In this paper, we propose a biologically inspired, global segmentation free methodology for manuscript noise reduction and classification. Our method consists of developing well adapted tools for writing enhancement, background noise, text and drawing separation and handwritten patterns characterization with orientation features. We have used here the spectral domain of handwritten images by frequency decompositions (Hermite transforms) and Gabor filtering for selective text information extraction. The method which has been developed for manuscript authentication has been tested on the 18th century Montesquieu’s corpus. The current results are very promising and show that it is possible to analyze handwritten drawings without any a priori grapheme segmentation.