CODEWITHSUNDEEP
c++opencv
Engine
Engine

Reputation: 5432

getting the mean value of image parts

I'm trying to subdivide a gray frame in multiple small squares and than caluculate for each one of them the mean color value of each one so I can build a result frame that display those values, here 's what I' done : 

int main (){
cv::Mat frame= cv::imread("test2.jpg",0), result, myROI; 
int key = 0; 
int roiSize =10;
cv::Scalar mean(0);
cv::Mat meanS;
meanS = cv::Mat::zeros (frame.rows/roiSize,frame.cols/roiSize,CV_32FC1)  ;
cv::Rect roi;
if(frame.channels()!=1)
    cv::cvtColor(frame,frame,CV_BGR2GRAY);
    for ( int i=0 ; i< frame.cols /roiSize; i++){
        for (int j = 0 ; j < frame.rows/roiSize; j++){
            roi.x= i*roiSize;
            roi.y= j*roiSize;
            roi.height=roiSize;
            roi.width= roiSize;
            myROI = frame(roi);
            cv::imshow("myRoi",myROI);
            mean = cv::mean(myROI);
            std::cout << mean[0] << std::endl;
            meanS.at<float>(j,i) = mean[0];

        }
    }

    //meanS *=1/255; // I've tried this one also, it didn't help ! 
    cv::imshow("the reuslt ",meanS);
    cv::waitKey(0);
    return 0;

}

in the console the values are correct but when I display the result with imshow I get only a white frame ! !! any Idea how can I solve this ? thanks in advance !

Upvotes: 0

Views: 123

Answers (1)

Zaw Lin
Zaw Lin

Reputation: 5708

your comment line is actually correct but it's doing integer division and thus multiplying by zero. just add a dot at the end like meanS *=1/255.; // I've tried this one also, it didn't help !

Upvotes: 3

Related Questions