If you have an image and you want to "reshape" it as a 3D matrix, I think that your image is a grayscale one. So if you want to transform your image to a 3D matrix, you have to options:
Convert your image from grayscale to BGR:
Mat color_dst;
cvtColor(gray, color_dst, CV_GRAY2BGR);
Convert your image from grayscale to BGR but with one channel information. So depending on what channel you put the information, you'll get a red, green or blue image:
Mat img, g, color;
img = imread("test.jpg",CV_LOAD_IMAGE_GRAYSCALE);
vector<Mat> channels;
g = Mat::zeros(Size(img.cols, img.rows), CV_8UC1);
imshow("grayscale", img);
channels.push_back(g); //blue information
channels.push_back(g); //green information
channels.push_back(img); //red information
merge(channels, color);
imshow("color_red", color);
//imwrite("result_color_red.jpg",color); //red image
what I have to do is this operation in the picture:
My code:
Mat img = imread("/home/elvio/workspace/aws/src/andaina.bmp");
if(img.empty()){
cout<<"The image is empty \n";
return -1;
}
rows=img.rows;
cols= img.cols;
nChannels= img.channels();
numColComp = nChannels ;
inCompImg = img.reshape(numColComp,rows*cols);
imgDecorr = colorDecorrelation(inCompImg,numColComp); //function colorDecorrelation -obtain: decorrelated image, same dimensions as the original one
rows=imgDecorr.rows;
cols=imgDecorr.cols;
nChannels= imgDecorr.channels();
imgDecorr.convertTo(rgbDecorrImg,CV_8UC3); //to convert the image in colorDecorrelation function
//Infact in colorDecorrelaion I have to convert because of gemm function need a float
//img.convertTo(img_f,CV_32FC3);
//img_f=img_f.reshape(1,img.rows*img.cols); // 1 channel
Now I have that problem. I have to do this operation in Matlab:
rgbDecorr= reshape(imgDecorr, r, c, numColComp);
RGBDECORR is the last operation to obtain this part of the work in the image. Thanks helping
Asked: 2013-11-26 02:48:10 -0600
Seen: 3,000 times
Last updated: Jan 08 '14