A Novel Logarithmic Mapping Algorithm for the Human IRIS Recognition using a MACH Filter
Huiqi Yvonne Lu, Peter Bone, Rupert C. D. Young, Chris R. Chatwin · 2007
In this paper, we solve the fully invariant problem of iris recognition by combining two existing techniques: logarithmic map and MACH filter; each capable of achieving invariance to several of the possible variations of a target object. The out-of-plane rotation invariance is achieved using a maximum average correlation height filter (MACH), and the scale and in-plane rotation invariance is achieved by using a Logarithmic r-thetas mapping (log map) of a localized region of the image space. A change in scale or rotation of the iris image will result a horizontal or vertical shift in the log map, which makes the iris detectable by correlation with the log map of the reference image.