MÁSCULO, F. M.; http://lattes.cnpq.br/0692607733714409; MÁSCULO, Felipe Maia.
Abstract:
In this work some of the most widely knwon techniques for automatic iris recognition
are described, with special focus on the processing steps proposed by John Daugman,
pioneer of the field. Particular attention was given to the investigation of the effects of the
supression of the noise masks during the comparison of 2 iris codes and to the
development of a method for the construction of the application points matrix with the
purpose of reducing the negative effects caused by occlusions of the iris by eyelashes and
eyelids, as well as by segmentation errors when the noise masks are not taken into
account. It was verified that the proposed method for the selection of the application points
performend better than expected, presenting results superior to those of the traditional
method, obtained with the use of noise masks and application points uniformely
distributed. Therefore, it follows that, from a judicious choice of the application points
matrix, it is possible to supress the noise mask of the process of comparison of iris codes
without significant loss of performance.