Two methodologies for iris detection and location in face images

• Author(s): Fernando Jorge Soares Carvalho ; João Manuel Ribeiro da Silva Tavares
• Date: 2007
• Persistent ID: http://hdl.handle.net/10216/319
• Origin: Repositório Aberto da Universidade do Porto
• Subject(s): Ciências tecnológicas; Tecnologia; Tecnologia de computadores
• Document Type: Conference Object
• Language: English

Description

In this paper two methodologies todetect and locate the iris of the eye in static images, are pre-sented. One uses the Hough’s transform and other is based on deformable templates. In both methodologies, the contour of the iris is represented geometrically by a circumference and its radius and centre are considered as its control parameters. The dynamic update of these parameters allows the determination of its final values, enabling the fully definition of the iris in the image. After a description of the considered methodologies, the advantages and disadvantages between them are presented, as well as some experimental results, conclusions and perspectives of future work.