Improving Robustness in Printed Side Communications using Higher Order Statistics

This paper uses higher order statistical moments to improve the detection in systems which transmit side information through the print and scan channel, such as 2-D bar codes and text luminance watermarking. When these systems print symbols using halftoning, symbol detection can be performed by evaluating second, third, and fourth order statistical moments of the transmitted symbol, in addition to the average luminance level. A print and scan channel model is described. The relationship between the modulated luminance and the higher order moments of a halftone image is analyzed. The detection error rate is reduced by merging the different moments into a single metric. Experiments validate the analyses and the applicability of the method.