Digital Image Processing Fundamentals

• The course will explore the fundamentals of grayscale and color images, along with various image formats
• It will also cover theoretical foundations of image processing techniques
• This includes their implementation through lab exercises
• The course aims to provide sufficient knowledge for learning advanced concepts and related fields through self-study or graduate programs

Course Pre-requisite

• CCCS314 Design and Analysis of Algorithms

Course Textbook

• Digital Image Processing, 4th Edition, Global Edition, by Rafael Gonalez and Richard Woods, (4th Ed) 2018, ISBN-10: 1292223049

Course Reference Materials

• Feature extraction and image processing for computer vision, 3rd Edition, 2012, by Mark Nixon, and Alberto Aguado, Academic Press, ISBN-10: 0123965497
• Documentation of MATLAB Image Processing Toolbox: https://www.mathworks.com/products/image.html
• Fundamentals of Digital Image Processing by Anil K. Jain, Prentice Hall, 1989
• Maria Petrou, Image Processing: The Fundamentals, 2nd Edition, Wiley

Course Evaluation

• Assignments due weeks 4, 7, and 10
• Assignments count for 10% of the final grade
• Quizzes due weeks 3, 6, 9, and 12; contribute 10% to the final grade
• The midterm exam is in week 10, accounting for 20% of evaluation
• Project presentations in week 11, weighing 10%
• Lab exam in week 12; 20% of the final grade
• Final exam in week 14; 30% of the final grade

Course Policies

• Arrive before class starts
• Absence in over 25% of classes results in a DN grade
• Permission to retake a missed quiz/midterm for acceptable reasons is granted only once
• Quizzes' purpose is to prepare students for the mid and semester exams
• Follow honor code and avoid copying; discussions are encouraged, but copying code/text is prohibited
• Laptops for lab work are recommended
• Group work might be necessary depending on the project
• Plagiarism is unacceptable, with strict measures applied; Blackboard can detect it, and cheating lacks honor

Introduction Topics

• What is an Image, Digital Image Processing
• Digital Image Processing Domain
• The Origins of Digital Image Processing
• Application Areas
• Motivation: Why Process Images?
• Examples of Fields that Use Digital Image Processing
• Imaging in The EM Spectrum
• Image Acquisition Methods
• Fundamental steps in Digital Image Processing
• Components of a General-Purpose Image Processing System
• Digital Image Fundamentals

What is a Digital Image, Digital Image Processing

• An Image is a visual representation in form of a 2-dimensional function f(x, y)
• x and y are spatial coordinates - each (x, y) pair has an intensity (gray level) on the image
• Digital Images are composed of discrete elements (location and value) called Pixels or image elements
• Pixel intensity values in range f(x, y) ∈ [0, 255]
• Digital images can be represented as an M (rows) × N (columns) array

Digital Image Processing Domain

• Involves using computers to process, manipulate, analyze, or recognize images
• Includes image processing, image analysis and computer vision

Image Processing

• Low-level processing: noise reduction, image sharpening/smoothing, compression
• Low level processing usually does not require knowledge about the objects
• Input is Image and Output is Image

Image Analysis

• Mid-level processing: segmenting images into regions/objects, extracting meaningful information
• Mid level processing objects to reduce them to a suitable form for processing without knowledge of the objects
• Input is Image and Output is Image features/attributes

Computer Vision

• High-level processing: understanding and making sense from recognized objects; Detects faces, humans, cars
• Computer vision recognizes objects (person identification/verification), tracking in dense crowds, and is associated with artificial intelligence/machine learning
• Input is Features/Attributes extracted from an image and the Output is high level knowledge and information

The Origins Of Digital Image Processing

• The first picture was taken by Nicephore Niepce in 1826 with a primitive camera (camera obscura)
• This picture captured outside his window in Burgundy (historical region in France)

Early Applications

• Newspaper industry used digital pictures transferred by submarine cable between London and New York in the 1920s
• In 1921, digital pictures were produced from coded tape by a telegraph printer with special type faces
• The pictures had poor visual quality that was related to printing process and distribution of intensity levels (brightness levels)
• In 1922 the printing process was replaced with a new printing technique based on photographic reproduction, with Improvement on tonal quality and in resolution
• Images were initially coded with 5 gray levels that improve to the number of 15 levels in 1929
• Cable picture of Generals Pershing (right) and Foch was transmitted in 1929 from London to New York by 15-tone equipment that improved distinct brightness levels
• Digital image processing began because digital images require high computational power, with meaningful tasks starting in 1960s

Application Areas

• Medical Field: Diagnostics, X-ray (or other biomedical) image enhancement
• Industrial Applications: Computer-based product inspection
• Image Forensics: Fingerprint processing, camera processing, surveillance
• Defense Systems: Recognizing and tracking enemy movements, guiding a missile, borders control
• Aerial and Satellite Image Enhancement: Agriculture, weather prediction
• Other include: Transmitting Images, Storing Images, Remote Sensing, Astronomy, Geology, Hydrocarbon Exploration, Archeology, Entertainment Industry, Machine

Motivation: Why Process Images?

• Image processing to solve many problems in different fields; to process, one must: Facilitate storage and transmission, Store efficiently an image in a digital camera when sending an image from space
• To Prepare images for display or printing: E.g., half-toning, adjust image size, color mapping, gamma-correction
• Enhancement and restoration, Noise removal qualities, sharpness, color enhancement
• To Extract information from images: image understanding (e.g., read 2-d bar codes, character recognition) or comparing images to find changes

Examples of Fields that Use Digital Image Processing

• Biometric, Security and Recognition: fingerprint recognition, face and Number plate recognition for fast moving-speed cameras or automated toll
• Object and Edge Detection detects cars/human and edges using the Canny Algorithm
• Face Morphing, Face Blurring/Pixelating techniques for privacy protection
• Image segmenting
• ImageRestoration and Noise Removal examples include https://github.com/cszn/I-RCNN

Imaging in The EM Spectrum

• Images based on radiation from the EM spectrum are the most familiar, especially those in the X-ray and visual bands
• Electromagnetic waves are propagating sinusoidal waves of varying wavelengths or a stream of massless particles
• Each massless particle travels in a wavelike pattern at the speed of light containing a certain amount of energy

Image Acquisition Methods

• Image Acquisition is the first stage of vision systems
• Examples of industrial inspection using digital image processing include: Circuit board controllers, Packaged pills, Bottles and air bubbles in a clear plastic products

Visual Band Imaging methods

• Recognition and Control: Recognition and Control examples include of image include fingerprinting, paper currency of different denominations and automated license plate recognition systems

Gamma Ray Imaging methods

• Images can be used in Nuclear Medicine, where a a bone scan is composed of a complete bone scan using gamma-ray imaging to locate sites of bone pathology, such as infections or tumors
• An example of nuclear imaging is Positron emission tomography (PET)., the Cygnus Loop in the gamma-ray band and the nuclear band

X-Ray Imaging Methods

• Medical imagery in industry and astronomy for things like: images of Chest X-ray, generating aortic angiogram, CAT scan slice of a human head, electronic circuit board and Cygnus loop

Radio Band Imaging (MRI)

• Used for medical diagnostics
• Examples include: knee and spine images

Ultraviolet-band Imaging Methods

• Used in Fluorescence Microscopy
• Examples include: normal crops, fungi

Microwave (Radar) Imaging

• Used in Satellite Imagery
• The radar collects microwave data over virtual data regardless of weather and ambient lighting conditions

Infrared Imaging

• Used in Satellite Imagery
• Provides views of the Americas/the world

Multispectral imaging and Astronomy

• Used in Satellite and Astronical imagery in the Washington D.C. area and for studying black holes

Image Acquisition Methods

• Electron Microscopy: Electron microscopes are capable of very high magnification
• While light microscopy is limited to magnifications on the order of 1000, electron microscopes can achieve magnification of 10,000
• Synthetic Images: Fractals are striking examples of computer-generated images, in which the fractal image could be grown radially out of a cent

Fundamental Steps In Digital Image Processing

• Considers any image processing task as the steps depends on the objective and output of the image processing task, to the outputs of various processes
• The various processes generally are images or image attributes
• Core processes include: Color Image, Image restoration, filtering/enhancement, ,Wavelets and other image transforms, Compression and ,Segmentation, Feature extraction and classifications

Image Processing Steps

• Image acquisition: Capture and store images in a digital form e.g., X-ray machine to capture X-ray or a phone to capture personal image
• Image enhancement: Manipulating/improving an image so the result is more suitable than the original for a specific application
• Image restoration: Improves the appearance of an image
• Color image processing: This is an area that has been gaining importance due to the increased significant use of digital images over the internet
• Wavelets are for representing images in various degrees of resolution with Resolution as number of pixel per inch
• Morphological processing: Deals with tools for extracting image components (useful in representation and description of shape)
• Compression: Size reduction/minimize size without losing quality; important for storage, time for transmission
• Segmentation: A process of partitioning an image into groups of pixels based on similarity with respect to some criteria.
• Representation and description: Transforming the raw image into suitable form, extracting the attributes/features of interest
• Image pattern classification is also called object recognition, which is the process of assigning a label to an object based on description

Components of a general-purpose image processing system

• Two elements required to acquire digital images, physical sensors and a digitizer
• Specialized image processing hardware usually consists of the digitizer hardware that performs primitive operations (arithmetic logic unit or ALU)
• The computer in an image processing system is a general-purpose computer and can range from a PC to a supercomputer
• Image processing softwareis software consists of specialized modules that perform specific tasks and contains all the mechanisms and algorithms that are used
• Mass storage stores the pixels of images during processing and falls into three main classes: Short-term, On-line, and archival storage

Digital Image Fundamentals

• Images are large matrices
• A black and white image (binary image) has only black and white colors
• Grayscale images are 2D matrices, color images have 3 layers (R, G, B) and each cell of the matrix represents a pixel
• Each pixel is quantized to a set of values, e.g. from 0-255 (1 Byte per pixel)
• Raster images are compiled pixel based, with six main image file formats (BMP, JPEG, PNG, TIFF, GIF, and Raw)
• Vector images are mathematical formula based, with four main image file formats (SVG, DXF EPS and PDF)

Explore grayscale and color images, and various image formats. Learn the theoretical foundations and implementation of image processing techniques through lab exercises. Prepare for advanced concepts and related fields through self-study or graduate programs.