人脸面部特征提取是自动视觉翻译和人脸识别中的最关键的技术之一.本文对活动轮廓模型理论的基本原理、方法和技术等作了有益的探索和尝试,并在此基础上,提出了提取人脸面部特征的系列算法,特征是精确的轮廓描述而非简单的数字表达.实验中,我们用动态模板方法提取眼睛轮廓,用活动轮廓方法提取眉毛形状,结果证明,该方法具有较强的鲁棒性和自适应性.

Facial features extraction is one of the essential techniques in automated visual interpretation and recognition of human faces. According to the basic theory, methods and techniques of active contour/snake model, a series of algorithms are presented. In this paper, facial features are described accurately about contour but not express simpllely with datas. In the experiments, the contours of eye are captured by deformable template model, the shapes of the eyebrow are extracted by active contour/snake model and the results show that using the algorithms for facial feature extraction can get good adaptability and robustness.