灰度閥值變換及二值化

1、當圖像的像素點的灰度大于T的時候，設置這個點為全黑，要不然為全白。這樣可以只選擇我們感興趣的領域。im2bw(I,level); %閾值法從灰度圖、RGB圖創(chuàng)建二值圖。level為人工設定閾值(threshold value)，范圍為0 ,1最大類間方差法（OTSU算法）最大類間方差法是由日本學者大津(Nobuyuki Otsu)于1979年提出的,是一種自適應的閾值確定的方法,又叫大律法,簡稱OTSU。它是按圖像的灰度特性,將圖像分成背景和目標2部分。背景和目標之間的類間方差越大,說明構(gòu)成圖像的2部分的差別越大,當部分目標錯分為背景或部分背景錯分為目標都會導致2部分差別變小。因此,使類間方差

2、最大的分割意味著錯分概率最小。在Matlab中，graythresh函數(shù)使用最大類間方差法獲得圖像的閾值。(注意標點要換一下)I = imread(beauty_yellowflowers.jpg);thresh= graythresh(I);%自適應設置閥值bw1 = im2bw(I, thresh);bw2 = im2bw(I, 130/255);%手工設置閥值subplot(1,3,1);imshow(I);title(original)subplot(1,3,2);imshow(bw1);title(autoset_thresh);subplot(1,3,3);imshow(bw2);

3、title(thresh=130);最小分類錯誤全局二值化算法 (kittlerMet算法)函數(shù)源代碼：function imagBW = kittlerMet(imag)% KITTLERMET binarizes a gray scale image imag into a binary image% Input:% imag: the gray scale image, with black foreground(0), and white% background(255).% Output:% imagBW: the binary image of the gray scale ima

4、ge imag, with kittlers% minimum error thresholding algorithm.% Reference:% J. Kittler and J. Illingworth. Minimum Error Thresholding. Pattern% Recognition. 1986. 19(1):41-47MAXD = 100000;imag = imag(:,:,1);counts, x = imhist(imag); % counts are the histogram. x is the intensity level.GradeI = length

5、(x); % the resolusion of the intensity. i.e. 256 for uint8.J_t = zeros(GradeI, 1); % criterion functionprob = counts ./ sum(counts); % Probability distributionmeanT = x * prob; % Total mean level of the picture% Initializationw0 = prob(1); % Probability of the first classmiuK = 0; % First-order cumu

6、lative moments of the histogram up to the kth level.J_t(1) = MAXD;n = GradeI-1;for i = 1 : n w0 = w0 + prob(i+1); miuK = miuK + i * prob(i+1); % first-order cumulative moment if (w0 1-eps) J_t(i+1) = MAXD; % T = i else miu1 = miuK / w0; miu2 = (meanT-miuK) / (1-w0); var1 = (0 : i)-miu1).2) * prob(1

7、: i+1); var1 = var1 / w0; % variance var2 = (i+1 : n)-miu2).2) * prob(i+2 : n+1); var2 = var2 / (1-w0); if var1 eps & var2 eps % in case of var1=0 or var2 =0 J_t(i+1) = 1+w0 * log(var1)+(1-w0) * log(var2)-2*w0*log(w0)-2*(1-w0)*log(1-w0); else J_t(i+1) = MAXD; end endendminJ = min(J_t);index = find(J

8、_t = minJ);th = mean(index);th = (th-1)/nimagBW = im2bw(imag, th);% figure, imshow(imagBW), title(kittler binary);MATLAB程序：I = imread(beauty_yellowflowers.jpg);imagSW = kittlerMet(I);%Kittler 算法bw1 = im2bw(I, 130/255);%手工設置閥值subplot(1,3,1);imshow(I);title(original);subplot(1,3,2);imshow(imagSW);titl

9、e(kittler binary);subplot(1,3,3);imshow(bw1); title(thresh=130);結(jié)果：Niblack二值化算法：Niblack二值化算法是比較簡單的局部閾值方法，閾值的計算公式是T = m + k*v，其中m為以該像素點為中心的區(qū)域的平均灰度值，v是該區(qū)域的標準差，k是一個系數(shù)。matlab程序如下：I = imread( beauty_yellowflowers.jpg );I = rgb2gray(I);w = 2;% max = 0; min = 0; m,n = size(I); T = zeros(m ,n ); %for i = (w

10、 + 1):(m - w) for j = (w + 1):(n - w) sum = 0; for k = -w:w for l = -w:w sum = sum + uint32(I(i + k,j + l); end end average = double(sum) /(2*w+1)*(2*w+1); s = 0; for k = -w:w for l = -w:w s = s + (uint32(I(i + k,j + l) - average)*(uint32(I(i + k,j + l) - average); end end s= sqrt(double(s)/(2*w+1)*

11、(2*w+1); T(i,j) = average + 0.2*s; end end for i = 1:m for j = 1:n if I(i,j) T(i,j) I(i,j) = uint8(255); else I(i,j) = uint8(0); end end end imshow(I); 此種算法速度很慢，一直都沒等到結(jié)果，也有可能是程序中有死循環(huán)，費解改進的算法如下：（也挺費時間的，效果不好）I = imread( beauty_yellowflowers.jpg );I = rgb2gray(I);m,n = size(I);block = 10;ver = floor(m/

12、block);hor = floor(n/block);T = zeros(m,n);for b_ver = 1:block for b_hor = 1: block% T(ver * (b_ver - 1)+1) : (ver *b_ver),(hor *(b_hor - 1) + 1):(hor*b_hor) = otsu(I(ver * (b_ver - 1)+1) : (ver *b_ver),(hor *(b_hor - 1) + 1):(hor*b_hor); t = 0; for i = (ver * (b_ver - 1)+1) : (ver * b_ver) for j =

13、(hor * (b_hor - 1) + 1):(hor * b_hor) t = t + uint32(I(i,j); end end t = double(t)/(ver * hor); std_deviation = 0; for i = (ver * (b_ver - 1)+1) : (ver * b_ver) for j = (hor * (b_hor - 1) + 1):(hor * b_hor) std_deviation = std_deviation + (uint32(I(i,j) - t)*(uint32(I(i,j) - t); end end std_deviation = sqrt(double(std_deviation)/(ver*hor); thr = t +