1。手动设定阈值
3
/*
手动设置阀值
*/
4 IplImage * binaryImg = cvCreateImage(cvSize(w, h),IPL_DEPTH_8U, 1 );
5 cvThreshold(smoothImgGauss,binaryImg, 71 , 255 ,CV_THRESH_BINARY);
6 cvNamedWindow( " cvThreshold " , CV_WINDOW_AUTOSIZE );
7 cvShowImage( " cvThreshold " , binaryImg );
4 IplImage * binaryImg = cvCreateImage(cvSize(w, h),IPL_DEPTH_8U, 1 );
5 cvThreshold(smoothImgGauss,binaryImg, 71 , 255 ,CV_THRESH_BINARY);
6 cvNamedWindow( " cvThreshold " , CV_WINDOW_AUTOSIZE );
7 cvShowImage( " cvThreshold " , binaryImg );
（１）将图像平均分为若干行列的子图像，行数和列数可以根据应用需要进行选取；
（２）对每一个子图像，利用全局阈值化方法求最优阈值，所得阈值用图像表示；
（３）将表示阈值的图像进行必要的延拓后进行滤波；
（４）将滤波结果作为子图像的新阈值对每个子图像进行阈值化处理
10/*自适应阀值 //计算像域邻域的平均灰度，来决定二值化的值*/
11IplImage* adThresImg= cvCreateImage(cvSize(w, h),IPL_DEPTH_8U,1);
12double max_value=255;
13int adpative_method=CV_ADAPTIVE_THRESH_GAUSSIAN_C;//CV_ADAPTIVE_THRESH_MEAN_C
14 int threshold_type=CV_THRESH_BINARY;
15int block_size=3;//阈值的象素邻域大小
16 int offset=5;//窗口尺寸
17 cvAdaptiveThreshold(smoothImgGauss,adThresImg,max_value,adpative_method,threshold_type,block_size,offset);
18cvNamedWindow("cvAdaptiveThreshold", CV_WINDOW_AUTOSIZE );
19cvShowImage("cvAdaptiveThreshold", adThresImg );
20cvReleaseImage(&adThresImg);
11IplImage* adThresImg= cvCreateImage(cvSize(w, h),IPL_DEPTH_8U,1);
12double max_value=255;
13int adpative_method=CV_ADAPTIVE_THRESH_GAUSSIAN_C;//CV_ADAPTIVE_THRESH_MEAN_C
14 int threshold_type=CV_THRESH_BINARY;
15int block_size=3;//阈值的象素邻域大小
16 int offset=5;//窗口尺寸
17 cvAdaptiveThreshold(smoothImgGauss,adThresImg,max_value,adpative_method,threshold_type,block_size,offset);
18cvNamedWindow("cvAdaptiveThreshold", CV_WINDOW_AUTOSIZE );
19cvShowImage("cvAdaptiveThreshold", adThresImg );
20cvReleaseImage(&adThresImg);
3.
最大熵阀值分割法
图像的信息熵反映了图像的总体概貌。若图像中包含目标，则在目标与背景可分割的交界处信息量（即熵）最大。
22/*最大熵阀值分割法*/
23IplImage* imgMaxEntropy= cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
24MaxEntropy(smoothImgGauss,imgMaxEntropy);
25cvNamedWindow("MaxEntroyThreshold", CV_WINDOW_AUTOSIZE );
26cvShowImage("MaxEntroyThreshold", imgMaxEntropy );//显示图像
27 cvReleaseImage(&imgMaxEntropy );
23IplImage* imgMaxEntropy= cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
24MaxEntropy(smoothImgGauss,imgMaxEntropy);
25cvNamedWindow("MaxEntroyThreshold", CV_WINDOW_AUTOSIZE );
26cvShowImage("MaxEntroyThreshold", imgMaxEntropy );//显示图像
27 cvReleaseImage(&imgMaxEntropy );
///
1/*============================================================================
2= 代码内容：最大熵阈值分割
3= 修改日期:2009-3-3
4= 作者:crond123
5= 博客:http://blog.csdn.net/crond123/
6= E_Mail:crond123@163.com
7===============================================================================*/
8// 计算当前位置的能量熵
9double caculateCurrentEntropy(CvHistogram* Histogram1,int cur_threshold,entropy_state state)
10{
11int start,end;
12int total=0;
13double cur_entropy=0.0;
14if(state== back)
15{
16start =0;
17end = cur_threshold;
18}
19else
20{
21start = cur_threshold;
22end =256;
23}
24for(int i=start;i<end;i++)
25{
26total += (int)cvQueryHistValue_1D(Histogram1,i);//查询直方块的值 P304
27}
28for(int j=start;j<end;j++)
29{
30if((int)cvQueryHistValue_1D(Histogram1,j)==0)
31continue;
32double percentage= cvQueryHistValue_1D(Histogram1,j)/total;
33/*熵的定义公式*/
34cur_entropy+=-percentage*logf(percentage);
35/*根据泰勒展式去掉高次项得到的熵的近似计算公式
36cur_entropy += percentage*percentage;*/
37}
38return cur_entropy;
39// return (1-cur_entropy);
40}
41
42//寻找最大熵阈值并分割
43void MaxEntropy(IplImage*src,IplImage*dst)
44{
45assert(src!= NULL);
46assert(src->depth==8&& dst->depth==8);
47assert(src->nChannels==1);
48CvHistogram* hist= cvCreateHist(1,&HistogramBins,CV_HIST_ARRAY,HistogramRange);//创建一个指定尺寸的直方图
49//参数含义：直方图包含的维数、直方图维数尺寸的数组、直方图的表示格式、方块范围数组、归一化标志
50cvCalcHist(&src,hist);//计算直方图
51double maxentropy=-1.0;
52int max_index=-1;
53// 循环测试每个分割点，寻找到最大的阈值分割点
54for(int i=0;i<HistogramBins;i++)
55{
56double cur_entropy= caculateCurrentEntropy(hist,i,object)+caculateCurrentEntropy(hist,i,back);
57if(cur_entropy>maxentropy)
58{
59maxentropy= cur_entropy;
60max_index= i;
61}
62}
63cout<<"The Threshold of this Image in MaxEntropy is:"<<max_index<<endl;
64cvThreshold(src, dst, (double)max_index,255, CV_THRESH_BINARY);
65cvReleaseHist(&hist);
66}
2= 代码内容：最大熵阈值分割
3= 修改日期:2009-3-3
4= 作者:crond123
5= 博客:http://blog.csdn.net/crond123/
6= E_Mail:crond123@163.com
7===============================================================================*/
8// 计算当前位置的能量熵
9double caculateCurrentEntropy(CvHistogram* Histogram1,int cur_threshold,entropy_state state)
10{
11int start,end;
12int total=0;
13double cur_entropy=0.0;
14if(state== back)
15{
16start =0;
17end = cur_threshold;
18}
19else
20{
21start = cur_threshold;
22end =256;
23}
24for(int i=start;i<end;i++)
25{
26total += (int)cvQueryHistValue_1D(Histogram1,i);//查询直方块的值 P304
27}
28for(int j=start;j<end;j++)
29{
30if((int)cvQueryHistValue_1D(Histogram1,j)==0)
31continue;
32double percentage= cvQueryHistValue_1D(Histogram1,j)/total;
33/*熵的定义公式*/
34cur_entropy+=-percentage*logf(percentage);
35/*根据泰勒展式去掉高次项得到的熵的近似计算公式
36cur_entropy += percentage*percentage;*/
37}
38return cur_entropy;
39// return (1-cur_entropy);
40}
41
42//寻找最大熵阈值并分割
43void MaxEntropy(IplImage*src,IplImage*dst)
44{
45assert(src!= NULL);
46assert(src->depth==8&& dst->depth==8);
47assert(src->nChannels==1);
48CvHistogram* hist= cvCreateHist(1,&HistogramBins,CV_HIST_ARRAY,HistogramRange);//创建一个指定尺寸的直方图
49//参数含义：直方图包含的维数、直方图维数尺寸的数组、直方图的表示格式、方块范围数组、归一化标志
50cvCalcHist(&src,hist);//计算直方图
51double maxentropy=-1.0;
52int max_index=-1;
53// 循环测试每个分割点，寻找到最大的阈值分割点
54for(int i=0;i<HistogramBins;i++)
55{
56double cur_entropy= caculateCurrentEntropy(hist,i,object)+caculateCurrentEntropy(hist,i,back);
57if(cur_entropy>maxentropy)
58{
59maxentropy= cur_entropy;
60max_index= i;
61}
62}
63cout<<"The Threshold of this Image in MaxEntropy is:"<<max_index<<endl;
64cvThreshold(src, dst, (double)max_index,255, CV_THRESH_BINARY);
65cvReleaseHist(&hist);
66}
//
4.
基本全局阀值法
p(x,y)代表点(x,y)处的灰度值，q(x,y)代表点邻域的某种局部特性。
如果阈值的选取只与p(x,y)有关，则是全局阈值，全局阈值是利用全局信息（例如整个图像的灰度直方图）得到的阈值，它仅与全图各像素的本事性质有关，对全图使用；如果阈值的选取与p(x,y),q(x,y)有关，则是局部阈值。
29/*基本全局阀值法*/
30IplImage* imgBasicGlobalThreshold= cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
31cvCopyImage(srcImgGrey,imgBasicGlobalThreshold);
32int pg[256],i,thre;
33for (i=0;i<256;i++) pg[i]=0;
34for (i=0;i<imgBasicGlobalThreshold->imageSize;i++)// 直方图统计
35 pg[(BYTE)imgBasicGlobalThreshold->imageData[i]]++;
36thre = BasicGlobalThreshold(pg,0,256);// 确定阈值
37 cout<<"The Threshold of this Image in BasicGlobalThreshold is:"<<thre<<endl;//输出显示阀值
38 cvThreshold(imgBasicGlobalThreshold,imgBasicGlobalThreshold,thre,255,CV_THRESH_BINARY);// 二值化
39 cvNamedWindow("BasicGlobalThreshold", CV_WINDOW_AUTOSIZE );
40cvShowImage("BasicGlobalThreshold", imgBasicGlobalThreshold);//显示图像
41 cvReleaseImage(&imgBasicGlobalThreshold);
30IplImage* imgBasicGlobalThreshold= cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
31cvCopyImage(srcImgGrey,imgBasicGlobalThreshold);
32int pg[256],i,thre;
33for (i=0;i<256;i++) pg[i]=0;
34for (i=0;i<imgBasicGlobalThreshold->imageSize;i++)// 直方图统计
35 pg[(BYTE)imgBasicGlobalThreshold->imageData[i]]++;
36thre = BasicGlobalThreshold(pg,0,256);// 确定阈值
37 cout<<"The Threshold of this Image in BasicGlobalThreshold is:"<<thre<<endl;//输出显示阀值
38 cvThreshold(imgBasicGlobalThreshold,imgBasicGlobalThreshold,thre,255,CV_THRESH_BINARY);// 二值化
39 cvNamedWindow("BasicGlobalThreshold", CV_WINDOW_AUTOSIZE );
40cvShowImage("BasicGlobalThreshold", imgBasicGlobalThreshold);//显示图像
41 cvReleaseImage(&imgBasicGlobalThreshold);
//
1
/*
============================================================================
2
= 代码内容：基本全局阈值法
3
==============================================================================
*/
4
int
BasicGlobalThreshold(
int
*
pg,
int
start,
int
end)
5
{
//
基本全局阈值法
6
int
i,t,t1,t2,k1,k2;
7
double
u,u1,u2;
8
t
=
0
;
9
u
=
0
;
10
for
(i
=
start;i
<
end;i
++
)
11
{
12
t
+=
pg[i];
13
u
+=
i
*
pg[i];
14
}
15
k2
=
(
int
) (u
/
t);
//
计算此范围灰度的平均值
16
do
17
{
18
k1
=
k2;
19
t1
=
0
;
20
u1
=
0
;
21
for
(i
=
start;i
<=
k1;i
++
)
22
{
//
计算低灰度组的累加和
23
t1
+=
pg[i];
24
u1
+=
i
*
pg[i];
25
}
26
t2
=
t
-
t1;
27
u2
=
u
-
u1;
28
if
(t1)
29
u1
=
u1
/
t1;
//
计算低灰度组的平均值
30
else
31
u1
=
0
;
32
if
(t2)
33
u2
=
u2
/
t2;
//
计算高灰度组的平均值
34
else
35
u2
=
0
;
36
k2
=
(
int
) ((u1
+
u2)
/
2
);
//
得到新的阈值估计值
37
}
38
while
(k1
!=
k2);
//
数据未稳定，继续
39
//
cout<<"The Threshold of this Image in BasicGlobalThreshold is:"<<k1<<endl;
40
return
(k1);
//
返回阈值
41
}
///
5.
OTSU（最大类间差分）
其基本思
想是用阈值把 图像像素划分为两类，通过使划分
后得到的两类的类间方差最大来确定最佳阈值
44 IplImage* imgOtsu = cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
45 cvCopyImage(srcImgGrey,imgOtsu);
46 int thre2;
47 thre2 = otsu2(imgOtsu);
48 cout<<"The Threshold of this Image in Otsu is:"<<thre2<<endl;//输出显示阀值
49 cvThreshold(imgOtsu,imgOtsu,thre2,255,CV_THRESH_BINARY); // 二值化
50 cvNamedWindow("imgOtsu", CV_WINDOW_AUTOSIZE );
51 cvShowImage( "imgOtsu", imgOtsu);//显示图像
52 cvReleaseImage(&imgOtsu);
1 /*======================================================================*/
2 /* OTSU global thresholding routine */
3 /* takes a 2D unsigned char array pointer, number of rows, and */
4 /* number of cols in the array. returns the value of the threshold */
5 /*parameter:
6 *image --- buffer for image
7 rows, cols --- size of image
8 x0, y0, dx, dy --- region of vector used for computing threshold
9 vvv --- debug option, is 0, no debug information outputed
10 */
11 /*
12 OTSU 算法可以说是自适应计算单阈值（用来转换灰度图像为二值图像）的简单高效方法。
13 下面的代码最早由 Ryan Dibble提供，此后经过多人Joerg.Schulenburg, R.Z.Liu 等修改，补正。
14 算法对输入的灰度图像的直方图进行分析，将直方图分成两个部分，使得两部分之间的距离最大。
15 划分点就是求得的阈值。
16 */
17 /*======================================================================*/
18 int otsu (unsigned char *image, int rows, int cols, int x0, int y0, int dx, int dy, int vvv)
19 {
20
21 unsigned char *np; // 图像指针
22 int thresholdValue=1; // 阈值
23 int ihist[256]; // 图像直方图，256个点
24
25 int i, j, k; // various counters
26 int n, n1, n2, gmin, gmax;
27 double m1, m2, sum, csum, fmax, sb;
28
29 // 对直方图置零
30 memset(ihist, 0, sizeof(ihist));
31
32 gmin=255; gmax=0;
33 // 生成直方图
34 for (i = y0 + 1; i < y0 + dy - 1; i++)
35 {
36 np = (unsigned char*)image[i*cols+x0+1];
37 for (j = x0 + 1; j < x0 + dx - 1; j++)
38 {
39 ihist[*np]++;
40 if(*np > gmax) gmax=*np;
41 if(*np < gmin) gmin=*np;
42 np++; /* next pixel */
43 }
44 }
45
46 // set up everything
47 sum = csum = 0.0;
48 n = 0;
49
50 for (k = 0; k <= 255; k++)
51 {
52 sum += (double) k * (double) ihist[k]; /* x*f(x) 质量矩*/
53 n += ihist[k]; /* f(x) 质量 */
54 }
55
56 if (!n)
57 {
58 // if n has no value, there is problems...
59 fprintf (stderr, "NOT NORMAL thresholdValue = 160\n");
60 return (160);
61 }
62
63 // do the otsu global thresholding method
64 fmax = -1.0;
65 n1 = 0;
66 for (k = 0; k < 255; k++)
67 {
68 n1 += ihist[k];
69 if (!n1)
70 {
71 continue;
72 }
73 n2 = n - n1;
74 if (n2 == 0)
75 {
76 break;
77 }
78 csum += (double) k *ihist[k];
79 m1 = csum / n1;
80 m2 = (sum - csum) / n2;
81 sb = (double) n1 *(double) n2 *(m1 - m2) * (m1 - m2);
82 /* bbg: note: can be optimized. */
83 if (sb > fmax)
84 {
85 fmax = sb;
86 thresholdValue = k;
87 }
88 }
89
90 // at this point we have our thresholding value
91
92 // debug code to display thresholding values
93 if ( vvv & 1 )
94 fprintf(stderr,"# OTSU: thresholdValue = %d gmin=%d gmax=%d\n",
95 thresholdValue, gmin, gmax);
96
97 return(thresholdValue);
98 }
6.
/*上下阀值法：利用正态分布求可信区间*/
55IplImage* imgTopDown= cvCreateImage( cvGetSize(imgGrey), IPL_DEPTH_8U,1 );
56cvCopyImage(srcImgGrey,imgTopDown);
57CvScalar mean ,std_dev;//平均值、 标准差
58double u_threshold,d_threshold;
59cvAvgSdv(imgTopDown,&mean,&std_dev,NULL);
60u_threshold= mean.val[0]+2.5* std_dev.val[0];//上阀值
61d_threshold= mean.val[0]-2.5* std_dev.val[0];//下阀值
62//u_threshold = mean + 2.5 * std_dev;//错误
63//d_threshold = mean - 2.5 * std_dev;
64cout<<"The TopThreshold of this Image in TopDown is:"<<d_threshold<<endl;//输出显示阀值
65cout<<"The DownThreshold of this Image in TopDown is:"<<u_threshold<<endl;
66cvThreshold(imgTopDown,imgTopDown,d_threshold,u_threshold,CV_THRESH_BINARY_INV);//上下阀值
67cvNamedWindow("imgTopDown", CV_WINDOW_AUTOSIZE );
68cvShowImage("imgTopDown", imgTopDown);//显示图像
69cvReleaseImage(&imgTopDown);
56cvCopyImage(srcImgGrey,imgTopDown);
57CvScalar mean ,std_dev;//平均值、 标准差
58double u_threshold,d_threshold;
59cvAvgSdv(imgTopDown,&mean,&std_dev,NULL);
60u_threshold= mean.val[0]+2.5* std_dev.val[0];//上阀值
61d_threshold= mean.val[0]-2.5* std_dev.val[0];//下阀值
62//u_threshold = mean + 2.5 * std_dev;//错误
63//d_threshold = mean - 2.5 * std_dev;
64cout<<"The TopThreshold of this Image in TopDown is:"<<d_threshold<<endl;//输出显示阀值
65cout<<"The DownThreshold of this Image in TopDown is:"<<u_threshold<<endl;
66cvThreshold(imgTopDown,imgTopDown,d_threshold,u_threshold,CV_THRESH_BINARY_INV);//上下阀值
67cvNamedWindow("imgTopDown", CV_WINDOW_AUTOSIZE );
68cvShowImage("imgTopDown", imgTopDown);//显示图像
69cvReleaseImage(&imgTopDown);
/
7.
迭代法
迭代法是基于逼近的思想，其步骤如下：
1． 求出图象的最大灰度值和最小灰度值，分别记为ZMAX和ZMIN，令初始阈值T0=(ZMAX+ZMIN)/2；
2． 根据阈值TK将图象分割为前景和背景，分别求出两者的平均灰度值ZO和ZB
3． 求出新阈值TK+1=(ZO+ZB)/2；
4． 若TK=TK+1，则所得即为阈值；否则转2，迭代计算。
1． 求出图象的最大灰度值和最小灰度值，分别记为ZMAX和ZMIN，令初始阈值T0=(ZMAX+ZMIN)/2；
2． 根据阈值TK将图象分割为前景和背景，分别求出两者的平均灰度值ZO和ZB
3． 求出新阈值TK+1=(ZO+ZB)/2；
4． 若TK=TK+1，则所得即为阈值；否则转2，迭代计算。
当TK=TK+1时，说明阈值已经达到稳定了，不需在进行迭代，当TK！=TK+1时，说明阈值还没有稳定，需要用迭代再进一步来确定。
72IplImage* imgIteration= cvCreateImage( cvGetSize(imgGrey), IPL_DEPTH_8U,1 );
73cvCopyImage(srcImgGrey,imgIteration);
74int thre3,nDiffRec;
75thre3 =DetectThreshold(imgIteration,100, nDiffRec);
76cout<<"The Threshold of this Image in imgIteration is:"<<thre3<<endl;//输出显示阀值
77cvThreshold(imgIteration,imgIteration,thre3,255,CV_THRESH_BINARY_INV);//上下阀值
78cvNamedWindow("imgIteration", CV_WINDOW_AUTOSIZE );
79cvShowImage("imgIteration", imgIteration);
80cvReleaseImage(&imgIteration);
73cvCopyImage(srcImgGrey,imgIteration);
74int thre3,nDiffRec;
75thre3 =DetectThreshold(imgIteration,100, nDiffRec);
76cout<<"The Threshold of this Image in imgIteration is:"<<thre3<<endl;//输出显示阀值
77cvThreshold(imgIteration,imgIteration,thre3,255,CV_THRESH_BINARY_INV);//上下阀值
78cvNamedWindow("imgIteration", CV_WINDOW_AUTOSIZE );
79cvShowImage("imgIteration", imgIteration);
80cvReleaseImage(&imgIteration);
/
1/*======================================================================*/
2/* 迭代法*/
3/*======================================================================*/
4// nMaxIter：最大迭代次数；nDiffRec：使用给定阀值确定的亮区与暗区平均灰度差异值
5int DetectThreshold(IplImage*img,int nMaxIter,int& iDiffRec)//阀值分割：迭代法
6{
7//图像信息
8int height= img->height;
9int width= img->width;
10int step= img->widthStep/sizeof(uchar);
11uchar *data= (uchar*)img->imageData;
12
13iDiffRec =0;
14int F[256]={0 };//直方图数组
15int iTotalGray=0;//灰度值和
16int iTotalPixel=0;//像素数和
17byte bt;//某点的像素值
18
19uchar iThrehold,iNewThrehold;//阀值、新阀值
20uchar iMaxGrayValue=0,iMinGrayValue=255;//原图像中的最大灰度值和最小灰度值
21uchar iMeanGrayValue1,iMeanGrayValue2;
22
23//获取(i,j)的值，存于直方图数组F
24for(int i=0;i<width;i++)
25{
26for(int j=0;j<height;j++)
27{
28bt = data[i*step+j];
29if(bt<iMinGrayValue)
30iMinGrayValue= bt;
31if(bt>iMaxGrayValue)
32iMaxGrayValue= bt;
33F[bt]++;
34}
35}
36
37iThrehold=0;//
38iNewThrehold= (iMinGrayValue+iMaxGrayValue)/2;//初始阀值
39iDiffRec = iMaxGrayValue- iMinGrayValue;
40
41for(int a=0;(abs(iThrehold-iNewThrehold)>0.5)&&a<nMaxIter;a++)//迭代中止条件
42{
43iThrehold= iNewThrehold;
44//小于当前阀值部分的平均灰度值
45for(int i=iMinGrayValue;i<iThrehold;i++)
46{
47iTotalGray+= F[i]*i;//F[]存储图像信息
48iTotalPixel+= F[i];
49}
50iMeanGrayValue1= (uchar)(iTotalGray/iTotalPixel);
51//大于当前阀值部分的平均灰度值
52iTotalPixel=0;
53iTotalGray=0;
54for(int j=iThrehold+1;j<iMaxGrayValue;j++)
55{
56iTotalGray+= F[j]*j;//F[]存储图像信息
57iTotalPixel+= F[j];
58}
59iMeanGrayValue2= (uchar)(iTotalGray/iTotalPixel);
60
61iNewThrehold= (iMeanGrayValue2+iMeanGrayValue1)/2;//新阀值
62iDiffRec = abs(iMeanGrayValue2- iMeanGrayValue1);
63}
64
65//cout<<"The Threshold of this Image in imgIteration is:"<<iThrehold<<endl;
66return iThrehold;
67}
68
2/* 迭代法*/
3/*======================================================================*/
4// nMaxIter：最大迭代次数；nDiffRec：使用给定阀值确定的亮区与暗区平均灰度差异值
5int DetectThreshold(IplImage*img,int nMaxIter,int& iDiffRec)//阀值分割：迭代法
6{
7//图像信息
8int height= img->height;
9int width= img->width;
10int step= img->widthStep/sizeof(uchar);
11uchar *data= (uchar*)img->imageData;
12
13iDiffRec =0;
14int F[256]={0 };//直方图数组
15int iTotalGray=0;//灰度值和
16int iTotalPixel=0;//像素数和
17byte bt;//某点的像素值
18
19uchar iThrehold,iNewThrehold;//阀值、新阀值
20uchar iMaxGrayValue=0,iMinGrayValue=255;//原图像中的最大灰度值和最小灰度值
21uchar iMeanGrayValue1,iMeanGrayValue2;
22
23//获取(i,j)的值，存于直方图数组F
24for(int i=0;i<width;i++)
25{
26for(int j=0;j<height;j++)
27{
28bt = data[i*step+j];
29if(bt<iMinGrayValue)
30iMinGrayValue= bt;
31if(bt>iMaxGrayValue)
32iMaxGrayValue= bt;
33F[bt]++;
34}
35}
36
37iThrehold=0;//
38iNewThrehold= (iMinGrayValue+iMaxGrayValue)/2;//初始阀值
39iDiffRec = iMaxGrayValue- iMinGrayValue;
40
41for(int a=0;(abs(iThrehold-iNewThrehold)>0.5)&&a<nMaxIter;a++)//迭代中止条件
42{
43iThrehold= iNewThrehold;
44//小于当前阀值部分的平均灰度值
45for(int i=iMinGrayValue;i<iThrehold;i++)
46{
47iTotalGray+= F[i]*i;//F[]存储图像信息
48iTotalPixel+= F[i];
49}
50iMeanGrayValue1= (uchar)(iTotalGray/iTotalPixel);
51//大于当前阀值部分的平均灰度值
52iTotalPixel=0;
53iTotalGray=0;
54for(int j=iThrehold+1;j<iMaxGrayValue;j++)
55{
56iTotalGray+= F[j]*j;//F[]存储图像信息
57iTotalPixel+= F[j];
58}
59iMeanGrayValue2= (uchar)(iTotalGray/iTotalPixel);
60
61iNewThrehold= (iMeanGrayValue2+iMeanGrayValue1)/2;//新阀值
62iDiffRec = abs(iMeanGrayValue2- iMeanGrayValue1);
63}
64
65//cout<<"The Threshold of this Image in imgIteration is:"<<iThrehold<<endl;
66return iThrehold;
67}
68
///
扫一扫
1万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
