This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
yes, you can do that, but you have to be careful with the type there.
if you read an image with imread(), the depth will be uchar and it's values in [0..255] range.
operations like adding or multiplying will get saturated, meaning, that 130*2 will be 255, not 260, and 17 *-1 will be 0, not -17
if you want to avoid the saturation, you have to convert your images to CV_32S (integer) or CV_32F (float).
then note another pitfall here:
A * 5
will be translated internally as:
A * Scalar(5,0,0,0)
this will only multiply the 1st channel with 5, all others(if they exist) with 0 !!
so, for multi-channel images, you need to explicitly write:
A * Scalar(5,5,5)
| | 2 | No.2 Revision |
yes, you can do that, but you have to be careful with the type there.
if you read an image with imread(), the depth will be uchar and it's values in [0..255] range.
operations like adding or multiplying will get saturated, meaning, that 130*2 will be 255, not 260, and 17 *-1 will be 0, not -17
if you want to avoid the saturation, you have to convert your images to CV_32S (integer) or CV_32F (float).
then note another pitfall here:
A * 5
will be translated internally as:
A * Scalar(5,0,0,0)
this will only multiply the 1st channel with 5, all others(if they exist) with 0 !!
so, for multi-channel images, you need to explicitly write:
A * Scalar(5,5,5)