直方图匹配算法

        
        function results = HistogramMatching(I1,I2)
            [h1,w1] = size(I1);
            [h2,w2] = size(I2);
            histI1 = imhist(I1)/(h1*w1);               
            histI2 = imhist(I2)/(h2*w2);
            
            s1=zeros(1,256);
            s1(1)=histI1(1);
            s2=zeros(1,256);
            s2(1)=histI2(1);
            % 求原图图均衡的映射表,比如原图中像素值为100的点在新图中对应的像素值是多少
            for index = 2:256
                s1(index) = s1(index-1) + histI1(index);
            end
            % 求匹配图均衡的映射表
            for index = 2:256
                s2(index) = s2(index-1) + histI2(index);
            end
            
            M   = zeros(1,256);
            for i = 1 : 256
                [~,pix] = min(abs(s1(i) - s2));%匹配图均衡化后距离该像素最近的那个像素的对应的原像素
                M(i)    = pix-1;%最终的映射表
            end
            results = M(double(I1)+1);
            results = uint8(results);
        end

I1,I2为灰度图像

效果如下：