基于嘴角约束的嘴部轮廓检测
首发时间:2013-12-24
摘要:由于嘴部轮廓的多样性,传统的基于ASM的嘴部轮廓检测方法并不能够准确的获得期望的轮廓,本文提出了一种基于嘴角约束的嘴部轮廓检测算法,该方法首先使用MeanShift聚类对嘴部灰度图进行处理,得到两个嘴角的位置,之后在这两个点的约束下使用改变的ASM算法来检测嘴部轮廓。该方法对不同口型的轮廓都具有较好的检测结果,与传统的基于ASM的嘴部轮廓检测方法相比有更高的适用性和鲁棒性。
For information in English, please click here
Month Contour detection based on corners of the mouth
Abstract:Due to the diversity of the mouth contours in face, The traditional mouth contour detection method based on ASM is not effective,this paper proposes a mouth contour method, Firstly, we use MeanShift clustering to processed mouth grayscale image, and get the location of the mouth corners, then under the constraint of the mouth corners ,we using the traditional mouth contour detection method based on ASM to detect mouth contour. The method has high accuracy to different mouth. Compared with the traditional mouth contour detection method this method has more applicability and robustness.
Keywords: mouth contour detection;MeanShift;ASM
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于嘴角约束的嘴部轮廓检测
评论
全部评论0/1000