七种常见阈值分割代码(Otsu、最大熵、迭代法）

本文转自：http://www.cnblogs.com/skyseraph/archive/2010/12/21/1913058.html

主程序(核心部分) 

代码

1 
2 
3 
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 );
8 //cvReleaseImage(&binaryImg); 
9  
10 
11 IplImage* adThresImg = cvCreateImage(cvSize(w, h),IPL_DEPTH_8U, 1);
12 double max_value=255;
13 int adpative_method=CV_ADAPTIVE_THRESH_GAUSSIAN_C;//CV_ADAPTIVE_THRESH_MEAN_C
14  int threshold_type=CV_THRESH_BINARY;
15 int 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);
18 cvNamedWindow("cvAdaptiveThreshold", CV_WINDOW_AUTOSIZE );
19 cvShowImage( "cvAdaptiveThreshold", adThresImg );
20 cvReleaseImage(&adThresImg);
21 
22  
23 IplImage* imgMaxEntropy = cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
24 MaxEntropy(smoothImgGauss,imgMaxEntropy);
25 cvNamedWindow("MaxEntroyThreshold", CV_WINDOW_AUTOSIZE );
26 cvShowImage( "MaxEntroyThreshold", imgMaxEntropy );//显示图像
27   cvReleaseImage(&imgMaxEntropy ); 
28 
29 
30 IplImage* imgBasicGlobalThreshold = cvCreateImage(cvGetSize(imgGrey),IPL_DEPTH_8U,1);
31 cvCopyImage(srcImgGrey,imgBasicGlobalThreshold);
32 int pg[256],i,thre; 
33 for (i=0;i<256;i++) pg[i]=0;
34 for (i=0;i<imgBasicGlobalThreshold->imageSize;i++) // 直方图统计
35   pg[(BYTE)imgBasicGlobalThreshold->imageData[i]]++; 
36 thre = 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 );
40 cvShowImage( "BasicGlobalThreshold", imgBasicGlobalThreshold);//显示图像
41   cvReleaseImage(&imgBasicGlobalThreshold);
42 
43 
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);
53 
54 
55 IplImage* imgTopDown = cvCreateImage( cvGetSize(imgGrey), IPL_DEPTH_8U, 1 );
56 cvCopyImage(srcImgGrey,imgTopDown);
57 CvScalar mean ,std_dev;//平均值、 标准差
58 double u_threshold,d_threshold;
59 cvAvgSdv(imgTopDown,&mean,&std_dev,NULL); 
60 u_threshold = mean.val[0] +2.5* std_dev.val[0];//上阀值
61 d_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;
64 cout<<"The TopThreshold of this Image in TopDown is:"<<d_threshold<<endl;//输出显示阀值
65 cout<<"The DownThreshold of this Image in TopDown is:"<<u_threshold<<endl;
66 cvThreshold(imgTopDown,imgTopDown,d_threshold,u_threshold,CV_THRESH_BINARY_INV);//上下阀值
67 cvNamedWindow("imgTopDown", CV_WINDOW_AUTOSIZE );
68 cvShowImage( "imgTopDown", imgTopDown);//显示图像 
69 cvReleaseImage(&imgTopDown);
70 
71 
72 IplImage* imgIteration = cvCreateImage( cvGetSize(imgGrey), IPL_DEPTH_8U, 1 );
73 cvCopyImage(srcImgGrey,imgIteration);
74 int thre3,nDiffRec;
75 thre3 =DetectThreshold(imgIteration, 100, nDiffRec);
76 cout<<"The Threshold of this Image in imgIteration is:"<<thre3<<endl;//输出显示阀值
77 cvThreshold(imgIteration,imgIteration,thre3,255,CV_THRESH_BINARY_INV);//上下阀值
78 cvNamedWindow("imgIteration", CV_WINDOW_AUTOSIZE );
79 cvShowImage( "imgIteration", imgIteration);
80 cvReleaseImage(&imgIteration);

模块程序

迭代法

代码

1 
2 
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
5 int DetectThreshold(IplImage*img, int nMaxIter, int& iDiffRec) //阀值分割：迭代法
6 {
7 //图像信息
8 int height = img->height;
9 int width = img->width;
10 int step = img->widthStep/sizeof(uchar);
11 uchar *data = (uchar*)img->imageData;
12 
13 iDiffRec =0;
14 int F[256]={ 0 }; //直方图数组
15 int iTotalGray=0;//灰度值和
16 int iTotalPixel =0;//像素数和
17 byte bt;//某点的像素值
18 
19 uchar iThrehold,iNewThrehold;//阀值、新阀值
20 uchar iMaxGrayValue=0,iMinGrayValue=255;//原图像中的最大灰度值和最小灰度值
21 uchar iMeanGrayValue1,iMeanGrayValue2;
22 
23 //获取(i,j)的值，存于直方图数组F
24 for(int i=0;i<width;i++)
25 {
26 for(int j=0;j<height;j++)
27 {
28 bt = data[i*step+j];
29 if(bt<iMinGrayValue)
30 iMinGrayValue = bt;
31 if(bt>iMaxGrayValue)
32 iMaxGrayValue = bt;
33 F[bt]++;
34 }
35 }
36 
37 iThrehold =0;//
38 iNewThrehold = (iMinGrayValue+iMaxGrayValue)/2;//初始阀值
39 iDiffRec = iMaxGrayValue - iMinGrayValue;
40 
41 for(int a=0;(abs(iThrehold-iNewThrehold)>0.5)&&a<nMaxIter;a++)//迭代中止条件
42 {
43 iThrehold = iNewThrehold;
44 //小于当前阀值部分的平均灰度值
45 for(int i=iMinGrayValue;i<iThrehold;i++)
46 {
47 iTotalGray += F[i]*i;//F[]存储图像信息
48 iTotalPixel += F[i];
49 }
50 iMeanGrayValue1 = (uchar)(iTotalGray/iTotalPixel);
51 //大于当前阀值部分的平均灰度值
52 iTotalPixel =0;
53 iTotalGray =0;
54 for(int j=iThrehold+1;j<iMaxGrayValue;j++)
55 {
56 iTotalGray += F[j]*j;//F[]存储图像信息
57 iTotalPixel += F[j]; 
58 }
59 iMeanGrayValue2 = (uchar)(iTotalGray/iTotalPixel);
60 
61 iNewThrehold = (iMeanGrayValue2+iMeanGrayValue1)/2; //新阀值
62 iDiffRec = abs(iMeanGrayValue2 - iMeanGrayValue1);
63 }
64 
65 //cout<<"The Threshold of this Image in imgIteration is:"<<iThrehold<<endl;
66 return iThrehold;
67 }
68

  

 

Otsu代码一 

代码

1 
2 
3 
4 
5 
11 
17 
18 int otsu (unsigned char*image, int rows, int cols, int x0, int y0, int dx, int dy, intvvv)
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++; 
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]; 
53 n += ihist[k]; 
54 }
55 
56 if (!n) 
57 {
58 // if n has no value, there is problems...
59 fprintf (stderr, "NOT NORMAL thresholdValue = 160n");
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 
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=%dn",
95 thresholdValue, gmin, gmax);
96 
97 return(thresholdValue);
98 }

Otsu代码二 

代码

1

2

3

4 int otsu2 (IplImage *image)

5 {

6 int w = image->width;

7 int h = image->height;

8 

9 unsigned char*np; // 图像指针

10 unsigned char pixel;

11 int thresholdValue=1; // 阈值

12 int ihist[256]; // 图像直方图，256个点

13 

14 int i, j, k; // various counters

15 int n, n1, n2, gmin, gmax;

16 double m1, m2, sum, csum, fmax, sb;

17 

18 // 对直方图置零...

19 memset(ihist, 0, sizeof(ihist));

20 

21 gmin=255; gmax=0;

22 // 生成直方图

23 for (i =0; i < h; i++) 

24 {

25 np = (unsigned char*)(image->imageData + image->widthStep*i);

26 for (j =0; j < w; j++) 

27 {

28 pixel = np[j];

29 ihist[ pixel]++;

30 if(pixel > gmax) gmax= pixel;

31 if(pixel < gmin) gmin= pixel;

32 }

33 }

34 

35 // set up everything

36 sum = csum =0.0;

37 n =0;

38 

39 for (k =0; k <=255; k++) 

40 {

41 sum += k * ihist[k];

42 n += ihist[k];

43 }

44 

45 if (!n) 

46 {

47 // if n has no value, there is problems...

48 //fprintf (stderr, "NOT NORMAL thresholdValue = 160n");

49 thresholdValue =160;

50 goto L;

51 }

52 

53 // do the otsu global thresholding method

54 fmax =-1.0;

55 n1 =0;

56 for (k =0; k <255; k++) 

57 {

58 n1 += ihist[k];

59 if (!n1) { continue; }

60 n2 = n - n1;

61 if (n2 ==0) { break; }

62 csum += k *ihist[k];

63 m1 = csum / n1;

64 m2 = (sum - csum) / n2;

65 sb = n1 * n2 *(m1 - m2) * (m1 - m2);

66

67 if (sb > fmax)

68 {

69 fmax = sb;

70 thresholdValue = k;

71 }

72 }

73 

74 L:

75 for (i =0; i < h; i++) 

76 {

77 np = (unsigned char*)(image->imageData + image->widthStep*i);

78 for (j =0; j < w; j++) 

79 {

80 if(np[j] >= thresholdValue)

81 np[j] =255;

82 else np[j] =0;

83 }

84 }

85 

86 //cout<<"The Threshold of this Image in Otsu is:"<<thresholdValue<<endl;

87 return(thresholdValue);

88 }

 

最大熵阀值 

代码

1 
8 // 计算当前位置的能量熵
9 double caculateCurrentEntropy(CvHistogram * Histogram1,int cur_threshold,entropy_state state)
10 {
11 int start,end;
12 int total =0;
13 double cur_entropy =0.0;
14 if(state == back) 
15 {
16 start =0;
17 end = cur_threshold; 
18 }
19 else 
20 {
21 start = cur_threshold;
22 end =256; 
23 } 
24 for(int i=start;i<end;i++) 
25 {
26 total += (int)cvQueryHistValue_1D(Histogram1,i);//查询直方块的值 P304
27 }
28 for(int j=start;j<end;j++)
29 {
30 if((int)cvQueryHistValue_1D(Histogram1,j)==0)
31 continue;
32 double percentage = cvQueryHistValue_1D(Histogram1,j)/total;
33 
34 cur_entropy +=-percentage*logf(percentage);
35  
37 }
38 return cur_entropy;
39 // return (1-cur_entropy);
40 }
41 
42 //寻找最大熵阈值并分割
43 void MaxEntropy(IplImage *src,IplImage *dst)
44 {
45 assert(src != NULL);
46 assert(src->depth ==8&& dst->depth ==8);
47 assert(src->nChannels ==1);
48 CvHistogram * 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
50 cvCalcHist(&src,hist);//计算直方图
51 double maxentropy =-1.0;
52 int max_index =-1;
53 // 循环测试每个分割点，寻找到最大的阈值分割点
54 for(int i=0;i<HistogramBins;i++) 
55 {
56 double cur_entropy = caculateCurrentEntropy(hist,i,object)+caculateCurrentEntropy(hist,i,back);
57 if(cur_entropy>maxentropy)
58 {
59 maxentropy = cur_entropy;
60 max_index = i;
61 }
62 }
63 cout<<"The Threshold of this Image in MaxEntropy is:"<<max_index<<endl;
64 cvThreshold(src, dst, (double)max_index,255, CV_THRESH_BINARY);
65 cvReleaseHist(&hist);
66 }

 

 

基本全局阀值法 

代码

1 
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 }