文档介绍：中文摘要
人像特征价值的分析能够明确价值高、独立性强的特征,为人像识别软件和相关人像识别提供支撑资料,提高人像识别和鉴定的准确率和效率。本课题随机选择50名25-55岁汉族成年男性,按照第二代身份证证件照片的标准拍摄人像正面照片,作为实验研究样本。对人像17项形态指标特征、17项长短间距特征、6项方向角度特征进行测量、统计、分析,找出了人像形态特征、长短间距测量特征、方向角度特征的分布规律,以及测量特征之间相关性大小,论证了人像各特征在人像识别中的价值。为计算机识别提供基础资料支持,为人像检验鉴定提供评断特征质量的依据。
关键词人像识别人像特征测量相关性
外文摘要
Title Analysis Of Facial Features Value
Abstract
Through defining high value and high independence of facial features, the analysis of facial features value not only can provide supporting materials for the facial recognition software and the correlative facial recognition ,but also can improve the accuracy and efficiency of facial recognition and identification. In this subject, the writer chooses 50 adult men at random, whose ages range from 25 to 50 among the Han nationality. Their facial photos, which are taken according to the second standard of the ID card, are adapted as the research samples. On the basis of the meterage, the statistics and the analysis of 17 morphology index-features,17 length distance features and 6 direction-and-angle features, this subject finds out following research results --- the distributing rule of facial morphology features, length distance features,direction-and-angle features and the relativity between those features, which eventually prove the value of each facial feature in the facial recognition, offer the basic material support to the automatic facial recognition and provide the evidence of judging the quality of the features for the facial recognition.
Key words facial recognition; facial features; measurement; relativity
目次
1 引言 1
2 研究方法、步骤 4
研究样本照片的制备 4
拍摄图像 4
几何内容处理 4
保存图像 4
指标特征的选取 4
形态特征 6
长短间距测量特征 7
方向角度特征 9
特征的测量、统计方法 9
数据处理分析方法 10
数据预处理 10
制作频数分布表、绘图 10
相关性分析 11
3 实验研究结果及分析 11
各项指标特征测量、统计数据 11
人像17项形态指标特征统计数据 11
人像17项长短间距测量特征统计数据 14
人像6项方向角度特征测量、统计数据 19
各项指标特征频数分布及分析 21
人像17项形态指标特征频数分布表和频数分布图 21
人像1