Adaptive Contour Image Interpolation for Sign Language Interpretations

Abstract

A modified method for 2D adaptive interpolation of contour images of sign language interpreter is presented. The adaptation is based on the local information from the four neighborhood pixels of processed image and the interpolation kind is changed to zero or bi-linear one. The presented method is used in a system for transmission of contour images of sign language interpreter for deaf people training. The experimental results obtained by simulation of the developed algorithm show that the quality of interpolated images is better than that obtained with the classic low order (zero and first) 2D interpolation and the computational complexity is lower than that of the high order 2D interpolation (spline and cubic).