结合图像增强和肤色分割的人脸检测新方法

摘要

There usually exist some problems in color image such as noise, blur and background, which directly affect the face detection results.This paper puts forward a new method which combines image enhancement and skin segment with Ada-Boost algorithm for face detection.The enhancement operation of smoothing and sharpening is used to input image to eliminate noise and enhance the edge information of the image.And the skin area from background using the skin segment method is effectively departedThe AdaBoost algorithm is utilized to locate accurately the face position in candidate region.The experiments illustrate that this method has a better improvement to the problem of leak detection and error detection.%彩色图像中通常存在噪音、模糊、背景等问题,直接影响到人脸检测的结果.提出一种AdaBoost算法结合图像增强和肤色分割的人脸检测新方法.对输入图像进行平滑、锐化图像增强操作,较好地消除噪声干扰和增强图像的边缘信息；利用肤色分割,将肤色区域和背景有效地区分开；在候选区域用AdaBoost算法精确地定位出人脸位置.实验结果证明,该方法对“漏检”和“错检”问题均有较好的改善.