一、简介

人脸检测是人脸识别、人机交互、智能视觉监控等:工作的前提。近年来,在模式识别与计算机视觉领域,人脸检测已经成为一个受到普遍 重视、研究十分活跃的方向。本文针对复杂背景下的彩色正面人脸图像，将肤色分割、模板匹配与候选人脸图像块筛选结合起来，构建了人脸检测实验系统，并用自制的人脸图像数据库在该系统下进行了一系列的实验统计。本文首先介绍了人脸检测技术研究的背景和现状，阐明人脸检测技术发展的重要意义，对目前常用的一一些检测算法进行了总结,然后着重阐述了基于肤色分割和模板匹配验证的人脸检测算法。肤色是人脸重要特征，在通过肤色采样统计和聚类分析后，确立一种在YCbCr空间下的基于高斯模型的肤色分割方法。在YCbCr色彩空间中建立肤色分布的高斯模型,得到肤色概率似然图像,在最佳动态阈值选取算法下完成肤色区域的分割。采用 数学形态学和一些先验知识对所得到的肤色区域进行人脸特征筛选,进–步剔除非人脸肤色区域，减少候选人脸数量,简化后续检测过程的处理。本文使用平均模板匹配方法对候选人脸进行确认，并针对图像中的人脸通常有一定角度旋转和尺寸大小不确定的问题，通过计算候选人脸图像块的偏转角度和面积，并以此调整模板,优化模板配准，提高模板匹配的准确性，同时避免使用多尺度模板进行多次匹配运算，提高算法效率。提出利用候选人脸图像区域和模板质心作为配准的原点,抑制人脸图像噪声的干扰。最后构建了基于肤色分割和模板验证的人脸检测试验系统，并对该系统采用自制人脸图像数据库进行测试。实验结果表明，系统算法是有效的，具有较高的检测性能和低的误判率。

二、源代码

function varargout = untitled(varargin) % UNTITLED MATLAB code for untitled.fig %      UNTITLED, by itself, creates a new UNTITLED or raises the existing %      singleton*. % %      H = UNTITLED returns the handle to a new UNTITLED or the handle to %      the existing singleton*. % %      UNTITLED('CALLBACK',hObject,eventData,handles,...) calls the local %      function named CALLBACK in UNTITLED.M with the given input arguments. % %      UNTITLED('Property','Value',...) creates a new UNTITLED or raises the %      existing singleton*.  Starting from the left, property value pairs are %      applied to the GUI before untitled_OpeningFcn gets called.  An %      unrecognized property name or invalid value makes property application %      stop.  All inputs are passed to untitled_OpeningFcn via varargin. % %      *See GUI Options on GUIDE's Tools menu.  Choose "GUI allows only one %      instance to run (singleton)". % % See also: GUIDE, GUIDATA, GUIHANDLES % Edit the above text to modify the response to help untitled % Last Modified by GUIDE v2.5 19-May-2020 15:25:06 % Begin initialization code - DO NOT EDIT gui_Singleton = 1; gui_State = struct('gui_Name',       mfilename, ...    'gui_Singleton',  gui_Singleton, ...    'gui_OpeningFcn', @untitled_OutputFcn, ...    'gui_OutputFcn',  @untitled_OutputFcn, ...    'gui_LayoutFcn',  [] , ...    'gui_Callback',   []); if nargin && ischar(varargin{1})    gui_State.gui_Callback = str2func(varargin{1}); end if nargout    [varargout{1:nargout}] = gui_mainfcn(gui_State, varargin{:}); else    gui_mainfcn(gui_State, varargin{:}); end % End initialization code - DO NOT EDIT % --- Executes just before untitled is made visible. function untitled_OutputFcn(hObject, eventdata, handles, varargin) % This function has no output args, see OutputFcn. % hObject    handle to figure % eventdata  reserved - to be defined in a future version of MATLAB % handles    structure with handles and user data (see GUIDATA) % varargin   command line arguments to untitled (see VARARGIN) % Choose default command line output for untitled handles.output = hObject; % Update handles structure guidata(hObject, handles); % UIWAIT makes untitled wait for user response (see UIRESUME) % uiwait(handles.figure1); % --- Outputs from this function are returned to the command line. %-------------------------------------- pushbutton1_Callback(回调函数) % --- Executes on button press in pushbutton1. function pushbutton1_Callback(hObject, eventdata, handles) global im imycbcr skin1 skin2 lpf %----------------宣布为全域变量 % hObject handle to pushbutton1 (see GCBO) % eventdata reserved - to be defined in a future version of MATLAB % handles structure with handles and user data (see GUIDATA) if get (gcbo, 'Value' ) ==1;    im=imread('10.jpg' ) ; %------------------------------读入彩色图像    axes(handles.axes1)    set(handles.axes1, 'XMinorTick' , 'on' ) %-------------嵌入Axes(1) 回调函数    imshow(im) ; %------------------------------显示彩色图像 else    imformats end % Hint: get(hObject, ' Value' ) returns push state of togglebutton1 %-------------------------------------- pushbutton2_Callback(回调函数) % --- Executes on button press in pushbutton2. function pushbutton2_Callback(hObject, eventdata, handles) % hObject handle to pushbutton2 (see GCBO) % eventdata reserved - to be defined in a future version of MATLAB % handles structure with handles and user data (see GUIDATA) global im imycbcr skin1 skin2 lpf %------    cr = filter2(lpf, cr) ;    cr = reshape(cr, 1, prod(size(cr) ) ) ;    bmean = mean(cb) ; %--------------------------------求平均值    rmean = mean(cr) ;    brcov = cov(cb, cr) ;    skin1 = zeros(dim(1) , dim(2) ) ;    for i = 1: dim(1)        for j = 1: dim(2)            cb = double(imycbcr(i, j, 2) ) ;            cr = double(imycbcr(i, j, 3) ) ;            x =[(cb-bmean);(cr-rmean)] ;            skin1(i,j) = exp(-0.5*x'*inv(brcov)*x) ; %---------------------------------计算任意像素为皮肤的概率          skin1 = skin1./max(max(skin1) ) ;    axes(handles. axes3)    set(handles. axes3, 'XMinorTick' , 'on' ) %-------------嵌入Axes(3) 回调函数    imshow(skin1) ; %------------------------------显示皮肤概率图像 else    imformats end % Hint: get(hObject, ' Value' ) returns push state of pushbutton3 %-------------------------------------- pushbutton4_Callback(回调函数) % --- Executes on button press in pushbutton4. function pushbutton4_Callback(hObject, eventdata, handles) % hObject handle to pushbutton4 (see GCBO) % eventdata reserved - to be defined in a future version of MATLAB % handles structure with handles and user data (see GUIDATA) global im imycbcr skin1 skin2 lpf %----------------: 0.05        skin2 = zeros(size(skin1,1),size(skin1,2)) ;        skin2(find(skin1>threshold)) = 1;        change = sum(sum(skin2 - previousSkin2) ) ;        changelist = [changelist change] ;        previousSkin2 = skin2;    end    %----------------------------------找出最佳门限值    [C, I] = min(changelist) ;    optimalThreshold = (7-I) *0.1;    skin2 = ones(size(skin1,1),size(skin1,2)) ;    skin2(find(skin1>optimalThreshold))= 0;    skin2 = filter2(lpf, skin2) ; %------------------------中值滤波    axes(handles.axes5)    set(handles. axes5, 'XMinorTick','on' ) %-------------嵌入Axes(4) 回调函数    imshow(skin2); %------------------------------显示皮肤二值化图像 else    imformats end % Hint: get(hObject, ' Value' ) returns push state of pushbutton4 %-------------------------------------- pushbutton5_Callback(回调函数) % --- Executes on button press in pushbutton5. function pushbutton5_Callback(hObject, eventdata, handles) % hObject handle to pushbutton5 (see GCBO) % eventdata reserved - to be defined in a future version of MATLAB % handles structure with handles and user data (see GUIDATA) global im imycbcr skin1 skin2 lpf %----------------宣布为全域变量 if get (gcbo, 'Value' ) ==1;    sumarea = bwarea(skin2) ; %-----------计算对象的总面积， 为利用面积进行判别做准备    [L,numobj] =bwlabel(skin2, 8) ;    avearea = sumarea/numobj; %-----------计算出对象的平均面积    A = zeros(4, numobj) ;%179    n = 1;          % ------------------------获得一副只包含该区域的图像， 让图像其他区域为黑色        bwsegment = bwselect(skin2, y, x, 8) ;        % --------------------- 计算出该区域内分割块的数目        [L, numobjs] = bwlabel(bwsegment, 4) ;        % ---------------------------------得到区域孔的数目        numfeatures = bweuler(bwsegment, 4) ;        numholes = 1 - numfeatures;        % --------------------------------扫描得出区域坐标        top = -1;        buttom =-1;        left = -1;        right = -1;        dim=size(bwsegment);        for i=1:dim(1)            for j=1:dim(2)                if(bwsegment(i,j) ~= 0)                    top=i;                    break;                end;            end;            if (top ~= -1) break; end;        end;

三、运行结果