Digital Image Processing - COMP4173

Lecture 1: Course Introduction

Prof. Hongjian Shi (时红建）

Department of Computer Science and Technology
BNU-HKBU United International University
Email: [email protected]
Office: T3-601-R3
Office Hours: Tues., Wed., Thur. 9:00-11:50; Wed. 16:00-16:50
1
Concept of Images
• Photo, picture, drawing, painting, map, scenario, fax, pdf,
microscopy, electrocardiogram, negative, US, CT, MRI etc. – image
• Perception of human to natural and artificial environment from
images
Images play the single most important role in human perception.
One picture is worth more than ten thousand words.

2
 • 𝑖(𝑥, 𝑦)-the amount of illumination
What is an image? incident to the scene
• 𝑟(𝑥, 𝑦) – the reflectance from the
objects
• 0 ≤ 𝑓 𝑥, 𝑦 ≤ 255 Pixel

• What is an image?

3
• Reflection images
• Emission images
• Absorption images

4
 280-780 nm

Infrared Radiation

1 angstrom = 0.1 nm = 10-10 m

5
Categories of Applications
• Two principal applications of Digital Image Processing:
• Improve pictorial information including
• General picture quality
• Entertainment (movie, video, multimedia)
• Medical image diagnosis
• Product quality check
• Information extraction/expression/implication
• Process image data by a digital computer for
• Storage, transmission, and display
• Representation for autonomous machine perception (essential of AI)

6
Imaging machines cover almost all entire EM spectrum

7
Image Digitization and Quantization

1 bit, 2 bits, 6 bits, 8 bits, 16 bits images

8
Fingerprint

9
• Low level processing – noise reduction, contrast enhancement, image sharping,
resizing etc.
• Mid level processing – segmentation, registration, object recognition, edge detection,
information extraction etc.
• High level processing – fingerprint, face detection, tracking, image reconstruction etc.

10
 7 4
6

3 5

2
1

These materials will be covered

in class.

11
Why do we study digital image processing?
• Image and video are the major communication media that people
perceive the world – a picture is worth 10000 words
• For almost all applications, images and videos are indispensable in
our daily life such as medical diagnosis, TV broadcast, teleplays,
films, scene capture, classification, object identification, intelligent
information decision, storage etc.
• It is interesting and funny!

12
Origin of Image Processing
• Newspaper - transmission
• Printing of picture quality
• Birth of modern digital computers (1940s by John von Neumann) -
memory to hold programs and data, and conditional branching)
• Simple history:
• 1948, invention of transistors, Shannon information theory
• 1950-1960, high level programming languages COBOL, FORTRAN
• 1958, invention of integrated circuits (IC) at Texas Instruments
• 1960 – digitalization of images
• 1964 – improve image quality starting at Jet Propulsion Lab (Pasadena, CA)
• 1970-75, Intel microprocessors
• 1970s – medical image processing, remote image processing (79 Nobel Price, HU)
• 1980 – very large scale integrated circuits(VLSI)
13
Applications
• Business and entertainment
• Security and surveillance
• Radar and Sonar
• Remote sensing
• Medical imaging and image processing
• Robotics and navigation
• Automatic character recognition
• ………

14
Computer vision (images)
• Computer vision is the science and technology of machines that see.
• Building artificial systems that obtain information from images,
videos, multiple cameras, or medical scanners. For examples,
• Controlling process – robot, autonomous vehicle, unmanned helicopter etc.
• Detecting events – visual monitoring, counting
• Classification – labeling, indexing databases of images or videos
• Modeling objects
• Interaction of human to computer
• Medical computer vision or medical image processing
• Quality control of manufacturing process
• Visual Effects creation for cinema and broadcast
• …………..
15
Face detection
• Determine the locations, sizes of human faces in images, any
facial features but ignore any other features.
• Determine the number of faces inside the image
• Determine if a portion or the whole image contain a face or not
• Video surveillance for possible face identification

16
Feature detection
• Low level operations
• Edge detection
• Corner and interest point detection
• Blobs or regions of interest
• Ridges such as ridges supporting teeth

17
Application – driving and parking
• Departure warning system
• Guide of car parking – not available in China
• Recording cars in or out of a parking lot
• Automatic unmanned car driving
• Warning system in case of some objects are close to a car

18
Recognition
• Finger Prints
• Counterfeiting
• License Plate Reading

19
Non-photorealistic rendering
Cartoon Artistic style:

- Expressive styles for digital art

- Digitalization of painting pictures 20
Medical image processing (CT, MRI, US)

21
22
Molecule imaging

Cryoelectron microscope （冷冻电镜）

23
Business and entertainment

24
Security and law enforcement

Finger print and family origin? (just thinking)

25
Radar and sonar

26
Robotics and navigation

27
Automatic character recognition

Signature!

28
Remote sensing （1）

Aerial views of Frankfurt, Germany, 1909. Pictures taken by pigeon retrofitted with camera.
Frame rate: 1 picture every 5 minutes

29
Remote sensing (2)

30
What is video?
• A video is the continuous displaying of
still images.

31
Digital videos

VCR, DVD, Video Prefilters before

TV broadcasting Video codecs
players encoding

Video denoising:
Intrafilters inside Postfilters after deblocking,
Size conversion
codec decoding deblurring,
deringing, deflicking

Deinterlacing

32
 Image
Processing
System

33
Course Contents
• Human vision and image acquisition
• Digital image concepts, and structures
• Image processing in spatial domain
• Image processing in frequency domain
• Image restoration
• Morphological processing
• Segmentation and its applications
• Color image processing
• Image and video formats and standards

34
Main Reference Textbooks
• R. C. Gonzalez and R. E. Woods, Digital Image Processing, Third
edition 2008, Fourth edition 2017.
• Assignments are from this book, this book has English version at 京东
https://item.jd.com/12113248.html.
• A. C. Bovik, Handbook of Image and Video Processing, Academic
Press, Second Edition, 2005.
• J. Woods, Multidimensional Signal, Image, and Video Processing
and Coding, Second Edition, 2011

35
Laboratory Class
• Lab is essential and important!
• The concepts, algorithms or procedures lectured in class will be
implemented in laboratory class.
• Guideline instructions to perform class projects will be given at the
beginning of laboratory time.
• Some explanations or summaries will be given through
experiments.
• The laboratory time is only for doing labs for this course.

36
 General Information
• Time and location:
• Monday 16:00-17:50, T7-301 (Lecture time)
• Thursday 15:00-16:50 am, T29-101 (Lab and Tutoring)
• Office Hours: Tues., Wed., Thurs., 9:00-11:50, Wed. 16:00-16:50
• TA: Chengyan Lin –email: [email protected]
• Any programming language, preferred programming language: C++
with OpenCV (teaching) or MATLAB
• Pre-requisites:
• Linear Algebra
• Probability and Statistics

37
Grading Policy
• 10 Assignments (25%)
• Due at next week lecture time
• 10 labs (50%)
• Due at next week lab time
• Each one must write your own codes and word reports including
questions, algorithms or methods or procedures, implementation results,
conclusion
• Submit at iSpace for grading before 10 pm at the due dates
• One final project (25%)
• For course projects, 2-3 persons form a team to do the projects for
undergraduates and graduates must do their final project individually
• Submit both electronic and paper versions
38
Late Policy (Assignments, labs and project)
• One week: 80%
• Two weeks: 70%
• Three weeks: 60%
• Four weeks or later before class over, 50% but you must do yourself
• This policy is for labs, course project, and assignments
• Please save returned emails in case of communication problems
• Zero tolerance on any cheating on assignments and projects

39