Robust Image Analysis of Faces for Genetic Applications

Abstrakt

This paper is devoted to the automatic localization of objects in two-dimensional grey values of images of faces. It is motivated by a practical problem in human genetics and the output of the localization of objects in the given database of images is needed for other tasks in genetic research. The mouth and both eyes are jointly localized by means of templates. The paper compares different measures of correlation between two images and studies robustness aspects of these methods with respect to rotation, noise, occlusion and asymmetry in the image. The joint localization of the mouth and both eyes is invariant to rotation of the face.