Normalized Cross Correlation ,Normalized Mean Error , RMSE Calculation Matlab Code

Here We give the matlab code for Normalized Cross Correlation ,Normalized Mean Error and RMSE Calculations.

Normalized Cross Correlation matlab code

function NK = NormalizedCrossCorrelation(origImg, distImg)

if size(origImg,3)==1

    origImg = double(origImg);

    distImg = double(distImg);



    NK = sum(sum(origImg .* distImg)) / sum(sum(origImg^2));

else

    r_rm = NormalizedCrossCorrelation(origImg(:,:,1),distImg(:,:,1));

    g_rm = NormalizedCrossCorrelation(origImg(:,:,2),distImg(:,:,2));

    b_rm = NormalizedCrossCorrelation(origImg(:,:,3),distImg(:,:,3));

    NK = (r_rm+g_rm+b_rm)/3; 

end

 NormalizedMeanError matlab code

function NAE = NormalizedMeanError(origImg, distImg)

if size(origImg,3)==1

    origImg = double(origImg);

    distImg = double(distImg);



    error = origImg - distImg;



    NAE = sum(sum(abs(error))) / sum(sum(origImg));

else

    r_rm = NormalizedMeanError(origImg(:,:,1),distImg(:,:,1));

    g_rm = NormalizedMeanError(origImg(:,:,2),distImg(:,:,2));

    b_rm = NormalizedMeanError(origImg(:,:,3),distImg(:,:,3));

    NAE = (r_rm+g_rm+b_rm)/3; 



end

Root Mean Square Error (RMSE) Matlab code

function rmse_val = RMSE_calc(r,f)

if size(r,3)==1

    [m n] = size(r);

    rmse_val = sqrt(sum(double(r(:))-double(f(:))).^2)/(m*n);

else

    r_rm = RMSE_calc(r(:,:,1),f(:,:,1));

    g_rm = RMSE_calc(r(:,:,2),f(:,:,2));

    b_rm = RMSE_calc(r(:,:,3),f(:,:,3));

    rmse_val = (r_rm+g_rm+b_rm)/3;



end

Kindly Bookmark and Share it: