基于直方图统计学习的人脸检测方法  

Face Detection Method Based on Histogram Statistical Learning

作　　者：袁泉[1] 杨杰[1] 杜春华[1] 吴证[1] 

机构地区：[1]上海交通大学模式识别与图像处理研究所,上海200240

出　　处：《计算机工程》2008年第19期182-184,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目(2007AA01Z100);国家自然科学基金资助项目(60675023;60602012)

摘　　要：提出一种基于直方图统计学习的人脸检测方法,对人脸样本和非人脸样本进行小波变换,运用一组小波系数来表征各种人脸特征信息。统计每个训练样本的直方图分布,用于描述人脸和非人脸外观特征的概率分布,每个直方图表示一组小波系数与它们在人脸中位置的联合概率密度。该方法可以准确检测自然场景中的多幅人脸,对侧面人脸有很好的检测效果。This paper presents a face detection method based on histogram statistical learning. It does wavelet transform to the face samples and non-face samples and uses groups of wavelet coefficients to represent all kinds of attributes of face. The probability distribution of visual attributes of face and non-face is represented by statistics of distribution of histograms for each training sample, and each histogram represents the joint probability distribution of a subset of wavelet coefficients and their position on the face. This method can detect multiple faces in the natural scenes accurately. It gives a good detection performance for profile-view face detection.

关 键 词：人脸检测 小波变换 直方图统计 贝叶斯决策规则 ADABOOST算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]