Springe direkt zu Inhalt

2019

Diaz Cortes, Margarita Arimatea: Soft computing applied to optimization, computer vision and medicine

Betreuer
Raúl Rojas, Erik Cuevas
Abschluss
Dr. rer. nat.
Abgabedatum
04.07.2019
Sprache
eng

Artificial intelligence has permeated almost every area of life in modern society, and its significance continues to grow. As a result, in recent years, Soft Computing has emerged as a powerful set of methodologies that propose innovative and robust solutions to a variety of complex problems. Soft Computing methods, because of their broad range of application, have the potential to significantly improve human living conditions. The motivation for the present research emerged from this background and possibility. This research aims to accomplish two main objectives: On the one hand, it endeavors to bridge the gap between Soft Computing techniques and their application to intricate problems. On the other hand, it explores the hypothetical benefits of Soft Computing methodologies as novel effective tools for such problems. This thesis synthesizes the results of extensive research on Soft Computing methods and their applications to optimization, Computer Vision, and medicine. This work is composed of several individual projects, which employ classical and new optimization algorithms. The manuscript presented here intends to provide an overview of the different aspects of Soft Computing methods in order to enable the reader to reach a global understanding of the field. Therefore, this document is assembled as a monograph that summarizes the outcomes of these projects across 12 chapters. The chapters are structured so that they can be read independently. The key focus of this work is the application and design of Soft Computing approaches for solving problems in the following: Block Matching, Pattern Detection, Thresholding, Corner Detection, Template Matching, Circle Detection, Color Segmentation, Leukocyte Detection, and Breast Thermogram Analysis. One of the outcomes presented in this thesis involves the development of two evolutionary approaches for global optimization. These were tested over complex benchmark datasets and showed promising results, thus opening the debate for future applications. Moreover, the applications for Computer Vision and medicine presented in this work have highlighted the utility of different Soft Computing methodologies in the solution of problems in such subjects. A milestone in this area is the translation of the Computer Vision and medical issues into optimization problems. Additionally, this work also strives to provide tools for combating public health issues by expanding the concepts to automated detection and diagnosis aid for pathologies such as Leukemia and breast cancer. The application of Soft Computing techniques in this field has attracted great interest worldwide due to the exponential growth of these diseases. Lastly, the use of Fuzzy Logic, Artificial Neural Networks, and Expert Systems in many everyday domestic appliances, such as washing machines, cookers, and refrigerators is now a reality. Many other industrial and commercial applications of Soft Computing have also been integrated into everyday use, and this is expected to increase within the next decade. Therefore, the research conducted here contributes an important piece for expanding these developments. The applications presented in this work are intended to serve as technological tools that can then be used in the development of new devices.