摘要
嘴唇是人脸部的重要特征,对嘴唇轮廓的提取有着非常重要的应用价值.本文针对彩色的嘴唇图像,采用基于区域生长的分水岭算法对嘴唇的轮廓进行提取.算法首先通过RGB三个分量,进行灰度转换,之后对图像采用区域生长算法进行处理,然后用中值滤波进行去噪和平滑处理,最后,应用分水岭算法得到嘴唇的轮廓.实验证明,此方法实现简单,达到了预期效果,性能良好.
The lips are an important feature of human face, lips contour extraction has a very important application value. In this paper, a region growing - based watershed algorithm is proposed to extract the contours of the lips. Firstly, the three RGB components are converted into grayscale; Then, conduct region growing algorithm on the images, and perform median fikering for de - noising and smoothing, and finally, by applying the watershed algorithm to detect lip contours. Experimental results show that this method is simple, and achieve the desired effect, and excellent performance.
出处
《吉林建筑大学学报》
CAS
2013年第5期53-55,共3页
Journal of Jilin Jianzhu University
基金
吉林建筑工程学院青年科技发展基金资助项目(J 20111041)
关键词
嘴唇
轮廓提取
区域生长
分水岭
lips
contour extraction
region growing
watershed