Medical Image Processing

22MD63

–1–
Objectives of Biomedical Image Analysis
1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

1.10 Objectives of Biomedical Image Analysis

The representation of biomedical images in electronic form

facilitates the following:

Computer processing and analysis of the

data. Computer-aided diagnosis (CAD).
Image-guided surgery.
Image-guided therapy.

–127– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Components of a general-purpose image

processing system.

–127– ◯c G o n z a l e z
 1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Physiological system
(patient) Image data acquisition

Biomedical images
Imaging Analog-to- Picture archival and
Transducers
system digital communication system
conversion (PACS)

Probing signal or
radiation

Computer-aided
diagnosis
and therapy

Pattern recognition, Analysis of regions or Detection of Filtering and

classification, and diagnostic objects; regions or image
decision feature extraction objects enhancement

Physician or
medical
specialist
Image or pattern analysis Image processing

Figure 1.33: Computer-aided diagnosis and therapy based upon biomedical image analysis.

–128– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

The human–instrument system:

The subject (or patient): The main purpose of biomedical

imaging and image analysis is to provide a certain benefit
to the subject or patient.
All systems and procedures should be designed so as not
to cause undue inconvenience to the subject, and not to
cause any harm or danger.
In applying invasive or risky procedures, it is extremely
important to perform a risk–benefit analysis and determine if
the anticipated benefits of the procedure are worth placing
the subject at the risks involved.

–129– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Transducers: films, scintillation detectors, fluorescent screens,

solid-state detectors, piezoelectric crystals, X-ray generators,
ultrasound generators, EM coils, electrodes, sensors.
Signal-conditioning equipment: PMTs, amplifiers, filters.
Display equipment: oscilloscopes, strip-chart or paper
recorders, computer monitors, printers.
Recording, data processing, and transmission equipment:
films, analog-to-digital converters (ADCs), digital-to-analog
converters (DACs), digital tapes, compact disks (CDs),
diskettes, computers, telemetry systems, picture archival
and communication systems (PACS).
Control devices: power supply stabilizers and isolation
equipment, patient intervention systems.

–130– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Major objectives of biomedical instrumentation:

Information gathering — measurement of phenomena to

interpret an organ, a process, or a system.
Screening — investigating a large asymptomatic population
for the incidence of a certain disease (early detection).
Diagnosis — detection or confirmation of malfunction,
pathology, or abnormality.
Monitoring — obtaining periodic information about a
system.
Therapy and control — modification of the behavior of a
system based upon the outcome of the activities listed
above to ensure a specific result.
Evaluation — objective analysis to determine the ability to
meet functional requirements, obtain proof of performance,
perform quality control, or quantify the effect of treatment.
–131– ◯c R.M. Rangayyan, CRC
1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Invasive versus noninvasive procedures:

Invasive procedures involve the placement of devices or

materials inside the body, such as the insertion of

endoscopes,
catheter-tip sensors,
X-ray contrast
media.

Noninvasive procedures are desirable in order to minimize risk

to the subject.

–132– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Active versus passive procedures:

Active data acquisition procedures require external stimuli to

be applied to the subject, or require the subject to perform a
certain activity to stimulate the system of interest in order to
elicit the desired response.

For example, in SPECT investigations of myocardial ischemia,

the patient performs vigorous exercise on a treadmill.

An ischemic zone is better delineated in SPECT images

taken when the cardiac system is under stress than when at
rest.

–133– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Stressing an unwell system may cause pain, irreparable damage,

or death.

The investigator should be aware of such risks, perform a

risk–benefit analysis, and be prepared to handle or
manage adverse reactions.

Passive procedures do not require the subject to

perform any activity.

–134– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.10. OBJECTIVES OF BIOMEDICAL IMAGE ANALYSIS

Most organizations require ethical approval by specialized

committees for experimental procedures involving human or
animal subjects, with the aim of

minimizing the risk and discomfort to the subject, and

maximizing the benefits to both the subject and the
investigator.

–135– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

1.11 Computer-aided Diagnosis (CAD)

Radiologists, physicians, cardiologists, neuroscientists,

pathologists, and other health-care professionals are highly
trained and skilled practitioners.

Why then would we want to suggest the use of computers for

the analysis of biomedical images?

–136– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Humans are highly skilled and fast in the analysis of visual

patterns, but are slow (usually) in arithmetic operations
with large numbers of values.
Computers can perform millions of arithmetic operations
or computations per second.
However, the recognition of objects and patterns in images
using mathematical procedures requires huge numbers of
operations: slow response from low-level computers.
A trained human observer, can usually recognize an object or
a pattern in an instant.

–137– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Humans could be affected by fatigue, boredom, and

environmental factors: susceptible to committing errors.
Working with large numbers of images in one sitting, such as
in breast cancer screening, poses practical difficulties.
A human observer could be distracted by other events in the
surrounding areas and may miss uncommon signs present in
some images.
Computers, being inanimate but mathematically accurate and
consistent machines, can be designed to perform
computationally specific and repetitive tasks.

–138– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Analysis by humans is usually subjective and qualitative.

Computers can assist in quantitative and objective analysis.
Quantitative analysis becomes possible by the application
of computers to biomedical images.
The logic of medical or clinical diagnosis via image analysis
could then be objectively encoded and consistently applied
in routine or repetitive tasks.

–139– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Analysis by humans is subject to interobserver as well as

intraobserver variations (with time).
The former could be due to lack of diligence or due
to inconsistent application of knowledge;
the latter due to variations in training and the level
of understanding or competence.
Computers can apply a given procedure repeatedly
and whenever recalled in a consistent manner.
It is possible to encode the knowledge (the logical processes)
of many experts into a single computational procedure:
enable a computer with the collective “intelligence” of several
human experts in an area of interest.

–140– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Computer-aided diagnosis or automated diagnosis?

A physician or medical specialist typically uses other

information in addition to images, including the general physical
appearance and mental state of the patient, family history, and
socio-economic factors affecting the patient, many of which are
not amenable to quantification and logical rule-based processes.

Biomedical images are, at best, indirect indicators of the state

of the patient; many cases may lack a direct or unique
image-to-pathology relationship.

The results of image analysis need to be integrated with

other clinical signs, symptoms, and information by a
specialist.

–141– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

Above all, the intuition of the medical specialist plays

an important role in arriving at the final diagnosis.

For these reasons, and keeping in mind the realms of practice of

various licensed and regulated professions, liability, and legal
factors, the final diagnostic decision is best left to the physician
or medical specialist.

It could be expected that quantitative and objective analysis

facilitated by the application of computers to biomedical image
analysis will lead to a more accurate diagnostic decision
by the physician.

–142– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

On the importance of quantitative analysis:

“When you can measure what you are speaking about, and
express it in numbers, you know something about it; but
when you cannot measure it, when you cannot express it
in numbers, your knowledge is of a meagre and
unsatisfactory kind: it may be the beginning of knowledge,
but you have scarcely, in your thoughts, advanced to the
stage of science.”
Lord Kelvin (William Thomson, 1824 – 1907)

–143– ◯c R.M. Rangayyan, CRC

1. THE NATURE OF BIOMEDICAL IMAGES 1.11. COMPUTER-AIDED DIAGNOSIS (CAD)

On assumptions made in quantitative analysis:

“Things do not in general run around with their
measure stamped on them like the capacity of a freight
car; it
requires a certain amount of investigation to discover what
their measures are ... What most experimenters take for
granted before they begin their experiments is infinitely
more interesting than any results to which their
experiments lead.”
Norbert Wiener (1894 – 1964)

–144– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT

Image Quality and Information Content

Several factors affect image quality and information content.

Good understanding of such factors and appropriate

characterization of the concomitant loss in image quality essential
in order to design image processing techniques to remove the
degradation and/or improve image quality.

–147– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT

Inherent problem in characterizing image quality:

judged by human observers in a subjective
manner.
To quantify the notion of image quality is a difficult
proposition.

Multifaceted characteristics of information in terms of:

statistical,
structural,
perceptual,
semantic, and
diagnostic
connotations.

–148– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

2.1 Difficulties in Image Acquisition and

Analysis

Accessibility of the organ of interest:

Several organs of interest in imaging-based investigation are

situated well within the body, encased in protective and
difficult-to-access regions.

Brain: protected by the skull.

Prostate: at the base of the bladder near the pelvic outlet.

Visualization of the arteries in the brain requires the injection of

an X-ray contrast agent and the subtraction of a reference
image.

–149– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Special transrectal probes have been designed for 3D ultrasonic

imaging of the prostate.

Images obtained as above tend to be affected by severe

artifacts.

–150– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Variability of information: Biological systems exhibit great ranges

of inherent variability within their different categories.

The intrinsic and natural variability presented by biological

entities within a given class far exceeds the variability that we
may observe in engineering, physical, and manufactured samples.

The distinction between a normal pattern and an abnormal

pattern is often clouded by overlap between the ranges of the
features or variables used to characterize the two categories;

the problem is compounded when multiple abnormalities need to

be considered.

Imaging conditions and parameters could cause ambiguities

due to the effects of subject positioning and projection.

–151– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Most malignant breast tumors are irregular and spiculated

in shape, whereas benign masses are smooth and
round/oval.
However, some malignant tumors may present smooth shapes,
and some benign masses may have rough shapes.

A tumor may present a rough appearance in one view

or projection, but a smoother profile in another.

The notion of shape roughness is nonspecific and open-

ended.

–152– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Overlapping patterns caused by ligaments, ducts, and breast

tissue in other planes could also affect the appearance of
tumors and masses in projection images.

The use of multiple views and spot magnification imaging

could help resolve some of these ambiguities, but at the cost of
additional radiation dose to the subject.

–153– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Physiological artifacts and interference:

Physiological systems are dynamic and active.

Some activities, such as breathing, may be suspended voluntarily

by an adult for brief periods of time to permit improved
imaging.

However, cardiac activity, blood circulation, and peristaltic

movement are not under one’s volitional control.

An analyst should pay attention to potential

physiological artifacts when interpreting biomedical
images.

–154– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Energy limitations:

In X-ray mammography, considering the fact that the organ

imaged is mainly composed of soft tissues, a low kV p would
be
desired in order to maximize image contrast.

However, low-energy X-ray photons are absorbed more readily

than high-energy photons by the skin and breast tissues,
thereby increasing the radiation dose to the patient.

A compromise is required between these two considerations.

–155– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.1. DIFFICULTIES IN IMAGE ACQUISITION AND ANALYSIS

Patient safety:

The protection of the subject in a study from electrical shock,

radiation hazard, and other potentially dangerous conditions is an
unquestionable requirement of paramount importance.

Most organizations require ethical approval by specialized

committees for experimental procedures involving human or
animal subjects, with the aim of

minimizing the risk and discomfort to the subject, and

maximizing the benefits to both the subjects and the
investigator.

–156– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.2. CHARACTERIZATION OF IMAGE QUALITY

2.2 Characterization of Image Quality

Images are complex sources of several items of information.

Many measures available to represent quantitatively

several attributes of images related to impressions of
quality.

–157– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.2. CHARACTERIZATION OF IMAGE QUALITY

Changes in measures related to quality may be analyzed for:

comparison of images generated by different imaging systems;

comparison of images obtained using different imaging
parameter settings of a given system;
comparison of the results of image enhancement
algorithms;
assessment of the effect of the passage of an image through a
transmission channel or medium; and
assessment of images compressed by different data
compression techniques at different rates of loss.

–158– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

2.3 Digitization of Images

The representation of natural scenes and objects as digital images

for processing using computers requires two steps:

sampling, and
quantization.

Both of these steps could potentially cause loss of quality

and introduce artifacts.

–159– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

2.3.1 Sampling

Sampling is the process of representing a continuous-time or

continuous-space signal on a discrete grid, with samples that are
separated by (usually) uniform intervals.

A band-limited signal with the frequency of its fastest component

being f m H z may be represented without loss by its samples
obtained at the Nyquist rate of f s = 2 f m H z .

Sampling may be modeled as the multiplication of the

given analog signal with a periodic train of impulses.

The multiplication of two signals in the time domain corresponds

to the convolution of their Fourier spectra.

–160– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

The Fourier transform of a periodic train of impulses is another

periodic train of impulses with a period that is equal to the
inverse of the period in the time domain (that is, f s Hz ).

Therefore, the Fourier spectrum of the sampled signal is periodic,

with a period equal to f s H z .

A sampled signal has infinite bandwidth; however, the sampled

signal contains distinct or unique frequency components only
up to f m = ± f s /2 H z .

If the signal as above is sampled at a rate lower than f s H z , an

error known as aliasing occurs, where the frequency
components above f s /2 H z appear at lower frequencies.

It then becomes impossible to recover the original signal from

its sampled version.
–161– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

If sampled at a rate of at least f s = 2 f m H z , the original signal

may be recovered from its sampled version by lowpass filtering
and extracting the base-band component over the band ± f m H z .

If an ideal (rectangular) lowpass filter were to be used, the

equivalent operation in the time domain would be
convolution with a sinc function (which is of infinite
duration).

This operation is known as interpolation.

–162– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

In 1D DSP, in order to prevent aliasing errors, it is common to

use an anti-aliasing filter prior to the sampling of 1D signals,
with a pass-band that is close to f s /2 H z .

This requires the prior knowledge that the signal contains

no significant energy or information beyond f m ≤ f s /2
Hz.

However, in most real-life applications of imaging and image

processing, it is not possible to estimate the frequency content
of the images, and also not possible to apply anti-aliasing filters.

–163– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

Figure 2.1 illustrates the loss of quality associated with sampling an

image at lower and lower numbers of pixels.

(a) (b)

(c) (d)

Figure 2.1: Effect of sampling on the appearance and quality of an image: (a) 225 × 250 pixels; (b) 112 × 125
pixels; (c) 56 × 62 pixels; and (d) 28 × 31 pixels. All four images have 256 gray levels at 8 bits per pixel.

–164– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

2.3.2 Quantization

Quantization is the process of representing the values of a

sampled signal or image using a finite set of allowed
values.

Using n bits per sample and positive integers only, there

exist
2n possible quantized levels, spanning the range [0, 2n −
1].

If n = 8 bits are used to represent each pixel, there can

exist
256 values or gray levels in the range [0, 255].

–165– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

It is necessary to map appropriately the range of variation of the

given analog signal to the input dynamic range of the quantizer.

The decision levels of the quantizer should be optimized in

accordance with the probability density function (PDF) of
the original signal or image.

–166– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

Table 2.1: Relationships Between Tissue Type, Tissue Density, X- ray Attenuation Coefficient, Hounsfield Units
(HU ), Optical Density (OD ), and Gray Level. The X- ray Attenuation Coefficient was Measured at a Photon
Energy of 103.2 keV .

Tissue Density X- ray Hounsfield Optical Gray level Appearance

type gm/cm 3 atten. (cm − 1 ) units density (brightness) in image

lung < 0.001 lower low high low dark

[−700, −800]

liver 1.2 0.18 medium medium medium gray

[50, 70]

bone 1.9 higher high low high white

[+800, +1, 000]

–170– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

(a) (b)

(c) (d)

Figure 2.3: Effect of gray-level quantization on the appearance and quality of an image: (a) 64 gray levels (6 bits
per pixel); (b) 16 gray levels (4 bits per pixel); (c) four gray levels (2 bits per pixel); and (d) two gray levels (1 bit
per pixel) All four images have 225 × 250 pixels. Compare with the image in Figure 2.1 (a) with 256 gray levels
at 8 bits per pixel.

–171– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

2.3.3 Array and matrix representation of images

Images commonly represented as 2D functions of space: f (x,

y).

A digital image f (m, n) may be interpreted as a

discretized version of f (x, y) in a 2D array, or as a matrix.

Notational differences between the representation of an image as

a function of space and as a matrix: source of confusion!

An M × N matrix has M rows and N columns;

its height is M and width is N ;

numbering of the elements starts with (1, 1) at the top-left

corner and ends with (M, N ) at the lower-right corner of the
–172– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

A function of space f (x, y) that has been converted into a

digital representation f (m, n) is typically placed in the first
quadrant in the Cartesian coordinate system.

Then, an M × N will have a width of M and height of N ;

indexing of the elements starts with (0, 0) at the origin at

the bottom-left corner and ends with (M − 1, N − 1) at
the upper-right corner of the image.

The size of a matrix is expressed as rows × columns,

the size of an image is usually expressed as width × height.

–173– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.3. DIGITIZATION OF IMAGES

colu n n = 1 2 3 4

row f(1,1) f(1,2) f(1,3) f(1,4)

2 f(0,2) f(1,2) f(2,2) f(3,2)
m = 1

f(0,1) f(1,1) f(2,1) f(3,1) 2 f(2,1) f(2,2) f(2,3) f(2,4)

1

0 f(0,0) f(1,0) f(2,0) f(3,0) f(3,1) f(3,2) f(3,3) f(3,4)

y = 3

x = 0 1 2 3

f(x, y) as a 4x3 function of f(m, n) as a 3x4 matrix

space in the first quadrant (in the fourth quadrant)

Figure 2.4: Array and matrix representation of an image.

–174– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.4. OPTICAL DENSITY

2.4 Optical Density

The value of a picture element or cell — commonly known as

a
pixel, or occasionally as a pel — may be expressed in terms of

a physical attribute such as temperature, density, or X-ray

attenuation coefficient;

the intensity of light reflected from the body at the

location corresponding to the pixel;

or the transmittance at the corresponding location on a film

rendition of the image.

–175– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.4. OPTICAL DENSITY

The O D at a spot on a film is defined as

 

O D = log10 I i .


  (2.5)
Io
 

A perfectly clear spot will transmit all of the light that is

input and will have O D = 0;

a dark spot that reduces the intensity of the input light by a

factor of 1, 000 will have O D = 3.

X-ray films: O D ≈ 0 to O D ≈ 3.5.

–176– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.4. OPTICAL DENSITY

I
o
film with image
(transparency)
Ii

light
Figure 2.5: Measurement of the optical density at a spot on a film or transparency using
a laser microdensitometer.

–177– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.5. DYNAMIC RANGE

2.5 Dynamic Range

The dynamic range of an imaging system or a variable is its

range or gamut of operation,

usually limited to the portion of linear response,

expressed as the maximum — minimum value of the variable.

X-ray films for mammography: dynamic range of 0 − 3.5

O D . Modern CRT monitors provide dynamic range of the

order of
0 − 600 cd/m2 in luminance or 1 : 1, 000 in sampled
gray levels.
–178– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.5. DYNAMIC RANGE

Device A has a larger slope or “gamma” than Device B,

and hence can provide higher contrast.

Device B has a larger latitude, or breadth of exposure and

optical density over which it can operate, than Device A.

Plots of film density versus the log of (X-ray) exposure

are known as Hurter–Driffield or H-D curves.

–179– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.5. DYNAMIC RANGE

Air in the lungs and bowels, as well as fat in various organs

including the breast, tend to extend the dynamic range of
images toward the lower end of the density scale.

Bone, calcifications in the breast and in tumors, as well as

metallic implants such as screws in bones and surgical
clips contribute to high-density areas in images.

Mammograms: dynamic range of 0 − 3.5 O D .

CT images: dynamic range of −1, 000 to +1, 000 H U .

–180– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.5. DYNAMIC RANGE

3.0 Saturation
-
Shoulder

2.0 Device A Device B

-
Optical
density

1
.
0

Toe
- Background level
(base, fog, noise)

log (exposure)

Figure 2.6: Characteristic response curves of two hypothetical imaging devices.

–181– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

2.6 Contrast

C O D = f O D − b OD , (2.6)

where f O D and b OD represent the foreground ROI and

background O D , respectively.

Figure 2.7: Illustration of the notion of contrast, comparing a foreground region f with its background b.

–182– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

When the image parameter has not been normalized, the measure
of contrast will require normalization.

If, for example, f and b represent the average light

intensities emitted or reflected from the foreground ROI and
the background, respectively, contrast may be defined as
f − b
C = , (2.7)
f +
b

or as

f − b
C1 = . (2.8)
b
Due to the use of a reference background, the measures
defined above are often referred to as simultaneous contrast.

–183– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Example:

130 −
Cl = = −0.1333, (2.9)
15015
0

130 − 50
Cr = = +1.6. (2.10)
5
0

C l and C r using Equation 2.7: −0.0714 and +0.444;

advantage: values limited to [−1, 1].

–184– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Figure 2.8: Illustration of the effect of the background on the perception of an object (simultaneous contrast).
The two inner squares have the same gray level of 130, but are placed on different background levels of 150 on
the left and 50 on the right.

–185– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Just-noticeable difference:

( f − b)/b at the level of minimal perception of the object f for

the background b.

Weber’s law:

JND is almost constant ≈ 0.02 or 2% over a wide range

of background intensity.

–186– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Example: The five bars have intensity values

of
155, 175, 195, 215, and 235. Background:
150.

Contrast of the first bar

155 − 150
Cl = = +0.033. (2.11)
15
0

–187– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Figure 2.9: Illustration of the notion of just-noticeable difference. The five bars have intensity values of (from left
to right) 155, 175, 195, 215, and 235, and are placed on a background of 150. The first bar is barely noticeable;
the contrast of the bars increases from left to right.

–188– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Example: A calcification that appears against fat and low-density

tissue may possess high contrast and be easily visible.

A similar calcification against a background of high-density

breast tissue, or a calcification within a high-density tumor, could
possess low contrast, and be difficult to detect.

–189– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.6. CONTRAST

Figure 2.10: Part of a mammogram with several calcifications associated with malignant breast disease. The
density of the background affects the contrast and visibility of the calcifications. The image has 768 × 512 pixels
at a resolution of 62 µm; the true width of the image is about 32 mm.

–190– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

2.7 Histogram

Dynamic range: global information on the extent or spread of

intensity levels across the image.

Histogram: information on the spread of gray levels over the

complete dynamic range of the image across all pixels.

–191– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

Consider an image f (m, n) of size M × N pixels, with

gray levels l = 0, 1, 2, . . . , L − 1.

The histogram of the image may be defined

as

P
M −1 N − 1 δd[f (m, n) − l], l = 0, 1, 2, . . . , L −
(l) =
Σ 1,
f
m =0
Σ (2.12)
n =0

where the discrete unit impulse function or delta function

is
,
,,
,
δd (k) = , 1 i f k = 0 (2.13)
,
0 otherwise.
,,

–192– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

L−1
Σ
P f (l) = M N . (2.14)
l=
0

The area under the function P f (l) , when multiplied with an

appropriate scaling factor, provides the total intensity, density, or
brightness of the image,

depending upon the physical parameter represented by the

pixels.

–193– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

The normalized histogram may be taken to represent the

probability density function (PDF) p f (l) of the image-
generating
process:

1
(2.15)
p f (l) = P f (l).
MN

L−1
Σ
p f (l) = 1. (2.16)
l=
0

–194– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

–195– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

4
x 10

1.
5
Number of
pixels

0.
5

0
0 5 10 15 20 25
0 0 Gray level 0 0 0

Figure 2.11: Histogram of the image of the ventricular myocyte in Figure 1.3. The size of the image is 480 ×480 =
230, 400 pixels. Entropy H = 4.96 bits.

–196– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

–197– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

3000

2500

2000

Number of
1500

pixels
1000

500

0
0 50 100 150 200 250
Gray
level

(a)
1800

1600

1400

1200

1000
Number of
pixels

800

600

400

200

00 50 100 150 200 250

Gray
level

(b)

Figure 2.12: (a) Histogram of the image of the collagen fibers in Figure 1.5 (b); H = 7.0 bits. (b) Histogram of
the image after the application of the 3 × 3 mean filter and rounding the results to integers; H = 7.1 bits.

–198– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

Figure 2.13: Part of a mammogram with a malignant tumor (the relatively bright region along the upper-left
edge of the image). The size of the image is 700 × 700 = 490, 000 pixels. The pixel resolution of 62 µm; the width
of the image is about 44 mm. Image courtesy of Foothills Hospital, Calgary.

–199– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

0.02
5

0.0
2

0.01
Probability of

5
occurrence

0.0
1

0.00
5

0
0 50 10 15 20 25
0 Gray level 0 0 0

Figure 2.14: Normalized histogram of the mammogram in Figure 2.13. Entropy H = 6.92 bits.

–200– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

Figure 2.15: C T image of a patient with neuroblastoma. Only one sectional image out of a total of 75 images in
the study is shown. The size of the image is 512 × 512 = 262, 144 pixels. The tumor, which appears as a large
circular region on the left-hand side of the image, includes calcified tissues that appear as bright regions. The
H U range of [−200, 400] has been linearly mapped to the display range of [0, 255]; see also Figures 2.16 and 4.4.
Image courtesy of Alberta Children’s Hospital, Calgary.

–201– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.7. HISTOGRAM

2500

2000

1500

Number of
pixels
1000

500

0
−2 − 10 0 100 200 300 400
00 0 Hounsfield
Units

(a)
x 104
10

6
Number of

5
voxels

0 −10 0 100 20 30 40
−200 0 Hounsfield 0 0 0
Units

(b)

Figure 2.16: (a) Histogram of the C T section image in Figure 2.15. (b) Histogram of the entire C T study of the
patient, with 75 sectional images. The histograms are displayed for the range H U = [−200, 400] only.
–202– ◯c R.M. Rangayyan, CRC
An example of digital image acquisition.
(a) Illumination (energy) source. (b) A scene. (c) Imaging system. (d) Projection of the
scene onto the image plane. (e) Digitized image.

◯c Gonzalez
NEIGHBORS OF A PIXEL
When referring in the following discussion to particular pixels, we use lower case letters,
such as p and q.

A pixel p at coordinates (x ,y ) has two horizontal and two vertical neighbors with
coordinates

(x+1,y), (x-1,y), (x, y+1), (x, y-1)

This set of pixels, called the 4-neighbors of p, is denoted N4(p). The four diagonal
neighbors of p have coordinates

(x+1,y+1), (x+1, y-1), (x-1,y+1), (x-1, y-1)

and are denoted ND(p).

These neighbors, together with the 4-neighbors, are called the 8-neighbors of p,
denoted by N8(p).

The set of image locations of the neighbors of a point p is called the neighborhood of p.
The neighborhood is said to be closed if it contains p. Otherwise, the neighborhood is
said to be open.
◯c Gonzalez
ADJACENCY, CONNECTIVITY, REGIONS, AND BOUNDARIES

Let V be the set of intensity values used to define adjacency. In a binary image,
V ={ 1 } if we are referring to adjacency of pixels with value 1.

In a grayscale image, the idea is the same, but set V typically contains more elements.

For example, if we are dealing with the adjacency of pixels whose values are in the
range 0 to 255, set V could be any subset of these 256 values.

We consider three types of adjacency:

1. 4-adjacency. Two pixels p and q with values from V are 4-adjacent if q is in the set N 4(p
).
2. 8-adjacency. Two pixels p and q with values from V are 8-adjacent if q is in the set
N8(p).
3. m-adjacency (also called mixed adjacency). Two pixels p and q with values from V are
m-adjacent if
(a) q is in N4(p). Or
(b) q is in ND(p)) and the N4(p)ՈN4(q) has no pixels whose values are from V.

◯c Gonzalez
 (a) (b) (c) (d) (e) (f)

(a) An arrangement of pixels.

(b) Pixels that are 8-adjacent (adjacency is shown by dashed lines).
(c) m-adjacency.
(d) Two regions (of 1’s) that are 8-adjacent.
(e) The circled point is on the boundary of the 1-valued pixels only if 8-adjacency between
the region and background is used.
(f) The inner boundary of the 1-valued region does not form a closed path, but its outer
boundary does.

◯c Gonzalez
◯c Gonzalez
V={1}
p, p2, and p4 have a value of 1
If p1 and p3 are 0, the length of the shortest m-path (the
Dm distance) between p and p4 is 2.
If p1 is 1, then p2 and p will no longer be m-adjacent
(see the definition of m-adjacency given earlier) and
the length of the shortest m-path becomes 3 (the path
goes through the points pp1p2p4).

Similar comments apply if p3 is 1 (and p1 is 0); in this

case, the length of the shortest m-path also is 3.

Finally, if both p1 and p3 are 1, the length of the shortest

m-path between p and p4 is 4. In this case, the path
goes through the sequence of points pp1p2p3p4

◯c Gonzalez
INTRODUCTION TO THE BASIC MATHEMATICAL TOOLS USED IN
DIGITAL IMAGE PROCESSING
ELEMENTWISE VERSUS MATRIX OPERATIONS

An elementwise operation involving one or more images is carried out on a pixel-by pixel
basis
Consider the following 2 * 2 images (matrices):

The matrix product of the images is formed using the rules of matrix multiplication

◯c Gonzalez
LINEAR VERSUS NONLINEAR OPERATIONS

The first property is called the property of additivity, and the second is called the property of
homogeneity. By definition, an operator that fails to satisfy Eq. (2-23) is said to be nonlinear.

We can prove Sum operator is linear

◯c Gonzalez
◯c Gonzalez
 ARITHMETIC OPERATIONS
Arithmetic operations between two images f (x, y) and g (x, y) are denoted as

These are elementwise operations which means that they are performed between corresponding
pixel pairs in f and g for x = 0,1,2, , , , , … M-1 and y= 0,1,2, , , , , . … N-1.

As usual, M and N are the row and column sizes of the images. Clearly, s, d, p, and v are images
of size M x N
As K increases, as equation indicate that the variability (as measured by the variance or the
standard deviation) of the pixel values at each location (x , y ) decreases.

Because E{g(x , y)} = f(x ,y ) this means it approaches the noiseless image f (x , y ) as the
number of noisy images used in the averaging process increases. In order to avoid blurring
and other artifacts in the output (average) image, it is necessary that the images g i (x, y ) be
registered (i.e., spatially aligned).
2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

2.9 Blur and Spread

Functions

Several components of image acquisition systems cause

blurring due to intrinsic and practical limitations.

The simplest visualization of blurring is provided by using

a single, ideal point to represent the object being imaged.

–218– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

(a)

(b)

Figure 2.17: (a) An ideal point source. (b) A Gaussian-shaped point spread function.

–219– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

The image obtained when the input is a point or impulse

function is known as the impulse response or

point spread function (PSF).

–225– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

Some examples of the cause of blurring

are:

Focal spot.
Thickness of screen or crystal.
Scattering.

–227– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

Ideal Finite
point focal
source spot

X rays

Object
being
imaged

Umbra Umbra
Penumbra
(blur)

Figure 2.18: The effect of a finite focal spot (X- ray-generating portion of the target) on the sharpness of the image
of an object.

–228– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

Point, line, and edge spread functions:

It is often not possible to obtain an image of an ideal point.

However, it is possible to construct phantoms to represent

ideal lines or edges.

An image obtained of line function is known as the

line spread function (LSF) of the system.

A cross-section of an ideal straight line is a point (impulse).

The reconstruction of a cross-section of a line phantom provides

the PSF of the system.

–229– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

 (x,y) y f (x,y) y y
l fe
(x,y)

x x x
Integrate Integrate

Point Line Edge

Figure 2.19: The relationship between point (impulse function), line, and edge (step) images. The height of each
function represents its strength.

–230– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

A phantom representing an ideal edge may also be used.

A profile of the image of such a phantom across the ideal

edge provides the edge spread function (ESF).

The derivative of the ESF gives the LSF of the

system. The PSF may be estimated from the LSF.

Ideal (sharp) edge

f(b)

Blurred or
unsharp edge

Intensity
f(x)

f(a)

x=a x=b
Distance x

Figure 2.20: Blurring of an ideal sharp edge into an unsharp edge by an imaging system.

–231– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

Consider the Fourier transform of h l (x, y) .

H l (v) = ∞
h l (y) exp(−j2πvy) dy
∫

y =−∞

= ∞
dy ∞
dx h(x, y) exp[−j2π(ux +
∫ ∫

y =−∞ x =−∞
vy)]|u=0

= H(u, v)| u=0

= H(0, v),

(2.35)

–235– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

H(u, v) is the 2D Fourier transform of h(x, y):

the modulation transfer function (MTF).

The Fourier transform of the LSF gives the values of the

Fourier transform of the PSF along a line in the 2D Fourier
plane
(in this case, along the v axis).

–236– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

From the Fourier slice theorem:

The Fourier transform of a profile of the LSF is equal to the

radial profile of the Fourier transform of the PSF at the angle of
placement of the line source.

If the imaging system may be assumed to be isotropic in the

plane of the line source, a single radial profile is adequate to
reconstruct the complete 2D Fourier transform of the PSF.

An inverse 2D Fourier transform provides the PSF.

–240– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

Line source prepared using a plastic tube of internal radius

1 mm , filled with 1 m C i (milli Curie) of 99m Tc .

Figure 2.21: Nuclear medicine (planar) image of a line source obtained using a gamma camera. The size of the
image is 64 × 64 pixels, with an effective width of 100 mm. The pixel size is 1.56 mm.

–241– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.9. BLUR AND SPREAD FUNCTIONS

18
0

16
0

14
0

12
0
Scaled
counts

10
0

80

60

40
0 10 20 30 40 50 70 80 90 10
0
20 60
Distance in
Figure 2.22: Sample profile (dotted line) and averaged
profile
mm
(solid line)
obtained from the image in Figure 2.21.
Either profile may
0
be taken to represent the L S F of the gamma camera.

Full width at half the maximum (FWHM): 0.5 − 1.7

cm.

–242– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.10. RESOLUTION

2.10 Resolution

The spatial resolution of an imaging system or an image may

be expressed in terms of:

The sampling interval (in, for example, mm or µm).

The width of (a profile of) the PSF, usually FWHM (in
mm).
The size of the laser spot used to obtain the digital image by
scanning an original film, or the size of the solid-state
detector used to obtain the digital image (in µm).
The smallest visible object or separation between objects
in the image (in mm or µm).
The finest grid pattern that remains visible in the
image (in lp/mm ).
–243– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.10. RESOLUTION

Typical resolution limits of a few imaging systems:

X-ray film: 25 − 100 lp/mm .

screen-film combination: 5 − 10 lp/mm ;
mammography: up to 20 lp/mm .
CT: 0.7 lp/mm ;
µCT: 50 lp/mm or 10 µm;
SPECT: < 0.1 lp/mm .

–244– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

2.11 The Fourier Transform and Spectral Content

The Fourier transform is a linear, reversible transform that maps

an image from the space domain to the frequency domain.

Converting an image from the spatial to the frequency (Fourier)

domain helps in

assessing the spectral content,

assessing the energy distribution over frequency
bands, designing filters to remove noise,
designing filters to enhance the image,
extracting certain components that are better separated in the
frequency domain than in the space domain.

–245– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

2D Fourier transform of an image f (x, y) is denoted by F (u, v):

F (u, v) =
∫
∞
x =−∞ −∞ f (x, y) exp[−j 2π(ux+vy)] dxdy.
∫
∞
y =
(2.40)

u, v : frequency in the horizontal and vertical directions.

–246– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

exp[−j 2π (ux + vy)] (2.41)

= exp(−j 2π ux) exp(−j 2π vy)

= [cos(2π ux) − j sin(2π ux)] [cos(2π vy) − j sin(2π vy)].

–247– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Images are typically functions of space.

Units of measurement in the image domain:

m , cm, mm , µm , etc.

In the 2D Fourier domain, the unit of frequency is

cycles/mm , cycles/m , mm − 1 , etc.

Frequency is also expressed as lp/mm .

If the distance to the viewer is taken into account, frequency

could be expressed in terms of cycles/degree of the visual
angle subtended at the viewer’s eye.

The unit Hertz is not used in 2D Fourier analysis.

–248– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

It is common to use the discrete Fourier transform (DFT) via the

fast Fourier transform (FFT) algorithm.

2D DFT of a digital image f (m, n) of size M × N

pixels:

  
1 M −1 N − 1   mk nl
F (k, l) = f (m, n) exp  − j + .
MN Σ  
M N


2π
 
m =0
Σ (2.42)
n =0

For complete recovery of f (m, n) from F (k, l) , the

latter should be computed for k = 0, 1, . . . , M − 1, and
l = 0, 1, . . . , N − 1, at the minimum.

–249– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Then, the inverse transform gives back the original image with no
error or loss of information as

M −1   
Σ
mk nl 
Nk=0
−1 F (k, l) exp  +j 2π + , 
 
 
M N

f (m, n) = Σ
(2.43)
l= 0

for m = 0, 1, . . . , M − 1, and n = 0, 1, . . . , N − 1.

This expression may be interpreted as resolving the given

image into a weighted sum of mutually orthogonal exponential
(sinusoidal) basis functions.

–250– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Figure 2.23: The first eight sine basis functions of the 1D D F T; k = 0, 1, 2, . . . , 7 from top to bottom. Each
function was computed using 64 samples.

–251– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Figure 2.24: The first 64 cosine basis functions of the 2D D F T. Each function was computed using a 64 × 64
matrix.

–252– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Figure 2.25: The first 64 sine basis functions of the 2D D F T. Each function was computed using a 64 × 64 matrix.

–253– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

2.11.1 Important properties of the Fourier transform (FT)

1. The kernel of the FT is separable and symmetric.

Facilitates the evaluation of the 2D DFT as a set of 1D row
transforms, followed by a set of 1D column transforms.

F (k, l) = 1 × (2.52)
N
   
N−1  2π  N−1 2π  
exp  − j mk f (m, n) exp − j nl  .
Σ

m =0

N 

Σ

n =0
N  

–271– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

1D FFT routines may be used to obtain 2D/ MD FT:

 N−1 2π 
1
F (m, l) = N  f (m, n) exp  − N
  
 Σ 

N

n =0

1 N−1 2π
F (k, l) = Σ
F (m, l) exp − N
 

(2.54)

N m =0

Check if N1 is included in the forward or inverse 1D

FFT.

–272– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

2. Parseval’s theorem:
The Fourier transform is an energy-conserving transform.

∞ ∞
|f (x, y)|2 dxdy
∫ ∫

x =−∞ y =−∞

=
∫
∞
u =−∞
∫
∞
v =−∞ |F (u, v)|2 du (2.55)
dv.

–273– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

3. The inverse Fourier transform (IFT) may be performed using

the same FFT routine by taking the forward Fourier
transform of the complex conjugate of the given function,
and then taking the complex conjugate of the result:

f = I F T (F ) = [FT {F
}]
∗ ∗

–274– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

4. The Fourier transform is a linear transform.

Images are often corrupted by additive
noise:
g(x, y) = f (x, y) + η(x, (2.56)
y).

G(u, v) = F (u, v) + η(u, (2.57)

v).

–275– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Most real-life images have a large portion of their energy

concentrated at (u, v) = (0, 0) in a low-frequency region.
Edges, sharp features, and small-scale or fine details lead
to increased strength of high-frequency components.
Random noise has a spectrum that is equally spread all
over the frequency space (flat, uniform, or “white”
spectrum).
Indiscriminate removal of high-frequency components could
cause blurring of edges and the loss of the fine details.

–276– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

5. The DFT and its inverse are periodic signals:

F (k, l) = F (k ± αN, l) = F (k, l ±

αN )
(2.58)
= F (k ± αN, l ± β N ),

where α and β are integers.

–277– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

6. The Fourier transform is conjugate-symmetric for images with

real values:

F (−k, − l ) = F ∗ (k, l).

(2.59)

|F (−k, −l)| = |F (k, l)|,

/ F (−k, − l ) = − / F (k, l).

The magnitude spectrum is even

symmetric. The phase spectrum is odd
symmetric.

–278– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

7. A spatial shift or translation applied to an image leads to

an additional linear phase component in its Fourier
transform.
The
If magnitude
f (m, n) ⇔spectrum
F (k, l)isare
unaffected.
a Fourier-transform pair, we
have
 
2π  
f (m − mo , n − n o ) ⇔ F (k, l) exp  − j (km o + ln o) , 

N



(2.60)

where (m o , n o ) is the shift applied in the space domain.

–279– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Conversely, we also have

 
2π 
f (m, n) exp j (k om + l on)  ⇔ F (k − k o, l − l ).
N o 

(2.61)

This property has important implications in the modulation

of 1D signals for transmission and communication;
it does not have a similar application with 2D images.

–280– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

8. F (0, 0) gives the average value of the image;

a scale factor may be required depending upon the
definition of the DFT used.

–281– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

9. For display purposes, log10[1 + |F (k, l)|2] is often used;

the addition of unity (to avoid taking the log of zero), and
the squaring may sometimes be dropped.
It is also common to fold or shift the spectrum to bring the
(0, 0) frequency point (the “DC” point) to the center.
Folding of the spectrum could be achieved by multiplying
the image f (m, n) with (−1) ( m + n ) before the FFT is
computed. Because the indices m and n are integers, this
amounts to merely changing the signs of alternate pixels.

–282– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

This is related to the property in Equation 2.61 with

ko = lo = N/2 :
 
2π
exp  j (k om + l n)  = exp[jπ(m + n)] = ( −1) (m+n) ,
 


N o 

(2.62)

(m+n)
f (m, n) (−1) ⇔ F (k − N/2, l − N/2). (2.63)

–283– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

10. Rotation of an image leads to a corresponding rotation of the

Fourier spectrum.

f (m 1 , n 1 ) ⇔ F (k 1 , l1 ), (2.64)

m1 = m cos θ + n sin n1 = − m sin θ + n cos

θ; θ; (2.65)

k1 = k cos θ + l sin l1 = − k sin θ + l cos (2.66)

θ; θ.

–284– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

11. Scaling an image leads to an inverse scaling of its

Fourier transform:

1
f (am, bn) ⇔ (2.67)
 
k l 
| F  ,  ,

a b
ab|
where a and b are scalar scaling factors.
The shrinking of an image leads to an expansion of
its spectrum, with increased high-frequency content.
On the contrary, if an image is enlarged, its spectrum is
shrunk, with reduced high-frequency energy.

–285– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

12. Linear shift-invariant systems and convolution:

Most imaging systems may be modeled as linear
and shift-invariant or position-invariant systems that
are completely characterized by their PSFs.
The output of such a system is given as the convolution of
the input image with the PSF:

g(m, n) = h(m, n) ∗ f (m, n)

(2.68)
N−1 N−1
= Σ
h(α, β) f (m − α, n −
α =0 β).
Σ

β =0

–286– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Upon Fourier transformation, the convolution maps to the

multiplication of the two spectra:

G(k, l) = H ( k , l) F (k, l). (2.69)

h(x, y) ∗ f (x, y) ⇔ H(u, v) F (u, (2.70)

v),

expressed now in the continuous coordinates

(x, y)
The and (u, v)⇔
convolution . multiplication property with the DFT
implies periodic or circular convolution;
Circular convolution may be made to be equivalent to linear
convolution by zero-padding.

–287– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

13. Multiplication of images in the space domain is equivalent to the

convolution of their Fourier transforms:

f 1 (x, y) f 2 (x, y) ⇔ F 1 (u, v) ∗ F 2 (u, v).

(2.71)

Some types of noise get multiplied with the image.

When a transparency, such as an X-ray image on film, is
viewed using a light box, the resulting image g(x, y) may
be modeled as the product of the transparency or
transmittance function f (x, y) with the light source
intensity field s(x, y) : g(x, y) = f (x, y) s(x, y) .

–288– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

If s(x, y) is absolutely uniform with a value A , its

Fourier transform will be an impulse:
S(u, v) = A δ(u, v).
The convolution of F (u, v) with A δ(u, v) will have
no effect on the spectrum except scaling by the constant
A.
If the source is not uniform, the viewed image will be
a distorted version of the original:
G(u, v) = F (u, v) ∗ S(u, v).

–289– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

14. The correlation of two images f (m, n) and g(m, n) is

N−1 N−1
γ
f, g (α, β) = Σ

m =0 f (m, n) g(m+α, n + β ) . (2.72)

Σ

n =0

Correlation is useful in the comparison of images where

features that are common to the images may be present with
a spatial shift (α, β) .

Γ f , g (k, l) = F (k, l) G ∗ (k, l).

(2.73)

–290– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

A related measure, known as the correlation coefficient and

useful in template matching and image classification, is

N−1 Σ N−1
n=0 f (m, n) g(m,
Σ
m=0
γ= .
"
N−1 N−1 n)
2 (m, n) Σ N − 1 Σ N − 1 g2(m, n) 2
#1
Σ
m=0
Σ
n=0 f m=0 n=0
(2.74)

Here, it is assumed that the two images f and g are aligned

and registered, and are of the same scale and orientation.

–291– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

15. Differentiation of an image results in the extraction of edges

and highpass filtering:

∂ f (x,
⇔ j2πu F (u, v);
y) ∂x
∂ f (x,
⇔ j2πv F (u, v). (2.75)
y) ∂y

Gain of the filter increases linearly with frequency u or v.

–292– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Derivatives are approximated by differences:

′
f y (m, n) ≈ f (m, n) − f (m − 1,
n),
′
f x (m, n) ≈ f (m, n) − f (m, n − (2.76)
1),

(using matrix notation).

–293– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Operators based upon differences could cause negative pixel

values in the result.
In order to display the result as an image, it will be
necessary to map the full range of the pixel values, including
the negative values, to the display range available.
The magnitude of the result may also be displayed if the
sign of the result is not important.
Differentiation results in the removal of the intensity
information from the image.
The values of the spectrum for u = 0 or v = 0 are
set to zero.

–294– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.26: (a) Rectangle image, with total size 128 × 128 pixels and a rectangle (square) of size 40 × 40 pixels.
(b) Log-magnitude spectrum of the image in (a). (c) Rectangle size 20 × 20 pixels. (d) Log-magnitude spectrum
of the image in (c). (e) Rectangle size 10 × 10 pixels. (f) Log-magnitude spectrum of the image in (e). The
spectra have been scaled to map the range [5, 12] to the display range [0, 255]. See also Figures 2.28 and 2.29.

–259– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(0, 0) (U/2, 0) (U, 0) (0, V/2)

(0, V/2) (-U/2, 0) (0, 0) (U/2, 0)

(0, -V/2)

(0, V) (U, V)
v v

(a) (b)

Figure 2.27: Frequency coordinates in (a) the unshifted mode and (b) the shifted mode of display of image
spectra. U and V represent the sampling frequencies along the two axes. Spectra of images with real values
possess conjugate symmetry about U/2 and V/2. Spectra of sampled images are periodic, with the periods equal
to U and V along the two axes. It is common practice to display one complete period of the shifted spectrum,
including the conjugate symmetric parts, as in (b). See also Figure 2.28.

–260– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

The peak at the center of the plot is F(0,0), which is

the sum of all the values in f(m,n).
The plot also shows that F(ω1,ω2) has more
energy at high horizontal frequencies than at high
vertical frequencies.
This reflects the fact that horizontal cross sections
of f(m,n) are narrow pulses, while vertical cross
sections are broad pulses. Narrow pulses have
more high-frequency content than broad pulses

–261– ◯c Mathworks
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

more energy at high horizontal frequencies than at high vertical frequencies.

This reflects the fact that horizontal cross sections of f(m,n) are narrow pulses, while vertical
cross sections are broad pulses. Narrow pulses have more high-frequency content than
broad pulses

–261– ◯c Mathworks
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Another common way to visualize the Fourier transform is to display logF(ω1,ω2) as

an image, as shown.

ω1

ω2

–261– ◯c Mathworks
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.28: (a) Log-magnitude spectrum of the rectangle image in Figure 2.26 (e) without shifting. Most F F T
routines provide spectral data in this format. (b) The spectrum in (a) shifted or folded such that (u, v) = (0, 0)
is at the center. It is common practice to display one complete period of the shifted spectrum, including the
conjugate symmetric parts, as in (b). See also Figure 2.27.

–261– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.29: (a) Mesh plot of the rectangle image in Figure 2.26 (e), with total size 128×128 pixels and a rectangle
(square) of size 10 × 10 pixels. (b) Magnitude spectrum of the image in (a).

–262– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.30: (a) Rectangle image, with total size 128 × 128 pixels and a rectangle of size 10 × 40 pixels.
(b) Log-magnitude spectrum of the image in (a). (c) Rectangle size 40 × 10 pixels; this image may be considered
to be that in (a) rotated by 90o. (d) Log-magnitude spectrum of the image in (c). (e) Image in (c) rotated by 45o
using nearest-neighbor selection. (f) Log-magnitude spectrum of the image in (e). Spectra scaled to map [5, 12]
to the display range [0, 255].
–263– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

The circle function and its Fourier transform:

Circular apertures and functions are encountered often in

imaging and image processing.

,
, 1 if r < 1
circ(r) =  (2.51)
,,,,
0 if r > 1 ,

where r = (x 2 + y2).

The Fourier transform of circ (r) is 1ν J 1 (2πν) ,

where ν = (u 2 + v 2 ) represents radial frequency in the 2D

(u, v) plane, and J 1 is the first-order Bessel function of the
first kind.

–264– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.31: (a) Image of a circular disc. The radius of the disc is 10 pixels; the size of the image is 128 × 128
pixels. (b) Log-magnitude spectrum of the image in (a). See also Figures 2.32 and 2.33.

–265– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

4
x 10

250 8

7
200
6

5
150
4

100 3

2
50
1

0
120 120
100 100
120 120
80 100 80 100
60 80 60 80
40 60 40 60
40 40
20 20 20 20

(a) (b)

Figure 2.32: (a) Mesh plot of the circular disc in Figure 2.31 (a). The radius of the disc is 10 pixels; the size of the image is 128 × 128 pixels.
(b) Magnitude spectrum of the image in (a).

–266– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

11 11

10 10

9 9

8 8

log magnitude

log magnitude
7 7

spectrum

spectrum
6 6

5 5

4 4

3 3
20 40 60 80 100 120 10 20 30 40 50 60
sample sample
number number

(a) (b)

Figure 2.33: (a) Profile of the log-magnitude spectrum in Figure 2.31 (b) along the central horizontal axis.
(b) Profile in (a) shown only for positive frequencies. The frequency axis is indicated in samples; the true
frequency values depend upon the sampling frequency.

–267– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Profiles of 2D system transfer functions:

it is common to show only one half of the profile

for positive frequencies.

It is to be assumed that the system possesses axial or

rotational symmetry; that is, the system is isotropic.

–268– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

Figure 2.34: (a) T E M image of collagen fibers in a normal rabbit ligament sample. (b) Log-magnitude spectrum
of the image in (a). (c) T E M image of collagen fibers in a scar tissue sample. (d) Log-magnitude spectrum of the
image in (c). See also Figure 1.5 and Section 1.4.

–269– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

Figure 2.35: (a) SEM image of collagen fibers in a normal rabbit ligament sample. (b) Log-magnitude spectrum
of the image in (a). (c) SEM image of collagen fibers in a scar tissue sample. (d) Log-magnitude spectrum of the
image in (c). See also Figure 1.8 and Section 1.4.

–270– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Image addition (averaging) for noise reduction.

–272– ◯c Gonzalez
 2. IMAGE QUALITY AND INFORMATION CONTENT

Digital subtraction angiography

(a) (b)

g(x,y) = f (x,y) - h (x ,y)

(a) Mask image (h(x,y).

(b) A live image f(x,y)
(c) Difference between (a)
and (b).
(d) Enhanced difference
image. (Figures (a) and (b)
(c) (d)
courtesy of the Image:
Sciences Institute,
University Medical Center,
Utrecht, The Netherlands.)

–272– ◯c Gonzalez
 2. IMAGE QUALITY AND INFORMATION CONTENT

Using image multiplication and division for shading correction and for masking

–272– ◯c Gonzalez
 2. IMAGE QUALITY AND INFORMATION CONTENT

Using image multiplication and division for shading correction and

for masking

–272– ◯c Gonzalez
2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

–274– ◯c Gonzalez
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Neighbourhood operations

Eq. 2-43

–274– ◯c Gonazalez
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Geometric transformations of digital images consist of two basic operations:

1. Spatial transformation of coordinates.

2. Intensity interpolation that assigns intensity values to the spatially transformed
pixels. The transformation of coordinates may be expressed as

where ( x, y ) are pixel coordinates in the original image and (x’ , y’ ) are the
corresponding pixel coordinates of the transformed image.

For example, the transformation (x’, y’) = ( x/2, y/2 ) shrinks the original image to half
its size in both spatial directions.

It is possible to use homogeneous coordinates to express all four affine

transformations which include scaling, translation, rotation, and shearing using a
single 3 x 3 matrix in the following general form:

–274– ◯c Gozalez
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Geometric Transformations

–272– ◯c Gonzalez
2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

–274– ◯c Gonzalez
 2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

IMAGE TRANSFORMS

–274– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Image processing in the transform domain.

–274– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.36: (a) Image of a rectangular box. (c) Horizontal and (e) vertical derivatives of the image in (a),
respectively. (b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e), respectively. The images
in (c) and (e) were obtained by mapping the range [−200, 200] to the display range of [0, 255]. Negative differences
appear in black, positive differences in white. The spectra show values in the range [5, 12] mapped to [0, 255].

–295– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.37: (a) Image of a myocyte. (c) Horizontal and (e) vertical derivatives of the image in (a), respectively.
(b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e), respectively. Images in (c) and (e)
were obtained by mapping the range [−20, 20] to the display range of [0, 255]. The spectra show values in the
range [3, 12] mapped to [0, 255].

–296– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.38: (a) MR image of a knee. (c) Horizontal and (e) vertical derivatives of the image in (a), respectively.
(b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e), respectively. The images in (c) and
(e) were obtained by mapping the range [−50, 50] to the display range of [0, 255]. Negative differences appear in
black, positive differences in white. The spectra show values in the range [3, 12] mapped to [0, 255].

–297– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

16. The Laplacian of an image:

2 ∂2 f ∂2 f
(2.77)
∇ f (x, y) = + .
∂x 2 ∂y 2

2 2 2 2
∇ f (x, y) ⇔ −(2π) (u + v ) F (u, v).
(2.78)

The spectrum of the image is multiplied by (u2 + v2 ), which

is isotropic and increases quadratically with frequency.
High-frequency components are amplified.
Omnidirectional operator: detects edges in all directions.

–298– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

The second derivatives may be approximated as follows:

Taking the derivative of the expression for f y (m, n) in
′

Equation 2.76 for the second time, we

get
′′
f y (m, n)

≈ f (m, n) − f (m − 1, n) − [f (m − 1, n) − f (m − 2,
n)]

= f (m, n) − 2 f (m − 1, n) + f (m − 2, n)
(2.79)

(using matrix notation).

Causality is usually not of concern in image processing:
desirable to have operators use collections of pixels that
are centered about the pixel being processed.
–299– ◯c R.M. Rangayyan, CRC
2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Applying a shift of one pixel to the result above, we

get ′′
f y (m, n)
(2.80)
≈ f (m + 1, n) − 2 f (m, n) + f (m − 1,
n)

= f (m − 1, n) − 2 f (m, n) + f (m + 1,
n).

′′
f x (m, n) ≈ f (m, n − 1) − 2 f (m, n) + f (m, n +
1). (2.81)

–300– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

The Laplacian could then be implemented as

f L ( m , n) = f (m − 1, n) + f (m, n − 1) − 4f (m, n)

+ f (m + 1, n) + f (m, n + 1).

(2.82)

≡ convolving the image with the 3 × 3 mask or operator




1 −4 1



. (2.83)
 
 0 1  0  
0 
0 

–301– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.39: (a) Laplacian of the rectangle image in Figure 2.36 (a). (b) Log-magnitude spectrum of the image
in (a).

–302– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.40: (a) Laplacian of the myocyte image in Figure 2.37 (a). (b) Log-magnitude spectrum of the image in
(a).

–303– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a)
(b)

Figure 2.41: (a) Laplacian of the MR image in Figure 2.38 (a). (b) Log-magnitude spectrum of the image in (a).

–304– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

17. Integration of an image leads to smoothing or blurring, and

lowpass filtering:

1
∫
x
α=−∞ f (α, y) dα ⇔ F (u, v), (2.84)
j2πu

1
∫ y
β=−∞ f (x, β) dβ ⇔ F (u, v). (2.85)
j2πv

The weighting factors that apply to F (u, v) diminish with

increasing frequency, and hence high-frequency components
are attenuated by this operation.

–305– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

The integration of an image from − ∞ to the current x or y

position is seldom encountered in practice.
Instead, it is common to encounter the integration of an
image over a small region or aperture surrounding the current
position, in the form

∫ A /2 ∫ B/2
1 α =
g(x, y) = −A/2 f (x + α, y + β) dα
AB β=−B/2 dβ,
(2.86)
where the region of integration is a rectangle of size A × B .
The normalization factor A1B leads to the average intensity
being computed over the area of integration.
This may be interpreted as a moving-average (MA)
filter.

–306– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Averaging over a 3 × 3 aperture or neighborhood:

1 1Σ 1
g(m, n) = 9 f (m + α, n + (2.87)
Σ
α=
β=−1
1
— β).

g(m, n) = 1 ×
9

[ f (m − 1, n − 1) + f (m − 1, n) + f (m − 1, n + 1)
+ f (m, n − 1) + f (m, n) + f (m, n + 1)
+ f (m + 1, n − 1) + f (m + 1, n) + f (m + 1, n + 1)
].
(2.88)

–307– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Equivalent to convolution of the image f (m, n)

with


1

11

1 

1

11

(2.89)
9


, 

1 1 

1
which may be viewed as the PSF of a filter.
Equivalent to multiplication of the Fourier transform of the
image with a 2D sinc function.

–308– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

Integration or averaging only along the horizontal or vertical

directions may be performed via convolution with the arrays
1
[1, 1, 1] or 1 [1, 1, 1]T
.
3 3

–309– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.42: (a) Image of a rectangular box. Results of averaging using three pixels in the (c) horizontal and
(e) vertical directions, respectively. (b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e),
respectively. The spectra show values in the range [5, 12] mapped to [0, 255].

–310– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.43: (a) Result of 3 × 3 averaging of the rectangle image in Figure 2.42 (a). (b) Log-magnitude spectrum
of the image in (a).

–311– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.44: (a) Image of a myocyte. Results of averaging using three pixels in the (c) horizontal and (e) vertical
directions, respectively. (b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e), respectively.
The spectra show values in the range [3, 12] mapped to [0, 255].

–312– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.45: (a) Result of 3 × 3 averaging of the myocyte image in Figure 2.44 (a). (b) Log-magnitude spectrum
of the image in (a).

–313– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

(c) (d)

(e) (f)

Figure 2.46: (a) MR image of a knee. Results of averaging using three pixels in the (c) horizontal and (e) vertical
directions, respectively. (b), (d), and (f): Log-magnitude spectra of the images in (a), (c), and (e), respectively.
The spectra show values in the range [3, 12] mapped to [0, 255].

–314– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.11. THE FOURIER TRANSFORM AND SPECTRAL CONTENT

(a) (b)

Figure 2.47: (a) Result of 3 × 3 averaging of the knee MR image in Figure 2.46 (a). (b) Log-magnitude spectrum
of the image in (a).

–315– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.13. SIGNAL-TO-NOISE RATIO (SNR)

2.13 Signal-to-Noise Ratio (SNR)

g(x, y) = f (x, y) + η(x, (2.91)

y).

Assume noise process is additve and statistically independent

of (uncorrelated with) the image process.

–324– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.13. SIGNAL-TO-NOISE RATIO (SNR)

Mean:

µg = µ f + µ η , (2.92)

Usually, the mean of the noise process is zero: µg = µ f .

Variance:

σ2 = σ2 + σ2. (2.93)
g f
η

–325– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.13. SIGNAL-TO-NOISE RATIO (SNR)

 
2f
S N R 1 = 10 log10  σ 

(2.94)
 σ η2 

dB.

Variance of noise estimated by computing the sample variance of

pixels from background areas of the image.

Variance may be computed from the PDF (histogram).

–326– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.13. SIGNAL-TO-NOISE RATIO (SNR)

The variance of the image may not provide an

appropriate indication of the useful range of variation in
the image.

SNR based upon the dynamic range of the image:

 

S N R 2 = 20 log10  f max − f min

 
 dB. (2.95)
ση

Video signals in modern CRT monitors: SNR ≈ 60 − 70 dB

with noninterlaced frame repetition rate of 70 − 80 fps .

–327– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.13. SIGNAL-TO-NOISE RATIO (SNR)

Contrast-to-noise ratio (CNR) is a measure that combines the

contrast or the visibility of an object and the SNR:

µ f − µb
CNR = , (2.96)
σb

Simultaneous contrast uses a background that encircles the ROI;

CNR could use a background region located elsewhere.

CNR is well suited to the analysis of X-ray imaging

systems: density of an ROI on a film image depends upon

dose; visibility of an object dependent upon both dose and

–328– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.14. ERROR-BASED MEASURES

2.14 Error-based Measures

Mean-squared error:

1 MΣ −1 NΣ− 1 2
MSE = [f (m, n) − g(m, n)] (2.97)
MN m=0 n=0
.

Normalized MSE:

M −1 N−1 2
Σ
m=0
Σ
n=0 [f (m, n) − g(m,
NMSE = . (2.98)
Σ M −1
m=0
n)]
Σ N−1
n=0 [f (m, 2

n)]

–329– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.14. ERROR-BASED MEASURES

Normalized error:

M −1 N−1
Σ Σ
|f (m, n) − g(m,
m=0 Σ Mn=0 (2.99)
N E = −1
m=0
n)|
Σ N−1
n=0 |f (m,
.
n)|

Laplacian MSE:

Σ M −2
m=1
Σ N−2
n=1 [f L (m, n) − gL (m, n)]2
LMSE = Σ M −2 Σ N−2 2 .
m=1 n=1 [f L (m, n)] (2.100)

f L ( m, n) is the Laplacian of f (m,

n) .

–330– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

2.15 Application: Image Sharpness and

Acutance

Acutance from the edge spread function:

A = 1 ∫
b

d f (x)
2

a

 dx. (2.101)
f (b) − f (a)


dx

–332– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Ideal (sharp) edge

f(b)

Blurred or
unsharp edge

Intensity
f(x)

f(a)

x=a x=b
Distance x

–333– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Region-based measure of edge sharpness:

Instead of the traditional difference defined as

f ′ (n) = f (n) − f (n − 1),

(2.108)
Rangayyan and Elkadiki split the normal at each boundary
pixel into a foreground part f (n) and a background part b(n)
and defined an averaged gradient as
1 N f
( n ) − b(n) (2.109)
Σ
d
f (k) = N n=1 2n ,

where k is the index of the boundary pixel,

N is the number of pairs of pixels used along the normal.

–340– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Normal

b(3) Background
b(2)
b(1)

f(1)
f(2)
Boundary
f(3)

Foreground
(Object)

Figure 2.52: Computation of differences along the normals to a region in order to derive a measure of acutance.
Four sample normals are illustrated, with three pairs of pixels being used to compute differences along each
normal.

–341– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

The averaged gradient values over all boundary pixels were then
combined to obtain a single normalized value of acutance A :

1

1 
1 KΣ 2 
2
A = 
f d (k)  , (2.110)
 k=1
dmax K


where K is the number of pixels along the boundary,

dmax is the maximum possible gradient value.

–342– ◯c R.M. Rangayyan, CRC

2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Acutance:

reduced by blurring,
increased by sharpening,
not affected significantly
by noise,
correlates well with
sharpness as judged by
human observers.

–343– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT

The Fourier Transform and Spectral Content

Converting an image from the spatial to the frequency Fourier domain helps in
assessing the
 spectral content and energy distribution over frequency bands
 Sharp edges in the Biomedical Image Analysis image domain are associated with
large proportions of high frequency content
 Oriented patterns in the space domain correspond to increased energy in bands of
frequency in the spectral domain with the corresponding orientation
 Simple geometric patterns such as rectangles and circles map to recognizable
functions in the frequency domain such as the sinc and Bessel functions
respectively
 Transforming an image to the frequency domain assists in the application of
frequency domain filters to remove noise enhance the image or extract certain
components that are better separated in the frequency domain than in the space
domain

–344– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT

Medical Image Protocols and Standards

Medical imaging protocols and standards ensure consistency, accuracy, and
interoperability in acquiring, processing, storing, and sharing medical images across
healthcare systems. They help improve diagnosis, treatment planning, and research
while ensuring patient safety and regulatory compliance.

1. Medical Imaging Protocols

Protocols define how medical images should be acquired to maintain quality and
consistency across different modalities (e.g., X-ray, MRI, CT, Ultrasound). These
protocols include:

CT Scan Protocols: Define parameters such as slice thickness, contrast agent usage,
and radiation dose optimization.
MRI Protocols: Specify pulse sequences, field strength, and scan duration based on
the region of interest (e.g., Brain MRI vs. Spine MRI).
Ultrasound Protocols: Define probe type, scanning angles, and Doppler settings for
imaging various organs.
Nuclear Medicine Protocols: Specify radiopharmaceutical dosage, imaging time
points, and reconstruction techniques for PET and SPECT scans.

Example:
• Head CT without contrast: Used for stroke detection.
• T1-weighted MRI with gadolinium: Used for tumor detection in the brain

–344– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Medical Image Standards

Standards ensure compatibility and interoperability between different imaging devices,
software, and healthcare systems. Some key standards include:

A. DICOM (Digital Imaging and Communications in Medicine)

 The most widely used standard for storing, transmitting, and sharing medical
images.
 Defines file format, metadata, and communication protocols for modalities like CT,
MRI, and X-ray.
 Enables PACS (Picture Archiving and Communication System) integration for
centralized image storage and retrieval.

B. HL7 (Health Level 7)

 Facilitates the exchange of medical data between imaging systems and Electronic
Health Records (EHR).
 Ensures that patient information, reports, and imaging results are synchronized
across hospital networks.

C. IHE (Integrating the Healthcare Enterprise)

 Defines workflows for seamless integration of imaging data across different hospital
systems.
 Supports interoperability between DICOM, HL7, and EHR systems.

–344– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Medical Image Standards

D. PACS (Picture Archiving and Communication System)

 A system for storing, retrieving, and sharing medical images electronically.
 Eliminates the need for physical film-based storage.

E. ISO and FDA Standards for Medical Imaging Devices

 ISO 13485: Ensures quality management for medical imaging equipment
manufacturers.
 FDA (Food and Drug Administration) Guidelines: Regulate the safety and
performance of medical imaging systems.

–344– ◯c R.M. Rangayyan, CRC

 2. IMAGE QUALITY AND INFORMATION CONTENT 2.15. APPLICATION: IMAGE SHARPNESS AND ACUTANCE

Importance of Medical Imaging Protocols & Standards

Ensures Image Quality & Accuracy – Standardized protocols prevent imaging errors
and variability.
Enhances Patient Safety – Optimized radiation doses and proper scanning techniques
reduce risks.
Facilitates Data Sharing & Interoperability – Standards like DICOM and HL7 ensure
seamless communication between systems.
Supports AI & Machine Learning Applications – Structured imaging data helps in
developing automated diagnosis tools.
Regulatory Compliance – Adhering to ISO, FDA, and hospital-specific guidelines
ensures legal and ethical practices.

–344– ◯c R.M. Rangayyan, CRC

Integrating the Healthcare Enterprise:
IHE and Clinical Workflow