science and engineering that involves the
manipulation and analysis of digital images.
Computer Vision:
Focus: The primary goal of computer vision is to enable machines to interpret and
understand visual information, similar to the way humans do.
Higher-Level Tasks: Involves higher-level tasks such as object recognition, scene
understanding, and making decisions based on visual data.
Interdisciplinary: Encompasses a broader range of tasks beyond basic image
processing.
Example: Identifying objects in an image, recognizing faces, or understanding the
context of a scene.
 Computer Vision Vs image processing
Image processing:
Focus: The primary goal of image processing is to manipulate or enhance images
for a specific purpose, often at a lower level.
Low-Level Tasks: Primarily involves low-level tasks such as noise reduction,
contrast enhancement, and sharpening.
Foundational: Provides foundational techniques for preprocessing images before
higher-level analysis in computer vision.
Example: Adjusting brightness and contrast, applying filters, or removing noise
from an image.

Integration: While they have distinct objectives, computer vision often

integrates image processing techniques as a part of its pipeline for tasks like image
preprocessing before high-level analysis.
 Application of CV and IP
Computer Vision (CV) and Image Processing (IP) have a wide range of applications across
various industries due to their ability to analyze and interpret visual information.
Applications of Computer Vision (CV):
Autonomous Vehicles:
CV enablesvehicles to perceive their surroundings, recognize
obstacles, and navigate autonomously.
Facial Recognition:
CV is applied in security systems, authentication, and
surveillance for facial recognition and
analysis.
Gesture Recognition:
Used in human-computer interaction systems to interpret gestures for controlling devices.
Video Surveillance:
CV is utilized for real-time monitoring, anomaly detection, and tracking in video surveillance
systems.
 Application of CV and
Applications of ImageIP
Processing (IP):
Medical Imaging:
IP is essential in medical imaging for tasks such as noise reduction, image enhancement, and
feature extraction.
Satellite Image Processing:
IP is applied for analyzing satellite images for environmental monitoring, urban planning, and
disaster response.
Digital Photography Enhancement:
IP techniques improve image quality, remove noise, and enhance color in digital photography.
Document Image Processing:
IP is used in OCR (Optical Character Recognition) for extracting text from scanned documents.
Image Compression:
IP algorithms compress images for efficient storage and transmission in applications such
as
multimedia and the internet.
 Different Image processing examples
Image processing involves manipulating and enhancing images for various purposes.
Noise Reduction:
Technique: Applying filters like Gaussian or Median filters.
Application: Enhancing the quality of medical images or photographs by reducing random
noise.
Contrast Enhancement:
Technique: Histogram equalization or contrast stretching.
Application: Improving visibility in low-contrast images, enhancing details.
Image Smoothing:
Technique: Applying smoothing filters (e.g., averaging or blurring).
Application: Reducing sharpness in an image, creating a more visually appealing result.
 Different Image processing examples
Image Sharpening:
Technique: Using sharpening filters (e.g., Laplacian or Unsharp Mask).
Application: Emphasizing edges and details in images.
Image Rotation and Cropping:
Technique: Rotating or cropping the image.
Application: Adjusting the orientation or composition of an image.
Image Compression:
Technique: Using compression algorithms (e.g., JPEG or PNG).
Application: Efficiently storing and transmitting images while
minimizing file size.
Color Adjustment:
Technique: Altering color balance or adjusting brightness and saturation.
Application: Correcting color casts or enhancing the visual appeal of
photographs.
Image Segmentation:
Technique: Dividing an image into segments based on certain criteria.
Application: Medical image analysis, object recognition, and computer vision
 Different Image processing examples
Edge Detection:
Technique: Using algorithms like Sobel, Prewitt, or Canny.
Application: Identifying boundaries in images, important for object
recognition.
Histogram Analysis:
Technique: Analyzing the distribution of pixel intensities.
Application: Adjusting image contrast and brightness based on the histogram.
 Fundamental steps in image processing
Image processing typically involves a series of fundamental steps to manipulate and enhance
digital images. The exact steps may vary depending on the specific task or application.
Image Acquisition:
Definition: The process of capturing an image using devices such as cameras or
sensors. Note: Image quality and characteristics depend on the acquisition device and
settings.
Preprocessing:
Purpose: Enhance image quality, correct imperfections, and prepare for further
analysis.
Steps:
Noise Reduction: Remove or reduce unwanted noise using filters.
Image Smoothing: Apply smoothing filters to reduce sharpness.
Contrast Adjustment: Enhance or normalize image contrast.
 Fundamental steps in image processing
Image Enhancement:
Purpose: Improve the visual appearance of the image for better human perception or
analysis.
Techniques:
Contrast Stretching: Expand the range of pixel intensities.
Histogram Equalization: Adjust the distribution of pixel
values. Sharpness Enhancement: Emphasize edges and details.
Image Segmentation:
Purpose: Divide the image into meaningful regions or objects.
Techniques:
Thresholding: Convert grayscale image to binary based on
intensity.
Clustering: Group pixels with similar attributes.
Edge Detection: Identify boundaries between objects.
 Fundamental steps in image processing
Image Recognition and Classification:
Purpose: Assign labels or categories to objects or regions based on extracted
features. Techniques:
Machine Learning: Train models using features for
classification. Pattern Recognition: Identify patterns in feature
space.
Postprocessing:
Purpose: Refine and finalize the processed image for specific
applications.
Steps:
Filtering: Apply additional filters for specific
effects. Color Correction: Adjust color balance and
tone.
Resizing or Cropping: Modify image dimensions for
specific requirements.
Image Display or Output:
Purpose: Present the processed image for