面部朝向的识别毕业论文
2022-02-16 20:52:40
论文总字数:18387字
摘 要
头部姿态的估测已经成为如今人脸识别与视觉领域中的重要课题。目的是通过对输入图像的特征提取从而计算人的头部在空间中的方向变化,该技术在驾驶员的疲劳驾驶检测、学生课堂考勤、实时监控等方面有广泛应用。本文研究的面部朝向即是对头部姿态的一种表达。
本论文主要的研究内容是人脸识别的相关技术,另外从贴近实际和容易实现的角度,设计并实现了利用肤色信息和面部特征的面部朝向识别方法。识别面部朝向,首先,考虑到在彩色空间内,肤色的正态分布满足高斯分布的特性,利用高斯肤色模型确定图像中可能的肤色区域。在可能的肤色范围中,包含脸部区域和非脸部区域,利用人的脸部特征关系从可能的区域中确定人脸的边界,实现脸部的定位。在确定了脸部区域后,利用投影法和颜色信息的方法寻找眼睛和嘴巴位置。分析眼睛与嘴巴构成的人脸三角形的参数信息确定人的面部朝向。
本文在MATLAB平台上做面部朝向识别算法的研究。实验表示,本文的方法能精确地定位到人脸,眼睛和嘴巴,以及准确地描述了人的面部朝向。
关键词:人脸检测 肤色分割 眼睛定位 面部朝向 人脸三角形
The Recognition of Facial Orientation
Abstract
Face detection and recognition is one of the hotspots in the field of computer vision and pattern recognition because of its application prospect and important academic value in the field of security inspection. Face-oriented face recognition technology can also be widely applied to fatigue testing, work attendance, defense construction, security and other fields and have a great research prospects. In this paper, based on the practical and easy to implement point of view, use the method which is based on the skin color and facial features to achieve facial recognition.
First, the algorithm of dividing one’s face candidate region from the background by using skin color is studied. In the color space, according to the characteristics of skin color and Gaussian distribution, Gaussian model have needed in segment the color of skin region, and the segmented region is morphologically processed to determine the face candidate region. Part of the candidate area for skin color segmentation is the face area, and some of the non-face areas. Therefore, it is proposed to use the facial geometry to remove the non-human face from the candidate region and determine the face position. According to the eye and mouth to form the characteristics of the human face triangle, through the angle of the triangle deflection to identify the person's face orientation.
The paper studies face orientation recognition algorithm on MATLAB platform. Experiments show that this method can be accurately positioned to face, eyes and mouth, and accurately describe the human face orientation.
Key words: Face detection;Color division;Eye detection,;Facial orientation;Face triangle
目 录
摘 要 I
ABSTRACT II
第一章 绪论 1
1.1 研究背景及意义 1
1.2 国内外研究现状 1
1.3 论文的研究内容和规划 2
第二章 头部姿态估计基础 4
2.1 头部姿态的描述 4
2.1.1 面部朝向的描述 4
2.1.2 欧拉旋转角 4
2.1.3 面部的朝向与欧拉旋转角 5
2.2 头部姿态识别的基本过程 6
2.3 头部姿态识别的主要问题 7
第三章 基于肤色和几何特征的人脸分割 8
3.1 人脸检测 8
3.1.1 人脸检测方法概述 8
3.1.2 本文的方法 8
3.2 肤色区域检测 9
3.2.1 色彩空间 10
3.2.2 肤色模型 11
3.2.3 计算人脸候选区域 12
3.3 人脸区域的分割 13
3.3.1 人脸的外观 13
3.3.2 人脸边界的确定 14
第四章 基于面部特征三角形的朝向识别 16
4.1 特征的选取 16
4.2 眼睛和嘴巴的检测 16
4.2.1眼睛检测 17
4.2.2 嘴巴检测 20
4.3 面部特征三角形的计算 21
4.4 面部朝向的识别 22
4.4.1 点头、仰头的识别 22
4.4.2 转头的识别 23
4.5 实验结果及分析 26
第五章 总结及展望 29
5.1总结 29
5.2展望 29
参考文献 31
致谢 33
请支付后下载全文,论文总字数:18387字