Face Recognition Using Contour-Based Multiscale Distance Matrix

Face Recognition is an emerging approach in the recent years. In this paper, the formulation of a face recognition approach using contour-based shape descriptor named Multiscale Distance Matrix (MDM) is developed using the concept of inner-distance in the distance matrix instead of the Euclidean distance. In the proposed scheme, first the similarity in the shape of the face is found by taking a sample number of boundary points on the contour of the face image and the Multiscale Distance Matrix (MDM) is constructed and second the closest resemblance of the face image using the geometrical characteristics is expected to be found. MDM is a contour-based and global based approach that extracts global features of a shape. The values in the proposed distance matrix are the inner-distances calculated between two given sample boundary points on the contour. Inner distance is defined as the length of the shortest path between two points within the shape boundary. The MDM method avoids the time-consuming point-wise matching used in most of the previous shape recognition algorithms. For 100 test images as query, MDM gave a performance rate of 80 percent in the retrieval of face images. Keywords—Face recognition, Multiscale Distance Matrix (MDM); Distance Matrix (DM); Inner-distance; Nearest Neighbour Classifier.