This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
If you use Opencv 3.0 I would prefer extractChannel(input, output, channel) instead of split(). I don't now if this function exists in OpenCV 2.X. But instead of extracting a COI you can directly access to a pixel by using:
int pixelValue = (int)(img.at<vec3b>(Point(x, y))[channel]);
Where channel is the channel you like to read between 0 and 2 for a 3channel image. Consider that Vec3b is for 3 channel byte Mat, if you have an int Mat use Vec3i and so on.
| | 2 | No.2 Revision |
If you use Opencv 3.0 I would prefer extractChannel(input, output, channel) instead of split(). I don't now if this function exists in OpenCV 2.X. But instead of extracting a COI you can directly access to a pixel by using:
int uchar pixelValue = (int)(img.at<vec3b>(Point(x, y))[channel]);img.at<vec3b>(Point(x, y))[channel];
Where channel is the channel you like to read between 0 and 2 for a 3channel image. Consider that Vec3b is for 3 channel byte Mat, if you have an int Mat use Vec3i and so on.
