Automatic handwriting analysis for writer identification and verification

The work presented in this paper in the field of automatic analysis of handwritten documents for the recognition of individuals through their handwriting. We proposed a hybrid method (local and global) for writer recognition. The proposed method is based on the detection and localization of the points in the image of the Harris editor and the global method that Local Binary Models (LBP) on the thumbnails built around operating points extracted from the image, in order to construct the feature vectors. It is important to note in this regard that the classification is carried out using the famous classifier of large margin separators (SVM). The performances are clearly preliminary based on "CVL", and the experimental results were obtained.