文档介绍：彩色图像肤色区域分割算法设计
中文摘要
人脸自动识别系统是目前计算机视觉领域的一个重要研究课题,而人脸自动识别系统中作为定位人脸的人脸检测是整个人脸识别系统正常及高效工作的基础。由于人脸的非刚性,使人脸检测成为一个相当复杂的模式识别问题。近年来,由于人脸检测在安全监视、基于内容的图像检索等领域的潜在应用价值,人脸检测已成为一个独立课题并受到众多研究人员的普遍重视。
肤色是人脸非常重要的一个特性。研究表明:尽管不同种族、不同年龄、不同性别的人肤色看上去不同,但这种不同主要集中在亮度上,在去除亮度的色度空间里,不同的肤色分布具有聚类性。为了利用肤色在色度空间的聚类性,本文在多种色彩空间中选取YCbCr色彩空间进行肤色提取。
本文针对复杂背景下的彩色图像,提出了一种应用YCbCr颜色空间的色度分量,建立肤色分布模型,进行肤色区域的分割。即针对每幅图像,可自动优选对应的阈值,完成肤色区域的分割。通过大量数据实验证明,这种分割算法是有效和可靠的。分割后得到的肤色区域和仅利用选定的不变阈值分割方法相比,更加准确。
实验结果表明,本算法是有效的,具有较高的检测性能。
关键词人脸检测,彩色空间,肤色模型,图像分割
Title Color image color region segmentation algorithm
Abstract
Automatic recognition of human faces is an important research area puter vision and image understanding in recent years. Human face detection that acts as one part of the automatic human face recognition system and is responsible of locating the faces is the prerequisite for the whole system to ensure it works normally and high efficiently. Due to human face' s variability, human face detection es a plex issue of pattern recognition. Recently, because of the potential value in most areas, such as security surveillance, content-based image retrieval, etc, human face detection has e a single research topic and attracted broad attention of more and more researchers.
Facial skin color is a very important feature. The results show that: Although of different races, different ages, different sex color look different, but the difference is mainly concentrated in the brightness, the brightness of the color space in the removal, the different nature of the color distribution of cluster. In order to use color in the color space clustering, this paper in a variety of color space, YCbCr color space, select the color extraction.
In this paper, color images plex background, a YCbCr color space, color ponent, the establishment of skin color distribution model for skin color regions segmentation. That for each image, automatically optimal threshold corresponding plete the color region segmentation. Large amount