Automatic Authentication of Handwritten Documents via Low Density Pixel Measurements

the skilled forgeries, often subclassified into traced and simulated forgeries, involve attempting to mimic the style of the In this paper, we propose an effective approach of automatic writer and thus can be difficult to detect with only these static handwriting authentication via low density pixel measurements features. The main problem comes in designing a feature ex- with an adaptive threshold as decision boundary. For simplictraction method which gives stable features for the genuine ity, we deal only with handwritten signatures in the proposed written samples despite their inevitable variations, and salient scheme, which can well be extended to any piece of handwritfeatures for the forgeries even if the imitations are skillfully ten samples without the loss of generality. Towards this, the done. Even though a genuine writer can never produce ex- low and high density pixel percentages are computed and ten actly the same handwriting twice, e.g., signatures, and many effective features are proposed along with the ratio of above factors can affect signatures and other handwriting including mentioned density percentages. An adaptive decision criteria