Digital 

Image Processing (CS/ECE 545) 
Lecture 1: Introduction to Image 
Processing and ImageJ

Prof Emmanuel Agu
Computer Science Dept.
Worcester Polytechnic Institute (WPI)
What is an Image?
 2‐dimensional matrix of Intensity (gray or color) values
Image coordinates
Set of Intensity values are integers
Example of Digital Images
a) Natural landscape
b) Synthetically generated scene
c) Poster graphic
d) Computer screenshot
e) Black and white illustration
f) Barcode
g) Fingerprint
h) X‐ray
i) Microscope slide
j) Satellite Image
k) Radar image
l) Astronomical object
Imaging System

Example: a camera Credits: Gonzales and Woods

Converts light to image
 Digital Image?
Remember: digitization
causes a digital image to 
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

become an approximation of a real scene

1 pixel

Digital Image Digital Image

Real image Real image
(an approximation) (an approximation)
 Digital Image
Common image formats include:
 1 values per point/pixel (B&W or Grayscale)
 3 values per point/pixel (Red, Green, and Blue)
 4 values per point/pixel (Red, Green, Blue, + “Alpha” or Opacity)

Grayscale RGB RGBA

We will start with gray‐scale images, extend to color later
 What is image Processing?
 Algorithms that alter an input image to create new image
 Input is image, output is image

Image Processing
Algorithm
(e.g. Sobel Filter)

Original Image Processed Image

 Improves an image for human interpretation in ways including:
 Image display and printing

 Image editting

 Image enhancement

 Image compression
Example Operation: Noise Removal

Think of noise as white specks on a picture (random or non-random)

Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Examples: Noise Removal
Example: Contrast Adjustment
Example: Edge Detection
Example: Region Detection, 
Segmentation
Example: Image Compression
Example: Image Inpainting

Inpainting? Reconstruct corrupted/destroyed parts of an image

Examples: Artistic (Movie Special )Effects
Applications of Image Processing
 dd
Applications of Image Processing
 dd
 Applications of Image Processing: Medicine
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Original MRI Image of a Dog Heart Edge Detection Image
Applications of Image Processing
 dd
 Applications of Image Processing: 
Geographic Information Systems (GIS)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

 Terrain classification
 Meteorology (weather)
 Applications of Image Processing: Law 
Enforcement
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

 Number plate recognition for speed cameras or 
automated toll systems
 Fingerprint recognition
Applications of Image Processing: HCI

 Face recognition
 Gesture recognition
Relationship with other Fields
Key Stages in Digital Image Processing
Image Morphological
Restoration Processing

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Image Aquisition
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological
Restoration Processing

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Example: Take a picture Object

Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Image Enhancement
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological
Restoration Processing

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Example: Change contrast

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Image Restoration
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological
Restoration Processing

Example: Remove
Noise
Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Morphological Processing
Extract
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological attributes

Restoration Processing useful for
describing
image

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Segmentation
Divide
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological image into

Restoration Processing constituent
parts

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Object Recognition
Image
regions
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological transformed

Restoration Processing suitable for
computer
processing

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Representation & Description
Finds &
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Image Morphological Labels

Restoration Processing objects in
scene (e.g.
motorbike)

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
Key Stages in Digital Image Processing:
Image Compression
Reduce
Image Morphological
image size
Restoration Processing (e.g. JPEG)

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Colour Image Image
Processing Compression
 Key Stages in Digital Image Processing:
Colour Image Processing
Image Morphological
Restoration Processing

Image
Segmentation
Enhancement

Image Representation
Acquisition & Description

Object
Problem Domain
recognition
Consider color
Colour Image Image
images (color
models, etc) Processing Compression
Mathematics for Image Processing
 Calculus
 Linear algebra
 Probability and statistics
 Differential Equations (PDEs and ODEs)
 Differential Geometry
 Harmonic Analysis (Fourier, wavelet, etc)
About This Course
 Image Processing has many aspects
 Computer Scientists/Engineers develop tools (e.g. photoshop)
 Requires knowledge of maths, algorithms, programming
 Artists use image processing tools to modify pictures
 DOES NOT require knowledge of maths, algorithms, programming

Example: Knoll Light Factory photoshop plugin

Example: Portraiture photoshop plugin Example: ToonIt

photoshop plugin
 About This Course
 Most hobbyists follow artist path. Not much math!
 This Course: Image Processing for computer scientists and 
Engineers!!!
 Teaches concepts, uses ImageJ as concrete example
 ImageJ: Image processing library
 Includes lots of already working algorithms, 
 Can be extended by programming new image processing techniques
 Course is NOT
 just about programming ImageJ
 a comprehensive course in ImageJ. (Only parts of ImageJ covered)
 about using packages like Photoshop, GIMP
 About This Course
 Class is concerned with:
 How to implement image processing algorithms
 Underlying mathematics
 Underlying algorithms
 This course is a lot of work. Requires:
 Lots of programming in Java (maybe some MATLAB)
 Lots of math, linear systems, fourier analysis
Administrivia: Syllabus Summary
 2 Exams (50%), 5 Projects (50%)
 Projects:
 Develop ImageJ Java code on any platform but must work in Zoolab machine
 May discuss projects but turn in individual projects
 Class website:  http://web.cs.wpi.edu/~emmanuel/courses/cs545/S14/
 Text:
 Digital Image Processing: An Algorithmic Introduction using Java by Wilhelm Burger 
and Mark J. Burge, Springer Verlag, 2008
 Cheating: Immediate ‘F’ in the course
 My advice:
 Come to class
 Read the text
 Understand concepts before coding
Light And The Electromagnetic 
Spectrum
just a particular part of electromagnetic 
Light: 
spectrum that can be sensed by the human eye
The electromagnetic spectrum is split up according to 
the wavelengths of different forms of energy
Reflected Light
The colours humans perceive are determined by  
nature of light reflected from an object
For example, if white light
(contains all wavelengths)  
is shone onto green object 
it absorbs most wavelengths  Colours
Absorbed
absorbed except green 
wavelength (color)
Electromagnetic Spectrum and IP
 Images can be made from any form of EM radiation
Images from Different EM Radiation
 Radar imaging (radio waves)
 Magnetic Resonance Imaging (MRI) (Radio waves)
 Microwave imaging
 Infrared imaging
 Photographs
 Ultraviolet imaging telescopes
 X‐rays and Computed tomography
 Positron emission tomography (gamma rays)
 Ultrasound (not EM waves)
 Human Visual System: Structure Of 
The Human Eye
The lens focuses light from objects onto the retina
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Retina covered with 
light receptors called 
cones (6‐7 million) and
rods (75‐150 million)
Cones concentrated 
around fovea. Very  
sensitive to colour
Rods more spread out 
and sensitive to low illumination levels
Image Formation In The Eye
Muscles in eye can change the shape of the lens 
allowing us focus on near or far objects
An image is focused onto retina exciting the rods and 
cones and send signals to the brain
Image Formation
 The Pinhole Camera (abstraction)
 First described by ancient Chinese and Greeks (300‐400AD)
Thin Lens
Brightness Adaptation & 
Discrimination
The human visual system can perceive approximately 
1010 different light intensity levels
However, at any one time we can only discriminate 
between a much smaller number – brightness adaptation
Similarly, perceived intensity of a region is related to the 
light intensities of the regions surrounding it
 Brightness Adaptation & 
Discrimination: Mach Band Effect
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Perceived intensity
overshoots or undershoots
at areas of intensity change
 Brightness Adaptation & 
Discrimination
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

An example of simultaneous contrast

All inner squares have same intensity but appear darker as outer
square (surrounding area) gets lighter
 Image Acquisition
Images typically generated by illuminatinga scene
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

and absorbing energy reflected by scene objects 
 Image Sensing
Incoming energy (e.g. light) lands on a sensor material  
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

responsive to that type of energy, generating a voltage
Collections of sensors are arranged to capture images

Imaging Sensor

Line of Image Sensors

Array of Image Sensors
Spatial Sampling
 Cannot record image values for all (x,y)
 Sample/record image values at discrete (x,y)
 Sensors arranged in grid to sample image
Image (Spatial) Sampling
A digital sensor can only measure a limited number of 
samples at a discrete set of energy levels
 Sampling can be thought of as:
Continuous signal x  comb function
 Image Quantization
Quantization: process of converting continuous analog
signal into its digital representation
Discretize image I(u,v) values 
Limit values image can take
 Image Sampling And Quantization
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Sampling and quantization generates  
approximation of a real world scene
Image as Discrete Function
Image as a Function
 Representing Images
 Image data structure is 2D array of pixel values
 Pixel values are gray levels in range 0‐255 or RGB colors
 Array values can be any data type (bit, byte, int, float, 
double, etc.)
Spatial Resolution
The spatial resolutionof an image is determined by 
how fine/coarse sampling was carried out
Spatial resolution: smallest discernable image detail 
 Vision specialists  
talk about image resolution
 Graphic designers  
talk about dots per 
inch (DPI)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Spatial Resolution
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Spatial Resolution: Stretched Images
Intensity Level Resolution
Intensity level resolution: number of intensity levels 
used to represent the image
 The more intensity levels used, the finer the level of detail 
discernable in an image
 Intensity level resolution usually given in terms of number 
of bits used to store each intensity level

Number of Intensity
Number of Bits Examples
Levels
1 2 0, 1
2 4 00, 01, 10, 11
4 16 0000, 0101, 1111
8 256 00110011, 01010101
16 65,536 1010101010101010
 Intensity Level Resolution
256 grey levels (8 bits per pixel) 128 grey levels (7 bpp) 64 grey levels (6 bpp) 32 grey levels (5 bpp)
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

16 grey levels (4 bpp) 8 grey levels (3 bpp) 4 grey levels (2 bpp) 2 grey levels (1 bpp)
 Saturation & Noise
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Saturation: highest intensity

value above which color is
washed out

Noise: grainy texture pattern

Resolution: How Much Is Enough?
The big question with resolution is always how much 
is enough?
 Depends on what is in the image (details) and what 
you would like to do with it (applications)
 Key questions:
 Does image look aesthetically pleasing?
 Can you see what you need to see in image?
Resolution: How Much Is Enough?

Example: Picture on right okay for counting number 
of cars, but not for reading the number plate
Images taken from Gonzalez & Woods, Digital Image Processing (2002)

Low Detail
Medium Detail
Intensity Level Resolution

High Detail
Image File Formats
 Hundreds of image file formats. Examples
 Tagged Image File Format (TIFF)
 Graphics Interchange Format (GIF)
 Portable Network Graphics (PNG)
 JPEG, BMP, Portable Bitmap Format (PBM), etc
 Image pixel values can be 
 Grayscale: 0 – 255 range
 Binary: 0 or 1
 Color:  RGB colors in 0‐255 range (or other color model) 
 Application specific (e.g. floating point values in astronomy)
How many Bits Per Image Element?
 Introduction to ImageJ
 ImageJ: Open source Java Image processing software
 Developed by Wayne Rasband at Nat. Inst for Health (NIH)
 Many image processing algorithms already implemented
 New image processing algorithms can also be implemented easily 
 Nice click‐and‐drag interface

Wayne Rasband (right)

ImageJ: Key Features
 Interactive tools for image processing of images 
 Supports many image file formats (JPEG, PNG, GIF, TIFF, 
BMP, DICOM, FITS)

 Plug‐in mechanism for implementing new 
functionality, extending ImageJ

 Macro language + interpreter: Easy to implement 
large blocks from small pieces without knowing Java
 ImageJ Software Architecture
 ImageJ uses Java’s windowing system (AWT) for display
 Programmer writes plugins to extend ImageJ
 Already implemented plugins available through ImageJ’s 
plugins menu
ImageJ Plugins
 Plugins: Java classes that implement an interface 
defined by ImageJ
 Two types of plugins
 Plugin: Requires no image to be open first
 PlugInFilter: Passed currently open image, operates on it
 We will mostly focus on PlugInFilters
 Two methods defined 
 int setup(String arg, ImagePlus im):
 Does initialization, verifies plugin capabilities matches input image
 int run(ImageProcessor ip):
 Does actual work. Passed image (ip), modifies it, creates new images
First ImageJ Example: Invert Image
 Task: Invert 8‐bit grayscale (M x N) image
 Basically, replace each image pixel with its complement

 We shall call plugIn My_Inverter
 Name of Java Class: My_Inverter
 Name of source file: My_Inverter.java
 “_” underscore makes ImageJ recognize source file as plugin
 After compilation, automatically inserted into ImageJ menu 
First ImageJ Example: Invert Image

Indicates plugIn handles

8-bit grayscale images

Retrieves width and

height of input image

Loops over all image pixels

Sets each pixel to its compliment
(255 – original pixel value)
 Compiling ImageJ Plugins
1. Place plugIn source code (My_Inverter.java) in sub‐
directory of ImageJ install location <ij>/plugins/
2. Open grayscale image from samples (since plugin requires 
image to be open)
3. Compile in run plugin by going to menu 
Plugins->Compile and Run…

 Note: On startup, ImageJ loads all plugins in the <ij>plugins/ 
sub‐directory
 ImageJ can also be used with eclipse IDE (large programs)
 References
 Wilhelm Burger and Mark J. Burge, Digital Image 
Processing, Springer, 2008
 University of Utah, CS 4640: Image Processing Basics, 
Spring 2012
 Gonzales and Woods, Digital Image Processing (3rd
edition), Prentice Hall
 Digital Image Processing slides by Brian Mac Namee