Chapter 1
INTRODUCTION
Need of Digital Image Processing
Improvement of pictorial information for human interpretation
Processing of image data for tasks such as storage, transmission and extraction of
pictorial information
Objectives
To define the scope of the field that we call image processing.
To give a historical perspective of the origins of this field.
To present an overview of the state of the art in image processing by the proper
examining some of the principal areas in which it is applied.
To discuss briefly the principal approaches used in digital image processing.
To give an overview of the components contained in a typical, general-purpose
image processing system.
To provide direction to the literature where image processing work is reported.
1.1 What is Digital Image Processing? Some Important Terms
Image: Intensity or Gray Level: Digital Image:
An image may be defined The amplitude of f at any of When x, y, and the all other
as a two-dimensional pair of coordinates (x, y) on intensity values of f are all
function, f (x, y), where x is called the intensity or gray finite, discrete quantities,
and y are spatial (plane) level of the image at that we call the image a digital
coordinates point. image.
Digital Image Processing: Pixels:
The field of digital image Note that a digital image is composed of a finite number of
processing refers to elements, each of which has a particular location and value.
processing digital images These elements are called picture elements, image elements,
by means of machine i.e. pels, and pixels. Pixel is the term used most widely to denote
a digital computer. the elements of a digital image.
Human Vision vs Imaging Machines
• Vision is the most important of our senses. • Imaging machines cover almost the
• Images play the most important role in entire EM spectrum
human perception. • They can operate on images generated
• Human Vision is limited to visible by sources that humans can not.
spectrum only. • These include ultrasound, electron
microscopy etc.
Image Analysis vs Computer Vision
• Understanding or interpreting image • Advanced stage after image analysis
content is its main task • Emulate human vision and take
• Meaningful information from the image decisions.
(e.g., shapes, objects, intensity) • Intelligent behaviour based on image
• Examples: Object Detection, understanding.
Segmentation, Feature Extraction • Self-driving cars, facial recognition with
response, AI decision-making
Levels of Digital Image Processing
Low–Level Processing Mid-Level Processing High-Level Processing
• Input: Raw Image • Input: Enhanced Image • Input: Extracted features
• Output: Improved Image • Output: Image from mid-level
• Tasks: Attributes(Edges etc) • Output: Meaning, level,
• Noise • Tasks: class
Reduction(Filters) • Segmentation (divide • Tasks:
• Contrast Enhancement image into parts) • Object Recognition
• Improves image quality • Edge Detection • Scene Understanding
at pixel level Histogram • Feature Extraction • Matching Patterns
• Image Acquisition • Object Boundary • Image Interpretation
• Image Smoothening detection • Decision Making
and Sharpening • Region labelling • Context based image
• Brightness and colour • Uses canny software retrieval
adjustment • Converts image into • Knowledge
• Direct pixel based symbolic/structural • Requires AI, pattern
operations representation recognition, deep learning
What is Digital Image Processing:
• Digital Image Processing is a process under which both input and output are images.
• In addition, it contains processes that extract attributes from images up to, and including, the
recognition of individual objects.
Consider an example of Automated Analysis of Text
It’s the process of a computer reading and understanding text from images, like scanned
documents or photos of pages. So, here it consists of 6 simple steps, let’s see:
Step 1: Image Acquisition Step 2: Preprocessing Step 3: Segmentation
• Capture or scan a page with • Clean the image (Remove • Break the image into lines,
text noise, correct glit etc) then words, then characters
• Result: Get a Raw image of • Result: Makes the image easier • Result: Find end of each letter
page for computer and start of other
Step 6: Further Step 5: Text Understanding Step 4: Feature Extraction and
Analysis(Computer Vision/ NLP) • Understand what the text mean recognition
• Classify documents(legal, • Example: Krishna Gupta • Recognise shape of letter
scientific etc) means Krishna is name and • Match them to known
• Translate , summarize data Gupta surname characters(A, B,C ….)