Digital Image Processing

pancakesnightmuteAI and Robotics

Nov 5, 2013 (3 years and 9 months ago)

87 views

1
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
1/45
Digital Image Processing
Chapter 01
Instructor:Lih-Jen Kau(高立人)
Department of Electronic Engineering
National Taipei University of Technology
Sep. 23, 2010
2010/09/23
2/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
What’s a Digital Image

Definition

An image can be defined as a two-dimensional
function, f(x, y).

x and y are spatial (plane) coordinates.

Amplitude of f at any pair of coordinates (x, y) is called
the intensity or gray level of the image at that point.

Called a digital image when x, y, and the intensity
values of f are all finite, discrete quantities.

Composition of a Digital Image

Composed of a finite number of elements, called
picture elements, image elements, pels, and pixels.

Pixels is the term most widely used.
2
2010/09/23
3/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
What’s Digital Image Processing

Definition

To process digital images by means of a digital computer.

Generally speaking, both the input and output are images
when talking about image processing.

Image Processing Machines

Unlike human perception, who are limited to the visual
band of the electromagnetic (EM) spectrum.

Imaging machines can cover almost the entire EM
spectrum, ranging from gamma to radio waves.

Can operate on images generated by sources like
ultrasound, electron microscopy, and computer-
generated images.
2010/09/23
4/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Classification of Image Processing
(1/2)

Continuum from Image Processing to Computer
Vision – Three types classification

Low-level Image Processing(雜訊降低、對比增強…)

Involve primitive operations such as noise reduction, contrast
enhancement, and image sharpening.

Both the inputs and outputs are images.

Mid-level Image Processing(分割、分類)

Involves segmentation (i.e., partitioning image into regions or
objects), description, and classification (recognition)

The inputs are generally images, but its outputs are attributes
extracted from those images, e.g., edges, contours, and the
identity of individual objects.

High-level Image Processing(辨識、理解、認知)

Involves ”making sense” of an ensemble of recognized objects,
as in image analysis.

Performing cognitive functions associated with vision at the far
end.
3
2010/09/23
5/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Scope in this course

Encompasses processes whose inputs and outputs are
images (Low-level processes, like noise reduction…)

Encompasses processes that extract attributes from
images

Recognition of individual objects.
Classification of Image Processing
(2/2)
2010/09/23
6/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(1/7)

Application in newspaper industry

Pictures were sent by submarine cable between London
and New York.

The Bartlane cable picture transmission system in the
early 1920s reduced the time required to transport a
picture across the Atlantic Ocean from more than one
week to less than three hours.
Halftone image obtained by
using the system in 1920~1921
Image obtained by using the
improved techniques after 1921
4
2010/09/23
7/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(2/7)
Improved Bartlane system capable of coding images in 15 distinct levels of gray
2010/09/23
8/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Continuous tone

A continuous tone image is one where each color at
any point in the image is reproduced as a single tone.

Halftone Image

Halftone is the reprographic technique that simulates
continuous tone imagery through the use of dots, varying
either in size or in spacing.

Consider the two images in the right side

Left:Halftone dots.

Right:How the human eye would see this sort
of arrangement from a sufficient
distance.
Origins of Digital Image Processing
(3/7)
5
2010/09/23
9/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(4/7)

The Core of Digital Image Processing

Recall the definition of DIP in page P3.

The examples just cited are not considered digital image
processing results because computers were not involved
in their creation.

Much storage and computational power is required in
Digital Image Processing
2010/09/23
10/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(5/7)

The Space Program in Early 1960s(太空計畫)

The first computers powerful enough to carry out
meaningful image processing tasks was introduced.

JPL(Jet Propulsion Laboratory; Pasadena, California)
began the use of computer techniques for improving
images from a space probe in 1964.

Pictures of the moon transmitted by Ranger 7 were
processed by a computer to correct various types of
image distortion inherent in the on-board television
camera.
6
2010/09/23
11/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(6/7)
Reseau
(網狀組織)
is used for geometric corrections
2010/09/23
12/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Origins of Digital Image Processing
(7/7)

In the late 1960s and early 1970s

In parallel with space applications, digital image
processing techniques are used in medical imaging,
remote Earth resources observations, and astronomy
(天文學).

CAT(Computer Axial Tomography)

Called CT for short

Invented in the early 1970s

One of the most important events in the application of DIP in
medical diagnosis.
7
2010/09/23
13/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Electromagnetic Spectrum
(1/3)

The Electromagnetic Spectrum
High Frequency
With Higher energy
Low Frequency
With Lower energy
2010/09/23
14/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
遠紫外線
10nm
波長 名稱
r線
x線
極遠
紫外線
中紫外線
近紫外線
可見光線
近紅外線
中紅外線
微波
遠紅外線
極遠
紅外線
1nm
100nm
315nm
380nm
780nm
1000nm
1.5um
5um
100um
1mm
280nm

...

...


紫外線領域
紅外線領域
熱射線
人眼可見範圍
化學線

...

...

白色光
三稜鏡
︵分出七色光的光譜︶
往此方向波長愈長
由此方向波長愈短
Electromagnetic
Spectrum
(2/3)
8
2010/09/23
15/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Electromagnetic Spectrum
(3/3)
2010/09/23
16/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(1/15)

The Gamma-ray Images

Mainly used in nuclear
medicine and
astronomical
observations.

See right side figures

(a).Bone scan(骨骼掃瞄)

(b).Positron Emission
Tomography(PET;正子
攝影)

The small white masses
indicate a tumor in the
brain and one in the lung.
9
2010/09/23
17/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(2/15)

The X-ray Images

Mainly used in medical diagnostics
and industrial imaging.

See right side figures

(a)Chest X-ray(胸部X光)

(b)Aortic angiogram(動脈血管)

(c)Head CT(電腦斷層掃瞄)

(d)Circuit Board

(e)Cygnus Loop(天鵝座星環)
2010/09/23
18/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(3/15)

Image in the Ultraviolet Band

Mainly used in lithography(平
板印刷), industrial inspection,
microscopy(顯微鏡), lasers,
biological imaging(生物成像),
and astronomical observations.

See right side figures

(a)Normal corn(正常的玉米)

(b)Smut corn(異常的玉米)

(c)Cygnus loop(天鵝座星環)
10
2010/09/23
19/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(4/15)

Image in the Visible and
Infrared Band

Mainly used in light microscopy
(光顯微鏡), astronomy,
remote sensing, industry, and law
enforcement(執法).

See right side figures

Examples of light microcopy images

a. Taxol(Anticancer agent;抗癌因子)

b. Cholesterol(膽固醇)

c. Microprocessor(微處理機)

d. Nickel oxide thin film
(氧化鎳薄膜)

e. Surface of audio CD

f. Organic superconductor
(有機超導體)
2010/09/23
20/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Visible and Infrared Band

Remote sensing of visual processing.

LANDSAT satellite image of the Washington, D.C. area.

The numbers refer to the thematic bands in Table 1.1(see next page)
Examples of Fields that Use DIP
(5/15)
11
2010/09/23
21/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Visible and Infrared Band

LANDSAT satellite image of the Washington, D.C. area.

The numbers refer to the thematic bands in Table 1.1
Examples of Fields that Use DIP
(6/15)
2010/09/23
22/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Visible and Infrared Band

Weather observation and prediction

Figure1.11 is an image of Hurricane Katrina, one of the most devastating
storms in recent memory in the Western Hemisphere.
Examples of Fields that Use DIP
(7/15)
12
2010/09/23
23/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Visible and
Infrared Band

Infrared for Satellite Imaging

Figure1.12 shows infrared satellite
images of the Americas.

White dots provide the inventory of
human settlements.
Examples of Fields that Use DIP
(8/15)
2010/09/23
24/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Visible
and Infrared Band

Infrared for Satellite
Imaging

Figure1.13 shows infrared
satellite images of the
remaining populated part of
the word.
Examples of Fields that Use DIP
(9/15)
13
2010/09/23
25/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(10/15)

Image in the Visible and
Infrared Band

Visible Spectrum

Applications in automated
visual inspection of
manufactures goods
2010/09/23
26/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(11/15)

Image in the
Visible and
Infrared Band

Visible Spectrum

Applications in thumb
print processing –
either to enhance
them or to find
features that aid in
the automated
search of a database
for potential matches

Applications in paper
currency – for
automated counting
and law enforcement

Applications in
license plate reading
14
2010/09/23
27/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Microwave Band

Mainly used in radar applications.

See the figure below

Mountains in southeast Tibet(西藏東南方山區影像)
Examples of Fields that Use DIP
(12/15)
2010/09/23
28/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Image in the Radio Band

Mainly used in medicine and astronomy.

See the figure below – MRI images of a human

MRI – Magnetic Resonance Imaging(磁振造影)
Examples of Fields that Use DIP
(13/15)
15
2010/09/23
29/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology

Acoustic Imaging

Though imaging in the electromagnetic spectrum is
dominant by far, there are a number of other imaging
modalities that also are important, like acoustic imaging,
and computer-generated imaging.
Examples of Fields that Use DIP
(14/15)

Ultrasound Imaging

Mainly used in
medicine.

Millions of pulses and
echoes are sent and
received each second.

See the figure in the
right side – a baby in
mother’s uterus

Usually used to
determine the health
and gender of a baby
2010/09/23
30/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Examples of Fields that Use DIP
(15/15)

Computer-generated Imaging

See the figures below for some examples of computer-
generated images.
16
2010/09/23
31/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(1/9)

Two categories

Methods whose input
and output are images.

Methods whose input
may be images but
whose outputs are
attributes extracted
from those images.
2010/09/23
32/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(2/9)

Image Acquisition

Acquire a digital image, require an image sensor and an
image digitizer.

Image sensor, e.g., a monochrome or color camera that
produces an image frame every 1/30 sec.

Image Preprocessing
(filtering and enhancement)

To improve the images in ways that increase the
chances for success of the other processes or specific
applications.

Contrast enhancement, and noise reduction are usually
performed in this step.
17
2010/09/23
33/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(3/9)

Image Restoration

An area that also deals with improving the appearance
of an image.

Unlike enhancement, which is subjective, image
restoration is objective

Tend to be based on mathematical or probabilistic
models of image degradation.

Color Image Preprocessing

An area that has been gaining increasing importance
because of the significance increase in the use of digital
images over the Internet.

Color is also widely used as the basis for extracting
features of interest in an image.
2010/09/23
34/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(4/9)

Wavelets

Foundation for representing images in various degree of
resolution.

Particularly used for image data compression.
18
2010/09/23
35/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(5/9)

Compression

Deals with techniques for reducing the storage required
to save an image, or the bandwidth requires to transmit
it.

Although storage technology has improved significantly
over the past decade, the same cannot be said for
transmission capacity.

Morphological Processing

Deals with tools for extracting image components that
are useful in the representation and description of
shape.
2010/09/23
36/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(6/9)

Segmentation

Partition an image into its constituent parts or objects.

In general, autonomous segmentation (自動分割)is
one of the most difficult tasks in digital image
processing.

Usually, a basis for the success of image recognition.

Output of segmentation stage: usually is raw pixel data
constituting either the boundary of a region or all the
points in the region itself.
19
2010/09/23
37/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(7/9)

Image Representation and Description

Representation:

To transform raw data into a form suitable for
subsequent computer processing.

Must specify a method for describing the data
(called descriptor).

Description:

Also called feature selection.

Extract features that result in some quantitative information of
interest of features that are basic for differentiating on class of
objects from another.
2010/09/23
38/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(8/9)

Image Recognition and Interpretation

Recognition:

A process that assigns a label to an object based on the
information provided by its descriptor.

Interpretation:

Assign meaning to an ensemble of recognized objects.

Knowledge Base

Code the knowledge about a problem domain into an
image processing system

Guide the operation of each processing module

Control the interaction between modules
(communication between processing modules is
based on prior knowledge of what a result should be)
20
2010/09/23
39/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Fundamental Steps in DIP
(9/9)

Note

Not all modules in Fig. 1.23 are needed in some cases.
For example, image enhancement for human visual
interpretation seldom goes beyond the processing
stage.

Processing functions including recognition and
interpretation are generally associated with image
analysis applications(the goal is automatic or even
partially automatic extraction of information from an
image)
2010/09/23
40/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(1/6)
21
2010/09/23
41/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(2/6)

Basic Functional Elements

Acquisition, Specialized IP Hardware, Computer, IP Software,
Storage, Display, Hardcopy, and Communication.

Image Acquisition:

Two elements required for acquiring digital image

A physical sensing device

Which is sensitive to a band in the electromagnetic energy
spectrum(e.g., x-ray, ultraviolet, visible, or infrared
bands).

Produces an electrical signal output proportional to the
level of energy sensed.

A digitizer

Convert the electrical output of the physical sensing device
into digital form
2010/09/23
42/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(3/6)

Specialized Image Processing Hardware:

Usually consists of the digitizer/frame buffer
combination for image digitization and temporary
storage.

Contains hardware that performs other primitive
operations, such as an arithmetic logic unit, that
performs arithmetic and logic operations in parallel on
entire images.

Late 80s and early 90s, the market shifted to image
processing hardware in the form of single boards.

It is sometimes called a front-end subsystem, and its
most distinguishing characteristic is speed.

For example, a real time transformation is required when
converting an analog video signal to the YCbCr format.
22
2010/09/23
43/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(4/6)

Computer

A general-purpose computer, ranging from PC to a
supercomputer, is required in an image processing
system.

Almost an well-equipped PC-type machine is suitable
for off-line image processing tasks.

Software:

Consists of specialized modules that perform specific
tasks.

A well-designed package also includes the capability for the
user to write code that, as a minimum, utilizes the specialized
modules.

More sophisticated software packages allow the integration of
those modules and general-purpose software commands from
at least one computer language.
2010/09/23
44/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(5/6)

Storage:

An example

An image of size 1024x1024 pixels, in which the intensity of
each pixel is an 8-bit quantity, requires 1MB of storage space if
not compressed.

Three categories of digital storage

Short term storage for use during processing, i.e., computer
memory or frame buffers

On-line storage for fast recall

Archival storage of infrequent access

Frame buffers in a Specialized Image Processing
Hardware: Can provide instantaneous image PTZ(PSZ)

P: Pan(Horizontal Shift)

T/S: Tilt/Scroll(Vertical Shift)

Z: Zoom
23
2010/09/23
45/45
Lih-Jen Kau
Multimedia Communication Group
National Taipei Univ. of Technology
Components of an IP System
(6/6)

Display:

Monochrome/color TV monitors are the principal display
devices in modern image processing systems.

Hardcopy:

For recording images include laser printers, film
cameras, hear-sensitive devices, inkjet units, and digital
units, such as optical and CD-ROM disks.

Networking:

The key consideration in image transmission is
bandwidth.

The transmission bandwidth is not a problem in
dedicated network, but communications with remote
site via the Internet are not always as efficient.