Code covered by the BSD License  

Highlights from
Objective evaluation of binarization methods for document images

Objective evaluation of binarization methods for document images

by

Reza Farrahi Moghaddam

 

18 May 2010 (Updated )

Several measures are implemented to evaluate the output of the binarization methods.

mypublishtest.m
% test example
% December 27th, 2012, By Reza FARRAHI MOGHADDAM and Hossein ZIAEI NAFCHI, Synchromedia Lab, ETS, Montreal, Canada
% April 15th, 2010, By Reza FARRAHI MOGHADDAM, Synchromedia Lab, ETS, Montreal, Canada

% original file name
u_filename = 'P03.tif';

% GT file name
u_GT_filename = 'P03_GT.tif';

% Binarized file name
u_bw_filename = 'P03_adotsu.tif';

% read files
u = double(imread(u_filename)) / 255;
% figure, imshow(u)
u_GT = [(imread(u_GT_filename)) > 0 ];
u_bw = [(imread(u_bw_filename)) > 0 ];
% figure, imshow([u_GT, u_bw])


% calculate the measures
temp_obj_eval = objective_evaluation_core(u_bw, u_GT);
fprintf(' Precision = %9.5f \n Recall = %9.5f \n F-measure (%%) = %9.5f \n Sensitivity = %9.5f \n Specificity = %9.5f \n BCR = %9.5f \n BER (%%) = %9.5f \n F-measure of sens/spec (%%) = %9.5f\n Geometric Accuracy = %9.5f\n pFMeasure (%%) = %9.5f\n NRM  = %9.5f\n PSNR  = %9.5f\n DRD  = %9.5f\n MPM (x1000) = %9.5f \n\n' , ...
    temp_obj_eval.Precision, temp_obj_eval.Recall, temp_obj_eval.Fmeasure, temp_obj_eval.Sensitivity, temp_obj_eval.Specificity, ...
    temp_obj_eval.BCR, temp_obj_eval.BER, temp_obj_eval.SFmeasure, temp_obj_eval.GAccuracy, temp_obj_eval.P_Fmeasure, temp_obj_eval.NRM, temp_obj_eval.PSNR, temp_obj_eval.DRD, 1000 * temp_obj_eval.MPM);
	
	

Contact us