One of the emerging applications of the millimeter-wave imaging technology is its use in biometric recognition.
This is mainly due to some properties of the millimeter-waves such as their ability to penetrate through clothing
and other occlusions, their low obtrusiveness when collecting the image and the fact that they are harmless to
health. In this work we first describe the generation of a database comprising 1200 synthetic images at 94 GHz
obtained from the body of 50 people. Then we extract a small set of distance-based features from each image
and select the best feature subsets for person recognition using the SFFS feature selection algorithm. Finally
these features are used in body geometry authentication obtaining promising results.