{

g_nThresholdValue = 60;

g_nThresholdType = 1;

nShowMidRst = 1;

nshowcol = 640;

nshowrow = 480;

nCalcol = 800;

nCalrow = 600;

bGetImage =false;

fratio = 0.17;

fminRatio = 0.1;

fmaxRatio = 0.1;

memset(&result, 0, sizeof(result));

//void on_Theshold(int, void *); //回调函数

{

return 1;

{

#if 0 // No GUI

Size dsize = Size(nshowcol,nshowrow);

Mat image2show = Mat(dsize,CV_8U);

::resize(img, image2show,dsize);

cvNamedWindow(szWinName);

imshow(szWinName, image2show);

#endif

return 1;

{

cvtColor(image, img, CV_RGB2GRAY);

//反色处理

img = ~img;

if(img.empty())

return false;

else

{

bGetImage = true;

return true;

}

{

Mat imageROI ,imgResult;

int w = nCalcol, h = nCalrow;

imgResult = img;

if(img.cols >= w)

{

imageROI = img.colRange(img.cols/2 - w/2, img.cols/2 + w/2);

imgResult = img;

}

if(img.rows >= h)

{

imgResult = imageROI.rowRange(imageROI.rows/2 - h/2, imageROI.rows/2 + h/2);

}

return imgResult;

{

int blocksize = 81;

double offset =15;

int threshold=100;

int adaptive_method = CV_ADAPTIVE_THRESH_GAUSSIAN_C;

::threshold(CalCount.dst_gray,CalCount.dstimgbw,CalCount.g_nThresholdValue, 255, CalCount.g_nThresholdType);

//adaptiveThreshold(CalCount.img,CalCount.dstimgbw,255,adaptive_method,CV_THRESH_BINARY,blocksize,0);

//cvNamedWindow("src2");

if(CalCount.nShowMidRst)

CalCount.ShowImage(WINDOWNAME,CalCount.dstimgbw);

//imshow(WINDOWNAME, CalCount.dstimgbw);

{

Mat tmp, tmp_back;

// int col = 800, row = 600;

Size dsize = Size(nshowcol, nshowrow);

Mat image2show = Mat(dsize,CV_8U);

GaussianBlur(img,img,cv::Size(5,5), 1.5);

if(nShowMidRst)

ShowImage("after filtering", img);

//cvSmooth(img, img, CV_MEDIAN, 3, 0, 0, 0); //中值滤波，消除小的噪声；

//Mat element(9,9,CV_8U,Scalar(1));

Mat element = getStructuringElement(MORPH_RECT,Size(10,10));

erode(img, tmp,element);

//ShowImage("erode",tmp);

dilate(tmp,tmp_back,element);

::resize(tmp_back, image2show,dsize);

//cvNamedWindow("dilate");

//imshow("dilate", image2show);

//ShowImage("dilate",tmp_back);

morphologyEx(img,dst_gray,MORPH_TOPHAT,element);

//morphologyEx(img,dst_gray,MORPH_BLACKHAT,cv::Mat());

//dst_gray = img;

//绘制直方图

Histgram1D hist;

hist.stretch(dst_gray,0.01f);

if(nShowMidRst)

ShowImage("Histogram", hist.getHistogramImage(dst_gray, 1));

Mat tmpImage;

//::equalizeHist(dst_gray,tmpImage);

//ShowImage("拉伸后", tmpImage);

//ShowImage("拉伸后Histogram", hist.getHistogramImage(tmpImage, 1));

if(nShowMidRst)

{

ShowImage("Picture EQ", hist.stretch(dst_gray, 50));

ShowImage("after draw, Histogram", hist.getHistogramImage(hist.stretch(dst_gray,50), 1));

}

dst_gray = hist.stretch(dst_gray, 50);

//dst_gray = ::equalizeHist(dst_gray,dst_gray);

//dst_gray = tmp_back - img ;

//dst_gray = img ;

::resize(dst_gray, image2show,dsize);

#if 0 // No GUI

if(nShowMidRst)

cvNamedWindow(WINDOWNAME);

::createTrackbar("filter pattern",WINDOWNAME,&g_nThresholdType,4,on_Theshold);

createTrackbar("theshold value",WINDOWNAME,&g_nThresholdValue,255,on_Theshold);

#endif

//初始化自定义回调函数

on_Theshold(0,0);

//adaptiveThreshold(dst_gray,dstimgbw,255,adaptive_method,CV_THRESH_BINARY,blocksize,0);

//cvAdaptiveThreshold(img, dst_bw,255,adaptive_method,//自适应阀值，blocksize为奇数

// CV_THRESH_BINARY,blocksize,offset);

// ::resize(dstimgbw, image2show,dsize);

// cvNamedWindow("src2");

// imshow("src2", image2show);

// waitKey(0);

vector<vector<Point> > contours;

findContours(dstimgbw,contours,CV_RETR_LIST, CV_CHAIN_APPROX_SIMPLE);

Mat rst(img.size(),CV_8U,Scalar(0));

drawContours(rst,contours,-1,255,1);

int cmin = minArea;

int cmax = maxArea;

vector<vector<Point> >::iterator itc = contours.begin();

int i = 0;

result.LittleNum = 0;

result.AimNum = 0;

result.LargeNum = 0;

fallcount = contours.size();

while(itc!=contours.end())

{

if(itc->size() < cmin )

{

result.LittleNum ++;

itc = contours.erase(itc);

//i++;

}

else if(itc->size() > cmax)

{

itc = contours.erase(itc);

result.LargeNum ++;

}

else

{

result.AimNum ++;

cout <<i<<" = "<< itc->size()<<endl;

i++;

++itc;

}

}

Mat rst2(img.size(),CV_8UC3,Scalar(0,0,0));

drawContours(rst2,contours,-1,cv::Scalar(255,255,255),1);

char sz1[MAX_PATH],sz2[MAX_PATH],sz3[MAX_PATH];

char sz4[MAX_PATH];

char szError[MAX_PATH] = " ";

fcount = fratio * result.AimNum;

if(result.LittleNum/fallcount > fminRatio || result.LargeNum/fallcount > fmaxRatio)

sprintf(szError,"Sample is dirty");

sprintf(sz1,"Aim Num = %d", result.AimNum);

sprintf(sz2,"Little Num=%d", result.LittleNum);

sprintf(sz3,"Large Num=%d", result.LargeNum);

sprintf(sz4,"Count =%3.3f", fcount);

putText(rst2,sz1,cv::Point(10,10),cv::FONT_HERSHEY_PLAIN, 1.0, cv::Scalar(255,0,0), 2);

putText(rst2,sz2,cv::Point(10,30),cv::FONT_HERSHEY_PLAIN, 1.0, cv::Scalar(255,0,0), 2);

putText(rst2,sz3,cv::Point(10,50),cv::FONT_HERSHEY_PLAIN, 1.0, cv::Scalar(255,0,0), 2);

putText(rst2,sz4,cv::Point(10,70),cv::FONT_HERSHEY_PLAIN, 1.0, cv::Scalar(255,0,255), 2);

putText(rst2,szError,cv::Point(10,90),cv::FONT_HERSHEY_PLAIN, 1.0, cv::Scalar(0,0,255), 2);

cout << "Aim Num = "<<result.AimNum<<endl;

cout << "Little Num = "<<result.LittleNum<<endl;

cout << "Large Num = "<<result.LargeNum<<endl;

cout << "Count = "<<fcount<<endl;

cout << "all = "<<contours.size()<<endl;

ShowImage("result",rst2);

//waitKey(0);

return result;