I am getting concept checking errors when trying to compute the Laplacian of an unsigned char image. Indeed it is mentioned in the documentation that the input image must be float or double. Is there a reason for this limitation? David