作者: Aris Nasuha , Tri Arief Sardjono , Mauridhi Hery Purnomo
DOI: 10.1109/ICITEED.2013.6676230
关键词:
摘要: Lip reading has wide spread application, e.g. audio-visual Automatic Speech Recognition (AV-ASR), silent speech interface and person identification. segmentation is one of important step in lip reading, because it provides basic information to be processed subsequent steps. tracking a process locating associate consecutive video frames. Chan-Vese model region-based algorithm, which also can used as method. This algorithm detect boundary object not defined by gradient, while classical active contour can't applied. method any initial curve the image, necessarily surround object. paper investigates about based on model, preceded color segmentation.