Publication: Identificación y clasificación de mariposas con técnicas de visión computarizada
Loading...
Official URL
Full text at PDC
Publication Date
2015
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
Estando a las puertas de finalizar los estudios y tras un Erasmus muy fructífero, la necesidad de buscar un tema para desarrollar un proyecto de fin de carrera me llevó a un Trabajo Académicamente Dirigido (TAD) comenzado por una alumna de la Licenciatura en Matemáticas, inconcluso, sobre visión computarizada con aplicaciones a la biología, en el campo de las mariposas. Dicho proyecto carecía de un desarrollo a fondo de la interfaz gráfica, y de otros objetivos que creemos haber completado, gracias a nuevo material disponible, como son el haber podido procesar un mayor número de muestras, conseguir resultados más reales y automatizar casi al 100% el proceso de interpretación de las muestras. La visión computarizada o también conocida como visión artificial es un nuevo campo de investigación cuyo objetivo principal es que el ordenador entienda lo que ve a través de la programación; en nuestro caso le enseñamos a interpretar imágenes de mariposas. Esta nueva disciplina tiene muchos posibles campos de aplicación a parte de la biología, como por ejemplo en medicina, video-juegos, robótica, procesos industriales, etc. y por tanto muchos nuevos desafíos y problemas por resolver. Era fundamental conocer el campo de la visión artificial, y más concretamente, trabajar con técnicas y procedimientos básicos sobre procesamiento de imágenes para poder plantear y resolver el objetivo del proyecto: crear una base de datos de mariposas inteligente, que al introducir una nueva muestra, fuese capaz de reconocer, a través de técnicas de visión computarizada, si había mariposas parecidas o iguales a la misma.
After a productive Erasmus study period abroad and upon my return to Madrid to complete my degree, for my Final Project I chose to develop a TAD (Trabajo Academicamente Dirigido) which had been initiated by a student in the Mathmatics Degree Program, but remained unfinished. This TAD involved a computerized version with applications in the field of biology, specifically, in the study of butterflies. It lacked an in-depth graphic interface and several other objectives were also incomplete as well. The Project presented here has completed these aspects thanks to the availability of new material and strategies such as having been able to process a greater number of samples, obtain more real results and automate almost 100% of the sample interpretation process. Computer vision, also known as artificial vision, is a new field of research in which computers are programmed to process, analyze and understand images. In our case the computer is taught to interpret images of butterflies. This new field has many possible applications besides those of biology such as in medicine, video games, robotics, industrial processes, etc. Thus new challenges and have arisen and must be solved. The objective of the Project was to create an intelligent data base of butterflies which, when faced with a new sample, would be capable of recognizing through computer vision technology whether similar or same butterflies were already included in the data base. To fulfill this objective it was essential to become familiar with the field of artificial vision, and specifically, with basic techniques and procedures of image processing.
After a productive Erasmus study period abroad and upon my return to Madrid to complete my degree, for my Final Project I chose to develop a TAD (Trabajo Academicamente Dirigido) which had been initiated by a student in the Mathmatics Degree Program, but remained unfinished. This TAD involved a computerized version with applications in the field of biology, specifically, in the study of butterflies. It lacked an in-depth graphic interface and several other objectives were also incomplete as well. The Project presented here has completed these aspects thanks to the availability of new material and strategies such as having been able to process a greater number of samples, obtain more real results and automate almost 100% of the sample interpretation process. Computer vision, also known as artificial vision, is a new field of research in which computers are programmed to process, analyze and understand images. In our case the computer is taught to interpret images of butterflies. This new field has many possible applications besides those of biology such as in medicine, video games, robotics, industrial processes, etc. Thus new challenges and have arisen and must be solved. The objective of the Project was to create an intelligent data base of butterflies which, when faced with a new sample, would be capable of recognizing through computer vision technology whether similar or same butterflies were already included in the data base. To fulfill this objective it was essential to become familiar with the field of artificial vision, and specifically, with basic techniques and procedures of image processing.
Description
Proyecto de Sistemas Informáticos (Universidad Complutense, Facultad de Informática, curso 2014/2015)