This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
maybe like this:
// 1. convert to CV_16U
Mat u16;
bgr.convertTo(u16, CV_16U); // no scaling factor, keep values "as is" !
// 2. split it (and throw away channel 0):
Mat chan[3];
split(16u, chan);
// 3. join green & red channel.
// now you need to decide, which channel should be the "high",
// and which should be the "low" nibble of your 16bit img
// we cannot simply "shift" Mat's so multiply by 256 instead:
Mat res = chan[1] + chan[2]*256; // M = G | (R<<8)
| | 2 | No.2 Revision |
maybe like this:
// 1. convert to CV_16U
Mat u16;
bgr.convertTo(u16, CV_16U); // no scaling factor, keep values "as is" !
// 2. split it (and throw away channel 0):
Mat chan[3];
split(16u, chan);
// 3. join green & red channel.
// now you need to decide, which channel should one of them will be the "high",
// and which should be "high" 8bit nibble in a 16 bit value, the other the "low" nibble of your 16bit img
nibble.
// we cannot simply "shift" Mat's so multiply by 256 instead:
Mat res = chan[1] + chan[2]*256; // M = G | (R<<8)