自己调研之后,写的伽马校正:
TestImgNum =1 ;00 335
str = strcat('.\',num2str(TestImgNum),'.bmp');
L_FaceImg = imread(str);
figure,imshow(uint8(L_FaceImg));
temp = double(L_FaceImg);
temp = temp./256;
gamma = 1/2.2;
y = imadjust(temp,[0; 1],[0; 1],gamma)
% y=gamma(temp);%实现gamma校正
% y = gamma_filter(temp,gamma);
y = y.*256;
figure,imshow(uint8(y));
原理不明,只是会用而已。具体过程就是归一化人脸图像到[0,1]之间之后,运用imadjust函数,将其根据伽马值gamma转换到另一个范围在[0,1]的空间中去,然后乘以255加以倍数,输出即得伽马校正的结果。
xb调研完之后,写的代码:
L_FaceImg=double(L_FaceImg_org);
%L_FaceImg=L_FaceImg.^0.2;
gama=2;
p=255/255^(gama);p=(1/p)^(1/gama);
for i=1:imrow
for
j=1:imcol
%for l=1:k
% r(i,j,l)=floor(p*double(I(i,j,l))^(1/gama));
L_FaceImg_gmma(i,j)=floor(p*L_FaceImg(i,j)^(1/gama));
%end
end
end
完全是原理性的东西。
高斯查分滤波,在网上查DoG查了好久,最后终于终于在pudn上面找到相关c代码,结果被xb两句话解决了……具体见下面:
low=
imfilter(L_FaceImg_gmma,fspecial('gaussian',7,1),'same','replicate');
lowlow=imfilter(L_FaceImg_gmma,fspecial('gaussian',9,2),'same','replicate');
Resimage=low-lowlow;
其中L_FaceImg_gmma是经过伽马光线校正之后的图像,分别作滤波,之后差分相减,就得到高斯差分滤波的结果了……囧。明天去pudn上查查代码,看看是不是这样写的……
下面是fspecial函数的使用样例,写上来原因是蛮喜欢subplot函数,哇咔咔~
I = imread('cameraman.tif');
subplot(2,2,1);imshow(I);title('Original Image');
H = fspecial('motion',20,45);
MotionBlur = imfilter(I,H,'replicate');%replicate表示Input array
values outside the bounds of the array are
assumed to equal the nearest array border value
输入值自动对齐到最近邻
subplot(2,2,2);imshow(MotionBlur);title('Motion Blurred
Image');
H = fspecial('disk',10);
blurred = imfilter(I,H,'replicate');
subplot(2,2,3);imshow(blurred);title('Blurred Image');
H = fspecial('unsharp');
sharpened = imfilter(I,H,'replicate');
subplot(2,2,4);imshow(sharpened);title('Sharpened Image');
对比均衡化还没搞明白,先把代码贴出来,供慢慢研究……
%Contrast Equalization.
a=0.1;
for i=1:imrow
for j=1:imcol
Resimage_temp(i,j)=(abs(Resimage(i,j)))^a; end
end
mu=mean(mean(Resimage_temp));
Resimage=Resimage./(mu^(1/a));
tt=10;
for i=1:imrow
for j=1:imcol
Resimage_temp(i,j)=min(tt,(abs(Resimage(i,j)))^a); end
end
mu=mean(mean(Resimage_temp));
Resimage=Resimage./(mu^(1/a));
Resimage=tt*tanh(Resimage/tt);
Resimage_max=max(max(Resimage));
Resimage_min=min(min(Resimage));
for i=1:imrow
for j=1:imcol
Resimage(i,j)=(Resimage(i,j)-Resimage_min)/(Resimage_max-Resimage_min)*255; end
end
% imshow(uint8(Resimage));
%提取LBP模板
% [result , Lbpface] = lbp(L_FaceImg);
%figure,imshow(uint8(Lbpface));
L_FaceImg_bandpass=uint8(Resimage);
TestImage
= strcat('.\M00模糊脸_bandpass\',num2str(ii),'.bmp');
imwrite(uint8(L_FaceImg_bandpass),TestImage,'bmp');
% figure,imshow(uint8(blurred));
end
晕死,就俩公式:
其中a = 0.1,t= 10.看着挺简单,我咋就做不出来呢?
真是的啊我