I. Introduction
Immunization is the most cost effective medical intervention to reduce the morality rates for infants and children under five. We are developing a concept of mobile phone based child identification for improved immunization in developing countries under a collaborative project between the “Center for Life Science Automation, Rostock, Germany,” and “Shah Satnamji Green-S Welfare Force Wing, Sirsa, India.” The child’s ear photograph taken by a mobile phone must be matched with in-database existing ear photographs for an accurate identification of small children and infants. This is of significance in information technology supported health care delivery to the underprivileged population of developing countries where a reliable mechanism for the identification and registration of the masses is not in place. The problem of ear imaging based child identification is challenging due to the changes in the photographic conditions (such as illumination, distance of camera, viewpoint, background, and so on) and the presence of unwanted objects in the background of the photograph. Therefore, a robust matching of photographs is required.