Our face, voice, fingerprints, handwriting, iris, retinal, and the way we walk. These are only some among the features measured by recognizing automated methods based on physiological or behavioral characteristics.
There has been constant innovation in the short history of biometrics. We have biometrics in passports, residence permits, stores, banks… Almost everything we do is controlled, someway, by a computer – and we do not know what happens with the data collected about ourselves. Is this an issue that concerns to social security or to privacy?
Anyway, now scientists working on biometrics at the University of Southampton have found a way to identify ears with a success rate of almost 100 percent. In a paper entitled A Novel Ray Analogy for Enrolment of Ear Biometric, researchers describe how a ray, highlighting tubular structures, can identify them.
Ears have a rich and stable structure that is preserved from birth to old age, and instead of aging they just get bigger. They also do not suffer from changes in facial expression. So, it is indeed a signature.
Although, there are problems to be solved – and they are not computational ones. According to Pieter Spierenburg, professor of Erasmus School of Law, for example, despite growing concerns about violence, projects related to social control must respect privacy.
In his words: “In the past, informal control was based on the acceptance of neighborhood spying and on maintaining the norms regarding marriage and personal lives. Informal social control suffered, however, as a result of individualization and modern ideas about privacy: everyone should mind their own business and street life disappeared.”
Images and models of the ray transform technique are available at: http://users.ecs.soton.ac.uk/ahc08r/rt/index.html
Nenhum comentário:
Postar um comentário