The Preservation and Segmentation of Medical Images Using an Ant Colony Optimization-Based Approach

Resumen

Over the course of the last several years, image processing has developed into a prominent
topic that has been the focus of study across a broad variety of fields. Imaging is used in
many different fields, including the medical field, the field of satellite images, the field of
computer vision, and others. Any method of computer vision will need a high degree of
detection accuracy in order to complete the duties of edge detection and segmentation, which
are without a doubt the most labor-intensive and important activities in the field. These tasks
are also among the most challenging to do. Computer-aided segmentation is witnessing a
substantial rise in popularity in the area of medical imaging as a consequence of the
subjectivity and error-prone nature of human segmentation. This is because human
segmentation is performed by hand. Because of this, the efforts that are being put out right
now are focused on the two jobs that come up the most frequently, namely the identification
of edges and the segmentation of medical photos. These are the two activities that are now
receiving the most attention. However, due to the presence of noise and the high level of
variability that is present in medical pictures, it is a significant challenge to devise methods
that are accurate, trustworthy, and practically applicable for edge detection and medical
image segmentation. This is a challenge that is faced by a significant number of researchers.
This is an issue that has been there for a considerable amount of time. It is imperative that
this task be completed in the allotted time. Despite the substantial progress that has been done
in this field, there are still research issues that have not been resolved to a level of
satisfaction. The performance of segmentation on medical pictures and the extraction of
edges from images that have been contaminated by high-density impulse noise are the topics
that are being discussed in these questions. In light of this, the difficulties that were discussed
before are addressed and rectified in this thesis via the use of metaheuristic algorithms in
conjunction with a range of image processing approaches. When it comes to edge recognition
and medical picture segmentation, two problems that are notoriously difficult to address
using more conventional methods, metaheuristics algorithms are uniquely equipped to handle
the difficulties that arise from attempting to address these problems. In other words,
metaheuristics algorithms are able to handle the difficulties that arise from attempting to
address problems that are notoriously difficult to address. These issues include recognizing
borders in photos and segmenting medical images, both of which may be problematic. These issues are summed up as three key research objectives, and each of these aims is addressed in
one of the book's five chapters. This book is made up of one chapter each. The primary
purpose of the first goal is to identify edge mappings in pictures that include both low and
high concentrations of impulse noise. This is the primary objective of the first aim. The ant
colony optimization (ACO) technique and the birds swarm algorithm (BSA) approach are
employed, respectively, to achieve this goal. Both the standard image dataset and the
Berkeley Segmentation Dataset, sometimes known as BSD for short, are used in the process
of putting these methodologies into practice. In contrast to the methods that are currently
being used, this strategy, on the other hand, produces edges that are continuous and smooth
regardless of the fluctuations that are present in the noise. In comparison to the methods that
are being used at the present time, the numerous metaheuristic-based methods that are being
used in this work for edge detection and medical image segmentation have demonstrated a
significant improvement in terms of both the quality and the quantity of the results that they
produce. This improvement has been demonstrated in relation to the methods that are being
used in this work for edge detection and medical image segmentation. This is the case with
regard to the throughput as well as the accuracy of the process.