Image Processing Reference
In-Depth Information
2020 campaign [ 1 ] . The campaign has aimed to improve the prognosis, diagnosis, treatment,
and postreatment procedures.
The ophthalmological practices have always been greatly influenced as the result of the im-
provements in health system and biomedical approaches, leading to beter understanding of
the underlying causes of different diseases. The design and use of the image-capturing and
-processing algorithms have also aided in early disease detection and progression, assisting
the treatment and monitoring the postreatment results. Overall in the past two decades, the
procedures have become faster, more reliable, repeatable, and accurate.
The statistics provided by World Health Organization (WHO) in 2010 suggests that there
are about 285 million people in the world who are currently visually impaired, consisting of
39 million who are blind and 246 million people who have low vision, including both moder-
ate and severe visual impairment cases. It has also been indicated that the leading causes of
blindness have been diseases such as cataract, glaucoma, and age-related macular degenera-
tion [ 1 - 4 ] .
The above findings suggest that there is still need for further improvement. The study also
indicates that the majority of the observed complications occur in developing regions of the
world, while 80% of these visual impairments could have been avoided if detected early [ 1 , 5 ].
Despite recent advancements in technology and improvements in biomedical research and
ophthalmological findings, there is still the need for fast, affordable, and reliable diagnosis and
treatment procedures in developing regions.
Current interest in telemedicine and the need for developing new techniques and method-
ologies for retinal image processing as well as a beter, highly accurate, reliable, accessible,
repeatable, and inexpensive system for disease diagnosis has been the key motivation of this
1.1 Image Processing
For decades image processing has been used in industrial application. However, in the past
few years there has been a sudden trend and increase in use due to recognition of image-pro-
cessing capabilities biomedical applications.
Usually, image processing consists of a few main steps including, detection of the object
with a camera and then processing the obtained images using segmentation, normalization,
feature extraction, and matching. The outcome of this can then be interpreted or displayed
[ 6 , 7 ] .
For this study, the above steps have been slightly modified and the associated flow chart is
illustrated in Figure 1 .
FIGURE 1 Flow chart of the image-processing steps.
The initial step in any image-processing procedure is image acquisition. This is a very cru-
cial step as the obtained images directly affect the precision and accuracy of the final outcome.
A few points are usually considered while collecting images; some of which are the general
understanding problem, the required image and possible storage capability and capacity, cap-
turing time, resolution, lighting, camera, available resources, and available options.
The next stage would be the image preprocessing step, where the usual colored image is
manipulated and prepared for the next stages. Since in majority of cases, the overall reduc-
tion in processing time is of great interest, this stage mainly concentrates on manipulation of
images so that the overall complexity of the process is reduced. However, the implemented
Search WWH ::

Custom Search