THE USE OF CARDIOGRAMS AND THEIR CHARACTERISTICS FOR HUMAN IDENTIFICATION
Keywords
Abstract
The paper deals with the task of human identification by electrocardiogram. By using three orthogonal leads, electrocardiogram can be represented as a three-dimensional hodograph. A segment of hodograph, which is related to QRS complex, was approximated via canonical parametric splines. It gave us a set of points which describe the curve shape. An electrocardiogram is assigned to the certain class by comparison of the coordinates of the points which describe the shape of approximating spline. For classification the known statistical methods and neural network were applied.
