简单背景下人物头像轮廓识别的仿真设计介绍.docxVIP

目录内容摘要：1关 键 词：1Abstract：1Key words：11绪论21.1研究背景与现状21.2人物头像轮廓识别的应用前景31.3本文主要研究内容32简单背景下人物头像轮廓识别的理论和算法42.1色彩空间42.1.1RGB色彩空间42.1.2归一化的RGB色彩空间52.1.3 YCbCr色彩空间52.2基于肤色的人物头部检测系统62.2.1系统流程框图62.2.2系统算法分析73在Matlab上实现简单背景下人物头像轮廓识别仿真103.1Matlab数字图象处理103.1.1图像类型的转换103.1.2图像增强113.1.3边缘检测113.2人物头像轮廓提取实现过程123.2.1人物头像轮廓提取框架123.2.2人物头像轮廓提取的处理过程123.3算法实现133.4仿真实验结果144小结14参考文献：17内容摘要：人物头像轮廓识别是模式识别的关键课题，在模式识别和计算机视觉领域受到广泛关注和应用。人物头像轮廓识别与提取是决定人物的头部位置、图像处理的大小的过程。识别和提取人物头像轮廓是人机交互、智能监控等工作的先决条件。本文中，对于一副简单背景下彩色人物图像,第一个对其读取得到 RGB色彩空间数据值，进一步变换成YCrCb色彩空间的数据。在YCrCb色彩空间下确定像素点判断是否非肤色将图像二值化操作。肤色分割的基础上,进行开或闭运算进一步操作消除噪声。通过结合人物头部尺寸、眼睛以及人物长宽比等信息对肤色区域面积进行筛选、消除假区域;从而标记出人物头部区域，完成对人物头部轮廓提取。从实验中可以看到在简单背景下可以准确的提取人物的头部轮廓，成功率非常高，而在较复杂背景下也可以达到很好的效果，表明该方法具有一定的实用性。关 键 词：简单背景；头部轮廓；色彩空间；肤色模型Abstract：Portraits outline recognition is a key issue for pattern recognition, pattern recognition and computer vision attracted widespread attention and application. Portraits outline recognition and extraction is to determine the position of the head of the figure, the size of the image processing procedure. Identify and extract portraits contour is a prerequisite for human-computer interaction, intelligent monitoring and other work.Herein, the character color at a simple background image, the RGB color space to obtain a first data read its value, the data is further converted into YCrCb color space. Determining pixels to determine whether a non-color image binarization operation in YCrCb color space. On the basis of skin color segmentation, opening or closing operation further operation to eliminate noise. By combining the size of the persons head, eyes and character information such as the aspect ratio of an area of color filter to eliminate false region; thus mark the character of the head region, the completion of the persons head contour extraction. From the experiment can be seen in the background can be simple and accurate extraction head profile figures, the success rate is very high, while in the more complex background