Thanks to visit codestin.com
Credit goes to www.scribd.com

0% found this document useful (0 votes)
4 views4 pages

An Introduction To Digital Image Process

The document discusses digital image processing, highlighting its significance in various fields, particularly in microbiology. It explains the components of a digital imaging system, the techniques used for image enhancement, and the applications of these techniques in medical diagnostics. The Aladin system is presented as an example of how digital image processing can automate laboratory diagnostics for microbial identification and antimicrobial susceptibility testing.

Uploaded by

kjeuse
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
4 views4 pages

An Introduction To Digital Image Process

The document discusses digital image processing, highlighting its significance in various fields, particularly in microbiology. It explains the components of a digital imaging system, the techniques used for image enhancement, and the applications of these techniques in medical diagnostics. The Aladin system is presented as an example of how digital image processing can automate laboratory diagnostics for microbial identification and antimicrobial susceptibility testing.

Uploaded by

kjeuse
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Clinical

Microbiology
Newsletter
Vol. 12, No. 12 June 15, 1990

An Introduction to Digital Image Processing


Geoffrey A. McKinley, Ph.D. brightness. Each point is assigned a nation of antimicrobial susceptibility
Analytab Products numeric location within the image, and and biochemical endpoints.
Plainview, NY 11803
the computer can manipulate the
brightness of each point. Digital image A Simple Digital I m a g i n g System
Image processing is one of today's processing has become the technique of In its simplest form, a digital
most exciting branches of applied com- choice for image processing because of imaging system is a series of hardware
puter sciences. In general, image pro- the advent of inexpensive processors, devices to input, digitize, store, pro-
cessing is the alteration and analysis of dense memory devices, and special- cess, and display images. Figure 1 is
pictorial information. An everyday ex- purpose signal processing components. an illustration of such a system. The
ample of an image processing system The origin of image processing can system begins with a device, such as
is the system that is composed of the be found in NASA's attempt to charac- the television camera, which inputs a
eye and brain. Here, information is re- terize the lunar surface in the Ranger video signal to be digitized. Digitizing
ceived, collected, enhanced, analyzed, program during the early 1960s. The converts the analog electrical form into
and stored. The objective of image Ranger 7 mission transmitted thousands a digital form for storage. After pro-
processing as a discipline is to enhance of images back to earth that were pro- cessing, the digital signal is sent to a
or evaluate a particular aspect of an cessed to remove various distortions.
image that normally would not be These early successes in digital image
readily apparent. Three basic tech- processing supported a series of plane-
niques are available for implementing tary explorations that provided imagery
this objective. data for scientific studies involving a In This Issue
The first, the optical technique, number of planets.
which has been widely implemented by Digital image processing has devel-
photographers, uses an arrangement of
An Introduction to Digital Image
oped from what was once an unpredict-
Processing . . . . . . . . . . . . . . . . . . . 89
optics to carry out a particular process. able, expensive, and time-consuming
A discussion of a new technology
The other two techniques are electronic discipline, into what is now a sophisti-
for examining and analyzing vi-
in nature and can be divided into an- cated science. It has been applied to a
alog and digital. Analog processing, of suaUy presented data and its appli-
number of markets, including defense,
cation to microbiology
which a television image is a good ex- artificial vision, graphic arts, printing
ample, is the alteration of images and publishing, remote sensing (geo- The National Laboratory
through electrical means. The televi- graphical satellites), geophysics, and Training Network: The Role of
sion signal is a voltage that varies in medicine. Within the medical field, Training in Ensuring the
amplitude to represent a brightness. examples of image processing include Nation's Health . . . . . . . . . . . . . . . 92
The amplitude of the video signal can enhancement of human jaw x-rays, A new approach for CDC and the
be adjusted through the brightness and evaluation of tissue samples to deter- state laboratories in bringing
contrast controls resulting in a brighter mine muscle fiber size and distribution, training closer to the facilities that
or darker picture. Finally, the digital histologic measurement of drug effects, need it
computer has allowed for the creation measurement of cell optical densities,
of the third basic technique, digital ratio of nuclear area to cytoplasm area, Vertebral Osteomyelitis with
image processing. In the digital world, visualization of calcified bone areas, Paravertebral Abscess Due to
an image is converted into defined optical densities of brain sections, ag- Myobacterium xenopi . . . . . . . . . . . 94
points, each with its own associated glutination reactions, and the determi- Case report

CMNEEI 12(12)89-96,1990 Elsevier 0196-4399/90/$0.00 + 02.20


a digital value. The RS-170 signal has
USER I
INTERFACE a band-width of 4.2 megahertz (MHz).
The signal is digitized at 10 MHz;
however, a sampling rate of twice the
tl, largest frequency is required to obtain
an accurate representation of the orig-
HOST ~ M inal signal. At this rate each line is
COMPUTER I STORAGE
[ o,S~APE sampled 500 times. Because there are
480 active lines of video (although
RS-170 specifies 525 lines, only 480 of
these lines actually contain informa-
tion) in a frame, an array of 500 by 480
STORAGE is produced. A frame buffer, or high-
IMAGEDATA
speed memory, which has dimensions
VIDEOINPUT of 512 × 512 lines, is used to store
CAMERA
this data. Each point in the buffer is a
picture element or pixel. A pixel has
an associated value representing the in-
APPLICATION
SPECIFIC tensity of the characteristic detected by
~tAGE a sensor. The resolution of the digi-
PROCESSOR
i •
tizer, which is expressed in bits, deter-
Figure 1. Basic imaging system. mines the number of intensity levels
available for each pixel. For example,
converter where the signal is changed CAT scanners, ultrasound systems, and a video camera signal is digitized to
back to an analog signal for display on scanning electron microscopes, provide eight bits. This will produce 256 gray
a monitor. The host computer oversees images produced by the penetration of levels of intensity (2s), which range
the image processing system, allows a particular energy source into an ob- from black to white.
for user interface, and provides for ject. The digital image can be converted
long-term storage. The user decides Once the characteristics of an image back to analog signals via the display
when to digitize, display, or process an have been collected, the image is parti- circuitry. Here, the digital image is
image. The host computer can also tioned into small parts. Each part has moved from the frame buffer, pixel by
analyze a stored image if its speed is an associated value that describes the pixel through a D to A converter, thus
not too slow for a particular application. intensity level at which it was de- reconstructing the RS-170 signal. The
tected. This partitioning is achieved by electronic signal is then converted back
scanning the object in lines. A set of to a visual image with a video mon-
System Components standards has been established for video itor. The monitor uses conventional
To better understand how digital cameras that determines the number of scanning to produce an image on a
image processing can be useful over scan lines and the rate of the scan. screen. It emits light in various
such a wide range of diverse applica- (The RS-170 standard is commonly amounts to represent the video signal.
tions, the focus of this discussion will used; it specifies 525 lines of scan to a
now shift to a more in-depth analysis of frame at a rate of 30 frames per Data Extraction and Analysis
the components of an imaging system. second.) The scanning provides a Now, with the image acquired and
An image can be detected by a number video signal in analog form consisting stored in the frame buffer, data that are
of sensors capable of measuring dif- of a sequence of lines containing infor- important for a particular application
ferent image characteristics. Reflected mation. can be extracted using a number of
visible light can be collected by an or- This information must be converted methods. Pixel point processing is the
dinary video camera. Thermal emis- from an analog waveform to a set of most fundamental of image processing
sion can be collected by devices digital values. An analog-to-digital (A operations. In general, pixels are ma-
sensitized to infrared wavelengths. to D) converter samples the amplitude nipulated based on their respective gray
Finally, tomographic devices, such as of the analog signal and converts it into levels of intensity. Point processing

NOTE. No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use
or operation of any methods, products, instructions or ideas contained in the material herein. No suggested test or procedure should be carried out unless, in the reader's judgment, its
risk is justified. Because of rapid advances in the medical sciences, we recommend that the independent verification of diagnoses and drug dosages should be made. Discussions, views
and recommendations as to medical procedures, choice of drugs and drug dosages are the responsibility of the authors.

Clinical Microbiolosy Newsletter (ISSN 0196-4399) is issued twice monthly in one ind"xed volume by Elsevier Science Publishing Co., 655 Avenue of the Americas, New york, NY
10010. Subscription prices per year:. $98.00 including postage and handling in the United States, Canada, and Mexico. Add $43.00 for postage in the rest of the world. Second-class
postage paid at New York, NY, and at additional mailing offices. Postmaster: send address changes to Clinical Microbiology Newsletter, Elsevier Science Publishing Co., Inc., 655
Avenue of the Americas, New York, NY 10010.

90 0196-4399/90/$0.00 + 02.20 © 1990 Elsevier Science Publishing Co., Inc. Clinical Microbiology Newsletter 12:12,1990
can be handed easily with look-up traction. Many applications could useful when there are variations in light
tables (LUTs) in random access involve a number of these functions. and subject. The difference of the two
memory (RAM). The RAM has an ad- Spatial filtering implies the separa- images in this case will result in the
dress for each pixel intensity. At each tion of components based on their fre- enhancement of the subject.
address in the LUT, the user can place quencies within an image. These Finally, depending on the applica-
any desired value for a pixel of that in- spatial frequencies relate to the rapidity tion, several functions will usually be
tensity. When a pixel from the frame of change in gray levels over a fixed involved in the processing of an
buffer is presented to RAM, the pixel distance. High- or low-pass filters will image. To a basic subsystem, the user
value is replaced with the user-selected attenuate low or high frequency de- may need to add some of the pro-
replacement. These LUTs allow pixels tails. High-pass filtering will accen- cessing capabilities that have been dis-
from the image to be transformed in tuate edges or transition areas. cussed. The application will determine
many different ways. It is important to Additionally, details that were once whether or not the host computer can
remember at this point that adjoining barely visible, muted, or hidden by be used to manipulate image data or if
pixels can give valuable information background noise will now become ap- the processes should be incorporated
concerning brightness trends in an area parent. into hardware with an application-spe-
being processed. The processing of The enhancement of edges in an cific image processor. If real time (de-
pixel group information can be helpful image reduces an image to display only fined as the ability to process data
in identifying structures contained in an the edge data. This information is then within one frame or 1/3oof a second)
image. used for feature recognition through information is needed, such as in the
Sometimes, the image contrast is sophisticated algorithms. If desired, inspection of parts on an assembly line
poor. Under this situation, it is diffi- the edge-enhanced image could be or the installation of a windshield
cult for the system to identify and iso- added back to the original image, re- guided by cameras, the host computer
late structures of interest properly. An suiting in an image with sharper detail. is normally much too slow for image
image with poor contrast can be en- A number of geometric operations, manipulations. In this case, additional
hanced by a technique known as histo- when applied to an image, will allow processing elements are needed in the
gram equalization. Intuitively, contrast for the spatial reorientation of pixel hardware that will free the host com-
is defined as how sharp or dull an data. Depending on the spatial algo- puter for other tasks.
image appears. In image processing, rithm, data can be transformed into new On the other hand, if the subject is
however, contrast is defined with re- locations. There are three basic geo- captured or immobile, as in the case of
spect to gray tones. A low-contrast metric processes--image scaling, rota- tissue analysis or photomicrography,
image would occupy only a small tion, and translation--that allow for real-time information is not required,
number of the available 256 gray levels the sizing, orientation, and movement and the necessary algorithms may take
of intensity, while a high-contrast of an image. Image scaling enlarges or only a few seconds to complete via the
image would have two peaks at the shrinks an image or a portion of an host computer. In this case, the algo-
outer fringes of the 256 gray levels. image. Through this geometric process rithms are incorporated into software
The objective of histogram equalization an image can be zoomed and panned routines. Obviously, depending on the
is to space gray levels equally, so that over the entire image surface while application, there are specific cost con-
they can be more easily visualized. searching for a particular structural ele- straints with regard to the requirement
The gray level range of a low- or high- ment or abnormality. Image rotation for processing speed and the dollar
contrast image can be normalized or provides the ability to rotate an image value of the hardware.
linearly stretched to increase the con- about a central point, while image
trast between areas with narrow gray- translation allows for up-and-down or Use of Digital Image Processing
scale distributions. Obviously, this side-to-side movement. in Microbiology
manipulation must be done without Other functions involve operations on An example of digital image pro-
changing the relationship between the more than one image, such as compar- cessing as applied to microbiology is
gray levels. After histogram equaliza- isons between separate images or the the Aladin system (Analytab Products,
tion, the resultant specimen becomes accumulation of a result from several Plainview, NY). This system is an au-
more easily identifiable, and it is now images. Two such operations are tomated laboratory diagnostic instru-
possible to isolate a particular structure frame averaging and frame subtraction. ment for the interpretation of iden-
of interest using gray levels. Frame averaging results in an image tification and antimicrobial suscep-
Once an image has been processed to that is the average of several frames. tibility products. Briefly, the Aladin
improve contrast, structures of interest This operation is useful in reducing automatically schedules, incubates,
can be isolated. A number of tech- high levels of random noise often ex- processes, discards, tracks, and mon-
niques are available to facilitate separa- perienced with images from the elec- itors patient data and information
tion. These techniques include spatial tron microscope. On the other hand, through the UniScept Dezine-er System
filtering, edge enhancement, image frame subtraction is the resultant image (Analytab Products, Plainview, NY).
scaling, image rotation, image transla- after two images have been subtracted The heart of Aladin is a video image
tion, frame averaging, and frame sub- pixel by pixel. This technique is often processing unit. Aladin is the only in-

Clinical Microbiology Newsletter 12:12,1990 © 1990 Elsevier Science Publishing Co., Inc. 0196-4399/90/$0.00 + 02.20 91
strument in microbiology that incorpo- the field-of-view. If a color determi- lect, process, and display data for the
rates a video image processor for the nation is necessary, as with a biochem- general purpose of improving or en-
interpretation of identification and sus- ical in an indentification product, a hancing an image. This science is
ceptibility products. filter or combination of filters is used to complex and involves knowledge of
Following incubation, a carrier, con- enhance the separation capabilities of optics, electronics, computers, and
taining the patient's disposables, is the system above the 256 levels of mathematics. This paper should enable
moved from the incubator to the read gray. The specific results per well or the reader to understand the basics of
station. (If reagents had been needed, area of interest are called binary partial data acquisition, analysis, and display,
the carrier would have previously been results, which are converted into spe- as well as the hardware needed to con-
positioned at the reagent station.) At cific plus or minus reactions. These struct a simple system. As described,
the read station, the career is aligned in results are then used to determine anti- this science is being widely used for a
the field-of-view. Within the field-of- microbial end points or bacterial iden- variety of applications and is now be-
view, the video imaging system parti- tifications. coming a fundamental tool in many
tions the image into approximately 512 The advantages of using video image areas of the medical field. For further
by 512 pixels. The areas-of-interest processing in microbiology are two information on digital image pro-
within this matrix are the specific wells fold. First, many more points of infor- cessing, refer to the texts listed as ad-
of a susceptibility tray or cupules of an mation can be collected per reaction as ditional reading.
identification product. Each well of a compared with the single data point out-
susceptibility tray, for example, con- put of other instruments (e.g., spectro-
tains approximately 100 to 150 pixels. photometric readers), and secondly, the
The output of the video camera is a instrument is not locked into a specific Additional Reading
specific voltage for each pixel. Each disposable. As long as the disposable 1. Baxes, G. A. 1984. Digital image pro-
pixel is assigned a gray scale value fits into the field-of-view, the image cessing: A practical primer. Cascade
from 0 to 256 that depicts its intensity processing software can be modified to Press, Denver.
of gray. Zero voltage is white while interpret different areas-of-interest. 2. Castleman, K. R. 1979. Digital image
full voltage is black. The image pro- processing. Prentice-Hall, Englewood
Cliffs, NJ.
cessor determines the number of pixels Summary 3. Gonzalez, R. C. and P. Wintz, 1977.
that exceed a given threshold voltage Digital image processing is a science Digital image processing. Addison-
for each well or area-of-interest within that uses hardware and software to col- Wesley, Reading, MA.

Editorial

The National Laboratory Public Health," public health is "what ganism is one that has not been well
we as a society do collectively to assure characterized or does not cause disease
Training Network: the conditions in which people can be under normal host circumstances? In
The Role of Training in healthy," and this is accomplished reality a number of events might pre-
Ensuring the through the application of scientific and vent or impede a laboratory from ful-
technical knowledge for the purposes of filling its contribution to health--
Nation's Health preventing disease and promoting despite the competence of its staff or
health (1). their commitment to the practice of
Ronald O. Valdiserri, M.D., M.P.H.
Division of Laboratory Systems If we adopt this definition, a clinical clinical microbiology. That is why
Public Health Practice Program Office microbiology laboratory contributes to training is an essential component of all
U.S. Department of Health and Human health by its ability to distinguish or- public health activities. It is a means
Services ganisms that can cause disease in of ensuring that scientific and technical
Public Health Service humans from those that do not. Infor- knowledge is appropriately and ade-
Centers for Disease Control
Atlanta, GA 30333 mation obtained through the efforts of quately applied to public health
the microbiology laboratory is used by problems.
clinicians to diagnose and treat infec- Groups and individuals will provide
How, exactly, does one define tious diseases, thereby preventing or different answers when asked to explain
health? Certainly, it can be character- halting serious illness. the reason for training laboratorians.
ized as the absence of disease. But this The paradigm outlined above may The laboratory manager is likely to see
definition, although functional, does seem straightforward, when in fact it is training as a means of ensuring that
not begin to express the extent of not. What if the laboratory is called staff are familiar with routine operating
meaning conveyed in that simple upon to culture an organism that is un- procedures within a laboratory. Manu-
phrase. According to the Institute of usual or very difficult to isolate? facturers may view training as a way to
Medicine (IOM) report "The Future of Worse yet, what happens if the or- convince consumers that their products

92 0196-4399/90/$0.00 + 02.20 © 1990 Elsevier Science Publishing Co., Inc. Clinical Microbiology Newsletter 12:12,1990

You might also like