Improvement of Iris Image Authentication using Moment Based Features

The iris structure offers many properties like stability,uniqueness, easy measurement and high recognition rate, which make the iris authentication so accurate. In this book, an iris verification system is proposed using local textural features to achieve high authentication accuracy, so, two different sets of textural features have been suggested to represent the iris texture attributes; they depend on the spatial variation of the intensity.The introduced features sets are: (i) Spatial Distribution of the Local Average for First Gradient Density (SDLAFGD), and (ii) Spatial Distribution of the Local Average for Second Gradient Density (SDLASGD). As well, this work focuses on the selection of proper features to distinguish between individuals. The weighted Euclidean distance measure was used to make the similarity matching decision. The system was tested using public databases CASIA v1.0 and CASIA v4.0.