pmse

Synopsis

Description

pmse measures the Mean Squared Error (MSE) between the initial image im_in1 and the restored or enhanced image im_in2.
A lower value for MSE means lesser error. However, the MSE depends on the maximum values of the input images. For example, a MSE=100.0 for Uchar image is very high whilst a MSE=100.0 for Long images with value in [0...65535] is very low.

MSE is defined as follows:

   MSE=1/N * sum { (im_in1(pixel)-im_in2(pixel))^2 }

where N is the total number of pixel in the input image im_in1.

Input images im_in1 and im_in2 must have the same dimensions and the same type.

For color images or multispectral images, the definition of MSE is the same except that the sum over all squared value differences is also divided by the number of bands (ie. 3 for the color images).

Inputs

• im_in1: an image.
• im_in2: an image (restored or enhanced version of im_in1).

Result

Returns the value as a positive real value.
(Use pstatus to get this value).

Examples

Computes the MSE for the meanfilter smoothing operator:

   pmeanfiltering 2 tangram.pan i1.pan
   pmse tangram.pan i1.pan
   pstatus

See Also

C++ prototype

Version française

Calcul de l'Erreur Quadique Moyenne (Mean Square Error).