Recognition of images through networks – Artificial neurons applied to iridology

In this work we do not intend to prove the validity of the method, but provide an automatic diagnostic tools.
In the development of this work we have used different techniques to improve
nipular digital images, which involve several areas of the science of
computation: digital topology, pattern recognition, processing
images, algorithms and neural networks, as well as iridological techniques.
In the first chapter, a general description of the work is given. In a second
We show the classifications of possible lesions of the human iris on which it is based
the iridology. In chapters three and four the basic elements are developed
about image processing that we have used. Chapter five pre-
the basic aspects of neural networks with emphasis on those
They used to recognize lesions in the iris. Chapter six is ​​the synthesis of what has been seen
in the previous chapters applied to the iridological diagnosis, and from our
perspective, in it the most important contributions of the work are described.
Finally in chapter seven we give our conclusions.
The results obtained have been satisfactory from the point of view of
we were able to correctly identify the lesions present in iris images
human beings, as well as their status and the organs that affect them. Of course, this
does not replace the work of the iridologist because there are characteristics that without being
injuries in the iris at any given time could be identified as such.

Ana Eugenia Romo Gonz · lez
to obtain the degree of Master of Science in
Electric engineering
with option in Computing
February 2005

Download full abstract (Spanish): tesisAnaEugeniaR

 

Leave a Reply

Your email address will not be published.

Please slide to verify that you are human *