Hi, i want to sum a 3 channel (lab) matrix to a one channel matrix. Is there a simple way to do this or must i split the image into 3 different CvMat-Objects?
It would to be nice, if there is a parameter for cvCvtColor that add all 3 channels with the same weight.
greetings
Hi,
here's a code sample without cv::reshape() or cv::reduce() :
cv::transform(m_in, m_out, cv::Matx13f(1,1,1))
You can reshape() it to a 3 cols* (initial_rows*initial_cols) rows, then use cv::reduce() by summing up the cols, then reshape it back to initial_rows-by-initial_cols. You won't need the split.
Note that reshape() does not reallocate or move elements, just inteprets their order in a different way, so it it very fast.
This took me way too long to figure out:
cv::Mat reduceChannels(cv::InputArray _src, int _rtype)
{
cv::Mat src = _src.getMat();
cv::Mat tmp;
cv::reduce(src.reshape(1, src.rows*src.cols), tmp, 1, _rtype);
return tmp.reshape(1, src.rows);
}
Note that cv::reduce() does not work on all data types! CV_32S (32bit int) for example is not supported, as there is no data type that would be able to hold the result.
One way to do it would be to split the channels with split and to add them with addWeighted.
yeah in time i splitt the image and use the cvAdd-feature. i thought there is a faster method
use cvMerge after converting the CvMat to CvArr
cvMerge is not performing sum operation. It is only rearranges data from multiple single-channel images to one multi-channel.
Asked: 2012-10-13 09:32:27 -0600
Seen: 13,679 times
Last updated: Dec 04 '16
http://stackoverflow.com/questions/34...